Blame view
mm/gup.c
65.8 KB
457c89965 treewide: Add SPD... |
1 |
// SPDX-License-Identifier: GPL-2.0-only |
4bbd4c776 mm: move get_user... |
2 3 4 5 |
#include <linux/kernel.h> #include <linux/errno.h> #include <linux/err.h> #include <linux/spinlock.h> |
4bbd4c776 mm: move get_user... |
6 |
#include <linux/mm.h> |
3565fce3a mm, x86: get_user... |
7 |
#include <linux/memremap.h> |
4bbd4c776 mm: move get_user... |
8 9 10 11 |
#include <linux/pagemap.h> #include <linux/rmap.h> #include <linux/swap.h> #include <linux/swapops.h> |
174cd4b1e sched/headers: Pr... |
12 |
#include <linux/sched/signal.h> |
2667f50e8 mm: introduce a g... |
13 |
#include <linux/rwsem.h> |
f30c59e92 mm: Update generi... |
14 |
#include <linux/hugetlb.h> |
9a4e9f3b2 mm: update get_us... |
15 16 17 |
#include <linux/migrate.h> #include <linux/mm_inline.h> #include <linux/sched/mm.h> |
1027e4436 mm: make GUP hand... |
18 |
|
33a709b25 mm/gup, x86/mm/pk... |
19 |
#include <asm/mmu_context.h> |
2667f50e8 mm: introduce a g... |
20 |
#include <asm/pgtable.h> |
1027e4436 mm: make GUP hand... |
21 |
#include <asm/tlbflush.h> |
2667f50e8 mm: introduce a g... |
22 |
|
4bbd4c776 mm: move get_user... |
23 |
#include "internal.h" |
df06b37ff mm/gup: cache dev... |
24 25 26 27 |
struct follow_page_context { struct dev_pagemap *pgmap; unsigned int page_mask; }; |
fc1d8e7cc mm: introduce put... |
28 |
/** |
2d15eb31b mm/gup: add make_... |
29 30 |
* put_user_pages_dirty_lock() - release and optionally dirty gup-pinned pages * @pages: array of pages to be maybe marked dirty, and definitely released. |
fc1d8e7cc mm: introduce put... |
31 |
* @npages: number of pages in the @pages array. |
2d15eb31b mm/gup: add make_... |
32 |
* @make_dirty: whether to mark the pages dirty |
fc1d8e7cc mm: introduce put... |
33 34 35 36 37 |
* * "gup-pinned page" refers to a page that has had one of the get_user_pages() * variants called on that page. * * For each page in the @pages array, make that page (or its head page, if a |
2d15eb31b mm/gup: add make_... |
38 39 40 |
* compound page) dirty, if @make_dirty is true, and if the page was previously * listed as clean. In any case, releases all pages using put_user_page(), * possibly via put_user_pages(), for the non-dirty case. |
fc1d8e7cc mm: introduce put... |
41 42 43 |
* * Please see the put_user_page() documentation for details. * |
2d15eb31b mm/gup: add make_... |
44 45 46 47 |
* set_page_dirty_lock() is used internally. If instead, set_page_dirty() is * required, then the caller should a) verify that this is really correct, * because _lock() is usually required, and b) hand code it: * set_page_dirty_lock(), put_user_page(). |
fc1d8e7cc mm: introduce put... |
48 49 |
* */ |
2d15eb31b mm/gup: add make_... |
50 51 |
void put_user_pages_dirty_lock(struct page **pages, unsigned long npages, bool make_dirty) |
fc1d8e7cc mm: introduce put... |
52 |
{ |
2d15eb31b mm/gup: add make_... |
53 |
unsigned long index; |
fc1d8e7cc mm: introduce put... |
54 |
|
2d15eb31b mm/gup: add make_... |
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 |
/* * TODO: this can be optimized for huge pages: if a series of pages is * physically contiguous and part of the same compound page, then a * single operation to the head page should suffice. */ if (!make_dirty) { put_user_pages(pages, npages); return; } for (index = 0; index < npages; index++) { struct page *page = compound_head(pages[index]); /* * Checking PageDirty at this point may race with * clear_page_dirty_for_io(), but that's OK. Two key * cases: * * 1) This code sees the page as already dirty, so it * skips the call to set_page_dirty(). That could happen * because clear_page_dirty_for_io() called * page_mkclean(), followed by set_page_dirty(). * However, now the page is going to get written back, * which meets the original intention of setting it * dirty, so all is well: clear_page_dirty_for_io() goes * on to call TestClearPageDirty(), and write the page * back. * * 2) This code sees the page as clean, so it calls * set_page_dirty(). The page stays dirty, despite being * written back, so it gets written back again in the * next writeback cycle. This is harmless. */ if (!PageDirty(page)) set_page_dirty_lock(page); put_user_page(page); } |
fc1d8e7cc mm: introduce put... |
92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 |
} EXPORT_SYMBOL(put_user_pages_dirty_lock); /** * put_user_pages() - release an array of gup-pinned pages. * @pages: array of pages to be marked dirty and released. * @npages: number of pages in the @pages array. * * For each page in the @pages array, release the page using put_user_page(). * * Please see the put_user_page() documentation for details. */ void put_user_pages(struct page **pages, unsigned long npages) { unsigned long index; /* * TODO: this can be optimized for huge pages: if a series of pages is * physically contiguous and part of the same compound page, then a * single operation to the head page should suffice. */ for (index = 0; index < npages; index++) put_user_page(pages[index]); } EXPORT_SYMBOL(put_user_pages); |
050a9adc6 mm: consolidate t... |
117 |
#ifdef CONFIG_MMU |
69e68b4f0 mm: cleanup follo... |
118 119 |
static struct page *no_page_table(struct vm_area_struct *vma, unsigned int flags) |
4bbd4c776 mm: move get_user... |
120 |
{ |
69e68b4f0 mm: cleanup follo... |
121 122 123 124 125 126 127 128 129 130 131 132 |
/* * When core dumping an enormous anonymous area that nobody * has touched so far, we don't want to allocate unnecessary pages or * page tables. Return error instead of NULL to skip handle_mm_fault, * then get_dump_page() will return NULL to leave a hole in the dump. * But we can only make this optimization where a hole would surely * be zero-filled if handle_mm_fault() actually did handle it. */ if ((flags & FOLL_DUMP) && (!vma->vm_ops || !vma->vm_ops->fault)) return ERR_PTR(-EFAULT); return NULL; } |
4bbd4c776 mm: move get_user... |
133 |
|
1027e4436 mm: make GUP hand... |
134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 |
static int follow_pfn_pte(struct vm_area_struct *vma, unsigned long address, pte_t *pte, unsigned int flags) { /* No page to get reference */ if (flags & FOLL_GET) return -EFAULT; if (flags & FOLL_TOUCH) { pte_t entry = *pte; if (flags & FOLL_WRITE) entry = pte_mkdirty(entry); entry = pte_mkyoung(entry); if (!pte_same(*pte, entry)) { set_pte_at(vma->vm_mm, address, pte, entry); update_mmu_cache(vma, address, pte); } } /* Proper page table entry exists, but no corresponding struct page */ return -EEXIST; } |
19be0eaff mm: remove gup_fl... |
157 158 159 160 161 162 |
/* * FOLL_FORCE can write to even unwritable pte's, but only * after we've gone through a COW cycle and they are dirty. */ static inline bool can_follow_write_pte(pte_t pte, unsigned int flags) { |
f6f373216 Revert "mm: repla... |
163 |
return pte_write(pte) || |
19be0eaff mm: remove gup_fl... |
164 165 |
((flags & FOLL_FORCE) && (flags & FOLL_COW) && pte_dirty(pte)); } |
69e68b4f0 mm: cleanup follo... |
166 |
static struct page *follow_page_pte(struct vm_area_struct *vma, |
df06b37ff mm/gup: cache dev... |
167 168 |
unsigned long address, pmd_t *pmd, unsigned int flags, struct dev_pagemap **pgmap) |
69e68b4f0 mm: cleanup follo... |
169 170 171 172 173 |
{ struct mm_struct *mm = vma->vm_mm; struct page *page; spinlock_t *ptl; pte_t *ptep, pte; |
4bbd4c776 mm: move get_user... |
174 |
|
69e68b4f0 mm: cleanup follo... |
175 |
retry: |
4bbd4c776 mm: move get_user... |
176 |
if (unlikely(pmd_bad(*pmd))) |
69e68b4f0 mm: cleanup follo... |
177 |
return no_page_table(vma, flags); |
4bbd4c776 mm: move get_user... |
178 179 |
ptep = pte_offset_map_lock(mm, pmd, address, &ptl); |
4bbd4c776 mm: move get_user... |
180 181 182 183 184 185 186 187 188 189 |
pte = *ptep; if (!pte_present(pte)) { swp_entry_t entry; /* * KSM's break_ksm() relies upon recognizing a ksm page * even while it is being migrated, so for that case we * need migration_entry_wait(). */ if (likely(!(flags & FOLL_MIGRATION))) goto no_page; |
0661a3361 mm: remove rest u... |
190 |
if (pte_none(pte)) |
4bbd4c776 mm: move get_user... |
191 192 193 194 195 196 |
goto no_page; entry = pte_to_swp_entry(pte); if (!is_migration_entry(entry)) goto no_page; pte_unmap_unlock(ptep, ptl); migration_entry_wait(mm, pmd, address); |
69e68b4f0 mm: cleanup follo... |
197 |
goto retry; |
4bbd4c776 mm: move get_user... |
198 |
} |
8a0516ed8 mm: convert p[te|... |
199 |
if ((flags & FOLL_NUMA) && pte_protnone(pte)) |
4bbd4c776 mm: move get_user... |
200 |
goto no_page; |
19be0eaff mm: remove gup_fl... |
201 |
if ((flags & FOLL_WRITE) && !can_follow_write_pte(pte, flags)) { |
69e68b4f0 mm: cleanup follo... |
202 203 204 |
pte_unmap_unlock(ptep, ptl); return NULL; } |
4bbd4c776 mm: move get_user... |
205 206 |
page = vm_normal_page(vma, address, pte); |
3565fce3a mm, x86: get_user... |
207 208 209 210 211 |
if (!page && pte_devmap(pte) && (flags & FOLL_GET)) { /* * Only return device mapping pages in the FOLL_GET case since * they are only valid while holding the pgmap reference. */ |
df06b37ff mm/gup: cache dev... |
212 213 |
*pgmap = get_dev_pagemap(pte_pfn(pte), *pgmap); if (*pgmap) |
3565fce3a mm, x86: get_user... |
214 215 216 217 |
page = pte_page(pte); else goto no_page; } else if (unlikely(!page)) { |
1027e4436 mm: make GUP hand... |
218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 |
if (flags & FOLL_DUMP) { /* Avoid special (like zero) pages in core dumps */ page = ERR_PTR(-EFAULT); goto out; } if (is_zero_pfn(pte_pfn(pte))) { page = pte_page(pte); } else { int ret; ret = follow_pfn_pte(vma, address, ptep, flags); page = ERR_PTR(ret); goto out; } |
4bbd4c776 mm: move get_user... |
233 |
} |
6742d293c mm: adjust FOLL_S... |
234 235 236 237 238 239 240 241 242 243 244 245 |
if (flags & FOLL_SPLIT && PageTransCompound(page)) { int ret; get_page(page); pte_unmap_unlock(ptep, ptl); lock_page(page); ret = split_huge_page(page); unlock_page(page); put_page(page); if (ret) return ERR_PTR(ret); goto retry; } |
8fde12ca7 mm: prevent get_u... |
246 247 248 249 250 251 |
if (flags & FOLL_GET) { if (unlikely(!try_get_page(page))) { page = ERR_PTR(-ENOMEM); goto out; } } |
4bbd4c776 mm: move get_user... |
252 253 254 255 256 257 258 259 260 261 262 |
if (flags & FOLL_TOUCH) { if ((flags & FOLL_WRITE) && !pte_dirty(pte) && !PageDirty(page)) set_page_dirty(page); /* * pte_mkyoung() would be more correct here, but atomic care * is needed to avoid losing the dirty bit: it is easier to use * mark_page_accessed(). */ mark_page_accessed(page); } |
de60f5f10 mm: introduce VM_... |
263 |
if ((flags & FOLL_MLOCK) && (vma->vm_flags & VM_LOCKED)) { |
e90309c9f thp: allow mlocke... |
264 265 266 |
/* Do not mlock pte-mapped THP */ if (PageTransCompound(page)) goto out; |
4bbd4c776 mm: move get_user... |
267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 |
/* * The preliminary mapping check is mainly to avoid the * pointless overhead of lock_page on the ZERO_PAGE * which might bounce very badly if there is contention. * * If the page is already locked, we don't need to * handle it now - vmscan will handle it later if and * when it attempts to reclaim the page. */ if (page->mapping && trylock_page(page)) { lru_add_drain(); /* push cached pages to LRU */ /* * Because we lock page here, and migration is * blocked by the pte's page reference, and we * know the page is still mapped, we don't even * need to check for file-cache page truncation. */ mlock_vma_page(page); unlock_page(page); } } |
1027e4436 mm: make GUP hand... |
288 |
out: |
4bbd4c776 mm: move get_user... |
289 |
pte_unmap_unlock(ptep, ptl); |
4bbd4c776 mm: move get_user... |
290 |
return page; |
4bbd4c776 mm: move get_user... |
291 292 293 |
no_page: pte_unmap_unlock(ptep, ptl); if (!pte_none(pte)) |
69e68b4f0 mm: cleanup follo... |
294 295 296 |
return NULL; return no_page_table(vma, flags); } |
080dbb618 mm/follow_page_ma... |
297 298 |
static struct page *follow_pmd_mask(struct vm_area_struct *vma, unsigned long address, pud_t *pudp, |
df06b37ff mm/gup: cache dev... |
299 300 |
unsigned int flags, struct follow_page_context *ctx) |
69e68b4f0 mm: cleanup follo... |
301 |
{ |
688272809 mm, gup: prevent ... |
302 |
pmd_t *pmd, pmdval; |
69e68b4f0 mm: cleanup follo... |
303 304 305 |
spinlock_t *ptl; struct page *page; struct mm_struct *mm = vma->vm_mm; |
080dbb618 mm/follow_page_ma... |
306 |
pmd = pmd_offset(pudp, address); |
688272809 mm, gup: prevent ... |
307 308 309 310 311 312 |
/* * The READ_ONCE() will stabilize the pmdval in a register or * on the stack so that it will stop changing under the code. */ pmdval = READ_ONCE(*pmd); if (pmd_none(pmdval)) |
69e68b4f0 mm: cleanup follo... |
313 |
return no_page_table(vma, flags); |
688272809 mm, gup: prevent ... |
314 |
if (pmd_huge(pmdval) && vma->vm_flags & VM_HUGETLB) { |
e66f17ff7 mm/hugetlb: take ... |
315 316 317 318 |
page = follow_huge_pmd(mm, address, pmd, flags); if (page) return page; return no_page_table(vma, flags); |
69e68b4f0 mm: cleanup follo... |
319 |
} |
688272809 mm, gup: prevent ... |
320 |
if (is_hugepd(__hugepd(pmd_val(pmdval)))) { |
4dc71451a mm/follow_page_ma... |
321 |
page = follow_huge_pd(vma, address, |
688272809 mm, gup: prevent ... |
322 |
__hugepd(pmd_val(pmdval)), flags, |
4dc71451a mm/follow_page_ma... |
323 324 325 326 327 |
PMD_SHIFT); if (page) return page; return no_page_table(vma, flags); } |
84c3fc4e9 mm: thp: check pm... |
328 |
retry: |
688272809 mm, gup: prevent ... |
329 |
if (!pmd_present(pmdval)) { |
84c3fc4e9 mm: thp: check pm... |
330 331 332 |
if (likely(!(flags & FOLL_MIGRATION))) return no_page_table(vma, flags); VM_BUG_ON(thp_migration_supported() && |
688272809 mm, gup: prevent ... |
333 334 |
!is_pmd_migration_entry(pmdval)); if (is_pmd_migration_entry(pmdval)) |
84c3fc4e9 mm: thp: check pm... |
335 |
pmd_migration_entry_wait(mm, pmd); |
688272809 mm, gup: prevent ... |
336 337 338 339 340 341 342 |
pmdval = READ_ONCE(*pmd); /* * MADV_DONTNEED may convert the pmd to null because * mmap_sem is held in read mode */ if (pmd_none(pmdval)) return no_page_table(vma, flags); |
84c3fc4e9 mm: thp: check pm... |
343 344 |
goto retry; } |
688272809 mm, gup: prevent ... |
345 |
if (pmd_devmap(pmdval)) { |
3565fce3a mm, x86: get_user... |
346 |
ptl = pmd_lock(mm, pmd); |
df06b37ff mm/gup: cache dev... |
347 |
page = follow_devmap_pmd(vma, address, pmd, flags, &ctx->pgmap); |
3565fce3a mm, x86: get_user... |
348 349 350 351 |
spin_unlock(ptl); if (page) return page; } |
688272809 mm, gup: prevent ... |
352 |
if (likely(!pmd_trans_huge(pmdval))) |
df06b37ff mm/gup: cache dev... |
353 |
return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap); |
6742d293c mm: adjust FOLL_S... |
354 |
|
688272809 mm, gup: prevent ... |
355 |
if ((flags & FOLL_NUMA) && pmd_protnone(pmdval)) |
db08f2030 mm/gup: check for... |
356 |
return no_page_table(vma, flags); |
84c3fc4e9 mm: thp: check pm... |
357 |
retry_locked: |
6742d293c mm: adjust FOLL_S... |
358 |
ptl = pmd_lock(mm, pmd); |
688272809 mm, gup: prevent ... |
359 360 361 362 |
if (unlikely(pmd_none(*pmd))) { spin_unlock(ptl); return no_page_table(vma, flags); } |
84c3fc4e9 mm: thp: check pm... |
363 364 365 366 367 368 369 |
if (unlikely(!pmd_present(*pmd))) { spin_unlock(ptl); if (likely(!(flags & FOLL_MIGRATION))) return no_page_table(vma, flags); pmd_migration_entry_wait(mm, pmd); goto retry_locked; } |
6742d293c mm: adjust FOLL_S... |
370 371 |
if (unlikely(!pmd_trans_huge(*pmd))) { spin_unlock(ptl); |
df06b37ff mm/gup: cache dev... |
372 |
return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap); |
6742d293c mm: adjust FOLL_S... |
373 |
} |
bfe7b00de mm, thp: introduc... |
374 |
if (flags & (FOLL_SPLIT | FOLL_SPLIT_PMD)) { |
6742d293c mm: adjust FOLL_S... |
375 376 377 378 379 |
int ret; page = pmd_page(*pmd); if (is_huge_zero_page(page)) { spin_unlock(ptl); ret = 0; |
78ddc5347 thp: rename split... |
380 |
split_huge_pmd(vma, pmd, address); |
337d9abf1 mm: thp: check pm... |
381 382 |
if (pmd_trans_unstable(pmd)) ret = -EBUSY; |
bfe7b00de mm, thp: introduc... |
383 |
} else if (flags & FOLL_SPLIT) { |
8fde12ca7 mm: prevent get_u... |
384 385 386 387 |
if (unlikely(!try_get_page(page))) { spin_unlock(ptl); return ERR_PTR(-ENOMEM); } |
69e68b4f0 mm: cleanup follo... |
388 |
spin_unlock(ptl); |
6742d293c mm: adjust FOLL_S... |
389 390 391 392 |
lock_page(page); ret = split_huge_page(page); unlock_page(page); put_page(page); |
baa355fd3 thp: file pages s... |
393 394 |
if (pmd_none(*pmd)) return no_page_table(vma, flags); |
bfe7b00de mm, thp: introduc... |
395 396 397 398 |
} else { /* flags & FOLL_SPLIT_PMD */ spin_unlock(ptl); split_huge_pmd(vma, pmd, address); ret = pte_alloc(mm, pmd) ? -ENOMEM : 0; |
6742d293c mm: adjust FOLL_S... |
399 400 401 |
} return ret ? ERR_PTR(ret) : |
df06b37ff mm/gup: cache dev... |
402 |
follow_page_pte(vma, address, pmd, flags, &ctx->pgmap); |
69e68b4f0 mm: cleanup follo... |
403 |
} |
6742d293c mm: adjust FOLL_S... |
404 405 |
page = follow_trans_huge_pmd(vma, address, pmd, flags); spin_unlock(ptl); |
df06b37ff mm/gup: cache dev... |
406 |
ctx->page_mask = HPAGE_PMD_NR - 1; |
6742d293c mm: adjust FOLL_S... |
407 |
return page; |
4bbd4c776 mm: move get_user... |
408 |
} |
080dbb618 mm/follow_page_ma... |
409 410 |
static struct page *follow_pud_mask(struct vm_area_struct *vma, unsigned long address, p4d_t *p4dp, |
df06b37ff mm/gup: cache dev... |
411 412 |
unsigned int flags, struct follow_page_context *ctx) |
080dbb618 mm/follow_page_ma... |
413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 |
{ pud_t *pud; spinlock_t *ptl; struct page *page; struct mm_struct *mm = vma->vm_mm; pud = pud_offset(p4dp, address); if (pud_none(*pud)) return no_page_table(vma, flags); if (pud_huge(*pud) && vma->vm_flags & VM_HUGETLB) { page = follow_huge_pud(mm, address, pud, flags); if (page) return page; return no_page_table(vma, flags); } |
4dc71451a mm/follow_page_ma... |
428 429 430 431 432 433 434 435 |
if (is_hugepd(__hugepd(pud_val(*pud)))) { page = follow_huge_pd(vma, address, __hugepd(pud_val(*pud)), flags, PUD_SHIFT); if (page) return page; return no_page_table(vma, flags); } |
080dbb618 mm/follow_page_ma... |
436 437 |
if (pud_devmap(*pud)) { ptl = pud_lock(mm, pud); |
df06b37ff mm/gup: cache dev... |
438 |
page = follow_devmap_pud(vma, address, pud, flags, &ctx->pgmap); |
080dbb618 mm/follow_page_ma... |
439 440 441 442 443 444 |
spin_unlock(ptl); if (page) return page; } if (unlikely(pud_bad(*pud))) return no_page_table(vma, flags); |
df06b37ff mm/gup: cache dev... |
445 |
return follow_pmd_mask(vma, address, pud, flags, ctx); |
080dbb618 mm/follow_page_ma... |
446 |
} |
080dbb618 mm/follow_page_ma... |
447 448 |
static struct page *follow_p4d_mask(struct vm_area_struct *vma, unsigned long address, pgd_t *pgdp, |
df06b37ff mm/gup: cache dev... |
449 450 |
unsigned int flags, struct follow_page_context *ctx) |
080dbb618 mm/follow_page_ma... |
451 452 |
{ p4d_t *p4d; |
4dc71451a mm/follow_page_ma... |
453 |
struct page *page; |
080dbb618 mm/follow_page_ma... |
454 455 456 457 458 459 460 |
p4d = p4d_offset(pgdp, address); if (p4d_none(*p4d)) return no_page_table(vma, flags); BUILD_BUG_ON(p4d_huge(*p4d)); if (unlikely(p4d_bad(*p4d))) return no_page_table(vma, flags); |
4dc71451a mm/follow_page_ma... |
461 462 463 464 465 466 467 468 |
if (is_hugepd(__hugepd(p4d_val(*p4d)))) { page = follow_huge_pd(vma, address, __hugepd(p4d_val(*p4d)), flags, P4D_SHIFT); if (page) return page; return no_page_table(vma, flags); } |
df06b37ff mm/gup: cache dev... |
469 |
return follow_pud_mask(vma, address, p4d, flags, ctx); |
080dbb618 mm/follow_page_ma... |
470 471 472 473 474 475 476 |
} /** * follow_page_mask - look up a page descriptor from a user-virtual address * @vma: vm_area_struct mapping @address * @address: virtual address to look up * @flags: flags modifying lookup behaviour |
78179556e mm/gup.c: fix fol... |
477 478 |
* @ctx: contains dev_pagemap for %ZONE_DEVICE memory pinning and a * pointer to output page_mask |
080dbb618 mm/follow_page_ma... |
479 480 481 |
* * @flags can have FOLL_ flags set, defined in <linux/mm.h> * |
78179556e mm/gup.c: fix fol... |
482 483 484 485 486 487 |
* When getting pages from ZONE_DEVICE memory, the @ctx->pgmap caches * the device's dev_pagemap metadata to avoid repeating expensive lookups. * * On output, the @ctx->page_mask is set according to the size of the page. * * Return: the mapped (struct page *), %NULL if no mapping exists, or |
080dbb618 mm/follow_page_ma... |
488 489 490 |
* an error pointer if there is a mapping to something not represented * by a page descriptor (see also vm_normal_page()). */ |
a7030aea2 mm/gup.c: make fo... |
491 |
static struct page *follow_page_mask(struct vm_area_struct *vma, |
080dbb618 mm/follow_page_ma... |
492 |
unsigned long address, unsigned int flags, |
df06b37ff mm/gup: cache dev... |
493 |
struct follow_page_context *ctx) |
080dbb618 mm/follow_page_ma... |
494 495 496 497 |
{ pgd_t *pgd; struct page *page; struct mm_struct *mm = vma->vm_mm; |
df06b37ff mm/gup: cache dev... |
498 |
ctx->page_mask = 0; |
080dbb618 mm/follow_page_ma... |
499 500 501 502 503 504 505 506 507 508 509 510 |
/* make this handle hugepd */ page = follow_huge_addr(mm, address, flags & FOLL_WRITE); if (!IS_ERR(page)) { BUG_ON(flags & FOLL_GET); return page; } pgd = pgd_offset(mm, address); if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd))) return no_page_table(vma, flags); |
faaa5b62d mm/follow_page_ma... |
511 512 513 514 515 516 |
if (pgd_huge(*pgd)) { page = follow_huge_pgd(mm, address, pgd, flags); if (page) return page; return no_page_table(vma, flags); } |
4dc71451a mm/follow_page_ma... |
517 518 519 520 521 522 523 524 |
if (is_hugepd(__hugepd(pgd_val(*pgd)))) { page = follow_huge_pd(vma, address, __hugepd(pgd_val(*pgd)), flags, PGDIR_SHIFT); if (page) return page; return no_page_table(vma, flags); } |
faaa5b62d mm/follow_page_ma... |
525 |
|
df06b37ff mm/gup: cache dev... |
526 527 528 529 530 531 532 533 534 535 536 537 538 |
return follow_p4d_mask(vma, address, pgd, flags, ctx); } struct page *follow_page(struct vm_area_struct *vma, unsigned long address, unsigned int foll_flags) { struct follow_page_context ctx = { NULL }; struct page *page; page = follow_page_mask(vma, address, foll_flags, &ctx); if (ctx.pgmap) put_dev_pagemap(ctx.pgmap); return page; |
080dbb618 mm/follow_page_ma... |
539 |
} |
f2b495ca8 mm: extract in_ga... |
540 541 542 543 544 |
static int get_gate_page(struct mm_struct *mm, unsigned long address, unsigned int gup_flags, struct vm_area_struct **vma, struct page **page) { pgd_t *pgd; |
c2febafc6 mm: convert gener... |
545 |
p4d_t *p4d; |
f2b495ca8 mm: extract in_ga... |
546 547 548 549 550 551 552 553 554 555 556 557 |
pud_t *pud; pmd_t *pmd; pte_t *pte; int ret = -EFAULT; /* user gate pages are read-only */ if (gup_flags & FOLL_WRITE) return -EFAULT; if (address > TASK_SIZE) pgd = pgd_offset_k(address); else pgd = pgd_offset_gate(mm, address); |
b5d1c39f3 mm/gup.c: remove ... |
558 559 |
if (pgd_none(*pgd)) return -EFAULT; |
c2febafc6 mm: convert gener... |
560 |
p4d = p4d_offset(pgd, address); |
b5d1c39f3 mm/gup.c: remove ... |
561 562 |
if (p4d_none(*p4d)) return -EFAULT; |
c2febafc6 mm: convert gener... |
563 |
pud = pud_offset(p4d, address); |
b5d1c39f3 mm/gup.c: remove ... |
564 565 |
if (pud_none(*pud)) return -EFAULT; |
f2b495ca8 mm: extract in_ga... |
566 |
pmd = pmd_offset(pud, address); |
84c3fc4e9 mm: thp: check pm... |
567 |
if (!pmd_present(*pmd)) |
f2b495ca8 mm: extract in_ga... |
568 569 570 571 572 573 574 575 576 577 578 579 580 581 |
return -EFAULT; VM_BUG_ON(pmd_trans_huge(*pmd)); pte = pte_offset_map(pmd, address); if (pte_none(*pte)) goto unmap; *vma = get_gate_vma(mm); if (!page) goto out; *page = vm_normal_page(*vma, address, *pte); if (!*page) { if ((gup_flags & FOLL_DUMP) || !is_zero_pfn(pte_pfn(*pte))) goto unmap; *page = pte_page(*pte); } |
8fde12ca7 mm: prevent get_u... |
582 583 584 585 |
if (unlikely(!try_get_page(*page))) { ret = -ENOMEM; goto unmap; } |
f2b495ca8 mm: extract in_ga... |
586 587 588 589 590 591 |
out: ret = 0; unmap: pte_unmap(pte); return ret; } |
9a95f3cf7 mm: describe mmap... |
592 593 594 595 596 |
/* * mmap_sem must be held on entry. If @nonblocking != NULL and * *@flags does not include FOLL_NOWAIT, the mmap_sem may be released. * If it is, *@nonblocking will be set to 0 and -EBUSY returned. */ |
167444834 mm: extract code ... |
597 598 599 |
static int faultin_page(struct task_struct *tsk, struct vm_area_struct *vma, unsigned long address, unsigned int *flags, int *nonblocking) { |
167444834 mm: extract code ... |
600 |
unsigned int fault_flags = 0; |
2b7403035 mm: Change return... |
601 |
vm_fault_t ret; |
167444834 mm: extract code ... |
602 |
|
de60f5f10 mm: introduce VM_... |
603 604 605 |
/* mlock all present pages, but do not fault in new pages */ if ((*flags & (FOLL_POPULATE | FOLL_MLOCK)) == FOLL_MLOCK) return -ENOENT; |
167444834 mm: extract code ... |
606 607 |
if (*flags & FOLL_WRITE) fault_flags |= FAULT_FLAG_WRITE; |
1b2ee1266 mm/core: Do not e... |
608 609 |
if (*flags & FOLL_REMOTE) fault_flags |= FAULT_FLAG_REMOTE; |
167444834 mm: extract code ... |
610 611 612 613 |
if (nonblocking) fault_flags |= FAULT_FLAG_ALLOW_RETRY; if (*flags & FOLL_NOWAIT) fault_flags |= FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_RETRY_NOWAIT; |
234b239be kvm: Faults which... |
614 615 616 617 |
if (*flags & FOLL_TRIED) { VM_WARN_ON_ONCE(fault_flags & FAULT_FLAG_ALLOW_RETRY); fault_flags |= FAULT_FLAG_TRIED; } |
167444834 mm: extract code ... |
618 |
|
dcddffd41 mm: do not pass m... |
619 |
ret = handle_mm_fault(vma, address, fault_flags); |
167444834 mm: extract code ... |
620 |
if (ret & VM_FAULT_ERROR) { |
9a291a7c9 mm/hugetlb: repor... |
621 622 623 624 |
int err = vm_fault_to_errno(ret, *flags); if (err) return err; |
167444834 mm: extract code ... |
625 626 627 628 629 630 631 632 633 634 635 |
BUG(); } if (tsk) { if (ret & VM_FAULT_MAJOR) tsk->maj_flt++; else tsk->min_flt++; } if (ret & VM_FAULT_RETRY) { |
96312e612 mm/gup.c: teach g... |
636 |
if (nonblocking && !(fault_flags & FAULT_FLAG_RETRY_NOWAIT)) |
167444834 mm: extract code ... |
637 638 639 640 641 642 643 644 645 646 647 648 649 650 |
*nonblocking = 0; return -EBUSY; } /* * The VM_FAULT_WRITE bit tells us that do_wp_page has broken COW when * necessary, even if maybe_mkwrite decided not to set pte_write. We * can thus safely do subsequent page lookups as if they were reads. * But only do so when looping for pte_write is futile: in some cases * userspace may also be wanting to write to the gotten user page, * which a read fault here might prevent (a readonly page might get * reCOWed by userspace write). */ if ((ret & VM_FAULT_WRITE) && !(vma->vm_flags & VM_WRITE)) |
2923117b7 mm/gup.c: fix cod... |
651 |
*flags |= FOLL_COW; |
167444834 mm: extract code ... |
652 653 |
return 0; } |
fa5bb2093 mm: cleanup __get... |
654 655 656 |
static int check_vma_flags(struct vm_area_struct *vma, unsigned long gup_flags) { vm_flags_t vm_flags = vma->vm_flags; |
1b2ee1266 mm/core: Do not e... |
657 658 |
int write = (gup_flags & FOLL_WRITE); int foreign = (gup_flags & FOLL_REMOTE); |
fa5bb2093 mm: cleanup __get... |
659 660 661 |
if (vm_flags & (VM_IO | VM_PFNMAP)) return -EFAULT; |
7f7ccc2cc proc: do not acce... |
662 663 |
if (gup_flags & FOLL_ANON && !vma_is_anonymous(vma)) return -EFAULT; |
1b2ee1266 mm/core: Do not e... |
664 |
if (write) { |
fa5bb2093 mm: cleanup __get... |
665 666 667 668 669 670 671 672 673 674 675 676 |
if (!(vm_flags & VM_WRITE)) { if (!(gup_flags & FOLL_FORCE)) return -EFAULT; /* * We used to let the write,force case do COW in a * VM_MAYWRITE VM_SHARED !VM_WRITE vma, so ptrace could * set a breakpoint in a read-only mapping of an * executable, without corrupting the file (yet only * when that file had been opened for writing!). * Anon pages in shared mappings are surprising: now * just reject it. */ |
464353647 mm: retire GUP WA... |
677 |
if (!is_cow_mapping(vm_flags)) |
fa5bb2093 mm: cleanup __get... |
678 |
return -EFAULT; |
fa5bb2093 mm: cleanup __get... |
679 680 681 682 683 684 685 686 687 688 689 |
} } else if (!(vm_flags & VM_READ)) { if (!(gup_flags & FOLL_FORCE)) return -EFAULT; /* * Is there actually any vma we can reach here which does not * have VM_MAYREAD set? */ if (!(vm_flags & VM_MAYREAD)) return -EFAULT; } |
d61172b4b mm/core, x86/mm/p... |
690 691 692 693 694 |
/* * gups are always data accesses, not instruction * fetches, so execute=false here */ if (!arch_vma_access_permitted(vma, write, false, foreign)) |
33a709b25 mm/gup, x86/mm/pk... |
695 |
return -EFAULT; |
fa5bb2093 mm: cleanup __get... |
696 697 |
return 0; } |
4bbd4c776 mm: move get_user... |
698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 |
/** * __get_user_pages() - pin user pages in memory * @tsk: task_struct of target task * @mm: mm_struct of target mm * @start: starting user address * @nr_pages: number of pages from start to pin * @gup_flags: flags modifying pin behaviour * @pages: array that receives pointers to the pages pinned. * Should be at least nr_pages long. Or NULL, if caller * only intends to ensure the pages are faulted in. * @vmas: array of pointers to vmas corresponding to each page. * Or NULL if the caller does not require them. * @nonblocking: whether waiting for disk IO or mmap_sem contention * * Returns number of pages pinned. This may be fewer than the number * requested. If nr_pages is 0 or negative, returns 0. If no pages * were pinned, returns -errno. Each page returned must be released * with a put_page() call when it is finished with. vmas will only * remain valid while mmap_sem is held. * |
9a95f3cf7 mm: describe mmap... |
718 |
* Must be called with mmap_sem held. It may be released. See below. |
4bbd4c776 mm: move get_user... |
719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 |
* * __get_user_pages walks a process's page tables and takes a reference to * each struct page that each user address corresponds to at a given * instant. That is, it takes the page that would be accessed if a user * thread accesses the given user virtual address at that instant. * * This does not guarantee that the page exists in the user mappings when * __get_user_pages returns, and there may even be a completely different * page there in some cases (eg. if mmapped pagecache has been invalidated * and subsequently re faulted). However it does guarantee that the page * won't be freed completely. And mostly callers simply care that the page * contains data that was valid *at some point in time*. Typically, an IO * or similar operation cannot guarantee anything stronger anyway because * locks can't be held over the syscall boundary. * * If @gup_flags & FOLL_WRITE == 0, the page must not be written to. If * the page is written to, set_page_dirty (or set_page_dirty_lock, as * appropriate) must be called after the page is finished with, and * before put_page is called. * * If @nonblocking != NULL, __get_user_pages will not wait for disk IO * or mmap_sem contention, and if waiting is needed to pin all pages, |
9a95f3cf7 mm: describe mmap... |
741 742 743 744 745 746 747 748 |
* *@nonblocking will be set to 0. Further, if @gup_flags does not * include FOLL_NOWAIT, the mmap_sem will be released via up_read() in * this case. * * A caller using such a combination of @nonblocking and @gup_flags * must therefore hold the mmap_sem for reading only, and recognize * when it's been released. Otherwise, it must be held for either * reading or writing and will not be released. |
4bbd4c776 mm: move get_user... |
749 750 751 752 753 |
* * In most cases, get_user_pages or get_user_pages_fast should be used * instead of __get_user_pages. __get_user_pages should be used only if * you need some special @gup_flags. */ |
0d7317598 mm: unexport __ge... |
754 |
static long __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, |
4bbd4c776 mm: move get_user... |
755 756 757 758 |
unsigned long start, unsigned long nr_pages, unsigned int gup_flags, struct page **pages, struct vm_area_struct **vmas, int *nonblocking) { |
df06b37ff mm/gup: cache dev... |
759 |
long ret = 0, i = 0; |
fa5bb2093 mm: cleanup __get... |
760 |
struct vm_area_struct *vma = NULL; |
df06b37ff mm/gup: cache dev... |
761 |
struct follow_page_context ctx = { NULL }; |
4bbd4c776 mm: move get_user... |
762 763 764 |
if (!nr_pages) return 0; |
f96525941 mm: untag user po... |
765 |
start = untagged_addr(start); |
4bbd4c776 mm: move get_user... |
766 767 768 769 770 771 772 773 774 |
VM_BUG_ON(!!pages != !!(gup_flags & FOLL_GET)); /* * If FOLL_FORCE is set then do not force a full fault as the hinting * fault information is unrelated to the reference behaviour of a task * using the address space */ if (!(gup_flags & FOLL_FORCE)) gup_flags |= FOLL_NUMA; |
4bbd4c776 mm: move get_user... |
775 |
do { |
fa5bb2093 mm: cleanup __get... |
776 777 778 779 780 781 782 783 |
struct page *page; unsigned int foll_flags = gup_flags; unsigned int page_increm; /* first iteration or cross vma bound */ if (!vma || start >= vma->vm_end) { vma = find_extend_vma(mm, start); if (!vma && in_gate_area(mm, start)) { |
fa5bb2093 mm: cleanup __get... |
784 785 786 787 |
ret = get_gate_page(mm, start & PAGE_MASK, gup_flags, &vma, pages ? &pages[i] : NULL); if (ret) |
08be37b79 mm/gup: finish co... |
788 |
goto out; |
df06b37ff mm/gup: cache dev... |
789 |
ctx.page_mask = 0; |
fa5bb2093 mm: cleanup __get... |
790 791 |
goto next_page; } |
4bbd4c776 mm: move get_user... |
792 |
|
df06b37ff mm/gup: cache dev... |
793 794 795 796 |
if (!vma || check_vma_flags(vma, gup_flags)) { ret = -EFAULT; goto out; } |
fa5bb2093 mm: cleanup __get... |
797 798 799 |
if (is_vm_hugetlb_page(vma)) { i = follow_hugetlb_page(mm, vma, pages, vmas, &start, &nr_pages, i, |
87ffc118b userfaultfd: huge... |
800 |
gup_flags, nonblocking); |
fa5bb2093 mm: cleanup __get... |
801 |
continue; |
4bbd4c776 mm: move get_user... |
802 |
} |
fa5bb2093 mm: cleanup __get... |
803 804 805 806 807 808 |
} retry: /* * If we have a pending SIGKILL, don't keep faulting pages and * potentially allocating memory. */ |
fa45f1162 mm/: remove calle... |
809 |
if (fatal_signal_pending(current)) { |
df06b37ff mm/gup: cache dev... |
810 811 812 |
ret = -ERESTARTSYS; goto out; } |
fa5bb2093 mm: cleanup __get... |
813 |
cond_resched(); |
df06b37ff mm/gup: cache dev... |
814 815 |
page = follow_page_mask(vma, start, foll_flags, &ctx); |
fa5bb2093 mm: cleanup __get... |
816 |
if (!page) { |
fa5bb2093 mm: cleanup __get... |
817 818 819 820 821 |
ret = faultin_page(tsk, vma, start, &foll_flags, nonblocking); switch (ret) { case 0: goto retry; |
df06b37ff mm/gup: cache dev... |
822 823 824 |
case -EBUSY: ret = 0; /* FALLTHRU */ |
fa5bb2093 mm: cleanup __get... |
825 826 827 |
case -EFAULT: case -ENOMEM: case -EHWPOISON: |
df06b37ff mm/gup: cache dev... |
828 |
goto out; |
fa5bb2093 mm: cleanup __get... |
829 830 |
case -ENOENT: goto next_page; |
4bbd4c776 mm: move get_user... |
831 |
} |
fa5bb2093 mm: cleanup __get... |
832 |
BUG(); |
1027e4436 mm: make GUP hand... |
833 834 835 836 837 838 839 |
} else if (PTR_ERR(page) == -EEXIST) { /* * Proper page table entry exists, but no corresponding * struct page. */ goto next_page; } else if (IS_ERR(page)) { |
df06b37ff mm/gup: cache dev... |
840 841 |
ret = PTR_ERR(page); goto out; |
1027e4436 mm: make GUP hand... |
842 |
} |
fa5bb2093 mm: cleanup __get... |
843 844 845 846 |
if (pages) { pages[i] = page; flush_anon_page(vma, page, start); flush_dcache_page(page); |
df06b37ff mm/gup: cache dev... |
847 |
ctx.page_mask = 0; |
4bbd4c776 mm: move get_user... |
848 |
} |
4bbd4c776 mm: move get_user... |
849 |
next_page: |
fa5bb2093 mm: cleanup __get... |
850 851 |
if (vmas) { vmas[i] = vma; |
df06b37ff mm/gup: cache dev... |
852 |
ctx.page_mask = 0; |
fa5bb2093 mm: cleanup __get... |
853 |
} |
df06b37ff mm/gup: cache dev... |
854 |
page_increm = 1 + (~(start >> PAGE_SHIFT) & ctx.page_mask); |
fa5bb2093 mm: cleanup __get... |
855 856 857 858 859 |
if (page_increm > nr_pages) page_increm = nr_pages; i += page_increm; start += page_increm * PAGE_SIZE; nr_pages -= page_increm; |
4bbd4c776 mm: move get_user... |
860 |
} while (nr_pages); |
df06b37ff mm/gup: cache dev... |
861 862 863 864 |
out: if (ctx.pgmap) put_dev_pagemap(ctx.pgmap); return i ? i : ret; |
4bbd4c776 mm: move get_user... |
865 |
} |
4bbd4c776 mm: move get_user... |
866 |
|
771ab4302 mm/gup.c: make un... |
867 868 |
static bool vma_permits_fault(struct vm_area_struct *vma, unsigned int fault_flags) |
d4925e00d mm/gup: Factor ou... |
869 |
{ |
1b2ee1266 mm/core: Do not e... |
870 871 |
bool write = !!(fault_flags & FAULT_FLAG_WRITE); bool foreign = !!(fault_flags & FAULT_FLAG_REMOTE); |
33a709b25 mm/gup, x86/mm/pk... |
872 |
vm_flags_t vm_flags = write ? VM_WRITE : VM_READ; |
d4925e00d mm/gup: Factor ou... |
873 874 875 |
if (!(vm_flags & vma->vm_flags)) return false; |
33a709b25 mm/gup, x86/mm/pk... |
876 877 |
/* * The architecture might have a hardware protection |
1b2ee1266 mm/core: Do not e... |
878 |
* mechanism other than read/write that can deny access. |
d61172b4b mm/core, x86/mm/p... |
879 880 881 |
* * gup always represents data access, not instruction * fetches, so execute=false here: |
33a709b25 mm/gup, x86/mm/pk... |
882 |
*/ |
d61172b4b mm/core, x86/mm/p... |
883 |
if (!arch_vma_access_permitted(vma, write, false, foreign)) |
33a709b25 mm/gup, x86/mm/pk... |
884 |
return false; |
d4925e00d mm/gup: Factor ou... |
885 886 |
return true; } |
4bbd4c776 mm: move get_user... |
887 888 889 890 891 892 893 |
/* * fixup_user_fault() - manually resolve a user page fault * @tsk: the task_struct to use for page fault accounting, or * NULL if faults are not to be recorded. * @mm: mm_struct of target mm * @address: user address * @fault_flags:flags to pass down to handle_mm_fault() |
4a9e1cda2 mm: bring in addi... |
894 895 |
* @unlocked: did we unlock the mmap_sem while retrying, maybe NULL if caller * does not allow retry |
4bbd4c776 mm: move get_user... |
896 897 898 899 900 901 902 903 904 905 906 |
* * This is meant to be called in the specific scenario where for locking reasons * we try to access user memory in atomic context (within a pagefault_disable() * section), this returns -EFAULT, and we want to resolve the user fault before * trying again. * * Typically this is meant to be used by the futex code. * * The main difference with get_user_pages() is that this function will * unconditionally call handle_mm_fault() which will in turn perform all the * necessary SW fixup of the dirty and young bits in the PTE, while |
4a9e1cda2 mm: bring in addi... |
907 |
* get_user_pages() only guarantees to update these in the struct page. |
4bbd4c776 mm: move get_user... |
908 909 910 911 912 913 |
* * This is important for some architectures where those bits also gate the * access permission to the page because they are maintained in software. On * such architectures, gup() will not be enough to make a subsequent access * succeed. * |
4a9e1cda2 mm: bring in addi... |
914 915 |
* This function will not return with an unlocked mmap_sem. So it has not the * same semantics wrt the @mm->mmap_sem as does filemap_fault(). |
4bbd4c776 mm: move get_user... |
916 917 |
*/ int fixup_user_fault(struct task_struct *tsk, struct mm_struct *mm, |
4a9e1cda2 mm: bring in addi... |
918 919 |
unsigned long address, unsigned int fault_flags, bool *unlocked) |
4bbd4c776 mm: move get_user... |
920 921 |
{ struct vm_area_struct *vma; |
2b7403035 mm: Change return... |
922 |
vm_fault_t ret, major = 0; |
4a9e1cda2 mm: bring in addi... |
923 |
|
f96525941 mm: untag user po... |
924 |
address = untagged_addr(address); |
4a9e1cda2 mm: bring in addi... |
925 926 |
if (unlocked) fault_flags |= FAULT_FLAG_ALLOW_RETRY; |
4bbd4c776 mm: move get_user... |
927 |
|
4a9e1cda2 mm: bring in addi... |
928 |
retry: |
4bbd4c776 mm: move get_user... |
929 930 931 |
vma = find_extend_vma(mm, address); if (!vma || address < vma->vm_start) return -EFAULT; |
d4925e00d mm/gup: Factor ou... |
932 |
if (!vma_permits_fault(vma, fault_flags)) |
4bbd4c776 mm: move get_user... |
933 |
return -EFAULT; |
dcddffd41 mm: do not pass m... |
934 |
ret = handle_mm_fault(vma, address, fault_flags); |
4a9e1cda2 mm: bring in addi... |
935 |
major |= ret & VM_FAULT_MAJOR; |
4bbd4c776 mm: move get_user... |
936 |
if (ret & VM_FAULT_ERROR) { |
9a291a7c9 mm/hugetlb: repor... |
937 938 939 940 |
int err = vm_fault_to_errno(ret, 0); if (err) return err; |
4bbd4c776 mm: move get_user... |
941 942 |
BUG(); } |
4a9e1cda2 mm: bring in addi... |
943 944 945 946 947 948 949 950 951 952 |
if (ret & VM_FAULT_RETRY) { down_read(&mm->mmap_sem); if (!(fault_flags & FAULT_FLAG_TRIED)) { *unlocked = true; fault_flags &= ~FAULT_FLAG_ALLOW_RETRY; fault_flags |= FAULT_FLAG_TRIED; goto retry; } } |
4bbd4c776 mm: move get_user... |
953 |
if (tsk) { |
4a9e1cda2 mm: bring in addi... |
954 |
if (major) |
4bbd4c776 mm: move get_user... |
955 956 957 958 959 960 |
tsk->maj_flt++; else tsk->min_flt++; } return 0; } |
add6a0cd1 KVM: MMU: try to ... |
961 |
EXPORT_SYMBOL_GPL(fixup_user_fault); |
4bbd4c776 mm: move get_user... |
962 |
|
f0818f472 mm: gup: add get_... |
963 964 965 966 |
static __always_inline long __get_user_pages_locked(struct task_struct *tsk, struct mm_struct *mm, unsigned long start, unsigned long nr_pages, |
f0818f472 mm: gup: add get_... |
967 968 |
struct page **pages, struct vm_area_struct **vmas, |
e716712f8 __get_user_pages_... |
969 |
int *locked, |
0fd71a56f mm: gup: add __ge... |
970 |
unsigned int flags) |
f0818f472 mm: gup: add get_... |
971 |
{ |
f0818f472 mm: gup: add get_... |
972 973 974 975 976 977 978 979 980 981 982 983 |
long ret, pages_done; bool lock_dropped; if (locked) { /* if VM_FAULT_RETRY can be returned, vmas become invalid */ BUG_ON(vmas); /* check caller initialized locked */ BUG_ON(*locked != 1); } if (pages) flags |= FOLL_GET; |
f0818f472 mm: gup: add get_... |
984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 |
pages_done = 0; lock_dropped = false; for (;;) { ret = __get_user_pages(tsk, mm, start, nr_pages, flags, pages, vmas, locked); if (!locked) /* VM_FAULT_RETRY couldn't trigger, bypass */ return ret; /* VM_FAULT_RETRY cannot return errors */ if (!*locked) { BUG_ON(ret < 0); BUG_ON(ret >= nr_pages); } |
f0818f472 mm: gup: add get_... |
999 1000 1001 1002 1003 1004 1005 |
if (ret > 0) { nr_pages -= ret; pages_done += ret; if (!nr_pages) break; } if (*locked) { |
96312e612 mm/gup.c: teach g... |
1006 1007 1008 1009 |
/* * VM_FAULT_RETRY didn't trigger or it was a * FOLL_NOWAIT. */ |
f0818f472 mm: gup: add get_... |
1010 1011 1012 1013 |
if (!pages_done) pages_done = ret; break; } |
df17277b2 mm/gup: continue ... |
1014 1015 1016 1017 1018 1019 |
/* * VM_FAULT_RETRY triggered, so seek to the faulting offset. * For the prefault case (!pages) we only update counts. */ if (likely(pages)) pages += ret; |
f0818f472 mm: gup: add get_... |
1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 |
start += ret << PAGE_SHIFT; /* * Repeat on the address that fired VM_FAULT_RETRY * without FAULT_FLAG_ALLOW_RETRY but with * FAULT_FLAG_TRIED. */ *locked = 1; lock_dropped = true; down_read(&mm->mmap_sem); ret = __get_user_pages(tsk, mm, start, 1, flags | FOLL_TRIED, pages, NULL, NULL); if (ret != 1) { BUG_ON(ret > 1); if (!pages_done) pages_done = ret; break; } nr_pages--; pages_done++; if (!nr_pages) break; |
df17277b2 mm/gup: continue ... |
1042 1043 |
if (likely(pages)) pages++; |
f0818f472 mm: gup: add get_... |
1044 1045 |
start += PAGE_SIZE; } |
e716712f8 __get_user_pages_... |
1046 |
if (lock_dropped && *locked) { |
f0818f472 mm: gup: add get_... |
1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 |
/* * We must let the caller know we temporarily dropped the lock * and so the critical section protected by it was lost. */ up_read(&mm->mmap_sem); *locked = 0; } return pages_done; } /* |
1e9877902 mm/gup: Introduce... |
1058 |
* get_user_pages_remote() - pin user pages in memory |
4bbd4c776 mm: move get_user... |
1059 1060 1061 1062 1063 |
* @tsk: the task_struct to use for page fault accounting, or * NULL if faults are not to be recorded. * @mm: mm_struct of target mm * @start: starting user address * @nr_pages: number of pages from start to pin |
9beae1ea8 mm: replace get_u... |
1064 |
* @gup_flags: flags modifying lookup behaviour |
4bbd4c776 mm: move get_user... |
1065 1066 1067 1068 1069 |
* @pages: array that receives pointers to the pages pinned. * Should be at least nr_pages long. Or NULL, if caller * only intends to ensure the pages are faulted in. * @vmas: array of pointers to vmas corresponding to each page. * Or NULL if the caller does not require them. |
5b56d49fc mm: add locked pa... |
1070 1071 1072 |
* @locked: pointer to lock flag indicating whether lock is held and * subsequently whether VM_FAULT_RETRY functionality can be * utilised. Lock must initially be held. |
4bbd4c776 mm: move get_user... |
1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 |
* * Returns number of pages pinned. This may be fewer than the number * requested. If nr_pages is 0 or negative, returns 0. If no pages * were pinned, returns -errno. Each page returned must be released * with a put_page() call when it is finished with. vmas will only * remain valid while mmap_sem is held. * * Must be called with mmap_sem held for read or write. * * get_user_pages walks a process's page tables and takes a reference to * each struct page that each user address corresponds to at a given * instant. That is, it takes the page that would be accessed if a user * thread accesses the given user virtual address at that instant. * * This does not guarantee that the page exists in the user mappings when * get_user_pages returns, and there may even be a completely different * page there in some cases (eg. if mmapped pagecache has been invalidated * and subsequently re faulted). However it does guarantee that the page * won't be freed completely. And mostly callers simply care that the page * contains data that was valid *at some point in time*. Typically, an IO * or similar operation cannot guarantee anything stronger anyway because * locks can't be held over the syscall boundary. * |
9beae1ea8 mm: replace get_u... |
1096 1097 1098 |
* If gup_flags & FOLL_WRITE == 0, the page must not be written to. If the page * is written to, set_page_dirty (or set_page_dirty_lock, as appropriate) must * be called after the page is finished with, and before put_page is called. |
4bbd4c776 mm: move get_user... |
1099 1100 1101 1102 1103 1104 1105 1106 |
* * get_user_pages is typically used for fewer-copy IO operations, to get a * handle on the memory by some means other than accesses via the user virtual * addresses. The pages may be submitted for DMA to devices or accessed via * their kernel linear mapping (via the kmap APIs). Care should be taken to * use the correct cache flushing APIs. * * See also get_user_pages_fast, for performance critical applications. |
f0818f472 mm: gup: add get_... |
1107 1108 1109 1110 1111 |
* * get_user_pages should be phased out in favor of * get_user_pages_locked|unlocked or get_user_pages_fast. Nothing * should use get_user_pages because it cannot pass * FAULT_FLAG_ALLOW_RETRY to handle_mm_fault. |
4bbd4c776 mm: move get_user... |
1112 |
*/ |
1e9877902 mm/gup: Introduce... |
1113 1114 |
long get_user_pages_remote(struct task_struct *tsk, struct mm_struct *mm, unsigned long start, unsigned long nr_pages, |
9beae1ea8 mm: replace get_u... |
1115 |
unsigned int gup_flags, struct page **pages, |
5b56d49fc mm: add locked pa... |
1116 |
struct vm_area_struct **vmas, int *locked) |
4bbd4c776 mm: move get_user... |
1117 |
{ |
932f4a630 mm/gup: replace g... |
1118 1119 1120 1121 1122 1123 1124 1125 |
/* * FIXME: Current FOLL_LONGTERM behavior is incompatible with * FAULT_FLAG_ALLOW_RETRY because of the FS DAX check requirement on * vmas. As there are no users of this flag in this call we simply * disallow this option for now. */ if (WARN_ON_ONCE(gup_flags & FOLL_LONGTERM)) return -EINVAL; |
859110d74 mm: remove write/... |
1126 |
return __get_user_pages_locked(tsk, mm, start, nr_pages, pages, vmas, |
e716712f8 __get_user_pages_... |
1127 |
locked, |
9beae1ea8 mm: replace get_u... |
1128 |
gup_flags | FOLL_TOUCH | FOLL_REMOTE); |
1e9877902 mm/gup: Introduce... |
1129 1130 |
} EXPORT_SYMBOL(get_user_pages_remote); |
d3649f68b mm: reorder code ... |
1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 |
/** * populate_vma_page_range() - populate a range of pages in the vma. * @vma: target vma * @start: start address * @end: end address * @nonblocking: * * This takes care of mlocking the pages too if VM_LOCKED is set. * * return 0 on success, negative error code on error. * * vma->vm_mm->mmap_sem must be held. * * If @nonblocking is NULL, it may be held for read or write and will * be unperturbed. * * If @nonblocking is non-NULL, it must held for read only and may be * released. If it's released, *@nonblocking will be set to 0. */ long populate_vma_page_range(struct vm_area_struct *vma, unsigned long start, unsigned long end, int *nonblocking) { struct mm_struct *mm = vma->vm_mm; unsigned long nr_pages = (end - start) / PAGE_SIZE; int gup_flags; VM_BUG_ON(start & ~PAGE_MASK); VM_BUG_ON(end & ~PAGE_MASK); VM_BUG_ON_VMA(start < vma->vm_start, vma); VM_BUG_ON_VMA(end > vma->vm_end, vma); VM_BUG_ON_MM(!rwsem_is_locked(&mm->mmap_sem), mm); gup_flags = FOLL_TOUCH | FOLL_POPULATE | FOLL_MLOCK; if (vma->vm_flags & VM_LOCKONFAULT) gup_flags &= ~FOLL_POPULATE; /* * We want to touch writable mappings with a write fault in order * to break COW, except for shared mappings because these don't COW * and we would not want to dirty them for nothing. */ if ((vma->vm_flags & (VM_WRITE | VM_SHARED)) == VM_WRITE) gup_flags |= FOLL_WRITE; /* * We want mlock to succeed for regions that have any permissions * other than PROT_NONE. */ if (vma->vm_flags & (VM_READ | VM_WRITE | VM_EXEC)) gup_flags |= FOLL_FORCE; /* * We made sure addr is within a VMA, so the following will * not result in a stack expansion that recurses back here. */ return __get_user_pages(current, mm, start, nr_pages, gup_flags, NULL, NULL, nonblocking); } /* * __mm_populate - populate and/or mlock pages within a range of address space. * * This is used to implement mlock() and the MAP_POPULATE / MAP_LOCKED mmap * flags. VMAs must be already marked with the desired vm_flags, and * mmap_sem must not be held. */ int __mm_populate(unsigned long start, unsigned long len, int ignore_errors) { struct mm_struct *mm = current->mm; unsigned long end, nstart, nend; struct vm_area_struct *vma = NULL; int locked = 0; long ret = 0; end = start + len; for (nstart = start; nstart < end; nstart = nend) { /* * We want to fault in pages for [nstart; end) address range. * Find first corresponding VMA. */ if (!locked) { locked = 1; down_read(&mm->mmap_sem); vma = find_vma(mm, nstart); } else if (nstart >= vma->vm_end) vma = vma->vm_next; if (!vma || vma->vm_start >= end) break; /* * Set [nstart; nend) to intersection of desired address * range with the first VMA. Also, skip undesirable VMA types. */ nend = min(end, vma->vm_end); if (vma->vm_flags & (VM_IO | VM_PFNMAP)) continue; if (nstart < vma->vm_start) nstart = vma->vm_start; /* * Now fault in a range of pages. populate_vma_page_range() * double checks the vma flags, so that it won't mlock pages * if the vma was already munlocked. */ ret = populate_vma_page_range(vma, nstart, nend, &locked); if (ret < 0) { if (ignore_errors) { ret = 0; continue; /* continue at next VMA */ } break; } nend = nstart + ret * PAGE_SIZE; ret = 0; } if (locked) up_read(&mm->mmap_sem); return ret; /* 0 or negative error code */ } /** * get_dump_page() - pin user page in memory while writing it to core dump * @addr: user address * * Returns struct page pointer of user page pinned for dump, * to be freed afterwards by put_page(). * * Returns NULL on any kind of failure - a hole must then be inserted into * the corefile, to preserve alignment with its headers; and also returns * NULL wherever the ZERO_PAGE, or an anonymous pte_none, has been found - * allowing a hole to be left in the corefile to save diskspace. * * Called without mmap_sem, but after all other threads have been killed. */ #ifdef CONFIG_ELF_CORE struct page *get_dump_page(unsigned long addr) { struct vm_area_struct *vma; struct page *page; if (__get_user_pages(current, current->mm, addr, 1, FOLL_FORCE | FOLL_DUMP | FOLL_GET, &page, &vma, NULL) < 1) return NULL; flush_cache_page(vma, addr, page_to_pfn(page)); return page; } #endif /* CONFIG_ELF_CORE */ |
050a9adc6 mm: consolidate t... |
1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 |
#else /* CONFIG_MMU */ static long __get_user_pages_locked(struct task_struct *tsk, struct mm_struct *mm, unsigned long start, unsigned long nr_pages, struct page **pages, struct vm_area_struct **vmas, int *locked, unsigned int foll_flags) { struct vm_area_struct *vma; unsigned long vm_flags; int i; /* calculate required read or write permissions. * If FOLL_FORCE is set, we only require the "MAY" flags. */ vm_flags = (foll_flags & FOLL_WRITE) ? (VM_WRITE | VM_MAYWRITE) : (VM_READ | VM_MAYREAD); vm_flags &= (foll_flags & FOLL_FORCE) ? (VM_MAYREAD | VM_MAYWRITE) : (VM_READ | VM_WRITE); for (i = 0; i < nr_pages; i++) { vma = find_vma(mm, start); if (!vma) goto finish_or_fault; /* protect what we can, including chardevs */ if ((vma->vm_flags & (VM_IO | VM_PFNMAP)) || !(vm_flags & vma->vm_flags)) goto finish_or_fault; if (pages) { pages[i] = virt_to_page(start); if (pages[i]) get_page(pages[i]); } if (vmas) vmas[i] = vma; start = (start + PAGE_SIZE) & PAGE_MASK; } return i; finish_or_fault: return i ? : -EFAULT; } #endif /* !CONFIG_MMU */ |
d3649f68b mm: reorder code ... |
1322 |
|
9a4e9f3b2 mm: update get_us... |
1323 |
#if defined(CONFIG_FS_DAX) || defined (CONFIG_CMA) |
9a4e9f3b2 mm: update get_us... |
1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 |
static bool check_dax_vmas(struct vm_area_struct **vmas, long nr_pages) { long i; struct vm_area_struct *vma_prev = NULL; for (i = 0; i < nr_pages; i++) { struct vm_area_struct *vma = vmas[i]; if (vma == vma_prev) continue; vma_prev = vma; if (vma_is_fsdax(vma)) return true; } return false; } |
9a4e9f3b2 mm: update get_us... |
1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 |
#ifdef CONFIG_CMA static struct page *new_non_cma_page(struct page *page, unsigned long private) { /* * We want to make sure we allocate the new page from the same node * as the source page. */ int nid = page_to_nid(page); /* * Trying to allocate a page for migration. Ignore allocation * failure warnings. We don't force __GFP_THISNODE here because * this node here is the node where we have CMA reservation and * in some case these nodes will have really less non movable * allocation memory. */ gfp_t gfp_mask = GFP_USER | __GFP_NOWARN; if (PageHighMem(page)) gfp_mask |= __GFP_HIGHMEM; #ifdef CONFIG_HUGETLB_PAGE if (PageHuge(page)) { struct hstate *h = page_hstate(page); /* * We don't want to dequeue from the pool because pool pages will * mostly be from the CMA region. */ return alloc_migrate_huge_page(h, gfp_mask, nid, NULL); } #endif if (PageTransHuge(page)) { struct page *thp; /* * ignore allocation failure warnings */ gfp_t thp_gfpmask = GFP_TRANSHUGE | __GFP_NOWARN; /* * Remove the movable mask so that we don't allocate from * CMA area again. */ thp_gfpmask &= ~__GFP_MOVABLE; thp = __alloc_pages_node(nid, thp_gfpmask, HPAGE_PMD_ORDER); if (!thp) return NULL; prep_transhuge_page(thp); return thp; } return __alloc_pages_node(nid, gfp_mask, 0); } |
932f4a630 mm/gup: replace g... |
1394 1395 1396 1397 |
static long check_and_migrate_cma_pages(struct task_struct *tsk, struct mm_struct *mm, unsigned long start, unsigned long nr_pages, |
9a4e9f3b2 mm: update get_us... |
1398 |
struct page **pages, |
932f4a630 mm/gup: replace g... |
1399 1400 |
struct vm_area_struct **vmas, unsigned int gup_flags) |
9a4e9f3b2 mm: update get_us... |
1401 |
{ |
aa712399c mm/gup: speed up ... |
1402 1403 |
unsigned long i; unsigned long step; |
9a4e9f3b2 mm: update get_us... |
1404 1405 1406 1407 1408 |
bool drain_allow = true; bool migrate_allow = true; LIST_HEAD(cma_page_list); check_again: |
aa712399c mm/gup: speed up ... |
1409 1410 1411 1412 1413 1414 1415 1416 |
for (i = 0; i < nr_pages;) { struct page *head = compound_head(pages[i]); /* * gup may start from a tail page. Advance step by the left * part. */ |
d8c6546b1 mm: introduce com... |
1417 |
step = compound_nr(head) - (pages[i] - head); |
9a4e9f3b2 mm: update get_us... |
1418 1419 1420 1421 1422 |
/* * If we get a page from the CMA zone, since we are going to * be pinning these entries, we might as well move them out * of the CMA zone if possible. */ |
aa712399c mm/gup: speed up ... |
1423 1424 |
if (is_migrate_cma_page(head)) { if (PageHuge(head)) |
9a4e9f3b2 mm: update get_us... |
1425 |
isolate_huge_page(head, &cma_page_list); |
aa712399c mm/gup: speed up ... |
1426 |
else { |
9a4e9f3b2 mm: update get_us... |
1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 |
if (!PageLRU(head) && drain_allow) { lru_add_drain_all(); drain_allow = false; } if (!isolate_lru_page(head)) { list_add_tail(&head->lru, &cma_page_list); mod_node_page_state(page_pgdat(head), NR_ISOLATED_ANON + page_is_file_cache(head), hpage_nr_pages(head)); } } } |
aa712399c mm/gup: speed up ... |
1441 1442 |
i += step; |
9a4e9f3b2 mm: update get_us... |
1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 |
} if (!list_empty(&cma_page_list)) { /* * drop the above get_user_pages reference. */ for (i = 0; i < nr_pages; i++) put_page(pages[i]); if (migrate_pages(&cma_page_list, new_non_cma_page, NULL, 0, MIGRATE_SYNC, MR_CONTIG_RANGE)) { /* * some of the pages failed migration. Do get_user_pages * without migration. */ migrate_allow = false; if (!list_empty(&cma_page_list)) putback_movable_pages(&cma_page_list); } /* |
932f4a630 mm/gup: replace g... |
1464 1465 1466 |
* We did migrate all the pages, Try to get the page references * again migrating any new CMA pages which we failed to isolate * earlier. |
9a4e9f3b2 mm: update get_us... |
1467 |
*/ |
932f4a630 mm/gup: replace g... |
1468 1469 1470 |
nr_pages = __get_user_pages_locked(tsk, mm, start, nr_pages, pages, vmas, NULL, gup_flags); |
9a4e9f3b2 mm: update get_us... |
1471 1472 1473 1474 1475 1476 1477 1478 1479 |
if ((nr_pages > 0) && migrate_allow) { drain_allow = true; goto check_again; } } return nr_pages; } #else |
932f4a630 mm/gup: replace g... |
1480 1481 1482 1483 1484 1485 1486 |
static long check_and_migrate_cma_pages(struct task_struct *tsk, struct mm_struct *mm, unsigned long start, unsigned long nr_pages, struct page **pages, struct vm_area_struct **vmas, unsigned int gup_flags) |
9a4e9f3b2 mm: update get_us... |
1487 1488 1489 |
{ return nr_pages; } |
050a9adc6 mm: consolidate t... |
1490 |
#endif /* CONFIG_CMA */ |
9a4e9f3b2 mm: update get_us... |
1491 |
|
2bb6d2837 mm: introduce get... |
1492 |
/* |
932f4a630 mm/gup: replace g... |
1493 1494 |
* __gup_longterm_locked() is a wrapper for __get_user_pages_locked which * allows us to process the FOLL_LONGTERM flag. |
2bb6d2837 mm: introduce get... |
1495 |
*/ |
932f4a630 mm/gup: replace g... |
1496 1497 1498 1499 1500 1501 1502 |
static long __gup_longterm_locked(struct task_struct *tsk, struct mm_struct *mm, unsigned long start, unsigned long nr_pages, struct page **pages, struct vm_area_struct **vmas, unsigned int gup_flags) |
2bb6d2837 mm: introduce get... |
1503 |
{ |
932f4a630 mm/gup: replace g... |
1504 1505 |
struct vm_area_struct **vmas_tmp = vmas; unsigned long flags = 0; |
2bb6d2837 mm: introduce get... |
1506 |
long rc, i; |
932f4a630 mm/gup: replace g... |
1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 |
if (gup_flags & FOLL_LONGTERM) { if (!pages) return -EINVAL; if (!vmas_tmp) { vmas_tmp = kcalloc(nr_pages, sizeof(struct vm_area_struct *), GFP_KERNEL); if (!vmas_tmp) return -ENOMEM; } flags = memalloc_nocma_save(); |
2bb6d2837 mm: introduce get... |
1519 |
} |
932f4a630 mm/gup: replace g... |
1520 1521 |
rc = __get_user_pages_locked(tsk, mm, start, nr_pages, pages, vmas_tmp, NULL, gup_flags); |
2bb6d2837 mm: introduce get... |
1522 |
|
932f4a630 mm/gup: replace g... |
1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 |
if (gup_flags & FOLL_LONGTERM) { memalloc_nocma_restore(flags); if (rc < 0) goto out; if (check_dax_vmas(vmas_tmp, rc)) { for (i = 0; i < rc; i++) put_page(pages[i]); rc = -EOPNOTSUPP; goto out; } rc = check_and_migrate_cma_pages(tsk, mm, start, rc, pages, vmas_tmp, gup_flags); |
9a4e9f3b2 mm: update get_us... |
1537 |
} |
2bb6d2837 mm: introduce get... |
1538 |
|
2bb6d2837 mm: introduce get... |
1539 |
out: |
932f4a630 mm/gup: replace g... |
1540 1541 |
if (vmas_tmp != vmas) kfree(vmas_tmp); |
2bb6d2837 mm: introduce get... |
1542 1543 |
return rc; } |
932f4a630 mm/gup: replace g... |
1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 |
#else /* !CONFIG_FS_DAX && !CONFIG_CMA */ static __always_inline long __gup_longterm_locked(struct task_struct *tsk, struct mm_struct *mm, unsigned long start, unsigned long nr_pages, struct page **pages, struct vm_area_struct **vmas, unsigned int flags) { return __get_user_pages_locked(tsk, mm, start, nr_pages, pages, vmas, NULL, flags); } #endif /* CONFIG_FS_DAX || CONFIG_CMA */ /* * This is the same as get_user_pages_remote(), just with a * less-flexible calling convention where we assume that the task * and mm being operated on are the current task's and don't allow * passing of a locked parameter. We also obviously don't pass * FOLL_REMOTE in here. */ long get_user_pages(unsigned long start, unsigned long nr_pages, unsigned int gup_flags, struct page **pages, struct vm_area_struct **vmas) { return __gup_longterm_locked(current, current->mm, start, nr_pages, pages, vmas, gup_flags | FOLL_TOUCH); } EXPORT_SYMBOL(get_user_pages); |
2bb6d2837 mm: introduce get... |
1573 |
|
d3649f68b mm: reorder code ... |
1574 1575 1576 1577 |
/* * We can leverage the VM_FAULT_RETRY functionality in the page fault * paths better by using either get_user_pages_locked() or * get_user_pages_unlocked(). |
acc3c8d15 mm: move mm_popul... |
1578 |
* |
d3649f68b mm: reorder code ... |
1579 |
* get_user_pages_locked() is suitable to replace the form: |
acc3c8d15 mm: move mm_popul... |
1580 |
* |
d3649f68b mm: reorder code ... |
1581 1582 1583 1584 |
* down_read(&mm->mmap_sem); * do_something() * get_user_pages(tsk, mm, ..., pages, NULL); * up_read(&mm->mmap_sem); |
acc3c8d15 mm: move mm_popul... |
1585 |
* |
d3649f68b mm: reorder code ... |
1586 |
* to: |
acc3c8d15 mm: move mm_popul... |
1587 |
* |
d3649f68b mm: reorder code ... |
1588 1589 1590 1591 1592 1593 |
* int locked = 1; * down_read(&mm->mmap_sem); * do_something() * get_user_pages_locked(tsk, mm, ..., pages, &locked); * if (locked) * up_read(&mm->mmap_sem); |
acc3c8d15 mm: move mm_popul... |
1594 |
*/ |
d3649f68b mm: reorder code ... |
1595 1596 1597 |
long get_user_pages_locked(unsigned long start, unsigned long nr_pages, unsigned int gup_flags, struct page **pages, int *locked) |
acc3c8d15 mm: move mm_popul... |
1598 |
{ |
acc3c8d15 mm: move mm_popul... |
1599 |
/* |
d3649f68b mm: reorder code ... |
1600 1601 1602 1603 |
* FIXME: Current FOLL_LONGTERM behavior is incompatible with * FAULT_FLAG_ALLOW_RETRY because of the FS DAX check requirement on * vmas. As there are no users of this flag in this call we simply * disallow this option for now. |
acc3c8d15 mm: move mm_popul... |
1604 |
*/ |
d3649f68b mm: reorder code ... |
1605 1606 |
if (WARN_ON_ONCE(gup_flags & FOLL_LONGTERM)) return -EINVAL; |
acc3c8d15 mm: move mm_popul... |
1607 |
|
d3649f68b mm: reorder code ... |
1608 1609 1610 |
return __get_user_pages_locked(current, current->mm, start, nr_pages, pages, NULL, locked, gup_flags | FOLL_TOUCH); |
acc3c8d15 mm: move mm_popul... |
1611 |
} |
d3649f68b mm: reorder code ... |
1612 |
EXPORT_SYMBOL(get_user_pages_locked); |
acc3c8d15 mm: move mm_popul... |
1613 1614 |
/* |
d3649f68b mm: reorder code ... |
1615 |
* get_user_pages_unlocked() is suitable to replace the form: |
acc3c8d15 mm: move mm_popul... |
1616 |
* |
d3649f68b mm: reorder code ... |
1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 |
* down_read(&mm->mmap_sem); * get_user_pages(tsk, mm, ..., pages, NULL); * up_read(&mm->mmap_sem); * * with: * * get_user_pages_unlocked(tsk, mm, ..., pages); * * It is functionally equivalent to get_user_pages_fast so * get_user_pages_fast should be used instead if specific gup_flags * (e.g. FOLL_FORCE) are not required. |
acc3c8d15 mm: move mm_popul... |
1628 |
*/ |
d3649f68b mm: reorder code ... |
1629 1630 |
long get_user_pages_unlocked(unsigned long start, unsigned long nr_pages, struct page **pages, unsigned int gup_flags) |
acc3c8d15 mm: move mm_popul... |
1631 1632 |
{ struct mm_struct *mm = current->mm; |
d3649f68b mm: reorder code ... |
1633 1634 |
int locked = 1; long ret; |
acc3c8d15 mm: move mm_popul... |
1635 |
|
d3649f68b mm: reorder code ... |
1636 1637 1638 1639 1640 1641 1642 1643 |
/* * FIXME: Current FOLL_LONGTERM behavior is incompatible with * FAULT_FLAG_ALLOW_RETRY because of the FS DAX check requirement on * vmas. As there are no users of this flag in this call we simply * disallow this option for now. */ if (WARN_ON_ONCE(gup_flags & FOLL_LONGTERM)) return -EINVAL; |
acc3c8d15 mm: move mm_popul... |
1644 |
|
d3649f68b mm: reorder code ... |
1645 1646 1647 |
down_read(&mm->mmap_sem); ret = __get_user_pages_locked(current, mm, start, nr_pages, pages, NULL, &locked, gup_flags | FOLL_TOUCH); |
acc3c8d15 mm: move mm_popul... |
1648 1649 |
if (locked) up_read(&mm->mmap_sem); |
d3649f68b mm: reorder code ... |
1650 |
return ret; |
4bbd4c776 mm: move get_user... |
1651 |
} |
d3649f68b mm: reorder code ... |
1652 |
EXPORT_SYMBOL(get_user_pages_unlocked); |
2667f50e8 mm: introduce a g... |
1653 1654 |
/* |
67a929e09 mm: rename CONFIG... |
1655 |
* Fast GUP |
2667f50e8 mm: introduce a g... |
1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 |
* * get_user_pages_fast attempts to pin user pages by walking the page * tables directly and avoids taking locks. Thus the walker needs to be * protected from page table pages being freed from under it, and should * block any THP splits. * * One way to achieve this is to have the walker disable interrupts, and * rely on IPIs from the TLB flushing code blocking before the page table * pages are freed. This is unsuitable for architectures that do not need * to broadcast an IPI when invalidating TLBs. * * Another way to achieve this is to batch up page table containing pages * belonging to more than one mm_user, then rcu_sched a callback to free those * pages. Disabling interrupts will allow the fast_gup walker to both block * the rcu_sched callback, and an IPI that we broadcast for splitting THPs * (which is a relatively rare event). The code below adopts this strategy. * * Before activating this code, please be aware that the following assumptions * are currently made: * |
e585513b7 x86/mm/gup: Switc... |
1676 1677 |
* *) Either HAVE_RCU_TABLE_FREE is enabled, and tlb_remove_table() is used to * free pages containing page tables or TLB flushing requires IPI broadcast. |
2667f50e8 mm: introduce a g... |
1678 |
* |
2667f50e8 mm: introduce a g... |
1679 1680 1681 1682 1683 1684 1685 1686 |
* *) ptes can be read atomically by the architecture. * * *) access_ok is sufficient to validate userspace address ranges. * * The last two assumptions can be relaxed by the addition of helper functions. * * This code is based heavily on the PowerPC implementation by Nick Piggin. */ |
67a929e09 mm: rename CONFIG... |
1687 |
#ifdef CONFIG_HAVE_FAST_GUP |
39656e83d mm: lift the x86_... |
1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 |
#ifdef CONFIG_GUP_GET_PTE_LOW_HIGH /* * WARNING: only to be used in the get_user_pages_fast() implementation. * * With get_user_pages_fast(), we walk down the pagetables without taking any * locks. For this we would like to load the pointers atomically, but sometimes * that is not possible (e.g. without expensive cmpxchg8b on x86_32 PAE). What * we do have is the guarantee that a PTE will only either go from not present * to present, or present to not present or both -- it will not switch to a * completely different present page without a TLB flush in between; something * that we are blocking by holding interrupts off. * * Setting ptes from not present to present goes: * * ptep->pte_high = h; * smp_wmb(); * ptep->pte_low = l; * * And present to not present goes: * * ptep->pte_low = 0; * smp_wmb(); * ptep->pte_high = 0; * * We must ensure here that the load of pte_low sees 'l' IFF pte_high sees 'h'. * We load pte_high *after* loading pte_low, which ensures we don't see an older * value of pte_high. *Then* we recheck pte_low, which ensures that we haven't * picked up a changed pte high. We might have gotten rubbish values from * pte_low and pte_high, but we are guaranteed that pte_low will not have the * present bit set *unless* it is 'l'. Because get_user_pages_fast() only * operates on present ptes we're safe. */ static inline pte_t gup_get_pte(pte_t *ptep) { pte_t pte; |
2667f50e8 mm: introduce a g... |
1723 |
|
39656e83d mm: lift the x86_... |
1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 |
do { pte.pte_low = ptep->pte_low; smp_rmb(); pte.pte_high = ptep->pte_high; smp_rmb(); } while (unlikely(pte.pte_low != ptep->pte_low)); return pte; } #else /* CONFIG_GUP_GET_PTE_LOW_HIGH */ |
0005d20b2 mm/gup: Move page... |
1734 |
/* |
39656e83d mm: lift the x86_... |
1735 |
* We require that the PTE can be read atomically. |
0005d20b2 mm/gup: Move page... |
1736 1737 1738 1739 1740 |
*/ static inline pte_t gup_get_pte(pte_t *ptep) { return READ_ONCE(*ptep); } |
39656e83d mm: lift the x86_... |
1741 |
#endif /* CONFIG_GUP_GET_PTE_LOW_HIGH */ |
0005d20b2 mm/gup: Move page... |
1742 |
|
790c73690 mm/gup.c: mark un... |
1743 1744 |
static void __maybe_unused undo_dev_pagemap(int *nr, int nr_start, struct page **pages) |
b59f65fa0 mm/gup: Implement... |
1745 1746 1747 1748 1749 1750 1751 1752 |
{ while ((*nr) - nr_start) { struct page *page = pages[--(*nr)]; ClearPageReferenced(page); put_page(page); } } |
8fde12ca7 mm: prevent get_u... |
1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 |
/* * Return the compund head page with ref appropriately incremented, * or NULL if that failed. */ static inline struct page *try_get_compound_head(struct page *page, int refs) { struct page *head = compound_head(page); if (WARN_ON_ONCE(page_ref_count(head) < 0)) return NULL; if (unlikely(!page_cache_add_speculative(head, refs))) return NULL; return head; } |
3010a5ea6 mm: introduce ARC... |
1766 |
#ifdef CONFIG_ARCH_HAS_PTE_SPECIAL |
2667f50e8 mm: introduce a g... |
1767 |
static int gup_pte_range(pmd_t pmd, unsigned long addr, unsigned long end, |
b798bec47 mm/gup: change wr... |
1768 |
unsigned int flags, struct page **pages, int *nr) |
2667f50e8 mm: introduce a g... |
1769 |
{ |
b59f65fa0 mm/gup: Implement... |
1770 1771 |
struct dev_pagemap *pgmap = NULL; int nr_start = *nr, ret = 0; |
2667f50e8 mm: introduce a g... |
1772 |
pte_t *ptep, *ptem; |
2667f50e8 mm: introduce a g... |
1773 1774 1775 |
ptem = ptep = pte_offset_map(&pmd, addr); do { |
0005d20b2 mm/gup: Move page... |
1776 |
pte_t pte = gup_get_pte(ptep); |
7aef4172c mm: handle PTE-ma... |
1777 |
struct page *head, *page; |
2667f50e8 mm: introduce a g... |
1778 1779 1780 |
/* * Similar to the PMD case below, NUMA hinting must take slow |
8a0516ed8 mm: convert p[te|... |
1781 |
* path using the pte_protnone check. |
2667f50e8 mm: introduce a g... |
1782 |
*/ |
e7884f8ea mm/gup: Move perm... |
1783 1784 |
if (pte_protnone(pte)) goto pte_unmap; |
b798bec47 mm/gup: change wr... |
1785 |
if (!pte_access_permitted(pte, flags & FOLL_WRITE)) |
e7884f8ea mm/gup: Move perm... |
1786 |
goto pte_unmap; |
b59f65fa0 mm/gup: Implement... |
1787 |
if (pte_devmap(pte)) { |
7af75561e mm/gup: add FOLL_... |
1788 1789 |
if (unlikely(flags & FOLL_LONGTERM)) goto pte_unmap; |
b59f65fa0 mm/gup: Implement... |
1790 1791 1792 1793 1794 1795 |
pgmap = get_dev_pagemap(pte_pfn(pte), pgmap); if (unlikely(!pgmap)) { undo_dev_pagemap(nr, nr_start, pages); goto pte_unmap; } } else if (pte_special(pte)) |
2667f50e8 mm: introduce a g... |
1796 1797 1798 1799 |
goto pte_unmap; VM_BUG_ON(!pfn_valid(pte_pfn(pte))); page = pte_page(pte); |
8fde12ca7 mm: prevent get_u... |
1800 1801 |
head = try_get_compound_head(page, 1); if (!head) |
2667f50e8 mm: introduce a g... |
1802 1803 1804 |
goto pte_unmap; if (unlikely(pte_val(pte) != pte_val(*ptep))) { |
7aef4172c mm: handle PTE-ma... |
1805 |
put_page(head); |
2667f50e8 mm: introduce a g... |
1806 1807 |
goto pte_unmap; } |
7aef4172c mm: handle PTE-ma... |
1808 |
VM_BUG_ON_PAGE(compound_head(page) != head, page); |
e93480537 mm/gup: Mark all ... |
1809 1810 |
SetPageReferenced(page); |
2667f50e8 mm: introduce a g... |
1811 1812 1813 1814 1815 1816 1817 1818 |
pages[*nr] = page; (*nr)++; } while (ptep++, addr += PAGE_SIZE, addr != end); ret = 1; pte_unmap: |
832d7aa05 mm: optimize dev_... |
1819 1820 |
if (pgmap) put_dev_pagemap(pgmap); |
2667f50e8 mm: introduce a g... |
1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 |
pte_unmap(ptem); return ret; } #else /* * If we can't determine whether or not a pte is special, then fail immediately * for ptes. Note, we can still pin HugeTLB and THP as these are guaranteed not * to be special. * * For a futex to be placed on a THP tail page, get_futex_key requires a * __get_user_pages_fast implementation that can pin pages. Thus it's still * useful to have gup_huge_pmd even if we can't operate on ptes. */ static int gup_pte_range(pmd_t pmd, unsigned long addr, unsigned long end, |
b798bec47 mm/gup: change wr... |
1836 |
unsigned int flags, struct page **pages, int *nr) |
2667f50e8 mm: introduce a g... |
1837 1838 1839 |
{ return 0; } |
3010a5ea6 mm: introduce ARC... |
1840 |
#endif /* CONFIG_ARCH_HAS_PTE_SPECIAL */ |
2667f50e8 mm: introduce a g... |
1841 |
|
175967318 mm: introduce ARC... |
1842 |
#if defined(CONFIG_ARCH_HAS_PTE_DEVMAP) && defined(CONFIG_TRANSPARENT_HUGEPAGE) |
b59f65fa0 mm/gup: Implement... |
1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 |
static int __gup_device_huge(unsigned long pfn, unsigned long addr, unsigned long end, struct page **pages, int *nr) { int nr_start = *nr; struct dev_pagemap *pgmap = NULL; do { struct page *page = pfn_to_page(pfn); pgmap = get_dev_pagemap(pfn, pgmap); if (unlikely(!pgmap)) { undo_dev_pagemap(nr, nr_start, pages); return 0; } SetPageReferenced(page); pages[*nr] = page; get_page(page); |
b59f65fa0 mm/gup: Implement... |
1860 1861 1862 |
(*nr)++; pfn++; } while (addr += PAGE_SIZE, addr != end); |
832d7aa05 mm: optimize dev_... |
1863 1864 1865 |
if (pgmap) put_dev_pagemap(pgmap); |
b59f65fa0 mm/gup: Implement... |
1866 1867 |
return 1; } |
a9b6de77b mm: fix __gup_dev... |
1868 |
static int __gup_device_huge_pmd(pmd_t orig, pmd_t *pmdp, unsigned long addr, |
b59f65fa0 mm/gup: Implement... |
1869 1870 1871 |
unsigned long end, struct page **pages, int *nr) { unsigned long fault_pfn; |
a9b6de77b mm: fix __gup_dev... |
1872 1873 1874 1875 1876 |
int nr_start = *nr; fault_pfn = pmd_pfn(orig) + ((addr & ~PMD_MASK) >> PAGE_SHIFT); if (!__gup_device_huge(fault_pfn, addr, end, pages, nr)) return 0; |
b59f65fa0 mm/gup: Implement... |
1877 |
|
a9b6de77b mm: fix __gup_dev... |
1878 1879 1880 1881 1882 |
if (unlikely(pmd_val(orig) != pmd_val(*pmdp))) { undo_dev_pagemap(nr, nr_start, pages); return 0; } return 1; |
b59f65fa0 mm/gup: Implement... |
1883 |
} |
a9b6de77b mm: fix __gup_dev... |
1884 |
static int __gup_device_huge_pud(pud_t orig, pud_t *pudp, unsigned long addr, |
b59f65fa0 mm/gup: Implement... |
1885 1886 1887 |
unsigned long end, struct page **pages, int *nr) { unsigned long fault_pfn; |
a9b6de77b mm: fix __gup_dev... |
1888 1889 1890 1891 1892 |
int nr_start = *nr; fault_pfn = pud_pfn(orig) + ((addr & ~PUD_MASK) >> PAGE_SHIFT); if (!__gup_device_huge(fault_pfn, addr, end, pages, nr)) return 0; |
b59f65fa0 mm/gup: Implement... |
1893 |
|
a9b6de77b mm: fix __gup_dev... |
1894 1895 1896 1897 1898 |
if (unlikely(pud_val(orig) != pud_val(*pudp))) { undo_dev_pagemap(nr, nr_start, pages); return 0; } return 1; |
b59f65fa0 mm/gup: Implement... |
1899 1900 |
} #else |
a9b6de77b mm: fix __gup_dev... |
1901 |
static int __gup_device_huge_pmd(pmd_t orig, pmd_t *pmdp, unsigned long addr, |
b59f65fa0 mm/gup: Implement... |
1902 1903 1904 1905 1906 |
unsigned long end, struct page **pages, int *nr) { BUILD_BUG(); return 0; } |
a9b6de77b mm: fix __gup_dev... |
1907 |
static int __gup_device_huge_pud(pud_t pud, pud_t *pudp, unsigned long addr, |
b59f65fa0 mm/gup: Implement... |
1908 1909 1910 1911 1912 1913 |
unsigned long end, struct page **pages, int *nr) { BUILD_BUG(); return 0; } #endif |
cbd34da7d mm: move the powe... |
1914 1915 1916 1917 1918 1919 1920 1921 1922 |
#ifdef CONFIG_ARCH_HAS_HUGEPD static unsigned long hugepte_addr_end(unsigned long addr, unsigned long end, unsigned long sz) { unsigned long __boundary = (addr + sz) & ~(sz-1); return (__boundary - 1 < end - 1) ? __boundary : end; } static int gup_hugepte(pte_t *ptep, unsigned long sz, unsigned long addr, |
0cd22afdc mm/gup: fix a mis... |
1923 1924 |
unsigned long end, unsigned int flags, struct page **pages, int *nr) |
cbd34da7d mm: move the powe... |
1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 |
{ unsigned long pte_end; struct page *head, *page; pte_t pte; int refs; pte_end = (addr + sz) & ~(sz-1); if (pte_end < end) end = pte_end; pte = READ_ONCE(*ptep); |
0cd22afdc mm/gup: fix a mis... |
1936 |
if (!pte_access_permitted(pte, flags & FOLL_WRITE)) |
cbd34da7d mm: move the powe... |
1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 |
return 0; /* hugepages are never "special" */ VM_BUG_ON(!pfn_valid(pte_pfn(pte))); refs = 0; head = pte_page(pte); page = head + ((addr & (sz-1)) >> PAGE_SHIFT); do { VM_BUG_ON(compound_head(page) != head); pages[*nr] = page; (*nr)++; page++; refs++; } while (addr += PAGE_SIZE, addr != end); |
01a369160 mm: switch gup_hu... |
1953 1954 |
head = try_get_compound_head(head, refs); if (!head) { |
cbd34da7d mm: move the powe... |
1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 |
*nr -= refs; return 0; } if (unlikely(pte_val(pte) != pte_val(*ptep))) { /* Could be optimized better */ *nr -= refs; while (refs--) put_page(head); return 0; } |
520b4a449 mm: mark the page... |
1966 |
SetPageReferenced(head); |
cbd34da7d mm: move the powe... |
1967 1968 1969 1970 |
return 1; } static int gup_huge_pd(hugepd_t hugepd, unsigned long addr, |
0cd22afdc mm/gup: fix a mis... |
1971 |
unsigned int pdshift, unsigned long end, unsigned int flags, |
cbd34da7d mm: move the powe... |
1972 1973 1974 1975 1976 1977 1978 1979 1980 |
struct page **pages, int *nr) { pte_t *ptep; unsigned long sz = 1UL << hugepd_shift(hugepd); unsigned long next; ptep = hugepte_offset(hugepd, addr, pdshift); do { next = hugepte_addr_end(addr, end, sz); |
0cd22afdc mm/gup: fix a mis... |
1981 |
if (!gup_hugepte(ptep, sz, addr, end, flags, pages, nr)) |
cbd34da7d mm: move the powe... |
1982 1983 1984 1985 1986 1987 1988 |
return 0; } while (ptep++, addr = next, addr != end); return 1; } #else static inline int gup_huge_pd(hugepd_t hugepd, unsigned long addr, |
0cd22afdc mm/gup: fix a mis... |
1989 |
unsigned int pdshift, unsigned long end, unsigned int flags, |
cbd34da7d mm: move the powe... |
1990 1991 1992 1993 1994 |
struct page **pages, int *nr) { return 0; } #endif /* CONFIG_ARCH_HAS_HUGEPD */ |
2667f50e8 mm: introduce a g... |
1995 |
static int gup_huge_pmd(pmd_t orig, pmd_t *pmdp, unsigned long addr, |
0cd22afdc mm/gup: fix a mis... |
1996 1997 |
unsigned long end, unsigned int flags, struct page **pages, int *nr) |
2667f50e8 mm: introduce a g... |
1998 |
{ |
ddc58f27f mm: drop tail pag... |
1999 |
struct page *head, *page; |
2667f50e8 mm: introduce a g... |
2000 |
int refs; |
b798bec47 mm/gup: change wr... |
2001 |
if (!pmd_access_permitted(orig, flags & FOLL_WRITE)) |
2667f50e8 mm: introduce a g... |
2002 |
return 0; |
7af75561e mm/gup: add FOLL_... |
2003 2004 2005 |
if (pmd_devmap(orig)) { if (unlikely(flags & FOLL_LONGTERM)) return 0; |
a9b6de77b mm: fix __gup_dev... |
2006 |
return __gup_device_huge_pmd(orig, pmdp, addr, end, pages, nr); |
7af75561e mm/gup: add FOLL_... |
2007 |
} |
b59f65fa0 mm/gup: Implement... |
2008 |
|
2667f50e8 mm: introduce a g... |
2009 |
refs = 0; |
d63206ee3 mm, gup: ensure r... |
2010 |
page = pmd_page(orig) + ((addr & ~PMD_MASK) >> PAGE_SHIFT); |
2667f50e8 mm: introduce a g... |
2011 |
do { |
2667f50e8 mm: introduce a g... |
2012 2013 2014 2015 2016 |
pages[*nr] = page; (*nr)++; page++; refs++; } while (addr += PAGE_SIZE, addr != end); |
8fde12ca7 mm: prevent get_u... |
2017 2018 |
head = try_get_compound_head(pmd_page(orig), refs); if (!head) { |
2667f50e8 mm: introduce a g... |
2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 |
*nr -= refs; return 0; } if (unlikely(pmd_val(orig) != pmd_val(*pmdp))) { *nr -= refs; while (refs--) put_page(head); return 0; } |
e93480537 mm/gup: Mark all ... |
2029 |
SetPageReferenced(head); |
2667f50e8 mm: introduce a g... |
2030 2031 2032 2033 |
return 1; } static int gup_huge_pud(pud_t orig, pud_t *pudp, unsigned long addr, |
b798bec47 mm/gup: change wr... |
2034 |
unsigned long end, unsigned int flags, struct page **pages, int *nr) |
2667f50e8 mm: introduce a g... |
2035 |
{ |
ddc58f27f mm: drop tail pag... |
2036 |
struct page *head, *page; |
2667f50e8 mm: introduce a g... |
2037 |
int refs; |
b798bec47 mm/gup: change wr... |
2038 |
if (!pud_access_permitted(orig, flags & FOLL_WRITE)) |
2667f50e8 mm: introduce a g... |
2039 |
return 0; |
7af75561e mm/gup: add FOLL_... |
2040 2041 2042 |
if (pud_devmap(orig)) { if (unlikely(flags & FOLL_LONGTERM)) return 0; |
a9b6de77b mm: fix __gup_dev... |
2043 |
return __gup_device_huge_pud(orig, pudp, addr, end, pages, nr); |
7af75561e mm/gup: add FOLL_... |
2044 |
} |
b59f65fa0 mm/gup: Implement... |
2045 |
|
2667f50e8 mm: introduce a g... |
2046 |
refs = 0; |
d63206ee3 mm, gup: ensure r... |
2047 |
page = pud_page(orig) + ((addr & ~PUD_MASK) >> PAGE_SHIFT); |
2667f50e8 mm: introduce a g... |
2048 |
do { |
2667f50e8 mm: introduce a g... |
2049 2050 2051 2052 2053 |
pages[*nr] = page; (*nr)++; page++; refs++; } while (addr += PAGE_SIZE, addr != end); |
8fde12ca7 mm: prevent get_u... |
2054 2055 |
head = try_get_compound_head(pud_page(orig), refs); if (!head) { |
2667f50e8 mm: introduce a g... |
2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 |
*nr -= refs; return 0; } if (unlikely(pud_val(orig) != pud_val(*pudp))) { *nr -= refs; while (refs--) put_page(head); return 0; } |
e93480537 mm/gup: Mark all ... |
2066 |
SetPageReferenced(head); |
2667f50e8 mm: introduce a g... |
2067 2068 |
return 1; } |
f30c59e92 mm: Update generi... |
2069 |
static int gup_huge_pgd(pgd_t orig, pgd_t *pgdp, unsigned long addr, |
b798bec47 mm/gup: change wr... |
2070 |
unsigned long end, unsigned int flags, |
f30c59e92 mm: Update generi... |
2071 2072 2073 |
struct page **pages, int *nr) { int refs; |
ddc58f27f mm: drop tail pag... |
2074 |
struct page *head, *page; |
f30c59e92 mm: Update generi... |
2075 |
|
b798bec47 mm/gup: change wr... |
2076 |
if (!pgd_access_permitted(orig, flags & FOLL_WRITE)) |
f30c59e92 mm: Update generi... |
2077 |
return 0; |
b59f65fa0 mm/gup: Implement... |
2078 |
BUILD_BUG_ON(pgd_devmap(orig)); |
f30c59e92 mm: Update generi... |
2079 |
refs = 0; |
d63206ee3 mm, gup: ensure r... |
2080 |
page = pgd_page(orig) + ((addr & ~PGDIR_MASK) >> PAGE_SHIFT); |
f30c59e92 mm: Update generi... |
2081 |
do { |
f30c59e92 mm: Update generi... |
2082 2083 2084 2085 2086 |
pages[*nr] = page; (*nr)++; page++; refs++; } while (addr += PAGE_SIZE, addr != end); |
8fde12ca7 mm: prevent get_u... |
2087 2088 |
head = try_get_compound_head(pgd_page(orig), refs); if (!head) { |
f30c59e92 mm: Update generi... |
2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 |
*nr -= refs; return 0; } if (unlikely(pgd_val(orig) != pgd_val(*pgdp))) { *nr -= refs; while (refs--) put_page(head); return 0; } |
e93480537 mm/gup: Mark all ... |
2099 |
SetPageReferenced(head); |
f30c59e92 mm: Update generi... |
2100 2101 |
return 1; } |
2667f50e8 mm: introduce a g... |
2102 |
static int gup_pmd_range(pud_t pud, unsigned long addr, unsigned long end, |
b798bec47 mm/gup: change wr... |
2103 |
unsigned int flags, struct page **pages, int *nr) |
2667f50e8 mm: introduce a g... |
2104 2105 2106 2107 2108 2109 |
{ unsigned long next; pmd_t *pmdp; pmdp = pmd_offset(&pud, addr); do { |
38c5ce936 mm/gup: Replace A... |
2110 |
pmd_t pmd = READ_ONCE(*pmdp); |
2667f50e8 mm: introduce a g... |
2111 2112 |
next = pmd_addr_end(addr, end); |
84c3fc4e9 mm: thp: check pm... |
2113 |
if (!pmd_present(pmd)) |
2667f50e8 mm: introduce a g... |
2114 |
return 0; |
414fd080d mm/gup: fix gup_p... |
2115 2116 |
if (unlikely(pmd_trans_huge(pmd) || pmd_huge(pmd) || pmd_devmap(pmd))) { |
2667f50e8 mm: introduce a g... |
2117 2118 2119 2120 2121 |
/* * NUMA hinting faults need to be handled in the GUP * slowpath for accounting purposes and so that they * can be serialised against THP migration. */ |
8a0516ed8 mm: convert p[te|... |
2122 |
if (pmd_protnone(pmd)) |
2667f50e8 mm: introduce a g... |
2123 |
return 0; |
b798bec47 mm/gup: change wr... |
2124 |
if (!gup_huge_pmd(pmd, pmdp, addr, next, flags, |
2667f50e8 mm: introduce a g... |
2125 2126 |
pages, nr)) return 0; |
f30c59e92 mm: Update generi... |
2127 2128 2129 2130 2131 2132 |
} else if (unlikely(is_hugepd(__hugepd(pmd_val(pmd))))) { /* * architecture have different format for hugetlbfs * pmd format and THP pmd format */ if (!gup_huge_pd(__hugepd(pmd_val(pmd)), addr, |
b798bec47 mm/gup: change wr... |
2133 |
PMD_SHIFT, next, flags, pages, nr)) |
f30c59e92 mm: Update generi... |
2134 |
return 0; |
b798bec47 mm/gup: change wr... |
2135 |
} else if (!gup_pte_range(pmd, addr, next, flags, pages, nr)) |
2923117b7 mm/gup.c: fix cod... |
2136 |
return 0; |
2667f50e8 mm: introduce a g... |
2137 2138 2139 2140 |
} while (pmdp++, addr = next, addr != end); return 1; } |
c2febafc6 mm: convert gener... |
2141 |
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end, |
b798bec47 mm/gup: change wr... |
2142 |
unsigned int flags, struct page **pages, int *nr) |
2667f50e8 mm: introduce a g... |
2143 2144 2145 |
{ unsigned long next; pud_t *pudp; |
c2febafc6 mm: convert gener... |
2146 |
pudp = pud_offset(&p4d, addr); |
2667f50e8 mm: introduce a g... |
2147 |
do { |
e37c69827 mm: replace ACCES... |
2148 |
pud_t pud = READ_ONCE(*pudp); |
2667f50e8 mm: introduce a g... |
2149 2150 2151 2152 |
next = pud_addr_end(addr, end); if (pud_none(pud)) return 0; |
f30c59e92 mm: Update generi... |
2153 |
if (unlikely(pud_huge(pud))) { |
b798bec47 mm/gup: change wr... |
2154 |
if (!gup_huge_pud(pud, pudp, addr, next, flags, |
f30c59e92 mm: Update generi... |
2155 2156 2157 2158 |
pages, nr)) return 0; } else if (unlikely(is_hugepd(__hugepd(pud_val(pud))))) { if (!gup_huge_pd(__hugepd(pud_val(pud)), addr, |
b798bec47 mm/gup: change wr... |
2159 |
PUD_SHIFT, next, flags, pages, nr)) |
2667f50e8 mm: introduce a g... |
2160 |
return 0; |
b798bec47 mm/gup: change wr... |
2161 |
} else if (!gup_pmd_range(pud, addr, next, flags, pages, nr)) |
2667f50e8 mm: introduce a g... |
2162 2163 2164 2165 2166 |
return 0; } while (pudp++, addr = next, addr != end); return 1; } |
c2febafc6 mm: convert gener... |
2167 |
static int gup_p4d_range(pgd_t pgd, unsigned long addr, unsigned long end, |
b798bec47 mm/gup: change wr... |
2168 |
unsigned int flags, struct page **pages, int *nr) |
c2febafc6 mm: convert gener... |
2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 |
{ unsigned long next; p4d_t *p4dp; p4dp = p4d_offset(&pgd, addr); do { p4d_t p4d = READ_ONCE(*p4dp); next = p4d_addr_end(addr, end); if (p4d_none(p4d)) return 0; BUILD_BUG_ON(p4d_huge(p4d)); if (unlikely(is_hugepd(__hugepd(p4d_val(p4d))))) { if (!gup_huge_pd(__hugepd(p4d_val(p4d)), addr, |
b798bec47 mm/gup: change wr... |
2183 |
P4D_SHIFT, next, flags, pages, nr)) |
c2febafc6 mm: convert gener... |
2184 |
return 0; |
b798bec47 mm/gup: change wr... |
2185 |
} else if (!gup_pud_range(p4d, addr, next, flags, pages, nr)) |
c2febafc6 mm: convert gener... |
2186 2187 2188 2189 2190 |
return 0; } while (p4dp++, addr = next, addr != end); return 1; } |
5b65c4677 mm, x86/mm: Fix p... |
2191 |
static void gup_pgd_range(unsigned long addr, unsigned long end, |
b798bec47 mm/gup: change wr... |
2192 |
unsigned int flags, struct page **pages, int *nr) |
5b65c4677 mm, x86/mm: Fix p... |
2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 |
{ unsigned long next; pgd_t *pgdp; pgdp = pgd_offset(current->mm, addr); do { pgd_t pgd = READ_ONCE(*pgdp); next = pgd_addr_end(addr, end); if (pgd_none(pgd)) return; if (unlikely(pgd_huge(pgd))) { |
b798bec47 mm/gup: change wr... |
2205 |
if (!gup_huge_pgd(pgd, pgdp, addr, next, flags, |
5b65c4677 mm, x86/mm: Fix p... |
2206 2207 2208 2209 |
pages, nr)) return; } else if (unlikely(is_hugepd(__hugepd(pgd_val(pgd))))) { if (!gup_huge_pd(__hugepd(pgd_val(pgd)), addr, |
b798bec47 mm/gup: change wr... |
2210 |
PGDIR_SHIFT, next, flags, pages, nr)) |
5b65c4677 mm, x86/mm: Fix p... |
2211 |
return; |
b798bec47 mm/gup: change wr... |
2212 |
} else if (!gup_p4d_range(pgd, addr, next, flags, pages, nr)) |
5b65c4677 mm, x86/mm: Fix p... |
2213 2214 2215 |
return; } while (pgdp++, addr = next, addr != end); } |
050a9adc6 mm: consolidate t... |
2216 2217 2218 2219 2220 2221 |
#else static inline void gup_pgd_range(unsigned long addr, unsigned long end, unsigned int flags, struct page **pages, int *nr) { } #endif /* CONFIG_HAVE_FAST_GUP */ |
5b65c4677 mm, x86/mm: Fix p... |
2222 2223 2224 2225 2226 2227 |
#ifndef gup_fast_permitted /* * Check if it's allowed to use __get_user_pages_fast() for the range, or * we need to fall back to the slow version: */ |
26f4c3280 mm: simplify gup_... |
2228 |
static bool gup_fast_permitted(unsigned long start, unsigned long end) |
5b65c4677 mm, x86/mm: Fix p... |
2229 |
{ |
26f4c3280 mm: simplify gup_... |
2230 |
return true; |
5b65c4677 mm, x86/mm: Fix p... |
2231 2232 |
} #endif |
2667f50e8 mm: introduce a g... |
2233 2234 |
/* * Like get_user_pages_fast() except it's IRQ-safe in that it won't fall back to |
d08110786 mm/gup.c: documen... |
2235 2236 2237 |
* the regular GUP. * Note a difference with get_user_pages_fast: this always returns the * number of pages pinned, 0 if no pages were pinned. |
050a9adc6 mm: consolidate t... |
2238 2239 2240 |
* * If the architecture does not support this function, simply return with no * pages pinned. |
2667f50e8 mm: introduce a g... |
2241 2242 2243 2244 |
*/ int __get_user_pages_fast(unsigned long start, int nr_pages, int write, struct page **pages) { |
d4faa4025 mm: remove unnece... |
2245 |
unsigned long len, end; |
5b65c4677 mm, x86/mm: Fix p... |
2246 |
unsigned long flags; |
2667f50e8 mm: introduce a g... |
2247 |
int nr = 0; |
f455c8548 mm: use untagged_... |
2248 |
start = untagged_addr(start) & PAGE_MASK; |
2667f50e8 mm: introduce a g... |
2249 2250 |
len = (unsigned long) nr_pages << PAGE_SHIFT; end = start + len; |
26f4c3280 mm: simplify gup_... |
2251 2252 |
if (end <= start) return 0; |
96d4f267e Remove 'type' arg... |
2253 |
if (unlikely(!access_ok((void __user *)start, len))) |
2667f50e8 mm: introduce a g... |
2254 2255 2256 2257 2258 2259 2260 |
return 0; /* * Disable interrupts. We use the nested form as we can already have * interrupts disabled by get_futex_key. * * With interrupts disabled, we block page table pages from being |
2ebe82288 mm/gup.c: fix __g... |
2261 2262 |
* freed from under us. See struct mmu_table_batch comments in * include/asm-generic/tlb.h for more details. |
2667f50e8 mm: introduce a g... |
2263 2264 2265 2266 |
* * We do not adopt an rcu_read_lock(.) here as we also want to * block IPIs that come from THPs splitting. */ |
050a9adc6 mm: consolidate t... |
2267 2268 |
if (IS_ENABLED(CONFIG_HAVE_FAST_GUP) && gup_fast_permitted(start, end)) { |
5b65c4677 mm, x86/mm: Fix p... |
2269 |
local_irq_save(flags); |
b798bec47 mm/gup: change wr... |
2270 |
gup_pgd_range(start, end, write ? FOLL_WRITE : 0, pages, &nr); |
5b65c4677 mm, x86/mm: Fix p... |
2271 2272 |
local_irq_restore(flags); } |
2667f50e8 mm: introduce a g... |
2273 2274 2275 |
return nr; } |
050a9adc6 mm: consolidate t... |
2276 |
EXPORT_SYMBOL_GPL(__get_user_pages_fast); |
2667f50e8 mm: introduce a g... |
2277 |
|
7af75561e mm/gup: add FOLL_... |
2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 |
static int __gup_longterm_unlocked(unsigned long start, int nr_pages, unsigned int gup_flags, struct page **pages) { int ret; /* * FIXME: FOLL_LONGTERM does not work with * get_user_pages_unlocked() (see comments in that function) */ if (gup_flags & FOLL_LONGTERM) { down_read(¤t->mm->mmap_sem); ret = __gup_longterm_locked(current, current->mm, start, nr_pages, pages, NULL, gup_flags); up_read(¤t->mm->mmap_sem); } else { ret = get_user_pages_unlocked(start, nr_pages, pages, gup_flags); } return ret; } |
2667f50e8 mm: introduce a g... |
2300 2301 2302 2303 |
/** * get_user_pages_fast() - pin user pages in memory * @start: starting user address * @nr_pages: number of pages from start to pin |
73b0140bf mm/gup: change GU... |
2304 |
* @gup_flags: flags modifying pin behaviour |
2667f50e8 mm: introduce a g... |
2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 |
* @pages: array that receives pointers to the pages pinned. * Should be at least nr_pages long. * * Attempt to pin user pages in memory without taking mm->mmap_sem. * If not successful, it will fall back to taking the lock and * calling get_user_pages(). * * Returns number of pages pinned. This may be fewer than the number * requested. If nr_pages is 0 or negative, returns 0. If no pages * were pinned, returns -errno. */ |
73b0140bf mm/gup: change GU... |
2316 2317 |
int get_user_pages_fast(unsigned long start, int nr_pages, unsigned int gup_flags, struct page **pages) |
2667f50e8 mm: introduce a g... |
2318 |
{ |
5b65c4677 mm, x86/mm: Fix p... |
2319 |
unsigned long addr, len, end; |
73e10a618 mm/gup: Provide c... |
2320 |
int nr = 0, ret = 0; |
2667f50e8 mm: introduce a g... |
2321 |
|
8cb5db61a mm/gup: allow FOL... |
2322 2323 |
if (WARN_ON_ONCE(gup_flags & ~(FOLL_WRITE | FOLL_LONGTERM | FOLL_FORCE))) |
817be129e mm: validate get_... |
2324 |
return -EINVAL; |
f455c8548 mm: use untagged_... |
2325 |
start = untagged_addr(start) & PAGE_MASK; |
5b65c4677 mm, x86/mm: Fix p... |
2326 2327 2328 |
addr = start; len = (unsigned long) nr_pages << PAGE_SHIFT; end = start + len; |
26f4c3280 mm: simplify gup_... |
2329 |
if (end <= start) |
c61611f70 get_user_pages_fa... |
2330 |
return 0; |
96d4f267e Remove 'type' arg... |
2331 |
if (unlikely(!access_ok((void __user *)start, len))) |
c61611f70 get_user_pages_fa... |
2332 |
return -EFAULT; |
73e10a618 mm/gup: Provide c... |
2333 |
|
050a9adc6 mm: consolidate t... |
2334 2335 |
if (IS_ENABLED(CONFIG_HAVE_FAST_GUP) && gup_fast_permitted(start, end)) { |
5b65c4677 mm, x86/mm: Fix p... |
2336 |
local_irq_disable(); |
73b0140bf mm/gup: change GU... |
2337 |
gup_pgd_range(addr, end, gup_flags, pages, &nr); |
5b65c4677 mm, x86/mm: Fix p... |
2338 |
local_irq_enable(); |
73e10a618 mm/gup: Provide c... |
2339 2340 |
ret = nr; } |
2667f50e8 mm: introduce a g... |
2341 2342 2343 2344 2345 |
if (nr < nr_pages) { /* Try to get the remaining pages with get_user_pages */ start += nr << PAGE_SHIFT; pages += nr; |
7af75561e mm/gup: add FOLL_... |
2346 2347 |
ret = __gup_longterm_unlocked(start, nr_pages - nr, gup_flags, pages); |
2667f50e8 mm: introduce a g... |
2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 |
/* Have to be a bit careful with return values */ if (nr > 0) { if (ret < 0) ret = nr; else ret += nr; } } return ret; } |
050a9adc6 mm: consolidate t... |
2360 |
EXPORT_SYMBOL_GPL(get_user_pages_fast); |