gup.c 87.2 KB
Newer Older
1
// SPDX-License-Identifier: GPL-2.0-only
2 3 4 5 6 7
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/err.h>
#include <linux/spinlock.h>

#include <linux/mm.h>
8
#include <linux/memremap.h>
9 10 11 12 13
#include <linux/pagemap.h>
#include <linux/rmap.h>
#include <linux/swap.h>
#include <linux/swapops.h>

14
#include <linux/sched/signal.h>
15
#include <linux/rwsem.h>
16
#include <linux/hugetlb.h>
17 18 19
#include <linux/migrate.h>
#include <linux/mm_inline.h>
#include <linux/sched/mm.h>
20

21
#include <asm/mmu_context.h>
22
#include <asm/tlbflush.h>
23

24 25
#include "internal.h"

26 27 28 29 30
struct follow_page_context {
	struct dev_pagemap *pgmap;
	unsigned int page_mask;
};

31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
static void hpage_pincount_add(struct page *page, int refs)
{
	VM_BUG_ON_PAGE(!hpage_pincount_available(page), page);
	VM_BUG_ON_PAGE(page != compound_head(page), page);

	atomic_add(refs, compound_pincount_ptr(page));
}

static void hpage_pincount_sub(struct page *page, int refs)
{
	VM_BUG_ON_PAGE(!hpage_pincount_available(page), page);
	VM_BUG_ON_PAGE(page != compound_head(page), page);

	atomic_sub(refs, compound_pincount_ptr(page));
}

47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
/*
 * Return the compound 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;
}

J
John Hubbard 已提交
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
/*
 * try_grab_compound_head() - attempt to elevate a page's refcount, by a
 * flags-dependent amount.
 *
 * "grab" names in this file mean, "look at flags to decide whether to use
 * FOLL_PIN or FOLL_GET behavior, when incrementing the page's refcount.
 *
 * Either FOLL_PIN or FOLL_GET (or neither) must be set, but not both at the
 * same time. (That's true throughout the get_user_pages*() and
 * pin_user_pages*() APIs.) Cases:
 *
 *    FOLL_GET: page's refcount will be incremented by 1.
 *    FOLL_PIN: page's refcount will be incremented by GUP_PIN_COUNTING_BIAS.
 *
 * Return: head page (with refcount appropriately incremented) for success, or
 * NULL upon failure. If neither FOLL_GET nor FOLL_PIN was set, that's
 * considered failure, and furthermore, a likely bug in the caller, so a warning
 * is also emitted.
 */
static __maybe_unused struct page *try_grab_compound_head(struct page *page,
							  int refs,
							  unsigned int flags)
{
	if (flags & FOLL_GET)
		return try_get_compound_head(page, refs);
	else if (flags & FOLL_PIN) {
88 89
		int orig_refs = refs;

90 91 92 93 94 95 96 97
		/*
		 * Can't do FOLL_LONGTERM + FOLL_PIN with CMA in the gup fast
		 * path, so fail and let the caller fall back to the slow path.
		 */
		if (unlikely(flags & FOLL_LONGTERM) &&
				is_migrate_cma_page(page))
			return NULL;

98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115
		/*
		 * When pinning a compound page of order > 1 (which is what
		 * hpage_pincount_available() checks for), use an exact count to
		 * track it, via hpage_pincount_add/_sub().
		 *
		 * However, be sure to *also* increment the normal page refcount
		 * field at least once, so that the page really is pinned.
		 */
		if (!hpage_pincount_available(page))
			refs *= GUP_PIN_COUNTING_BIAS;

		page = try_get_compound_head(page, refs);
		if (!page)
			return NULL;

		if (hpage_pincount_available(page))
			hpage_pincount_add(page, refs);

116 117 118
		mod_node_page_state(page_pgdat(page), NR_FOLL_PIN_ACQUIRED,
				    orig_refs);

119
		return page;
J
John Hubbard 已提交
120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153
	}

	WARN_ON_ONCE(1);
	return NULL;
}

/**
 * try_grab_page() - elevate a page's refcount by a flag-dependent amount
 *
 * This might not do anything at all, depending on the flags argument.
 *
 * "grab" names in this file mean, "look at flags to decide whether to use
 * FOLL_PIN or FOLL_GET behavior, when incrementing the page's refcount.
 *
 * @page:    pointer to page to be grabbed
 * @flags:   gup flags: these are the FOLL_* flag values.
 *
 * Either FOLL_PIN or FOLL_GET (or neither) may be set, but not both at the same
 * time. Cases:
 *
 *    FOLL_GET: page's refcount will be incremented by 1.
 *    FOLL_PIN: page's refcount will be incremented by GUP_PIN_COUNTING_BIAS.
 *
 * Return: true for success, or if no action was required (if neither FOLL_PIN
 * nor FOLL_GET was set, nothing is done). False for failure: FOLL_GET or
 * FOLL_PIN was set, but the page could not be grabbed.
 */
bool __must_check try_grab_page(struct page *page, unsigned int flags)
{
	WARN_ON_ONCE((flags & (FOLL_GET | FOLL_PIN)) == (FOLL_GET | FOLL_PIN));

	if (flags & FOLL_GET)
		return try_get_page(page);
	else if (flags & FOLL_PIN) {
154 155
		int refs = 1;

J
John Hubbard 已提交
156 157 158 159 160
		page = compound_head(page);

		if (WARN_ON_ONCE(page_ref_count(page) <= 0))
			return false;

161 162 163 164 165 166 167 168 169 170 171 172
		if (hpage_pincount_available(page))
			hpage_pincount_add(page, 1);
		else
			refs = GUP_PIN_COUNTING_BIAS;

		/*
		 * Similar to try_grab_compound_head(): even if using the
		 * hpage_pincount_add/_sub() routines, be sure to
		 * *also* increment the normal page refcount field at least
		 * once, so that the page really is pinned.
		 */
		page_ref_add(page, refs);
173 174

		mod_node_page_state(page_pgdat(page), NR_FOLL_PIN_ACQUIRED, 1);
J
John Hubbard 已提交
175 176 177 178 179 180 181 182
	}

	return true;
}

#ifdef CONFIG_DEV_PAGEMAP_OPS
static bool __unpin_devmap_managed_user_page(struct page *page)
{
183
	int count, refs = 1;
J
John Hubbard 已提交
184 185 186 187

	if (!page_is_devmap_managed(page))
		return false;

188 189 190 191 192 193
	if (hpage_pincount_available(page))
		hpage_pincount_sub(page, 1);
	else
		refs = GUP_PIN_COUNTING_BIAS;

	count = page_ref_sub_return(page, refs);
J
John Hubbard 已提交
194

195
	mod_node_page_state(page_pgdat(page), NR_FOLL_PIN_RELEASED, 1);
J
John Hubbard 已提交
196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225
	/*
	 * devmap page refcounts are 1-based, rather than 0-based: if
	 * refcount is 1, then the page is free and the refcount is
	 * stable because nobody holds a reference on the page.
	 */
	if (count == 1)
		free_devmap_managed_page(page);
	else if (!count)
		__put_page(page);

	return true;
}
#else
static bool __unpin_devmap_managed_user_page(struct page *page)
{
	return false;
}
#endif /* CONFIG_DEV_PAGEMAP_OPS */

/**
 * unpin_user_page() - release a dma-pinned page
 * @page:            pointer to page to be released
 *
 * Pages that were pinned via pin_user_pages*() must be released via either
 * unpin_user_page(), or one of the unpin_user_pages*() routines. This is so
 * that such pages can be separately tracked and uniquely handled. In
 * particular, interactions with RDMA and filesystems need special handling.
 */
void unpin_user_page(struct page *page)
{
226 227
	int refs = 1;

J
John Hubbard 已提交
228 229 230 231 232 233 234 235 236 237 238
	page = compound_head(page);

	/*
	 * For devmap managed pages we need to catch refcount transition from
	 * GUP_PIN_COUNTING_BIAS to 1, when refcount reach one it means the
	 * page is free and we need to inform the device driver through
	 * callback. See include/linux/memremap.h and HMM for details.
	 */
	if (__unpin_devmap_managed_user_page(page))
		return;

239 240 241 242 243 244
	if (hpage_pincount_available(page))
		hpage_pincount_sub(page, 1);
	else
		refs = GUP_PIN_COUNTING_BIAS;

	if (page_ref_sub_and_test(page, refs))
J
John Hubbard 已提交
245
		__put_page(page);
246 247

	mod_node_page_state(page_pgdat(page), NR_FOLL_PIN_RELEASED, 1);
J
John Hubbard 已提交
248 249 250
}
EXPORT_SYMBOL(unpin_user_page);

251
/**
252
 * unpin_user_pages_dirty_lock() - release and optionally dirty gup-pinned pages
253
 * @pages:  array of pages to be maybe marked dirty, and definitely released.
254
 * @npages: number of pages in the @pages array.
255
 * @make_dirty: whether to mark the pages dirty
256 257 258 259 260
 *
 * "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
261
 * compound page) dirty, if @make_dirty is true, and if the page was previously
262 263
 * listed as clean. In any case, releases all pages using unpin_user_page(),
 * possibly via unpin_user_pages(), for the non-dirty case.
264
 *
265
 * Please see the unpin_user_page() documentation for details.
266
 *
267 268 269
 * 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:
270
 * set_page_dirty_lock(), unpin_user_page().
271 272
 *
 */
273 274
void unpin_user_pages_dirty_lock(struct page **pages, unsigned long npages,
				 bool make_dirty)
275
{
276
	unsigned long index;
277

278 279 280 281 282 283 284
	/*
	 * 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) {
285
		unpin_user_pages(pages, npages);
286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312
		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);
313
		unpin_user_page(page);
314
	}
315
}
316
EXPORT_SYMBOL(unpin_user_pages_dirty_lock);
317 318

/**
319
 * unpin_user_pages() - release an array of gup-pinned pages.
320 321 322
 * @pages:  array of pages to be marked dirty and released.
 * @npages: number of pages in the @pages array.
 *
323
 * For each page in the @pages array, release the page using unpin_user_page().
324
 *
325
 * Please see the unpin_user_page() documentation for details.
326
 */
327
void unpin_user_pages(struct page **pages, unsigned long npages)
328 329 330 331 332 333 334 335 336
{
	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++)
337
		unpin_user_page(pages[index]);
338
}
339
EXPORT_SYMBOL(unpin_user_pages);
340

341
#ifdef CONFIG_MMU
342 343
static struct page *no_page_table(struct vm_area_struct *vma,
		unsigned int flags)
344
{
345 346 347 348 349 350 351 352
	/*
	 * 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.
	 */
353 354
	if ((flags & FOLL_DUMP) &&
			(vma_is_anonymous(vma) || !vma->vm_ops->fault))
355 356 357
		return ERR_PTR(-EFAULT);
	return NULL;
}
358

359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382
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;
}

383
/*
384 385
 * FOLL_FORCE or a forced COW break can write even to unwritable pte's,
 * but only after we've gone through a COW cycle and they are dirty.
386 387 388
 */
static inline bool can_follow_write_pte(pte_t pte, unsigned int flags)
{
389 390 391 392 393 394 395 396 397 398 399
	return pte_write(pte) || ((flags & FOLL_COW) && pte_dirty(pte));
}

/*
 * A (separate) COW fault might break the page the other way and
 * get_user_pages() would return the page from what is now the wrong
 * VM. So we need to force a COW break at GUP time even for reads.
 */
static inline bool should_force_cow_break(struct vm_area_struct *vma, unsigned int flags)
{
	return is_cow_mapping(vma->vm_flags) && (flags & (FOLL_GET | FOLL_PIN));
400 401
}

402
static struct page *follow_page_pte(struct vm_area_struct *vma,
403 404
		unsigned long address, pmd_t *pmd, unsigned int flags,
		struct dev_pagemap **pgmap)
405 406 407 408 409
{
	struct mm_struct *mm = vma->vm_mm;
	struct page *page;
	spinlock_t *ptl;
	pte_t *ptep, pte;
410
	int ret;
411

412 413 414 415
	/* FOLL_GET and FOLL_PIN are mutually exclusive. */
	if (WARN_ON_ONCE((flags & (FOLL_PIN | FOLL_GET)) ==
			 (FOLL_PIN | FOLL_GET)))
		return ERR_PTR(-EINVAL);
416
retry:
417
	if (unlikely(pmd_bad(*pmd)))
418
		return no_page_table(vma, flags);
419 420 421 422 423 424 425 426 427 428 429 430

	ptep = pte_offset_map_lock(mm, pmd, address, &ptl);
	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;
431
		if (pte_none(pte))
432 433 434 435 436 437
			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);
438
		goto retry;
439
	}
440
	if ((flags & FOLL_NUMA) && pte_protnone(pte))
441
		goto no_page;
442
	if ((flags & FOLL_WRITE) && !can_follow_write_pte(pte, flags)) {
443 444 445
		pte_unmap_unlock(ptep, ptl);
		return NULL;
	}
446 447

	page = vm_normal_page(vma, address, pte);
J
John Hubbard 已提交
448
	if (!page && pte_devmap(pte) && (flags & (FOLL_GET | FOLL_PIN))) {
449
		/*
J
John Hubbard 已提交
450 451 452
		 * Only return device mapping pages in the FOLL_GET or FOLL_PIN
		 * case since they are only valid while holding the pgmap
		 * reference.
453
		 */
454 455
		*pgmap = get_dev_pagemap(pte_pfn(pte), *pgmap);
		if (*pgmap)
456 457 458 459
			page = pte_page(pte);
		else
			goto no_page;
	} else if (unlikely(!page)) {
460 461 462 463 464 465 466 467 468 469 470 471 472
		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 {
			ret = follow_pfn_pte(vma, address, ptep, flags);
			page = ERR_PTR(ret);
			goto out;
		}
473 474
	}

475 476 477 478 479 480 481 482 483 484 485 486
	if (flags & FOLL_SPLIT && PageTransCompound(page)) {
		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;
	}

J
John Hubbard 已提交
487 488 489 490
	/* try_grab_page() does nothing unless FOLL_GET or FOLL_PIN is set. */
	if (unlikely(!try_grab_page(page, flags))) {
		page = ERR_PTR(-ENOMEM);
		goto out;
491
	}
492 493 494 495 496 497 498 499 500 501 502 503 504
	/*
	 * We need to make the page accessible if and only if we are going
	 * to access its content (the FOLL_PIN case).  Please see
	 * Documentation/core-api/pin_user_pages.rst for details.
	 */
	if (flags & FOLL_PIN) {
		ret = arch_make_page_accessible(page);
		if (ret) {
			unpin_user_page(page);
			page = ERR_PTR(ret);
			goto out;
		}
	}
505 506 507 508 509 510 511 512 513 514 515
	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);
	}
E
Eric B Munson 已提交
516
	if ((flags & FOLL_MLOCK) && (vma->vm_flags & VM_LOCKED)) {
517 518 519 520
		/* Do not mlock pte-mapped THP */
		if (PageTransCompound(page))
			goto out;

521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541
		/*
		 * 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);
		}
	}
542
out:
543 544 545 546 547
	pte_unmap_unlock(ptep, ptl);
	return page;
no_page:
	pte_unmap_unlock(ptep, ptl);
	if (!pte_none(pte))
548 549 550 551
		return NULL;
	return no_page_table(vma, flags);
}

552 553
static struct page *follow_pmd_mask(struct vm_area_struct *vma,
				    unsigned long address, pud_t *pudp,
554 555
				    unsigned int flags,
				    struct follow_page_context *ctx)
556
{
557
	pmd_t *pmd, pmdval;
558 559 560 561
	spinlock_t *ptl;
	struct page *page;
	struct mm_struct *mm = vma->vm_mm;

562
	pmd = pmd_offset(pudp, address);
563 564 565 566 567 568
	/*
	 * 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))
569
		return no_page_table(vma, flags);
570
	if (pmd_huge(pmdval) && is_vm_hugetlb_page(vma)) {
571 572 573 574
		page = follow_huge_pmd(mm, address, pmd, flags);
		if (page)
			return page;
		return no_page_table(vma, flags);
575
	}
576
	if (is_hugepd(__hugepd(pmd_val(pmdval)))) {
577
		page = follow_huge_pd(vma, address,
578
				      __hugepd(pmd_val(pmdval)), flags,
579 580 581 582 583
				      PMD_SHIFT);
		if (page)
			return page;
		return no_page_table(vma, flags);
	}
584
retry:
585
	if (!pmd_present(pmdval)) {
586 587 588
		if (likely(!(flags & FOLL_MIGRATION)))
			return no_page_table(vma, flags);
		VM_BUG_ON(thp_migration_supported() &&
589 590
				  !is_pmd_migration_entry(pmdval));
		if (is_pmd_migration_entry(pmdval))
591
			pmd_migration_entry_wait(mm, pmd);
592 593 594 595 596 597 598
		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);
599 600
		goto retry;
	}
601
	if (pmd_devmap(pmdval)) {
602
		ptl = pmd_lock(mm, pmd);
603
		page = follow_devmap_pmd(vma, address, pmd, flags, &ctx->pgmap);
604 605 606 607
		spin_unlock(ptl);
		if (page)
			return page;
	}
608
	if (likely(!pmd_trans_huge(pmdval)))
609
		return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
610

611
	if ((flags & FOLL_NUMA) && pmd_protnone(pmdval))
612 613
		return no_page_table(vma, flags);

614
retry_locked:
615
	ptl = pmd_lock(mm, pmd);
616 617 618 619
	if (unlikely(pmd_none(*pmd))) {
		spin_unlock(ptl);
		return no_page_table(vma, flags);
	}
620 621 622 623 624 625 626
	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;
	}
627 628
	if (unlikely(!pmd_trans_huge(*pmd))) {
		spin_unlock(ptl);
629
		return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
630
	}
S
Song Liu 已提交
631
	if (flags & (FOLL_SPLIT | FOLL_SPLIT_PMD)) {
632 633 634 635 636
		int ret;
		page = pmd_page(*pmd);
		if (is_huge_zero_page(page)) {
			spin_unlock(ptl);
			ret = 0;
637
			split_huge_pmd(vma, pmd, address);
638 639
			if (pmd_trans_unstable(pmd))
				ret = -EBUSY;
S
Song Liu 已提交
640
		} else if (flags & FOLL_SPLIT) {
641 642 643 644
			if (unlikely(!try_get_page(page))) {
				spin_unlock(ptl);
				return ERR_PTR(-ENOMEM);
			}
645
			spin_unlock(ptl);
646 647 648 649
			lock_page(page);
			ret = split_huge_page(page);
			unlock_page(page);
			put_page(page);
650 651
			if (pmd_none(*pmd))
				return no_page_table(vma, flags);
S
Song Liu 已提交
652 653 654 655
		} else {  /* flags & FOLL_SPLIT_PMD */
			spin_unlock(ptl);
			split_huge_pmd(vma, pmd, address);
			ret = pte_alloc(mm, pmd) ? -ENOMEM : 0;
656 657 658
		}

		return ret ? ERR_PTR(ret) :
659
			follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
660
	}
661 662
	page = follow_trans_huge_pmd(vma, address, pmd, flags);
	spin_unlock(ptl);
663
	ctx->page_mask = HPAGE_PMD_NR - 1;
664
	return page;
665 666
}

667 668
static struct page *follow_pud_mask(struct vm_area_struct *vma,
				    unsigned long address, p4d_t *p4dp,
669 670
				    unsigned int flags,
				    struct follow_page_context *ctx)
671 672 673 674 675 676 677 678 679
{
	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);
680
	if (pud_huge(*pud) && is_vm_hugetlb_page(vma)) {
681 682 683 684 685
		page = follow_huge_pud(mm, address, pud, flags);
		if (page)
			return page;
		return no_page_table(vma, flags);
	}
686 687 688 689 690 691 692 693
	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);
	}
694 695
	if (pud_devmap(*pud)) {
		ptl = pud_lock(mm, pud);
696
		page = follow_devmap_pud(vma, address, pud, flags, &ctx->pgmap);
697 698 699 700 701 702 703
		spin_unlock(ptl);
		if (page)
			return page;
	}
	if (unlikely(pud_bad(*pud)))
		return no_page_table(vma, flags);

704
	return follow_pmd_mask(vma, address, pud, flags, ctx);
705 706 707 708
}

static struct page *follow_p4d_mask(struct vm_area_struct *vma,
				    unsigned long address, pgd_t *pgdp,
709 710
				    unsigned int flags,
				    struct follow_page_context *ctx)
711 712
{
	p4d_t *p4d;
713
	struct page *page;
714 715 716 717 718 719 720 721

	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);

722 723 724 725 726 727 728 729
	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);
	}
730
	return follow_pud_mask(vma, address, p4d, flags, ctx);
731 732 733 734 735 736 737
}

/**
 * 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
738 739
 * @ctx: contains dev_pagemap for %ZONE_DEVICE memory pinning and a
 *       pointer to output page_mask
740 741 742
 *
 * @flags can have FOLL_ flags set, defined in <linux/mm.h>
 *
743 744 745 746 747 748
 * 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
749 750 751
 * an error pointer if there is a mapping to something not represented
 * by a page descriptor (see also vm_normal_page()).
 */
752
static struct page *follow_page_mask(struct vm_area_struct *vma,
753
			      unsigned long address, unsigned int flags,
754
			      struct follow_page_context *ctx)
755 756 757 758 759
{
	pgd_t *pgd;
	struct page *page;
	struct mm_struct *mm = vma->vm_mm;

760
	ctx->page_mask = 0;
761 762 763 764

	/* make this handle hugepd */
	page = follow_huge_addr(mm, address, flags & FOLL_WRITE);
	if (!IS_ERR(page)) {
J
John Hubbard 已提交
765
		WARN_ON_ONCE(flags & (FOLL_GET | FOLL_PIN));
766 767 768 769 770 771 772 773
		return page;
	}

	pgd = pgd_offset(mm, address);

	if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd)))
		return no_page_table(vma, flags);

774 775 776 777 778 779
	if (pgd_huge(*pgd)) {
		page = follow_huge_pgd(mm, address, pgd, flags);
		if (page)
			return page;
		return no_page_table(vma, flags);
	}
780 781 782 783 784 785 786 787
	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);
	}
788

789 790 791 792 793 794 795 796 797 798 799 800 801
	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;
802 803
}

804 805 806 807 808
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;
809
	p4d_t *p4d;
810 811 812 813 814 815 816 817 818 819 820 821
	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);
822 823
	if (pgd_none(*pgd))
		return -EFAULT;
824
	p4d = p4d_offset(pgd, address);
825 826
	if (p4d_none(*p4d))
		return -EFAULT;
827
	pud = pud_offset(p4d, address);
828 829
	if (pud_none(*pud))
		return -EFAULT;
830
	pmd = pmd_offset(pud, address);
831
	if (!pmd_present(*pmd))
832 833 834 835 836 837 838 839 840 841 842 843 844 845
		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);
	}
846 847 848 849
	if (unlikely(!try_get_page(*page))) {
		ret = -ENOMEM;
		goto unmap;
	}
850 851 852 853 854 855 856
out:
	ret = 0;
unmap:
	pte_unmap(pte);
	return ret;
}

857
/*
858 859 860
 * mmap_sem must be held on entry.  If @locked != NULL and *@flags
 * does not include FOLL_NOWAIT, the mmap_sem may be released.  If it
 * is, *@locked will be set to 0 and -EBUSY returned.
861
 */
862
static int faultin_page(struct task_struct *tsk, struct vm_area_struct *vma,
863
		unsigned long address, unsigned int *flags, int *locked)
864 865
{
	unsigned int fault_flags = 0;
866
	vm_fault_t ret;
867

E
Eric B Munson 已提交
868 869 870
	/* mlock all present pages, but do not fault in new pages */
	if ((*flags & (FOLL_POPULATE | FOLL_MLOCK)) == FOLL_MLOCK)
		return -ENOENT;
871 872
	if (*flags & FOLL_WRITE)
		fault_flags |= FAULT_FLAG_WRITE;
873 874
	if (*flags & FOLL_REMOTE)
		fault_flags |= FAULT_FLAG_REMOTE;
875
	if (locked)
876
		fault_flags |= FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE;
877 878
	if (*flags & FOLL_NOWAIT)
		fault_flags |= FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_RETRY_NOWAIT;
879
	if (*flags & FOLL_TRIED) {
880 881 882 883
		/*
		 * Note: FAULT_FLAG_ALLOW_RETRY and FAULT_FLAG_TRIED
		 * can co-exist
		 */
884 885
		fault_flags |= FAULT_FLAG_TRIED;
	}
886

887
	ret = handle_mm_fault(vma, address, fault_flags);
888
	if (ret & VM_FAULT_ERROR) {
889 890 891 892
		int err = vm_fault_to_errno(ret, *flags);

		if (err)
			return err;
893 894 895 896 897 898 899 900 901 902 903
		BUG();
	}

	if (tsk) {
		if (ret & VM_FAULT_MAJOR)
			tsk->maj_flt++;
		else
			tsk->min_flt++;
	}

	if (ret & VM_FAULT_RETRY) {
904 905
		if (locked && !(fault_flags & FAULT_FLAG_RETRY_NOWAIT))
			*locked = 0;
906 907 908 909 910 911 912 913 914 915 916 917 918
		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))
919
		*flags |= FOLL_COW;
920 921 922
	return 0;
}

923 924 925
static int check_vma_flags(struct vm_area_struct *vma, unsigned long gup_flags)
{
	vm_flags_t vm_flags = vma->vm_flags;
926 927
	int write = (gup_flags & FOLL_WRITE);
	int foreign = (gup_flags & FOLL_REMOTE);
928 929 930 931

	if (vm_flags & (VM_IO | VM_PFNMAP))
		return -EFAULT;

932 933 934
	if (gup_flags & FOLL_ANON && !vma_is_anonymous(vma))
		return -EFAULT;

935
	if (write) {
936 937 938 939 940 941 942 943 944 945 946 947
		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.
			 */
948
			if (!is_cow_mapping(vm_flags))
949 950 951 952 953 954 955 956 957 958 959 960
				return -EFAULT;
		}
	} 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;
	}
961 962 963 964 965
	/*
	 * gups are always data accesses, not instruction
	 * fetches, so execute=false here
	 */
	if (!arch_vma_access_permitted(vma, write, false, foreign))
966
		return -EFAULT;
967 968 969
	return 0;
}

970 971 972 973 974 975 976 977 978 979 980 981
/**
 * __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.
982
 * @locked:     whether we're still with the mmap_sem held
983
 *
984 985 986 987 988 989 990
 * Returns either number of pages pinned (which may be less than the
 * number requested), or an error. Details about the return value:
 *
 * -- If nr_pages is 0, returns 0.
 * -- If nr_pages is >0, but no pages were pinned, returns -errno.
 * -- If nr_pages is >0, and some pages were pinned, returns the number of
 *    pages pinned. Again, this may be less than nr_pages.
991
 * -- 0 return value is possible when the fault would need to be retried.
992 993 994 995
 *
 * The caller is responsible for releasing returned @pages, via put_page().
 *
 * @vmas are valid only as long as mmap_sem is held.
996
 *
997
 * Must be called with mmap_sem held.  It may be released.  See below.
998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017
 *
 * __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.
 *
1018 1019 1020
 * If @locked != NULL, *@locked will be set to 0 when mmap_sem is
 * released by an up_read().  That can happen if @gup_flags does not
 * have FOLL_NOWAIT.
1021
 *
1022
 * A caller using such a combination of @locked and @gup_flags
1023 1024 1025
 * 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.
1026 1027 1028 1029 1030
 *
 * 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.
 */
L
Lorenzo Stoakes 已提交
1031
static long __get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
1032 1033
		unsigned long start, unsigned long nr_pages,
		unsigned int gup_flags, struct page **pages,
1034
		struct vm_area_struct **vmas, int *locked)
1035
{
1036
	long ret = 0, i = 0;
1037
	struct vm_area_struct *vma = NULL;
1038
	struct follow_page_context ctx = { NULL };
1039 1040 1041 1042

	if (!nr_pages)
		return 0;

1043 1044
	start = untagged_addr(start);

1045
	VM_BUG_ON(!!pages != !!(gup_flags & (FOLL_GET | FOLL_PIN)));
1046 1047 1048 1049 1050 1051 1052 1053 1054 1055

	/*
	 * 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;

	do {
1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067
		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)) {
				ret = get_gate_page(mm, start & PAGE_MASK,
						gup_flags, &vma,
						pages ? &pages[i] : NULL);
				if (ret)
1068
					goto out;
1069
				ctx.page_mask = 0;
1070 1071
				goto next_page;
			}
1072

1073 1074 1075 1076
			if (!vma || check_vma_flags(vma, gup_flags)) {
				ret = -EFAULT;
				goto out;
			}
1077
			if (is_vm_hugetlb_page(vma)) {
1078 1079
				if (should_force_cow_break(vma, foll_flags))
					foll_flags |= FOLL_WRITE;
1080 1081
				i = follow_hugetlb_page(mm, vma, pages, vmas,
						&start, &nr_pages, i,
1082
						foll_flags, locked);
1083 1084 1085 1086 1087 1088 1089 1090 1091 1092
				if (locked && *locked == 0) {
					/*
					 * We've got a VM_FAULT_RETRY
					 * and we've lost mmap_sem.
					 * We must stop here.
					 */
					BUG_ON(gup_flags & FOLL_NOWAIT);
					BUG_ON(ret != 0);
					goto out;
				}
1093
				continue;
1094
			}
1095
		}
1096 1097 1098 1099

		if (should_force_cow_break(vma, foll_flags))
			foll_flags |= FOLL_WRITE;

1100 1101 1102 1103 1104
retry:
		/*
		 * If we have a pending SIGKILL, don't keep faulting pages and
		 * potentially allocating memory.
		 */
1105
		if (fatal_signal_pending(current)) {
1106
			ret = -EINTR;
1107 1108
			goto out;
		}
1109
		cond_resched();
1110 1111

		page = follow_page_mask(vma, start, foll_flags, &ctx);
1112 1113
		if (!page) {
			ret = faultin_page(tsk, vma, start, &foll_flags,
1114
					   locked);
1115 1116 1117
			switch (ret) {
			case 0:
				goto retry;
1118 1119
			case -EBUSY:
				ret = 0;
J
Joe Perches 已提交
1120
				fallthrough;
1121 1122 1123
			case -EFAULT:
			case -ENOMEM:
			case -EHWPOISON:
1124
				goto out;
1125 1126
			case -ENOENT:
				goto next_page;
1127
			}
1128
			BUG();
1129 1130 1131 1132 1133 1134 1135
		} else if (PTR_ERR(page) == -EEXIST) {
			/*
			 * Proper page table entry exists, but no corresponding
			 * struct page.
			 */
			goto next_page;
		} else if (IS_ERR(page)) {
1136 1137
			ret = PTR_ERR(page);
			goto out;
1138
		}
1139 1140 1141 1142
		if (pages) {
			pages[i] = page;
			flush_anon_page(vma, page, start);
			flush_dcache_page(page);
1143
			ctx.page_mask = 0;
1144 1145
		}
next_page:
1146 1147
		if (vmas) {
			vmas[i] = vma;
1148
			ctx.page_mask = 0;
1149
		}
1150
		page_increm = 1 + (~(start >> PAGE_SHIFT) & ctx.page_mask);
1151 1152 1153 1154 1155
		if (page_increm > nr_pages)
			page_increm = nr_pages;
		i += page_increm;
		start += page_increm * PAGE_SIZE;
		nr_pages -= page_increm;
1156
	} while (nr_pages);
1157 1158 1159 1160
out:
	if (ctx.pgmap)
		put_dev_pagemap(ctx.pgmap);
	return i ? i : ret;
1161 1162
}

1163 1164
static bool vma_permits_fault(struct vm_area_struct *vma,
			      unsigned int fault_flags)
1165
{
1166 1167
	bool write   = !!(fault_flags & FAULT_FLAG_WRITE);
	bool foreign = !!(fault_flags & FAULT_FLAG_REMOTE);
1168
	vm_flags_t vm_flags = write ? VM_WRITE : VM_READ;
1169 1170 1171 1172

	if (!(vm_flags & vma->vm_flags))
		return false;

1173 1174
	/*
	 * The architecture might have a hardware protection
1175
	 * mechanism other than read/write that can deny access.
1176 1177 1178
	 *
	 * gup always represents data access, not instruction
	 * fetches, so execute=false here:
1179
	 */
1180
	if (!arch_vma_access_permitted(vma, write, false, foreign))
1181 1182
		return false;

1183 1184 1185
	return true;
}

1186
/**
1187 1188 1189 1190 1191 1192
 * 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()
1193
 * @unlocked:	did we unlock the mmap_sem while retrying, maybe NULL if caller
1194 1195
 *		does not allow retry. If NULL, the caller must guarantee
 *		that fault_flags does not contain FAULT_FLAG_ALLOW_RETRY.
1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206
 *
 * 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
1207
 * get_user_pages() only guarantees to update these in the struct page.
1208 1209 1210 1211 1212 1213
 *
 * 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.
 *
1214 1215
 * 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().
1216 1217
 */
int fixup_user_fault(struct task_struct *tsk, struct mm_struct *mm,
1218 1219
		     unsigned long address, unsigned int fault_flags,
		     bool *unlocked)
1220 1221
{
	struct vm_area_struct *vma;
1222
	vm_fault_t ret, major = 0;
1223

1224 1225
	address = untagged_addr(address);

1226
	if (unlocked)
1227
		fault_flags |= FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE;
1228

1229
retry:
1230 1231 1232 1233
	vma = find_extend_vma(mm, address);
	if (!vma || address < vma->vm_start)
		return -EFAULT;

1234
	if (!vma_permits_fault(vma, fault_flags))
1235 1236
		return -EFAULT;

1237 1238 1239 1240
	if ((fault_flags & FAULT_FLAG_KILLABLE) &&
	    fatal_signal_pending(current))
		return -EINTR;

1241
	ret = handle_mm_fault(vma, address, fault_flags);
1242
	major |= ret & VM_FAULT_MAJOR;
1243
	if (ret & VM_FAULT_ERROR) {
1244 1245 1246 1247
		int err = vm_fault_to_errno(ret, 0);

		if (err)
			return err;
1248 1249
		BUG();
	}
1250 1251

	if (ret & VM_FAULT_RETRY) {
1252
		mmap_read_lock(mm);
1253 1254 1255
		*unlocked = true;
		fault_flags |= FAULT_FLAG_TRIED;
		goto retry;
1256 1257
	}

1258
	if (tsk) {
1259
		if (major)
1260 1261 1262 1263 1264 1265
			tsk->maj_flt++;
		else
			tsk->min_flt++;
	}
	return 0;
}
1266
EXPORT_SYMBOL_GPL(fixup_user_fault);
1267

1268 1269 1270 1271
/*
 * Please note that this function, unlike __get_user_pages will not
 * return 0 for nr_pages > 0 without FOLL_NOWAIT
 */
1272 1273 1274 1275 1276 1277
static __always_inline 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,
1278
						int *locked,
1279
						unsigned int flags)
1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290
{
	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);
	}

1291 1292 1293 1294 1295 1296 1297 1298 1299 1300
	/*
	 * FOLL_PIN and FOLL_GET are mutually exclusive. Traditional behavior
	 * is to set FOLL_GET if the caller wants pages[] filled in (but has
	 * carelessly failed to specify FOLL_GET), so keep doing that, but only
	 * for FOLL_GET, not for the newer FOLL_PIN.
	 *
	 * FOLL_PIN always expects pages to be non-null, but no need to assert
	 * that here, as any failures will be obvious enough.
	 */
	if (pages && !(flags & FOLL_PIN))
1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324
		flags |= FOLL_GET;

	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);
		}

		if (ret > 0) {
			nr_pages -= ret;
			pages_done += ret;
			if (!nr_pages)
				break;
		}
		if (*locked) {
1325 1326 1327 1328
			/*
			 * VM_FAULT_RETRY didn't trigger or it was a
			 * FOLL_NOWAIT.
			 */
1329 1330 1331 1332
			if (!pages_done)
				pages_done = ret;
			break;
		}
1333 1334 1335 1336 1337 1338
		/*
		 * VM_FAULT_RETRY triggered, so seek to the faulting offset.
		 * For the prefault case (!pages) we only update counts.
		 */
		if (likely(pages))
			pages += ret;
1339
		start += ret << PAGE_SHIFT;
1340
		lock_dropped = true;
1341

1342
retry:
1343 1344
		/*
		 * Repeat on the address that fired VM_FAULT_RETRY
1345 1346 1347 1348
		 * with both FAULT_FLAG_ALLOW_RETRY and
		 * FAULT_FLAG_TRIED.  Note that GUP can be interrupted
		 * by fatal signals, so we need to check it before we
		 * start trying again otherwise it can loop forever.
1349
		 */
1350

1351 1352 1353
		if (fatal_signal_pending(current)) {
			if (!pages_done)
				pages_done = -EINTR;
1354
			break;
1355
		}
1356

1357
		ret = mmap_read_lock_killable(mm);
1358 1359 1360 1361 1362 1363
		if (ret) {
			BUG_ON(ret > 0);
			if (!pages_done)
				pages_done = ret;
			break;
		}
1364

1365
		*locked = 1;
1366
		ret = __get_user_pages(tsk, mm, start, 1, flags | FOLL_TRIED,
1367 1368 1369 1370 1371 1372
				       pages, NULL, locked);
		if (!*locked) {
			/* Continue to retry until we succeeded */
			BUG_ON(ret != 0);
			goto retry;
		}
1373 1374 1375 1376 1377 1378 1379 1380 1381 1382
		if (ret != 1) {
			BUG_ON(ret > 1);
			if (!pages_done)
				pages_done = ret;
			break;
		}
		nr_pages--;
		pages_done++;
		if (!nr_pages)
			break;
1383 1384
		if (likely(pages))
			pages++;
1385 1386
		start += PAGE_SIZE;
	}
1387
	if (lock_dropped && *locked) {
1388 1389 1390 1391
		/*
		 * We must let the caller know we temporarily dropped the lock
		 * and so the critical section protected by it was lost.
		 */
1392
		mmap_read_unlock(mm);
1393 1394 1395 1396 1397
		*locked = 0;
	}
	return pages_done;
}

1398 1399 1400 1401 1402
/**
 * populate_vma_page_range() -  populate a range of pages in the vma.
 * @vma:   target vma
 * @start: start address
 * @end:   end address
1403
 * @locked: whether the mmap_sem is still held
1404 1405 1406 1407 1408 1409 1410
 *
 * 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.
 *
1411
 * If @locked is NULL, it may be held for read or write and will
1412 1413
 * be unperturbed.
 *
1414 1415
 * If @locked is non-NULL, it must held for read only and may be
 * released.  If it's released, *@locked will be set to 0.
1416 1417
 */
long populate_vma_page_range(struct vm_area_struct *vma,
1418
		unsigned long start, unsigned long end, int *locked)
1419 1420 1421 1422 1423 1424 1425 1426 1427
{
	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);
1428
	mmap_assert_locked(mm);
1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444

	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.
	 */
1445
	if (vma_is_accessible(vma))
1446 1447 1448 1449 1450 1451 1452
		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,
1453
				NULL, NULL, locked);
1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479
}

/*
 * __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;
1480
			mmap_read_lock(mm);
1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511
			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)
1512
		mmap_read_unlock(mm);
1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543
	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 */
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 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588
#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 */
1589

1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661
#if defined(CONFIG_FS_DAX) || defined (CONFIG_CMA)
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;
}

#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);
}

1662 1663 1664 1665
static long check_and_migrate_cma_pages(struct task_struct *tsk,
					struct mm_struct *mm,
					unsigned long start,
					unsigned long nr_pages,
1666
					struct page **pages,
1667 1668
					struct vm_area_struct **vmas,
					unsigned int gup_flags)
1669
{
1670 1671
	unsigned long i;
	unsigned long step;
1672 1673 1674
	bool drain_allow = true;
	bool migrate_allow = true;
	LIST_HEAD(cma_page_list);
1675
	long ret = nr_pages;
1676 1677

check_again:
1678 1679 1680 1681 1682 1683 1684 1685
	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.
		 */
1686
		step = compound_nr(head) - (pages[i] - head);
1687 1688 1689 1690 1691
		/*
		 * 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.
		 */
1692 1693
		if (is_migrate_cma_page(head)) {
			if (PageHuge(head))
1694
				isolate_huge_page(head, &cma_page_list);
1695
			else {
1696 1697 1698 1699 1700 1701 1702 1703 1704
				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 +
H
Huang Ying 已提交
1705
							    page_is_file_lru(head),
1706 1707 1708 1709
							    hpage_nr_pages(head));
				}
			}
		}
1710 1711

		i += step;
1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732
	}

	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);
		}
		/*
1733 1734 1735
		 * We did migrate all the pages, Try to get the page references
		 * again migrating any new CMA pages which we failed to isolate
		 * earlier.
1736
		 */
1737
		ret = __get_user_pages_locked(tsk, mm, start, nr_pages,
1738 1739 1740
						   pages, vmas, NULL,
						   gup_flags);

1741 1742
		if ((ret > 0) && migrate_allow) {
			nr_pages = ret;
1743 1744 1745 1746 1747
			drain_allow = true;
			goto check_again;
		}
	}

1748
	return ret;
1749 1750
}
#else
1751 1752 1753 1754 1755 1756 1757
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)
1758 1759 1760
{
	return nr_pages;
}
1761
#endif /* CONFIG_CMA */
1762

1763
/*
1764 1765
 * __gup_longterm_locked() is a wrapper for __get_user_pages_locked which
 * allows us to process the FOLL_LONGTERM flag.
1766
 */
1767 1768 1769 1770 1771 1772 1773
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)
1774
{
1775 1776
	struct vm_area_struct **vmas_tmp = vmas;
	unsigned long flags = 0;
1777 1778
	long rc, i;

1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790
	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();
1791 1792
	}

1793 1794
	rc = __get_user_pages_locked(tsk, mm, start, nr_pages, pages,
				     vmas_tmp, NULL, gup_flags);
1795

1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809
	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);
1810
	}
1811 1812

out:
1813 1814
	if (vmas_tmp != vmas)
		kfree(vmas_tmp);
1815 1816
	return rc;
}
1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830
#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 */

1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861
#ifdef CONFIG_MMU
static long __get_user_pages_remote(struct task_struct *tsk,
				    struct mm_struct *mm,
				    unsigned long start, unsigned long nr_pages,
				    unsigned int gup_flags, struct page **pages,
				    struct vm_area_struct **vmas, int *locked)
{
	/*
	 * Parts of FOLL_LONGTERM behavior are incompatible with
	 * FAULT_FLAG_ALLOW_RETRY because of the FS DAX check requirement on
	 * vmas. However, this only comes up if locked is set, and there are
	 * callers that do request FOLL_LONGTERM, but do not set locked. So,
	 * allow what we can.
	 */
	if (gup_flags & FOLL_LONGTERM) {
		if (WARN_ON_ONCE(locked))
			return -EINVAL;
		/*
		 * This will check the vmas (even if our vmas arg is NULL)
		 * and return -ENOTSUPP if DAX isn't allowed in this case:
		 */
		return __gup_longterm_locked(tsk, mm, start, nr_pages, pages,
					     vmas, gup_flags | FOLL_TOUCH |
					     FOLL_REMOTE);
	}

	return __get_user_pages_locked(tsk, mm, start, nr_pages, pages, vmas,
				       locked,
				       gup_flags | FOLL_TOUCH | FOLL_REMOTE);
}

1862
/**
1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892
 * get_user_pages_remote() - pin user pages in memory
 * @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
 * @gup_flags:	flags modifying lookup 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.
 * @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.
 *
 * Returns either number of pages pinned (which may be less than the
 * number requested), or an error. Details about the return value:
 *
 * -- If nr_pages is 0, returns 0.
 * -- If nr_pages is >0, but no pages were pinned, returns -errno.
 * -- If nr_pages is >0, and some pages were pinned, returns the number of
 *    pages pinned. Again, this may be less than nr_pages.
 *
 * The caller is responsible for releasing returned @pages, via put_page().
 *
 * @vmas are valid only as long as mmap_sem is held.
 *
 * Must be called with mmap_sem held for read or write.
 *
1893 1894
 * get_user_pages_remote walks a process's page tables and takes a reference
 * to each struct page that each user address corresponds to at a given
1895 1896 1897 1898
 * 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
1899
 * get_user_pages_remote returns, and there may even be a completely different
1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910
 * 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.
 *
1911 1912 1913 1914 1915
 * get_user_pages_remote 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.
1916 1917 1918
 *
 * See also get_user_pages_fast, for performance critical applications.
 *
1919
 * get_user_pages_remote should be phased out in favor of
1920
 * get_user_pages_locked|unlocked or get_user_pages_fast. Nothing
1921
 * should use get_user_pages_remote because it cannot pass
1922 1923 1924 1925 1926 1927 1928
 * FAULT_FLAG_ALLOW_RETRY to handle_mm_fault.
 */
long get_user_pages_remote(struct task_struct *tsk, struct mm_struct *mm,
		unsigned long start, unsigned long nr_pages,
		unsigned int gup_flags, struct page **pages,
		struct vm_area_struct **vmas, int *locked)
{
1929 1930 1931 1932 1933 1934 1935
	/*
	 * FOLL_PIN must only be set internally by the pin_user_pages*() APIs,
	 * never directly by the caller, so enforce that with an assertion:
	 */
	if (WARN_ON_ONCE(gup_flags & FOLL_PIN))
		return -EINVAL;

1936 1937
	return __get_user_pages_remote(tsk, mm, start, nr_pages, gup_flags,
				       pages, vmas, locked);
1938 1939 1940
}
EXPORT_SYMBOL(get_user_pages_remote);

1941 1942 1943 1944 1945 1946 1947 1948
#else /* CONFIG_MMU */
long get_user_pages_remote(struct task_struct *tsk, struct mm_struct *mm,
			   unsigned long start, unsigned long nr_pages,
			   unsigned int gup_flags, struct page **pages,
			   struct vm_area_struct **vmas, int *locked)
{
	return 0;
}
J
John Hubbard 已提交
1949 1950 1951 1952 1953 1954 1955 1956 1957

static long __get_user_pages_remote(struct task_struct *tsk,
				    struct mm_struct *mm,
				    unsigned long start, unsigned long nr_pages,
				    unsigned int gup_flags, struct page **pages,
				    struct vm_area_struct **vmas, int *locked)
{
	return 0;
}
1958 1959
#endif /* !CONFIG_MMU */

1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970
/**
 * get_user_pages() - pin user pages in memory
 * @start:      starting user address
 * @nr_pages:   number of pages from start to pin
 * @gup_flags:  flags modifying lookup 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.
 *
1971 1972 1973 1974 1975 1976 1977 1978 1979 1980
 * 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)
{
1981 1982 1983 1984 1985 1986 1987
	/*
	 * FOLL_PIN must only be set internally by the pin_user_pages*() APIs,
	 * never directly by the caller, so enforce that with an assertion:
	 */
	if (WARN_ON_ONCE(gup_flags & FOLL_PIN))
		return -EINVAL;

1988 1989 1990 1991
	return __gup_longterm_locked(current, current->mm, start, nr_pages,
				     pages, vmas, gup_flags | FOLL_TOUCH);
}
EXPORT_SYMBOL(get_user_pages);
1992

1993
/**
1994
 * get_user_pages_locked() is suitable to replace the form:
1995
 *
1996 1997 1998 1999
 *      down_read(&mm->mmap_sem);
 *      do_something()
 *      get_user_pages(tsk, mm, ..., pages, NULL);
 *      up_read(&mm->mmap_sem);
2000
 *
2001
 *  to:
2002
 *
2003 2004 2005 2006 2007 2008
 *      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);
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023
 *
 * @start:      starting user address
 * @nr_pages:   number of pages from start to pin
 * @gup_flags:  flags modifying lookup 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.
 * @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.
 *
 * 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().
 *
2024
 */
2025 2026 2027
long get_user_pages_locked(unsigned long start, unsigned long nr_pages,
			   unsigned int gup_flags, struct page **pages,
			   int *locked)
2028 2029
{
	/*
2030 2031 2032 2033
	 * 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.
2034
	 */
2035 2036
	if (WARN_ON_ONCE(gup_flags & FOLL_LONGTERM))
		return -EINVAL;
2037 2038 2039 2040 2041 2042
	/*
	 * FOLL_PIN must only be set internally by the pin_user_pages*() APIs,
	 * never directly by the caller, so enforce that:
	 */
	if (WARN_ON_ONCE(gup_flags & FOLL_PIN))
		return -EINVAL;
2043

2044 2045 2046
	return __get_user_pages_locked(current, current->mm, start, nr_pages,
				       pages, NULL, locked,
				       gup_flags | FOLL_TOUCH);
2047
}
2048
EXPORT_SYMBOL(get_user_pages_locked);
2049 2050

/*
2051
 * get_user_pages_unlocked() is suitable to replace the form:
2052
 *
2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063
 *      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.
2064
 */
2065 2066
long get_user_pages_unlocked(unsigned long start, unsigned long nr_pages,
			     struct page **pages, unsigned int gup_flags)
2067 2068
{
	struct mm_struct *mm = current->mm;
2069 2070
	int locked = 1;
	long ret;
2071

2072 2073 2074 2075 2076 2077 2078 2079
	/*
	 * 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;
2080

2081
	mmap_read_lock(mm);
2082 2083
	ret = __get_user_pages_locked(current, mm, start, nr_pages, pages, NULL,
				      &locked, gup_flags | FOLL_TOUCH);
2084
	if (locked)
2085
		mmap_read_unlock(mm);
2086
	return ret;
2087
}
2088
EXPORT_SYMBOL(get_user_pages_unlocked);
2089 2090

/*
2091
 * Fast GUP
2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111
 *
 * 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:
 *
2112
 *  *) Either MMU_GATHER_RCU_TABLE_FREE is enabled, and tlb_remove_table() is used to
2113
 *  free pages containing page tables or TLB flushing requires IPI broadcast.
2114 2115 2116 2117 2118 2119 2120 2121 2122
 *
 *  *) 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.
 */
2123
#ifdef CONFIG_HAVE_FAST_GUP
J
John Hubbard 已提交
2124 2125 2126

static void put_compound_head(struct page *page, int refs, unsigned int flags)
{
2127
	if (flags & FOLL_PIN) {
2128 2129 2130
		mod_node_page_state(page_pgdat(page), NR_FOLL_PIN_RELEASED,
				    refs);

2131 2132 2133 2134 2135
		if (hpage_pincount_available(page))
			hpage_pincount_sub(page, refs);
		else
			refs *= GUP_PIN_COUNTING_BIAS;
	}
J
John Hubbard 已提交
2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146

	VM_BUG_ON_PAGE(page_ref_count(page) < refs, page);
	/*
	 * Calling put_page() for each ref is unnecessarily slow. Only the last
	 * ref needs a put_page().
	 */
	if (refs > 1)
		page_ref_sub(page, refs - 1);
	put_page(page);
}

2147
#ifdef CONFIG_GUP_GET_PTE_LOW_HIGH
J
John Hubbard 已提交
2148

2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182
/*
 * 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;
2183

2184 2185 2186 2187 2188 2189 2190 2191 2192 2193
	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 */
2194
/*
2195
 * We require that the PTE can be read atomically.
2196 2197 2198 2199 2200
 */
static inline pte_t gup_get_pte(pte_t *ptep)
{
	return READ_ONCE(*ptep);
}
2201
#endif /* CONFIG_GUP_GET_PTE_LOW_HIGH */
2202

2203
static void __maybe_unused undo_dev_pagemap(int *nr, int nr_start,
2204
					    unsigned int flags,
2205
					    struct page **pages)
2206 2207 2208 2209 2210
{
	while ((*nr) - nr_start) {
		struct page *page = pages[--(*nr)];

		ClearPageReferenced(page);
J
John Hubbard 已提交
2211 2212 2213 2214
		if (flags & FOLL_PIN)
			unpin_user_page(page);
		else
			put_page(page);
2215 2216 2217
	}
}

2218
#ifdef CONFIG_ARCH_HAS_PTE_SPECIAL
2219
static int gup_pte_range(pmd_t pmd, unsigned long addr, unsigned long end,
2220
			 unsigned int flags, struct page **pages, int *nr)
2221
{
2222 2223
	struct dev_pagemap *pgmap = NULL;
	int nr_start = *nr, ret = 0;
2224 2225 2226 2227
	pte_t *ptep, *ptem;

	ptem = ptep = pte_offset_map(&pmd, addr);
	do {
2228
		pte_t pte = gup_get_pte(ptep);
2229
		struct page *head, *page;
2230 2231 2232

		/*
		 * Similar to the PMD case below, NUMA hinting must take slow
2233
		 * path using the pte_protnone check.
2234
		 */
2235 2236 2237
		if (pte_protnone(pte))
			goto pte_unmap;

2238
		if (!pte_access_permitted(pte, flags & FOLL_WRITE))
2239 2240
			goto pte_unmap;

2241
		if (pte_devmap(pte)) {
2242 2243 2244
			if (unlikely(flags & FOLL_LONGTERM))
				goto pte_unmap;

2245 2246
			pgmap = get_dev_pagemap(pte_pfn(pte), pgmap);
			if (unlikely(!pgmap)) {
2247
				undo_dev_pagemap(nr, nr_start, flags, pages);
2248 2249 2250
				goto pte_unmap;
			}
		} else if (pte_special(pte))
2251 2252 2253 2254 2255
			goto pte_unmap;

		VM_BUG_ON(!pfn_valid(pte_pfn(pte)));
		page = pte_page(pte);

J
John Hubbard 已提交
2256
		head = try_grab_compound_head(page, 1, flags);
2257
		if (!head)
2258 2259 2260
			goto pte_unmap;

		if (unlikely(pte_val(pte) != pte_val(*ptep))) {
J
John Hubbard 已提交
2261
			put_compound_head(head, 1, flags);
2262 2263 2264
			goto pte_unmap;
		}

2265
		VM_BUG_ON_PAGE(compound_head(page) != head, page);
2266

2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279
		/*
		 * We need to make the page accessible if and only if we are
		 * going to access its content (the FOLL_PIN case).  Please
		 * see Documentation/core-api/pin_user_pages.rst for
		 * details.
		 */
		if (flags & FOLL_PIN) {
			ret = arch_make_page_accessible(page);
			if (ret) {
				unpin_user_page(page);
				goto pte_unmap;
			}
		}
2280
		SetPageReferenced(page);
2281 2282 2283 2284 2285 2286 2287 2288
		pages[*nr] = page;
		(*nr)++;

	} while (ptep++, addr += PAGE_SIZE, addr != end);

	ret = 1;

pte_unmap:
2289 2290
	if (pgmap)
		put_dev_pagemap(pgmap);
2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301
	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
2302
 * get_user_pages_fast_only implementation that can pin pages. Thus it's still
2303 2304 2305
 * 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,
2306
			 unsigned int flags, struct page **pages, int *nr)
2307 2308 2309
{
	return 0;
}
2310
#endif /* CONFIG_ARCH_HAS_PTE_SPECIAL */
2311

R
Robin Murphy 已提交
2312
#if defined(CONFIG_ARCH_HAS_PTE_DEVMAP) && defined(CONFIG_TRANSPARENT_HUGEPAGE)
2313
static int __gup_device_huge(unsigned long pfn, unsigned long addr,
2314 2315
			     unsigned long end, unsigned int flags,
			     struct page **pages, int *nr)
2316 2317 2318 2319 2320 2321 2322 2323 2324
{
	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)) {
2325
			undo_dev_pagemap(nr, nr_start, flags, pages);
2326 2327 2328 2329
			return 0;
		}
		SetPageReferenced(page);
		pages[*nr] = page;
J
John Hubbard 已提交
2330 2331 2332 2333
		if (unlikely(!try_grab_page(page, flags))) {
			undo_dev_pagemap(nr, nr_start, flags, pages);
			return 0;
		}
2334 2335 2336
		(*nr)++;
		pfn++;
	} while (addr += PAGE_SIZE, addr != end);
2337 2338 2339

	if (pgmap)
		put_dev_pagemap(pgmap);
2340 2341 2342
	return 1;
}

2343
static int __gup_device_huge_pmd(pmd_t orig, pmd_t *pmdp, unsigned long addr,
2344 2345
				 unsigned long end, unsigned int flags,
				 struct page **pages, int *nr)
2346 2347
{
	unsigned long fault_pfn;
2348 2349 2350
	int nr_start = *nr;

	fault_pfn = pmd_pfn(orig) + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
2351
	if (!__gup_device_huge(fault_pfn, addr, end, flags, pages, nr))
2352
		return 0;
2353

2354
	if (unlikely(pmd_val(orig) != pmd_val(*pmdp))) {
2355
		undo_dev_pagemap(nr, nr_start, flags, pages);
2356 2357 2358
		return 0;
	}
	return 1;
2359 2360
}

2361
static int __gup_device_huge_pud(pud_t orig, pud_t *pudp, unsigned long addr,
2362 2363
				 unsigned long end, unsigned int flags,
				 struct page **pages, int *nr)
2364 2365
{
	unsigned long fault_pfn;
2366 2367 2368
	int nr_start = *nr;

	fault_pfn = pud_pfn(orig) + ((addr & ~PUD_MASK) >> PAGE_SHIFT);
2369
	if (!__gup_device_huge(fault_pfn, addr, end, flags, pages, nr))
2370
		return 0;
2371

2372
	if (unlikely(pud_val(orig) != pud_val(*pudp))) {
2373
		undo_dev_pagemap(nr, nr_start, flags, pages);
2374 2375 2376
		return 0;
	}
	return 1;
2377 2378
}
#else
2379
static int __gup_device_huge_pmd(pmd_t orig, pmd_t *pmdp, unsigned long addr,
2380 2381
				 unsigned long end, unsigned int flags,
				 struct page **pages, int *nr)
2382 2383 2384 2385 2386
{
	BUILD_BUG();
	return 0;
}

2387
static int __gup_device_huge_pud(pud_t pud, pud_t *pudp, unsigned long addr,
2388 2389
				 unsigned long end, unsigned int flags,
				 struct page **pages, int *nr)
2390 2391 2392 2393 2394 2395
{
	BUILD_BUG();
	return 0;
}
#endif

2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406
static int record_subpages(struct page *page, unsigned long addr,
			   unsigned long end, struct page **pages)
{
	int nr;

	for (nr = 0; addr != end; addr += PAGE_SIZE)
		pages[nr++] = page++;

	return nr;
}

2407 2408 2409 2410 2411 2412 2413 2414 2415
#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,
2416 2417
		       unsigned long end, unsigned int flags,
		       struct page **pages, int *nr)
2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429
{
	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);

2430
	if (!pte_access_permitted(pte, flags & FOLL_WRITE))
2431 2432 2433 2434 2435 2436 2437
		return 0;

	/* hugepages are never "special" */
	VM_BUG_ON(!pfn_valid(pte_pfn(pte)));

	head = pte_page(pte);
	page = head + ((addr & (sz-1)) >> PAGE_SHIFT);
2438
	refs = record_subpages(page, addr, end, pages + *nr);
2439

J
John Hubbard 已提交
2440
	head = try_grab_compound_head(head, refs, flags);
2441
	if (!head)
2442 2443 2444
		return 0;

	if (unlikely(pte_val(pte) != pte_val(*ptep))) {
2445
		put_compound_head(head, refs, flags);
2446 2447 2448
		return 0;
	}

2449
	*nr += refs;
2450
	SetPageReferenced(head);
2451 2452 2453 2454
	return 1;
}

static int gup_huge_pd(hugepd_t hugepd, unsigned long addr,
2455
		unsigned int pdshift, unsigned long end, unsigned int flags,
2456 2457 2458 2459 2460 2461 2462 2463 2464
		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);
2465
		if (!gup_hugepte(ptep, sz, addr, end, flags, pages, nr))
2466 2467 2468 2469 2470 2471 2472
			return 0;
	} while (ptep++, addr = next, addr != end);

	return 1;
}
#else
static inline int gup_huge_pd(hugepd_t hugepd, unsigned long addr,
2473
		unsigned int pdshift, unsigned long end, unsigned int flags,
2474 2475 2476 2477 2478 2479
		struct page **pages, int *nr)
{
	return 0;
}
#endif /* CONFIG_ARCH_HAS_HUGEPD */

2480
static int gup_huge_pmd(pmd_t orig, pmd_t *pmdp, unsigned long addr,
2481 2482
			unsigned long end, unsigned int flags,
			struct page **pages, int *nr)
2483
{
2484
	struct page *head, *page;
2485 2486
	int refs;

2487
	if (!pmd_access_permitted(orig, flags & FOLL_WRITE))
2488 2489
		return 0;

2490 2491 2492
	if (pmd_devmap(orig)) {
		if (unlikely(flags & FOLL_LONGTERM))
			return 0;
2493 2494
		return __gup_device_huge_pmd(orig, pmdp, addr, end, flags,
					     pages, nr);
2495
	}
2496

2497
	page = pmd_page(orig) + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
2498
	refs = record_subpages(page, addr, end, pages + *nr);
2499

J
John Hubbard 已提交
2500
	head = try_grab_compound_head(pmd_page(orig), refs, flags);
2501
	if (!head)
2502 2503 2504
		return 0;

	if (unlikely(pmd_val(orig) != pmd_val(*pmdp))) {
2505
		put_compound_head(head, refs, flags);
2506 2507 2508
		return 0;
	}

2509
	*nr += refs;
2510
	SetPageReferenced(head);
2511 2512 2513 2514
	return 1;
}

static int gup_huge_pud(pud_t orig, pud_t *pudp, unsigned long addr,
2515 2516
			unsigned long end, unsigned int flags,
			struct page **pages, int *nr)
2517
{
2518
	struct page *head, *page;
2519 2520
	int refs;

2521
	if (!pud_access_permitted(orig, flags & FOLL_WRITE))
2522 2523
		return 0;

2524 2525 2526
	if (pud_devmap(orig)) {
		if (unlikely(flags & FOLL_LONGTERM))
			return 0;
2527 2528
		return __gup_device_huge_pud(orig, pudp, addr, end, flags,
					     pages, nr);
2529
	}
2530

2531
	page = pud_page(orig) + ((addr & ~PUD_MASK) >> PAGE_SHIFT);
2532
	refs = record_subpages(page, addr, end, pages + *nr);
2533

J
John Hubbard 已提交
2534
	head = try_grab_compound_head(pud_page(orig), refs, flags);
2535
	if (!head)
2536 2537 2538
		return 0;

	if (unlikely(pud_val(orig) != pud_val(*pudp))) {
2539
		put_compound_head(head, refs, flags);
2540 2541 2542
		return 0;
	}

2543
	*nr += refs;
2544
	SetPageReferenced(head);
2545 2546 2547
	return 1;
}

2548
static int gup_huge_pgd(pgd_t orig, pgd_t *pgdp, unsigned long addr,
2549
			unsigned long end, unsigned int flags,
2550 2551 2552
			struct page **pages, int *nr)
{
	int refs;
2553
	struct page *head, *page;
2554

2555
	if (!pgd_access_permitted(orig, flags & FOLL_WRITE))
2556 2557
		return 0;

2558
	BUILD_BUG_ON(pgd_devmap(orig));
2559

2560
	page = pgd_page(orig) + ((addr & ~PGDIR_MASK) >> PAGE_SHIFT);
2561
	refs = record_subpages(page, addr, end, pages + *nr);
2562

J
John Hubbard 已提交
2563
	head = try_grab_compound_head(pgd_page(orig), refs, flags);
2564
	if (!head)
2565 2566 2567
		return 0;

	if (unlikely(pgd_val(orig) != pgd_val(*pgdp))) {
2568
		put_compound_head(head, refs, flags);
2569 2570 2571
		return 0;
	}

2572
	*nr += refs;
2573
	SetPageReferenced(head);
2574 2575 2576
	return 1;
}

2577
static int gup_pmd_range(pud_t pud, unsigned long addr, unsigned long end,
2578
		unsigned int flags, struct page **pages, int *nr)
2579 2580 2581 2582 2583 2584
{
	unsigned long next;
	pmd_t *pmdp;

	pmdp = pmd_offset(&pud, addr);
	do {
2585
		pmd_t pmd = READ_ONCE(*pmdp);
2586 2587

		next = pmd_addr_end(addr, end);
2588
		if (!pmd_present(pmd))
2589 2590
			return 0;

Y
Yu Zhao 已提交
2591 2592
		if (unlikely(pmd_trans_huge(pmd) || pmd_huge(pmd) ||
			     pmd_devmap(pmd))) {
2593 2594 2595 2596 2597
			/*
			 * NUMA hinting faults need to be handled in the GUP
			 * slowpath for accounting purposes and so that they
			 * can be serialised against THP migration.
			 */
2598
			if (pmd_protnone(pmd))
2599 2600
				return 0;

2601
			if (!gup_huge_pmd(pmd, pmdp, addr, next, flags,
2602 2603 2604
				pages, nr))
				return 0;

2605 2606 2607 2608 2609 2610
		} 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,
2611
					 PMD_SHIFT, next, flags, pages, nr))
2612
				return 0;
2613
		} else if (!gup_pte_range(pmd, addr, next, flags, pages, nr))
2614
			return 0;
2615 2616 2617 2618 2619
	} while (pmdp++, addr = next, addr != end);

	return 1;
}

2620
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
2621
			 unsigned int flags, struct page **pages, int *nr)
2622 2623 2624 2625
{
	unsigned long next;
	pud_t *pudp;

2626
	pudp = pud_offset(&p4d, addr);
2627
	do {
2628
		pud_t pud = READ_ONCE(*pudp);
2629 2630

		next = pud_addr_end(addr, end);
Q
Qiujun Huang 已提交
2631
		if (unlikely(!pud_present(pud)))
2632
			return 0;
2633
		if (unlikely(pud_huge(pud))) {
2634
			if (!gup_huge_pud(pud, pudp, addr, next, flags,
2635 2636 2637 2638
					  pages, nr))
				return 0;
		} else if (unlikely(is_hugepd(__hugepd(pud_val(pud))))) {
			if (!gup_huge_pd(__hugepd(pud_val(pud)), addr,
2639
					 PUD_SHIFT, next, flags, pages, nr))
2640
				return 0;
2641
		} else if (!gup_pmd_range(pud, addr, next, flags, pages, nr))
2642 2643 2644 2645 2646 2647
			return 0;
	} while (pudp++, addr = next, addr != end);

	return 1;
}

2648
static int gup_p4d_range(pgd_t pgd, unsigned long addr, unsigned long end,
2649
			 unsigned int flags, struct page **pages, int *nr)
2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663
{
	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,
2664
					 P4D_SHIFT, next, flags, pages, nr))
2665
				return 0;
2666
		} else if (!gup_pud_range(p4d, addr, next, flags, pages, nr))
2667 2668 2669 2670 2671 2672
			return 0;
	} while (p4dp++, addr = next, addr != end);

	return 1;
}

2673
static void gup_pgd_range(unsigned long addr, unsigned long end,
2674
		unsigned int flags, struct page **pages, int *nr)
2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686
{
	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))) {
2687
			if (!gup_huge_pgd(pgd, pgdp, addr, next, flags,
2688 2689 2690 2691
					  pages, nr))
				return;
		} else if (unlikely(is_hugepd(__hugepd(pgd_val(pgd))))) {
			if (!gup_huge_pd(__hugepd(pgd_val(pgd)), addr,
2692
					 PGDIR_SHIFT, next, flags, pages, nr))
2693
				return;
2694
		} else if (!gup_p4d_range(pgd, addr, next, flags, pages, nr))
2695 2696 2697
			return;
	} while (pgdp++, addr = next, addr != end);
}
2698 2699 2700 2701 2702 2703
#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 */
2704 2705 2706

#ifndef gup_fast_permitted
/*
2707
 * Check if it's allowed to use get_user_pages_fast_only() for the range, or
2708 2709
 * we need to fall back to the slow version:
 */
2710
static bool gup_fast_permitted(unsigned long start, unsigned long end)
2711
{
2712
	return true;
2713 2714 2715
}
#endif

2716 2717 2718 2719 2720 2721 2722 2723 2724 2725
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) {
2726
		mmap_read_lock(current->mm);
2727 2728 2729
		ret = __gup_longterm_locked(current, current->mm,
					    start, nr_pages,
					    pages, NULL, gup_flags);
2730
		mmap_read_unlock(current->mm);
2731 2732 2733 2734 2735 2736 2737 2738
	} else {
		ret = get_user_pages_unlocked(start, nr_pages,
					      pages, gup_flags);
	}

	return ret;
}

2739 2740 2741
static int internal_get_user_pages_fast(unsigned long start, int nr_pages,
					unsigned int gup_flags,
					struct page **pages)
2742
{
2743
	unsigned long addr, len, end;
2744
	unsigned long flags;
2745
	int nr_pinned = 0, ret = 0;
2746

2747
	if (WARN_ON_ONCE(gup_flags & ~(FOLL_WRITE | FOLL_LONGTERM |
2748 2749
				       FOLL_FORCE | FOLL_PIN | FOLL_GET |
				       FOLL_FAST_ONLY)))
2750 2751
		return -EINVAL;

2752 2753 2754
	if (!(gup_flags & FOLL_FAST_ONLY))
		might_lock_read(&current->mm->mmap_sem);

2755
	start = untagged_addr(start) & PAGE_MASK;
2756 2757 2758 2759
	addr = start;
	len = (unsigned long) nr_pages << PAGE_SHIFT;
	end = start + len;

2760
	if (end <= start)
2761
		return 0;
2762
	if (unlikely(!access_ok((void __user *)start, len)))
2763
		return -EFAULT;
2764

2765 2766 2767 2768 2769 2770
	/*
	 * The FAST_GUP case requires FOLL_WRITE even for pure reads,
	 * because get_user_pages() may need to cause an early COW in
	 * order to avoid confusing the normal COW routines. So only
	 * targets that are already writable are safe to do by just
	 * looking at the page tables.
2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787
	 *
	 * NOTE! With FOLL_FAST_ONLY we allow read-only gup_fast() here,
	 * because there is no slow path to fall back on. But you'd
	 * better be careful about possible COW pages - you'll get _a_
	 * COW page, but not necessarily the one you intended to get
	 * depending on what COW event happens after this. COW may break
	 * the page copy in a random direction.
	 *
	 * Disable interrupts. The nested form is used, in order to allow
	 * full, general purpose use of this routine.
	 *
	 * With interrupts disabled, we block page table pages from being
	 * freed from under us. See struct mmu_table_batch comments in
	 * include/asm-generic/tlb.h for more details.
	 *
	 * We do not adopt an rcu_read_lock(.) here as we also want to
	 * block IPIs that come from THPs splitting.
2788
	 */
2789 2790 2791 2792 2793 2794 2795 2796
	if (IS_ENABLED(CONFIG_HAVE_FAST_GUP) && gup_fast_permitted(start, end)) {
		unsigned long fast_flags = gup_flags;
		if (!(gup_flags & FOLL_FAST_ONLY))
			fast_flags |= FOLL_WRITE;

		local_irq_save(flags);
		gup_pgd_range(addr, end, fast_flags, pages, &nr_pinned);
		local_irq_restore(flags);
2797
		ret = nr_pinned;
2798
	}
2799

2800
	if (nr_pinned < nr_pages && !(gup_flags & FOLL_FAST_ONLY)) {
2801
		/* Try to get the remaining pages with get_user_pages */
2802 2803
		start += nr_pinned << PAGE_SHIFT;
		pages += nr_pinned;
2804

2805
		ret = __gup_longterm_unlocked(start, nr_pages - nr_pinned,
2806
					      gup_flags, pages);
2807 2808

		/* Have to be a bit careful with return values */
2809
		if (nr_pinned > 0) {
2810
			if (ret < 0)
2811
				ret = nr_pinned;
2812
			else
2813
				ret += nr_pinned;
2814 2815 2816 2817 2818
		}
	}

	return ret;
}
2819 2820 2821 2822 2823 2824 2825 2826
/**
 * get_user_pages_fast_only() - pin user pages in memory
 * @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.
 *
2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838
 * Like get_user_pages_fast() except it's IRQ-safe in that it won't fall back to
 * 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.
 *
 * If the architecture does not support this function, simply return with no
 * pages pinned.
 *
 * Careful, careful! COW breaking can go either way, so a non-write
 * access can get ambiguous page results. If you call this function without
 * 'write' set, you'd better be sure that you're ok with that ambiguity.
 */
2839 2840
int get_user_pages_fast_only(unsigned long start, int nr_pages,
			     unsigned int gup_flags, struct page **pages)
2841
{
2842
	int nr_pinned;
2843 2844 2845
	/*
	 * Internally (within mm/gup.c), gup fast variants must set FOLL_GET,
	 * because gup fast is always a "pin with a +1 page refcount" request.
2846 2847 2848
	 *
	 * FOLL_FAST_ONLY is required in order to match the API description of
	 * this routine: no fall back to regular ("slow") GUP.
2849
	 */
2850
	gup_flags |= FOLL_GET | FOLL_FAST_ONLY;
2851

2852 2853
	nr_pinned = internal_get_user_pages_fast(start, nr_pages, gup_flags,
						 pages);
2854 2855

	/*
2856 2857 2858 2859
	 * As specified in the API description above, this routine is not
	 * allowed to return negative values. However, the common core
	 * routine internal_get_user_pages_fast() *can* return -errno.
	 * Therefore, correct for that here:
2860
	 */
2861 2862
	if (nr_pinned < 0)
		nr_pinned = 0;
2863 2864 2865

	return nr_pinned;
}
2866
EXPORT_SYMBOL_GPL(get_user_pages_fast_only);
2867

2868 2869
/**
 * get_user_pages_fast() - pin user pages in memory
J
John Hubbard 已提交
2870 2871 2872 2873 2874
 * @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.
2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893
 *
 * 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.
 */
int get_user_pages_fast(unsigned long start, int nr_pages,
			unsigned int gup_flags, struct page **pages)
{
	/*
	 * FOLL_PIN must only be set internally by the pin_user_pages*() APIs,
	 * never directly by the caller, so enforce that:
	 */
	if (WARN_ON_ONCE(gup_flags & FOLL_PIN))
		return -EINVAL;

2894 2895 2896 2897 2898 2899 2900
	/*
	 * The caller may or may not have explicitly set FOLL_GET; either way is
	 * OK. However, internally (within mm/gup.c), gup fast variants must set
	 * FOLL_GET, because gup fast is always a "pin with a +1 page refcount"
	 * request.
	 */
	gup_flags |= FOLL_GET;
2901 2902
	return internal_get_user_pages_fast(start, nr_pages, gup_flags, pages);
}
2903
EXPORT_SYMBOL_GPL(get_user_pages_fast);
2904 2905 2906 2907

/**
 * pin_user_pages_fast() - pin user pages in memory without taking locks
 *
J
John Hubbard 已提交
2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918
 * @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.
 *
 * Nearly the same as get_user_pages_fast(), except that FOLL_PIN is set. See
 * get_user_pages_fast() for documentation on the function arguments, because
 * the arguments here are identical.
 *
 * FOLL_PIN means that the pages must be released via unpin_user_page(). Please
2919
 * see Documentation/core-api/pin_user_pages.rst for further details.
2920 2921 2922 2923
 */
int pin_user_pages_fast(unsigned long start, int nr_pages,
			unsigned int gup_flags, struct page **pages)
{
J
John Hubbard 已提交
2924 2925 2926 2927 2928 2929
	/* FOLL_GET and FOLL_PIN are mutually exclusive. */
	if (WARN_ON_ONCE(gup_flags & FOLL_GET))
		return -EINVAL;

	gup_flags |= FOLL_PIN;
	return internal_get_user_pages_fast(start, nr_pages, gup_flags, pages);
2930 2931 2932
}
EXPORT_SYMBOL_GPL(pin_user_pages_fast);

2933
/*
2934 2935
 * This is the FOLL_PIN equivalent of get_user_pages_fast_only(). Behavior
 * is the same, except that this one sets FOLL_PIN instead of FOLL_GET.
2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968
 *
 * The API rules are the same, too: no negative values may be returned.
 */
int pin_user_pages_fast_only(unsigned long start, int nr_pages,
			     unsigned int gup_flags, struct page **pages)
{
	int nr_pinned;

	/*
	 * FOLL_GET and FOLL_PIN are mutually exclusive. Note that the API
	 * rules require returning 0, rather than -errno:
	 */
	if (WARN_ON_ONCE(gup_flags & FOLL_GET))
		return 0;
	/*
	 * FOLL_FAST_ONLY is required in order to match the API description of
	 * this routine: no fall back to regular ("slow") GUP.
	 */
	gup_flags |= (FOLL_PIN | FOLL_FAST_ONLY);
	nr_pinned = internal_get_user_pages_fast(start, nr_pages, gup_flags,
						 pages);
	/*
	 * This routine is not allowed to return negative values. However,
	 * internal_get_user_pages_fast() *can* return -errno. Therefore,
	 * correct for that here:
	 */
	if (nr_pinned < 0)
		nr_pinned = 0;

	return nr_pinned;
}
EXPORT_SYMBOL_GPL(pin_user_pages_fast_only);

2969 2970 2971
/**
 * pin_user_pages_remote() - pin pages of a remote process (task != current)
 *
J
John Hubbard 已提交
2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991
 * @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
 * @gup_flags:	flags modifying lookup 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.
 * @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.
 *
 * Nearly the same as get_user_pages_remote(), except that FOLL_PIN is set. See
 * get_user_pages_remote() for documentation on the function arguments, because
 * the arguments here are identical.
 *
 * FOLL_PIN means that the pages must be released via unpin_user_page(). Please
2992
 * see Documentation/core-api/pin_user_pages.rst for details.
2993 2994 2995 2996 2997 2998
 */
long pin_user_pages_remote(struct task_struct *tsk, struct mm_struct *mm,
			   unsigned long start, unsigned long nr_pages,
			   unsigned int gup_flags, struct page **pages,
			   struct vm_area_struct **vmas, int *locked)
{
J
John Hubbard 已提交
2999 3000 3001 3002 3003 3004 3005
	/* FOLL_GET and FOLL_PIN are mutually exclusive. */
	if (WARN_ON_ONCE(gup_flags & FOLL_GET))
		return -EINVAL;

	gup_flags |= FOLL_PIN;
	return __get_user_pages_remote(tsk, mm, start, nr_pages, gup_flags,
				       pages, vmas, locked);
3006 3007 3008 3009 3010 3011
}
EXPORT_SYMBOL(pin_user_pages_remote);

/**
 * pin_user_pages() - pin user pages in memory for use by other devices
 *
J
John Hubbard 已提交
3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024
 * @start:	starting user address
 * @nr_pages:	number of pages from start to pin
 * @gup_flags:	flags modifying lookup 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.
 *
 * Nearly the same as get_user_pages(), except that FOLL_TOUCH is not set, and
 * FOLL_PIN is set.
 *
 * FOLL_PIN means that the pages must be released via unpin_user_page(). Please
3025
 * see Documentation/core-api/pin_user_pages.rst for details.
3026 3027 3028 3029 3030
 */
long pin_user_pages(unsigned long start, unsigned long nr_pages,
		    unsigned int gup_flags, struct page **pages,
		    struct vm_area_struct **vmas)
{
J
John Hubbard 已提交
3031 3032 3033 3034 3035 3036 3037
	/* FOLL_GET and FOLL_PIN are mutually exclusive. */
	if (WARN_ON_ONCE(gup_flags & FOLL_GET))
		return -EINVAL;

	gup_flags |= FOLL_PIN;
	return __gup_longterm_locked(current, current->mm, start, nr_pages,
				     pages, vmas, gup_flags);
3038 3039
}
EXPORT_SYMBOL(pin_user_pages);
3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056

/*
 * pin_user_pages_unlocked() is the FOLL_PIN variant of
 * get_user_pages_unlocked(). Behavior is the same, except that this one sets
 * FOLL_PIN and rejects FOLL_GET.
 */
long pin_user_pages_unlocked(unsigned long start, unsigned long nr_pages,
			     struct page **pages, unsigned int gup_flags)
{
	/* FOLL_GET and FOLL_PIN are mutually exclusive. */
	if (WARN_ON_ONCE(gup_flags & FOLL_GET))
		return -EINVAL;

	gup_flags |= FOLL_PIN;
	return get_user_pages_unlocked(start, nr_pages, pages, gup_flags);
}
EXPORT_SYMBOL(pin_user_pages_unlocked);
3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085

/*
 * pin_user_pages_locked() is the FOLL_PIN variant of get_user_pages_locked().
 * Behavior is the same, except that this one sets FOLL_PIN and rejects
 * FOLL_GET.
 */
long pin_user_pages_locked(unsigned long start, unsigned long nr_pages,
			   unsigned int gup_flags, struct page **pages,
			   int *locked)
{
	/*
	 * 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;

	/* FOLL_GET and FOLL_PIN are mutually exclusive. */
	if (WARN_ON_ONCE(gup_flags & FOLL_GET))
		return -EINVAL;

	gup_flags |= FOLL_PIN;
	return __get_user_pages_locked(current, current->mm, start, nr_pages,
				       pages, NULL, locked,
				       gup_flags | FOLL_TOUCH);
}
EXPORT_SYMBOL(pin_user_pages_locked);