gup.c 80.9 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/pgtable.h>
23
#include <asm/tlbflush.h>
24

25 26
#include "internal.h"

27 28 29 30 31
struct follow_page_context {
	struct dev_pagemap *pgmap;
	unsigned int page_mask;
};

32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
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));
}

48 49 50 51 52 53 54 55 56 57 58 59 60 61 62
/*
 * 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
63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88
/*
 * 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) {
89 90
		int orig_refs = refs;

91 92 93 94 95 96 97 98
		/*
		 * 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;

99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116
		/*
		 * 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);

117 118 119
		mod_node_page_state(page_pgdat(page), NR_FOLL_PIN_ACQUIRED,
				    orig_refs);

120
		return page;
J
John Hubbard 已提交
121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154
	}

	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) {
155 156
		int refs = 1;

J
John Hubbard 已提交
157 158 159 160 161
		page = compound_head(page);

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

162 163 164 165 166 167 168 169 170 171 172 173
		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);
174 175

		mod_node_page_state(page_pgdat(page), NR_FOLL_PIN_ACQUIRED, 1);
J
John Hubbard 已提交
176 177 178 179 180 181 182 183
	}

	return true;
}

#ifdef CONFIG_DEV_PAGEMAP_OPS
static bool __unpin_devmap_managed_user_page(struct page *page)
{
184
	int count, refs = 1;
J
John Hubbard 已提交
185 186 187 188

	if (!page_is_devmap_managed(page))
		return false;

189 190 191 192 193 194
	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
195

196
	mod_node_page_state(page_pgdat(page), NR_FOLL_PIN_RELEASED, 1);
J
John Hubbard 已提交
197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226
	/*
	 * 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)
{
227 228
	int refs = 1;

J
John Hubbard 已提交
229 230 231 232 233 234 235 236 237 238 239
	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;

240 241 242 243 244 245
	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
246
		__put_page(page);
247 248

	mod_node_page_state(page_pgdat(page), NR_FOLL_PIN_RELEASED, 1);
J
John Hubbard 已提交
249 250 251
}
EXPORT_SYMBOL(unpin_user_page);

252
/**
253
 * unpin_user_pages_dirty_lock() - release and optionally dirty gup-pinned pages
254
 * @pages:  array of pages to be maybe marked dirty, and definitely released.
255
 * @npages: number of pages in the @pages array.
256
 * @make_dirty: whether to mark the pages dirty
257 258 259 260 261
 *
 * "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
262
 * compound page) dirty, if @make_dirty is true, and if the page was previously
263 264
 * listed as clean. In any case, releases all pages using unpin_user_page(),
 * possibly via unpin_user_pages(), for the non-dirty case.
265
 *
266
 * Please see the unpin_user_page() documentation for details.
267
 *
268 269 270
 * 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:
271
 * set_page_dirty_lock(), unpin_user_page().
272 273
 *
 */
274 275
void unpin_user_pages_dirty_lock(struct page **pages, unsigned long npages,
				 bool make_dirty)
276
{
277
	unsigned long index;
278

279 280 281 282 283 284 285
	/*
	 * 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) {
286
		unpin_user_pages(pages, npages);
287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313
		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);
314
		unpin_user_page(page);
315
	}
316
}
317
EXPORT_SYMBOL(unpin_user_pages_dirty_lock);
318 319

/**
320
 * unpin_user_pages() - release an array of gup-pinned pages.
321 322 323
 * @pages:  array of pages to be marked dirty and released.
 * @npages: number of pages in the @pages array.
 *
324
 * For each page in the @pages array, release the page using unpin_user_page().
325
 *
326
 * Please see the unpin_user_page() documentation for details.
327
 */
328
void unpin_user_pages(struct page **pages, unsigned long npages)
329 330 331 332 333 334 335 336 337
{
	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++)
338
		unpin_user_page(pages[index]);
339
}
340
EXPORT_SYMBOL(unpin_user_pages);
341

342
#ifdef CONFIG_MMU
343 344
static struct page *no_page_table(struct vm_area_struct *vma,
		unsigned int flags)
345
{
346 347 348 349 350 351 352 353 354 355 356 357
	/*
	 * When core dumping an enormous anonymous area that nobody
	 * has touched so far, we don't want to allocate unnecessary pages or
	 * page tables.  Return error instead of NULL to skip handle_mm_fault,
	 * then get_dump_page() will return NULL to leave a hole in the dump.
	 * But we can only make this optimization where a hole would surely
	 * be zero-filled if handle_mm_fault() actually did handle it.
	 */
	if ((flags & FOLL_DUMP) && (!vma->vm_ops || !vma->vm_ops->fault))
		return ERR_PTR(-EFAULT);
	return NULL;
}
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;
}

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

393
static struct page *follow_page_pte(struct vm_area_struct *vma,
394 395
		unsigned long address, pmd_t *pmd, unsigned int flags,
		struct dev_pagemap **pgmap)
396 397 398 399 400
{
	struct mm_struct *mm = vma->vm_mm;
	struct page *page;
	spinlock_t *ptl;
	pte_t *ptep, pte;
401
	int ret;
402

403 404 405 406
	/* 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);
407
retry:
408
	if (unlikely(pmd_bad(*pmd)))
409
		return no_page_table(vma, flags);
410 411 412 413 414 415 416 417 418 419 420 421

	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;
422
		if (pte_none(pte))
423 424 425 426 427 428
			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);
429
		goto retry;
430
	}
431
	if ((flags & FOLL_NUMA) && pte_protnone(pte))
432
		goto no_page;
433
	if ((flags & FOLL_WRITE) && !can_follow_write_pte(pte, flags)) {
434 435 436
		pte_unmap_unlock(ptep, ptl);
		return NULL;
	}
437 438

	page = vm_normal_page(vma, address, pte);
J
John Hubbard 已提交
439
	if (!page && pte_devmap(pte) && (flags & (FOLL_GET | FOLL_PIN))) {
440
		/*
J
John Hubbard 已提交
441 442 443
		 * Only return device mapping pages in the FOLL_GET or FOLL_PIN
		 * case since they are only valid while holding the pgmap
		 * reference.
444
		 */
445 446
		*pgmap = get_dev_pagemap(pte_pfn(pte), *pgmap);
		if (*pgmap)
447 448 449 450
			page = pte_page(pte);
		else
			goto no_page;
	} else if (unlikely(!page)) {
451 452 453 454 455 456 457 458 459 460 461 462 463
		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;
		}
464 465
	}

466 467 468 469 470 471 472 473 474 475 476 477
	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
478 479 480 481
	/* 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;
482
	}
483 484 485 486 487 488 489 490 491 492 493 494 495
	/*
	 * 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;
		}
	}
496 497 498 499 500 501 502 503 504 505 506
	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
507
	if ((flags & FOLL_MLOCK) && (vma->vm_flags & VM_LOCKED)) {
508 509 510 511
		/* Do not mlock pte-mapped THP */
		if (PageTransCompound(page))
			goto out;

512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532
		/*
		 * 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);
		}
	}
533
out:
534 535 536 537 538
	pte_unmap_unlock(ptep, ptl);
	return page;
no_page:
	pte_unmap_unlock(ptep, ptl);
	if (!pte_none(pte))
539 540 541 542
		return NULL;
	return no_page_table(vma, flags);
}

543 544
static struct page *follow_pmd_mask(struct vm_area_struct *vma,
				    unsigned long address, pud_t *pudp,
545 546
				    unsigned int flags,
				    struct follow_page_context *ctx)
547
{
548
	pmd_t *pmd, pmdval;
549 550 551 552
	spinlock_t *ptl;
	struct page *page;
	struct mm_struct *mm = vma->vm_mm;

553
	pmd = pmd_offset(pudp, address);
554 555 556 557 558 559
	/*
	 * 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))
560
		return no_page_table(vma, flags);
561
	if (pmd_huge(pmdval) && is_vm_hugetlb_page(vma)) {
562 563 564 565
		page = follow_huge_pmd(mm, address, pmd, flags);
		if (page)
			return page;
		return no_page_table(vma, flags);
566
	}
567
	if (is_hugepd(__hugepd(pmd_val(pmdval)))) {
568
		page = follow_huge_pd(vma, address,
569
				      __hugepd(pmd_val(pmdval)), flags,
570 571 572 573 574
				      PMD_SHIFT);
		if (page)
			return page;
		return no_page_table(vma, flags);
	}
575
retry:
576
	if (!pmd_present(pmdval)) {
577 578 579
		if (likely(!(flags & FOLL_MIGRATION)))
			return no_page_table(vma, flags);
		VM_BUG_ON(thp_migration_supported() &&
580 581
				  !is_pmd_migration_entry(pmdval));
		if (is_pmd_migration_entry(pmdval))
582
			pmd_migration_entry_wait(mm, pmd);
583 584 585 586 587 588 589
		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);
590 591
		goto retry;
	}
592
	if (pmd_devmap(pmdval)) {
593
		ptl = pmd_lock(mm, pmd);
594
		page = follow_devmap_pmd(vma, address, pmd, flags, &ctx->pgmap);
595 596 597 598
		spin_unlock(ptl);
		if (page)
			return page;
	}
599
	if (likely(!pmd_trans_huge(pmdval)))
600
		return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
601

602
	if ((flags & FOLL_NUMA) && pmd_protnone(pmdval))
603 604
		return no_page_table(vma, flags);

605
retry_locked:
606
	ptl = pmd_lock(mm, pmd);
607 608 609 610
	if (unlikely(pmd_none(*pmd))) {
		spin_unlock(ptl);
		return no_page_table(vma, flags);
	}
611 612 613 614 615 616 617
	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;
	}
618 619
	if (unlikely(!pmd_trans_huge(*pmd))) {
		spin_unlock(ptl);
620
		return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
621
	}
S
Song Liu 已提交
622
	if (flags & (FOLL_SPLIT | FOLL_SPLIT_PMD)) {
623 624 625 626 627
		int ret;
		page = pmd_page(*pmd);
		if (is_huge_zero_page(page)) {
			spin_unlock(ptl);
			ret = 0;
628
			split_huge_pmd(vma, pmd, address);
629 630
			if (pmd_trans_unstable(pmd))
				ret = -EBUSY;
S
Song Liu 已提交
631
		} else if (flags & FOLL_SPLIT) {
632 633 634 635
			if (unlikely(!try_get_page(page))) {
				spin_unlock(ptl);
				return ERR_PTR(-ENOMEM);
			}
636
			spin_unlock(ptl);
637 638 639 640
			lock_page(page);
			ret = split_huge_page(page);
			unlock_page(page);
			put_page(page);
641 642
			if (pmd_none(*pmd))
				return no_page_table(vma, flags);
S
Song Liu 已提交
643 644 645 646
		} else {  /* flags & FOLL_SPLIT_PMD */
			spin_unlock(ptl);
			split_huge_pmd(vma, pmd, address);
			ret = pte_alloc(mm, pmd) ? -ENOMEM : 0;
647 648 649
		}

		return ret ? ERR_PTR(ret) :
650
			follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
651
	}
652 653
	page = follow_trans_huge_pmd(vma, address, pmd, flags);
	spin_unlock(ptl);
654
	ctx->page_mask = HPAGE_PMD_NR - 1;
655
	return page;
656 657
}

658 659
static struct page *follow_pud_mask(struct vm_area_struct *vma,
				    unsigned long address, p4d_t *p4dp,
660 661
				    unsigned int flags,
				    struct follow_page_context *ctx)
662 663 664 665 666 667 668 669 670
{
	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);
671
	if (pud_huge(*pud) && is_vm_hugetlb_page(vma)) {
672 673 674 675 676
		page = follow_huge_pud(mm, address, pud, flags);
		if (page)
			return page;
		return no_page_table(vma, flags);
	}
677 678 679 680 681 682 683 684
	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);
	}
685 686
	if (pud_devmap(*pud)) {
		ptl = pud_lock(mm, pud);
687
		page = follow_devmap_pud(vma, address, pud, flags, &ctx->pgmap);
688 689 690 691 692 693 694
		spin_unlock(ptl);
		if (page)
			return page;
	}
	if (unlikely(pud_bad(*pud)))
		return no_page_table(vma, flags);

695
	return follow_pmd_mask(vma, address, pud, flags, ctx);
696 697 698 699
}

static struct page *follow_p4d_mask(struct vm_area_struct *vma,
				    unsigned long address, pgd_t *pgdp,
700 701
				    unsigned int flags,
				    struct follow_page_context *ctx)
702 703
{
	p4d_t *p4d;
704
	struct page *page;
705 706 707 708 709 710 711 712

	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);

713 714 715 716 717 718 719 720
	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);
	}
721
	return follow_pud_mask(vma, address, p4d, flags, ctx);
722 723 724 725 726 727 728
}

/**
 * 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
729 730
 * @ctx: contains dev_pagemap for %ZONE_DEVICE memory pinning and a
 *       pointer to output page_mask
731 732 733
 *
 * @flags can have FOLL_ flags set, defined in <linux/mm.h>
 *
734 735 736 737 738 739
 * 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
740 741 742
 * an error pointer if there is a mapping to something not represented
 * by a page descriptor (see also vm_normal_page()).
 */
743
static struct page *follow_page_mask(struct vm_area_struct *vma,
744
			      unsigned long address, unsigned int flags,
745
			      struct follow_page_context *ctx)
746 747 748 749 750
{
	pgd_t *pgd;
	struct page *page;
	struct mm_struct *mm = vma->vm_mm;

751
	ctx->page_mask = 0;
752 753 754 755

	/* make this handle hugepd */
	page = follow_huge_addr(mm, address, flags & FOLL_WRITE);
	if (!IS_ERR(page)) {
J
John Hubbard 已提交
756
		WARN_ON_ONCE(flags & (FOLL_GET | FOLL_PIN));
757 758 759 760 761 762 763 764
		return page;
	}

	pgd = pgd_offset(mm, address);

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

765 766 767 768 769 770
	if (pgd_huge(*pgd)) {
		page = follow_huge_pgd(mm, address, pgd, flags);
		if (page)
			return page;
		return no_page_table(vma, flags);
	}
771 772 773 774 775 776 777 778
	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);
	}
779

780 781 782 783 784 785 786 787 788 789 790 791 792
	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;
793 794
}

795 796 797 798 799
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;
800
	p4d_t *p4d;
801 802 803 804 805 806 807 808 809 810 811 812
	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);
813 814
	if (pgd_none(*pgd))
		return -EFAULT;
815
	p4d = p4d_offset(pgd, address);
816 817
	if (p4d_none(*p4d))
		return -EFAULT;
818
	pud = pud_offset(p4d, address);
819 820
	if (pud_none(*pud))
		return -EFAULT;
821
	pmd = pmd_offset(pud, address);
822
	if (!pmd_present(*pmd))
823 824 825 826 827 828 829 830 831 832 833 834 835 836
		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);
	}
837 838 839 840
	if (unlikely(!try_get_page(*page))) {
		ret = -ENOMEM;
		goto unmap;
	}
841 842 843 844 845 846 847
out:
	ret = 0;
unmap:
	pte_unmap(pte);
	return ret;
}

848
/*
849 850 851
 * 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.
852
 */
853
static int faultin_page(struct task_struct *tsk, struct vm_area_struct *vma,
854
		unsigned long address, unsigned int *flags, int *locked)
855 856
{
	unsigned int fault_flags = 0;
857
	vm_fault_t ret;
858

E
Eric B Munson 已提交
859 860 861
	/* mlock all present pages, but do not fault in new pages */
	if ((*flags & (FOLL_POPULATE | FOLL_MLOCK)) == FOLL_MLOCK)
		return -ENOENT;
862 863
	if (*flags & FOLL_WRITE)
		fault_flags |= FAULT_FLAG_WRITE;
864 865
	if (*flags & FOLL_REMOTE)
		fault_flags |= FAULT_FLAG_REMOTE;
866
	if (locked)
867 868 869
		fault_flags |= FAULT_FLAG_ALLOW_RETRY;
	if (*flags & FOLL_NOWAIT)
		fault_flags |= FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_RETRY_NOWAIT;
870 871 872 873
	if (*flags & FOLL_TRIED) {
		VM_WARN_ON_ONCE(fault_flags & FAULT_FLAG_ALLOW_RETRY);
		fault_flags |= FAULT_FLAG_TRIED;
	}
874

875
	ret = handle_mm_fault(vma, address, fault_flags);
876
	if (ret & VM_FAULT_ERROR) {
877 878 879 880
		int err = vm_fault_to_errno(ret, *flags);

		if (err)
			return err;
881 882 883 884 885 886 887 888 889 890 891
		BUG();
	}

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

	if (ret & VM_FAULT_RETRY) {
892 893
		if (locked && !(fault_flags & FAULT_FLAG_RETRY_NOWAIT))
			*locked = 0;
894 895 896 897 898 899 900 901 902 903 904 905 906
		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))
907
		*flags |= FOLL_COW;
908 909 910
	return 0;
}

911 912 913
static int check_vma_flags(struct vm_area_struct *vma, unsigned long gup_flags)
{
	vm_flags_t vm_flags = vma->vm_flags;
914 915
	int write = (gup_flags & FOLL_WRITE);
	int foreign = (gup_flags & FOLL_REMOTE);
916 917 918 919

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

920 921 922
	if (gup_flags & FOLL_ANON && !vma_is_anonymous(vma))
		return -EFAULT;

923
	if (write) {
924 925 926 927 928 929 930 931 932 933 934 935
		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.
			 */
936
			if (!is_cow_mapping(vm_flags))
937 938 939 940 941 942 943 944 945 946 947 948
				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;
	}
949 950 951 952 953
	/*
	 * gups are always data accesses, not instruction
	 * fetches, so execute=false here
	 */
	if (!arch_vma_access_permitted(vma, write, false, foreign))
954
		return -EFAULT;
955 956 957
	return 0;
}

958 959 960 961 962 963 964 965 966 967 968 969
/**
 * __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.
970
 * @locked:     whether we're still with the mmap_sem held
971
 *
972 973 974 975 976 977 978 979 980 981 982
 * 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.
983
 *
984
 * Must be called with mmap_sem held.  It may be released.  See below.
985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004
 *
 * __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.
 *
1005 1006 1007
 * 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.
1008
 *
1009
 * A caller using such a combination of @locked and @gup_flags
1010 1011 1012
 * 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.
1013 1014 1015 1016 1017
 *
 * 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
1018
static long __get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
1019 1020
		unsigned long start, unsigned long nr_pages,
		unsigned int gup_flags, struct page **pages,
1021
		struct vm_area_struct **vmas, int *locked)
1022
{
1023
	long ret = 0, i = 0;
1024
	struct vm_area_struct *vma = NULL;
1025
	struct follow_page_context ctx = { NULL };
1026 1027 1028 1029

	if (!nr_pages)
		return 0;

1030 1031
	start = untagged_addr(start);

1032
	VM_BUG_ON(!!pages != !!(gup_flags & (FOLL_GET | FOLL_PIN)));
1033 1034 1035 1036 1037 1038 1039 1040 1041 1042

	/*
	 * 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 {
1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054
		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)
1055
					goto out;
1056
				ctx.page_mask = 0;
1057 1058
				goto next_page;
			}
1059

1060 1061 1062 1063
			if (!vma || check_vma_flags(vma, gup_flags)) {
				ret = -EFAULT;
				goto out;
			}
1064 1065 1066
			if (is_vm_hugetlb_page(vma)) {
				i = follow_hugetlb_page(mm, vma, pages, vmas,
						&start, &nr_pages, i,
1067
						gup_flags, locked);
1068
				continue;
1069
			}
1070 1071 1072 1073 1074 1075
		}
retry:
		/*
		 * If we have a pending SIGKILL, don't keep faulting pages and
		 * potentially allocating memory.
		 */
1076
		if (fatal_signal_pending(current)) {
1077 1078 1079
			ret = -ERESTARTSYS;
			goto out;
		}
1080
		cond_resched();
1081 1082

		page = follow_page_mask(vma, start, foll_flags, &ctx);
1083 1084
		if (!page) {
			ret = faultin_page(tsk, vma, start, &foll_flags,
1085
					   locked);
1086 1087 1088
			switch (ret) {
			case 0:
				goto retry;
1089 1090 1091
			case -EBUSY:
				ret = 0;
				/* FALLTHRU */
1092 1093 1094
			case -EFAULT:
			case -ENOMEM:
			case -EHWPOISON:
1095
				goto out;
1096 1097
			case -ENOENT:
				goto next_page;
1098
			}
1099
			BUG();
1100 1101 1102 1103 1104 1105 1106
		} else if (PTR_ERR(page) == -EEXIST) {
			/*
			 * Proper page table entry exists, but no corresponding
			 * struct page.
			 */
			goto next_page;
		} else if (IS_ERR(page)) {
1107 1108
			ret = PTR_ERR(page);
			goto out;
1109
		}
1110 1111 1112 1113
		if (pages) {
			pages[i] = page;
			flush_anon_page(vma, page, start);
			flush_dcache_page(page);
1114
			ctx.page_mask = 0;
1115 1116
		}
next_page:
1117 1118
		if (vmas) {
			vmas[i] = vma;
1119
			ctx.page_mask = 0;
1120
		}
1121
		page_increm = 1 + (~(start >> PAGE_SHIFT) & ctx.page_mask);
1122 1123 1124 1125 1126
		if (page_increm > nr_pages)
			page_increm = nr_pages;
		i += page_increm;
		start += page_increm * PAGE_SIZE;
		nr_pages -= page_increm;
1127
	} while (nr_pages);
1128 1129 1130 1131
out:
	if (ctx.pgmap)
		put_dev_pagemap(ctx.pgmap);
	return i ? i : ret;
1132 1133
}

1134 1135
static bool vma_permits_fault(struct vm_area_struct *vma,
			      unsigned int fault_flags)
1136
{
1137 1138
	bool write   = !!(fault_flags & FAULT_FLAG_WRITE);
	bool foreign = !!(fault_flags & FAULT_FLAG_REMOTE);
1139
	vm_flags_t vm_flags = write ? VM_WRITE : VM_READ;
1140 1141 1142 1143

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

1144 1145
	/*
	 * The architecture might have a hardware protection
1146
	 * mechanism other than read/write that can deny access.
1147 1148 1149
	 *
	 * gup always represents data access, not instruction
	 * fetches, so execute=false here:
1150
	 */
1151
	if (!arch_vma_access_permitted(vma, write, false, foreign))
1152 1153
		return false;

1154 1155 1156
	return true;
}

1157 1158 1159 1160 1161 1162 1163
/*
 * 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()
1164 1165
 * @unlocked:	did we unlock the mmap_sem while retrying, maybe NULL if caller
 *		does not allow retry
1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176
 *
 * 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
1177
 * get_user_pages() only guarantees to update these in the struct page.
1178 1179 1180 1181 1182 1183
 *
 * 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.
 *
1184 1185
 * 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().
1186 1187
 */
int fixup_user_fault(struct task_struct *tsk, struct mm_struct *mm,
1188 1189
		     unsigned long address, unsigned int fault_flags,
		     bool *unlocked)
1190 1191
{
	struct vm_area_struct *vma;
1192
	vm_fault_t ret, major = 0;
1193

1194 1195
	address = untagged_addr(address);

1196 1197
	if (unlocked)
		fault_flags |= FAULT_FLAG_ALLOW_RETRY;
1198

1199
retry:
1200 1201 1202 1203
	vma = find_extend_vma(mm, address);
	if (!vma || address < vma->vm_start)
		return -EFAULT;

1204
	if (!vma_permits_fault(vma, fault_flags))
1205 1206
		return -EFAULT;

1207
	ret = handle_mm_fault(vma, address, fault_flags);
1208
	major |= ret & VM_FAULT_MAJOR;
1209
	if (ret & VM_FAULT_ERROR) {
1210 1211 1212 1213
		int err = vm_fault_to_errno(ret, 0);

		if (err)
			return err;
1214 1215
		BUG();
	}
1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226

	if (ret & VM_FAULT_RETRY) {
		down_read(&mm->mmap_sem);
		if (!(fault_flags & FAULT_FLAG_TRIED)) {
			*unlocked = true;
			fault_flags &= ~FAULT_FLAG_ALLOW_RETRY;
			fault_flags |= FAULT_FLAG_TRIED;
			goto retry;
		}
	}

1227
	if (tsk) {
1228
		if (major)
1229 1230 1231 1232 1233 1234
			tsk->maj_flt++;
		else
			tsk->min_flt++;
	}
	return 0;
}
1235
EXPORT_SYMBOL_GPL(fixup_user_fault);
1236

1237 1238 1239 1240 1241 1242
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,
1243
						int *locked,
1244
						unsigned int flags)
1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255
{
	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);
	}

1256 1257 1258 1259 1260 1261 1262 1263 1264 1265
	/*
	 * 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))
1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289
		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) {
1290 1291 1292 1293
			/*
			 * VM_FAULT_RETRY didn't trigger or it was a
			 * FOLL_NOWAIT.
			 */
1294 1295 1296 1297
			if (!pages_done)
				pages_done = ret;
			break;
		}
1298 1299 1300 1301 1302 1303
		/*
		 * VM_FAULT_RETRY triggered, so seek to the faulting offset.
		 * For the prefault case (!pages) we only update counts.
		 */
		if (likely(pages))
			pages += ret;
1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325
		start += ret << PAGE_SHIFT;

		/*
		 * Repeat on the address that fired VM_FAULT_RETRY
		 * without FAULT_FLAG_ALLOW_RETRY but with
		 * FAULT_FLAG_TRIED.
		 */
		*locked = 1;
		lock_dropped = true;
		down_read(&mm->mmap_sem);
		ret = __get_user_pages(tsk, mm, start, 1, flags | FOLL_TRIED,
				       pages, NULL, NULL);
		if (ret != 1) {
			BUG_ON(ret > 1);
			if (!pages_done)
				pages_done = ret;
			break;
		}
		nr_pages--;
		pages_done++;
		if (!nr_pages)
			break;
1326 1327
		if (likely(pages))
			pages++;
1328 1329
		start += PAGE_SIZE;
	}
1330
	if (lock_dropped && *locked) {
1331 1332 1333 1334 1335 1336 1337 1338 1339 1340
		/*
		 * We must let the caller know we temporarily dropped the lock
		 * and so the critical section protected by it was lost.
		 */
		up_read(&mm->mmap_sem);
		*locked = 0;
	}
	return pages_done;
}

1341 1342 1343 1344 1345
/**
 * populate_vma_page_range() -  populate a range of pages in the vma.
 * @vma:   target vma
 * @start: start address
 * @end:   end address
1346
 * @locked: whether the mmap_sem is still held
1347 1348 1349 1350 1351 1352 1353
 *
 * 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.
 *
1354
 * If @locked is NULL, it may be held for read or write and will
1355 1356
 * be unperturbed.
 *
1357 1358
 * 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.
1359 1360
 */
long populate_vma_page_range(struct vm_area_struct *vma,
1361
		unsigned long start, unsigned long end, int *locked)
1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395
{
	struct mm_struct *mm = vma->vm_mm;
	unsigned long nr_pages = (end - start) / PAGE_SIZE;
	int gup_flags;

	VM_BUG_ON(start & ~PAGE_MASK);
	VM_BUG_ON(end   & ~PAGE_MASK);
	VM_BUG_ON_VMA(start < vma->vm_start, vma);
	VM_BUG_ON_VMA(end   > vma->vm_end, vma);
	VM_BUG_ON_MM(!rwsem_is_locked(&mm->mmap_sem), mm);

	gup_flags = FOLL_TOUCH | FOLL_POPULATE | FOLL_MLOCK;
	if (vma->vm_flags & VM_LOCKONFAULT)
		gup_flags &= ~FOLL_POPULATE;
	/*
	 * We want to touch writable mappings with a write fault in order
	 * to break COW, except for shared mappings because these don't COW
	 * and we would not want to dirty them for nothing.
	 */
	if ((vma->vm_flags & (VM_WRITE | VM_SHARED)) == VM_WRITE)
		gup_flags |= FOLL_WRITE;

	/*
	 * We want mlock to succeed for regions that have any permissions
	 * other than PROT_NONE.
	 */
	if (vma->vm_flags & (VM_READ | VM_WRITE | VM_EXEC))
		gup_flags |= FOLL_FORCE;

	/*
	 * We made sure addr is within a VMA, so the following will
	 * not result in a stack expansion that recurses back here.
	 */
	return __get_user_pages(current, mm, start, nr_pages, gup_flags,
1396
				NULL, NULL, locked);
1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 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 1480 1481 1482 1483 1484 1485 1486
}

/*
 * __mm_populate - populate and/or mlock pages within a range of address space.
 *
 * This is used to implement mlock() and the MAP_POPULATE / MAP_LOCKED mmap
 * flags. VMAs must be already marked with the desired vm_flags, and
 * mmap_sem must not be held.
 */
int __mm_populate(unsigned long start, unsigned long len, int ignore_errors)
{
	struct mm_struct *mm = current->mm;
	unsigned long end, nstart, nend;
	struct vm_area_struct *vma = NULL;
	int locked = 0;
	long ret = 0;

	end = start + len;

	for (nstart = start; nstart < end; nstart = nend) {
		/*
		 * We want to fault in pages for [nstart; end) address range.
		 * Find first corresponding VMA.
		 */
		if (!locked) {
			locked = 1;
			down_read(&mm->mmap_sem);
			vma = find_vma(mm, nstart);
		} else if (nstart >= vma->vm_end)
			vma = vma->vm_next;
		if (!vma || vma->vm_start >= end)
			break;
		/*
		 * Set [nstart; nend) to intersection of desired address
		 * range with the first VMA. Also, skip undesirable VMA types.
		 */
		nend = min(end, vma->vm_end);
		if (vma->vm_flags & (VM_IO | VM_PFNMAP))
			continue;
		if (nstart < vma->vm_start)
			nstart = vma->vm_start;
		/*
		 * Now fault in a range of pages. populate_vma_page_range()
		 * double checks the vma flags, so that it won't mlock pages
		 * if the vma was already munlocked.
		 */
		ret = populate_vma_page_range(vma, nstart, nend, &locked);
		if (ret < 0) {
			if (ignore_errors) {
				ret = 0;
				continue;	/* continue at next VMA */
			}
			break;
		}
		nend = nstart + ret * PAGE_SIZE;
		ret = 0;
	}
	if (locked)
		up_read(&mm->mmap_sem);
	return ret;	/* 0 or negative error code */
}

/**
 * get_dump_page() - pin user page in memory while writing it to core dump
 * @addr: user address
 *
 * Returns struct page pointer of user page pinned for dump,
 * to be freed afterwards by put_page().
 *
 * Returns NULL on any kind of failure - a hole must then be inserted into
 * the corefile, to preserve alignment with its headers; and also returns
 * NULL wherever the ZERO_PAGE, or an anonymous pte_none, has been found -
 * allowing a hole to be left in the corefile to save diskspace.
 *
 * Called without mmap_sem, but after all other threads have been killed.
 */
#ifdef CONFIG_ELF_CORE
struct page *get_dump_page(unsigned long addr)
{
	struct vm_area_struct *vma;
	struct page *page;

	if (__get_user_pages(current, current->mm, addr, 1,
			     FOLL_FORCE | FOLL_DUMP | FOLL_GET, &page, &vma,
			     NULL) < 1)
		return NULL;
	flush_cache_page(vma, addr, page_to_pfn(page));
	return page;
}
#endif /* CONFIG_ELF_CORE */
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 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531
#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 */
1532

1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 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 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604
#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);
}

1605 1606 1607 1608
static long check_and_migrate_cma_pages(struct task_struct *tsk,
					struct mm_struct *mm,
					unsigned long start,
					unsigned long nr_pages,
1609
					struct page **pages,
1610 1611
					struct vm_area_struct **vmas,
					unsigned int gup_flags)
1612
{
1613 1614
	unsigned long i;
	unsigned long step;
1615 1616 1617
	bool drain_allow = true;
	bool migrate_allow = true;
	LIST_HEAD(cma_page_list);
1618
	long ret = nr_pages;
1619 1620

check_again:
1621 1622 1623 1624 1625 1626 1627 1628
	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.
		 */
1629
		step = compound_nr(head) - (pages[i] - head);
1630 1631 1632 1633 1634
		/*
		 * 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.
		 */
1635 1636
		if (is_migrate_cma_page(head)) {
			if (PageHuge(head))
1637
				isolate_huge_page(head, &cma_page_list);
1638
			else {
1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652
				if (!PageLRU(head) && drain_allow) {
					lru_add_drain_all();
					drain_allow = false;
				}

				if (!isolate_lru_page(head)) {
					list_add_tail(&head->lru, &cma_page_list);
					mod_node_page_state(page_pgdat(head),
							    NR_ISOLATED_ANON +
							    page_is_file_cache(head),
							    hpage_nr_pages(head));
				}
			}
		}
1653 1654

		i += step;
1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675
	}

	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);
		}
		/*
1676 1677 1678
		 * We did migrate all the pages, Try to get the page references
		 * again migrating any new CMA pages which we failed to isolate
		 * earlier.
1679
		 */
1680
		ret = __get_user_pages_locked(tsk, mm, start, nr_pages,
1681 1682 1683
						   pages, vmas, NULL,
						   gup_flags);

1684 1685
		if ((ret > 0) && migrate_allow) {
			nr_pages = ret;
1686 1687 1688 1689 1690
			drain_allow = true;
			goto check_again;
		}
	}

1691
	return ret;
1692 1693
}
#else
1694 1695 1696 1697 1698 1699 1700
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)
1701 1702 1703
{
	return nr_pages;
}
1704
#endif /* CONFIG_CMA */
1705

1706
/*
1707 1708
 * __gup_longterm_locked() is a wrapper for __get_user_pages_locked which
 * allows us to process the FOLL_LONGTERM flag.
1709
 */
1710 1711 1712 1713 1714 1715 1716
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)
1717
{
1718 1719
	struct vm_area_struct **vmas_tmp = vmas;
	unsigned long flags = 0;
1720 1721
	long rc, i;

1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733
	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();
1734 1735
	}

1736 1737
	rc = __get_user_pages_locked(tsk, mm, start, nr_pages, pages,
				     vmas_tmp, NULL, gup_flags);
1738

1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752
	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);
1753
	}
1754 1755

out:
1756 1757
	if (vmas_tmp != vmas)
		kfree(vmas_tmp);
1758 1759
	return rc;
}
1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773
#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 */

1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804
#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);
}

1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871
/*
 * 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.
 *
 * 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.
 *
 * get_user_pages is typically used for fewer-copy IO operations, to get a
 * handle on the memory by some means other than accesses via the user virtual
 * addresses. The pages may be submitted for DMA to devices or accessed via
 * their kernel linear mapping (via the kmap APIs). Care should be taken to
 * use the correct cache flushing APIs.
 *
 * See also get_user_pages_fast, for performance critical applications.
 *
 * get_user_pages should be phased out in favor of
 * get_user_pages_locked|unlocked or get_user_pages_fast. Nothing
 * should use get_user_pages because it cannot pass
 * FAULT_FLAG_ALLOW_RETRY to handle_mm_fault.
 */
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)
{
1872 1873 1874 1875 1876 1877 1878
	/*
	 * 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;

1879 1880
	return __get_user_pages_remote(tsk, mm, start, nr_pages, gup_flags,
				       pages, vmas, locked);
1881 1882 1883
}
EXPORT_SYMBOL(get_user_pages_remote);

1884 1885 1886 1887 1888 1889 1890 1891
#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
1892 1893 1894 1895 1896 1897 1898 1899 1900

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;
}
1901 1902
#endif /* !CONFIG_MMU */

1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913
/*
 * 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)
{
1914 1915 1916 1917 1918 1919 1920
	/*
	 * 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;

1921 1922 1923 1924
	return __gup_longterm_locked(current, current->mm, start, nr_pages,
				     pages, vmas, gup_flags | FOLL_TOUCH);
}
EXPORT_SYMBOL(get_user_pages);
1925

1926 1927 1928 1929
/*
 * 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().
1930
 *
1931
 * get_user_pages_locked() is suitable to replace the form:
1932
 *
1933 1934 1935 1936
 *      down_read(&mm->mmap_sem);
 *      do_something()
 *      get_user_pages(tsk, mm, ..., pages, NULL);
 *      up_read(&mm->mmap_sem);
1937
 *
1938
 *  to:
1939
 *
1940 1941 1942 1943 1944 1945
 *      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);
1946
 */
1947 1948 1949
long get_user_pages_locked(unsigned long start, unsigned long nr_pages,
			   unsigned int gup_flags, struct page **pages,
			   int *locked)
1950 1951
{
	/*
1952 1953 1954 1955
	 * 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.
1956
	 */
1957 1958
	if (WARN_ON_ONCE(gup_flags & FOLL_LONGTERM))
		return -EINVAL;
1959

1960 1961 1962
	return __get_user_pages_locked(current, current->mm, start, nr_pages,
				       pages, NULL, locked,
				       gup_flags | FOLL_TOUCH);
1963
}
1964
EXPORT_SYMBOL(get_user_pages_locked);
1965 1966

/*
1967
 * get_user_pages_unlocked() is suitable to replace the form:
1968
 *
1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979
 *      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.
1980
 */
1981 1982
long get_user_pages_unlocked(unsigned long start, unsigned long nr_pages,
			     struct page **pages, unsigned int gup_flags)
1983 1984
{
	struct mm_struct *mm = current->mm;
1985 1986
	int locked = 1;
	long ret;
1987

1988 1989 1990 1991 1992 1993 1994 1995
	/*
	 * 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;
1996

1997 1998 1999
	down_read(&mm->mmap_sem);
	ret = __get_user_pages_locked(current, mm, start, nr_pages, pages, NULL,
				      &locked, gup_flags | FOLL_TOUCH);
2000 2001
	if (locked)
		up_read(&mm->mmap_sem);
2002
	return ret;
2003
}
2004
EXPORT_SYMBOL(get_user_pages_unlocked);
2005 2006

/*
2007
 * Fast GUP
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027
 *
 * 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:
 *
2028
 *  *) Either MMU_GATHER_RCU_TABLE_FREE is enabled, and tlb_remove_table() is used to
2029
 *  free pages containing page tables or TLB flushing requires IPI broadcast.
2030 2031 2032 2033 2034 2035 2036 2037 2038
 *
 *  *) 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.
 */
2039
#ifdef CONFIG_HAVE_FAST_GUP
J
John Hubbard 已提交
2040 2041 2042

static void put_compound_head(struct page *page, int refs, unsigned int flags)
{
2043
	if (flags & FOLL_PIN) {
2044 2045 2046
		mod_node_page_state(page_pgdat(page), NR_FOLL_PIN_RELEASED,
				    refs);

2047 2048 2049 2050 2051
		if (hpage_pincount_available(page))
			hpage_pincount_sub(page, refs);
		else
			refs *= GUP_PIN_COUNTING_BIAS;
	}
J
John Hubbard 已提交
2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062

	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);
}

2063
#ifdef CONFIG_GUP_GET_PTE_LOW_HIGH
J
John Hubbard 已提交
2064

2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098
/*
 * 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;
2099

2100 2101 2102 2103 2104 2105 2106 2107 2108 2109
	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 */
2110
/*
2111
 * We require that the PTE can be read atomically.
2112 2113 2114 2115 2116
 */
static inline pte_t gup_get_pte(pte_t *ptep)
{
	return READ_ONCE(*ptep);
}
2117
#endif /* CONFIG_GUP_GET_PTE_LOW_HIGH */
2118

2119
static void __maybe_unused undo_dev_pagemap(int *nr, int nr_start,
2120
					    unsigned int flags,
2121
					    struct page **pages)
2122 2123 2124 2125 2126
{
	while ((*nr) - nr_start) {
		struct page *page = pages[--(*nr)];

		ClearPageReferenced(page);
J
John Hubbard 已提交
2127 2128 2129 2130
		if (flags & FOLL_PIN)
			unpin_user_page(page);
		else
			put_page(page);
2131 2132 2133
	}
}

2134
#ifdef CONFIG_ARCH_HAS_PTE_SPECIAL
2135
static int gup_pte_range(pmd_t pmd, unsigned long addr, unsigned long end,
2136
			 unsigned int flags, struct page **pages, int *nr)
2137
{
2138 2139
	struct dev_pagemap *pgmap = NULL;
	int nr_start = *nr, ret = 0;
2140 2141 2142 2143
	pte_t *ptep, *ptem;

	ptem = ptep = pte_offset_map(&pmd, addr);
	do {
2144
		pte_t pte = gup_get_pte(ptep);
2145
		struct page *head, *page;
2146 2147 2148

		/*
		 * Similar to the PMD case below, NUMA hinting must take slow
2149
		 * path using the pte_protnone check.
2150
		 */
2151 2152 2153
		if (pte_protnone(pte))
			goto pte_unmap;

2154
		if (!pte_access_permitted(pte, flags & FOLL_WRITE))
2155 2156
			goto pte_unmap;

2157
		if (pte_devmap(pte)) {
2158 2159 2160
			if (unlikely(flags & FOLL_LONGTERM))
				goto pte_unmap;

2161 2162
			pgmap = get_dev_pagemap(pte_pfn(pte), pgmap);
			if (unlikely(!pgmap)) {
2163
				undo_dev_pagemap(nr, nr_start, flags, pages);
2164 2165 2166
				goto pte_unmap;
			}
		} else if (pte_special(pte))
2167 2168 2169 2170 2171
			goto pte_unmap;

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

J
John Hubbard 已提交
2172
		head = try_grab_compound_head(page, 1, flags);
2173
		if (!head)
2174 2175 2176
			goto pte_unmap;

		if (unlikely(pte_val(pte) != pte_val(*ptep))) {
J
John Hubbard 已提交
2177
			put_compound_head(head, 1, flags);
2178 2179 2180
			goto pte_unmap;
		}

2181
		VM_BUG_ON_PAGE(compound_head(page) != head, page);
2182

2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195
		/*
		 * 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;
			}
		}
2196
		SetPageReferenced(page);
2197 2198 2199 2200 2201 2202 2203 2204
		pages[*nr] = page;
		(*nr)++;

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

	ret = 1;

pte_unmap:
2205 2206
	if (pgmap)
		put_dev_pagemap(pgmap);
2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221
	pte_unmap(ptem);
	return ret;
}
#else

/*
 * If we can't determine whether or not a pte is special, then fail immediately
 * for ptes. Note, we can still pin HugeTLB and THP as these are guaranteed not
 * to be special.
 *
 * For a futex to be placed on a THP tail page, get_futex_key requires a
 * __get_user_pages_fast implementation that can pin pages. Thus it's still
 * useful to have gup_huge_pmd even if we can't operate on ptes.
 */
static int gup_pte_range(pmd_t pmd, unsigned long addr, unsigned long end,
2222
			 unsigned int flags, struct page **pages, int *nr)
2223 2224 2225
{
	return 0;
}
2226
#endif /* CONFIG_ARCH_HAS_PTE_SPECIAL */
2227

R
Robin Murphy 已提交
2228
#if defined(CONFIG_ARCH_HAS_PTE_DEVMAP) && defined(CONFIG_TRANSPARENT_HUGEPAGE)
2229
static int __gup_device_huge(unsigned long pfn, unsigned long addr,
2230 2231
			     unsigned long end, unsigned int flags,
			     struct page **pages, int *nr)
2232 2233 2234 2235 2236 2237 2238 2239 2240
{
	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)) {
2241
			undo_dev_pagemap(nr, nr_start, flags, pages);
2242 2243 2244 2245
			return 0;
		}
		SetPageReferenced(page);
		pages[*nr] = page;
J
John Hubbard 已提交
2246 2247 2248 2249
		if (unlikely(!try_grab_page(page, flags))) {
			undo_dev_pagemap(nr, nr_start, flags, pages);
			return 0;
		}
2250 2251 2252
		(*nr)++;
		pfn++;
	} while (addr += PAGE_SIZE, addr != end);
2253 2254 2255

	if (pgmap)
		put_dev_pagemap(pgmap);
2256 2257 2258
	return 1;
}

2259
static int __gup_device_huge_pmd(pmd_t orig, pmd_t *pmdp, unsigned long addr,
2260 2261
				 unsigned long end, unsigned int flags,
				 struct page **pages, int *nr)
2262 2263
{
	unsigned long fault_pfn;
2264 2265 2266
	int nr_start = *nr;

	fault_pfn = pmd_pfn(orig) + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
2267
	if (!__gup_device_huge(fault_pfn, addr, end, flags, pages, nr))
2268
		return 0;
2269

2270
	if (unlikely(pmd_val(orig) != pmd_val(*pmdp))) {
2271
		undo_dev_pagemap(nr, nr_start, flags, pages);
2272 2273 2274
		return 0;
	}
	return 1;
2275 2276
}

2277
static int __gup_device_huge_pud(pud_t orig, pud_t *pudp, unsigned long addr,
2278 2279
				 unsigned long end, unsigned int flags,
				 struct page **pages, int *nr)
2280 2281
{
	unsigned long fault_pfn;
2282 2283 2284
	int nr_start = *nr;

	fault_pfn = pud_pfn(orig) + ((addr & ~PUD_MASK) >> PAGE_SHIFT);
2285
	if (!__gup_device_huge(fault_pfn, addr, end, flags, pages, nr))
2286
		return 0;
2287

2288
	if (unlikely(pud_val(orig) != pud_val(*pudp))) {
2289
		undo_dev_pagemap(nr, nr_start, flags, pages);
2290 2291 2292
		return 0;
	}
	return 1;
2293 2294
}
#else
2295
static int __gup_device_huge_pmd(pmd_t orig, pmd_t *pmdp, unsigned long addr,
2296 2297
				 unsigned long end, unsigned int flags,
				 struct page **pages, int *nr)
2298 2299 2300 2301 2302
{
	BUILD_BUG();
	return 0;
}

2303
static int __gup_device_huge_pud(pud_t pud, pud_t *pudp, unsigned long addr,
2304 2305
				 unsigned long end, unsigned int flags,
				 struct page **pages, int *nr)
2306 2307 2308 2309 2310 2311
{
	BUILD_BUG();
	return 0;
}
#endif

2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322
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;
}

2323 2324 2325 2326 2327 2328 2329 2330 2331
#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,
2332 2333
		       unsigned long end, unsigned int flags,
		       struct page **pages, int *nr)
2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345
{
	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);

2346
	if (!pte_access_permitted(pte, flags & FOLL_WRITE))
2347 2348 2349 2350 2351 2352 2353
		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);
2354
	refs = record_subpages(page, addr, end, pages + *nr);
2355

J
John Hubbard 已提交
2356
	head = try_grab_compound_head(head, refs, flags);
2357
	if (!head)
2358 2359 2360
		return 0;

	if (unlikely(pte_val(pte) != pte_val(*ptep))) {
2361
		put_compound_head(head, refs, flags);
2362 2363 2364
		return 0;
	}

2365
	*nr += refs;
2366
	SetPageReferenced(head);
2367 2368 2369 2370
	return 1;
}

static int gup_huge_pd(hugepd_t hugepd, unsigned long addr,
2371
		unsigned int pdshift, unsigned long end, unsigned int flags,
2372 2373 2374 2375 2376 2377 2378 2379 2380
		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);
2381
		if (!gup_hugepte(ptep, sz, addr, end, flags, pages, nr))
2382 2383 2384 2385 2386 2387 2388
			return 0;
	} while (ptep++, addr = next, addr != end);

	return 1;
}
#else
static inline int gup_huge_pd(hugepd_t hugepd, unsigned long addr,
2389
		unsigned int pdshift, unsigned long end, unsigned int flags,
2390 2391 2392 2393 2394 2395
		struct page **pages, int *nr)
{
	return 0;
}
#endif /* CONFIG_ARCH_HAS_HUGEPD */

2396
static int gup_huge_pmd(pmd_t orig, pmd_t *pmdp, unsigned long addr,
2397 2398
			unsigned long end, unsigned int flags,
			struct page **pages, int *nr)
2399
{
2400
	struct page *head, *page;
2401 2402
	int refs;

2403
	if (!pmd_access_permitted(orig, flags & FOLL_WRITE))
2404 2405
		return 0;

2406 2407 2408
	if (pmd_devmap(orig)) {
		if (unlikely(flags & FOLL_LONGTERM))
			return 0;
2409 2410
		return __gup_device_huge_pmd(orig, pmdp, addr, end, flags,
					     pages, nr);
2411
	}
2412

2413
	page = pmd_page(orig) + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
2414
	refs = record_subpages(page, addr, end, pages + *nr);
2415

J
John Hubbard 已提交
2416
	head = try_grab_compound_head(pmd_page(orig), refs, flags);
2417
	if (!head)
2418 2419 2420
		return 0;

	if (unlikely(pmd_val(orig) != pmd_val(*pmdp))) {
2421
		put_compound_head(head, refs, flags);
2422 2423 2424
		return 0;
	}

2425
	*nr += refs;
2426
	SetPageReferenced(head);
2427 2428 2429 2430
	return 1;
}

static int gup_huge_pud(pud_t orig, pud_t *pudp, unsigned long addr,
2431 2432
			unsigned long end, unsigned int flags,
			struct page **pages, int *nr)
2433
{
2434
	struct page *head, *page;
2435 2436
	int refs;

2437
	if (!pud_access_permitted(orig, flags & FOLL_WRITE))
2438 2439
		return 0;

2440 2441 2442
	if (pud_devmap(orig)) {
		if (unlikely(flags & FOLL_LONGTERM))
			return 0;
2443 2444
		return __gup_device_huge_pud(orig, pudp, addr, end, flags,
					     pages, nr);
2445
	}
2446

2447
	page = pud_page(orig) + ((addr & ~PUD_MASK) >> PAGE_SHIFT);
2448
	refs = record_subpages(page, addr, end, pages + *nr);
2449

J
John Hubbard 已提交
2450
	head = try_grab_compound_head(pud_page(orig), refs, flags);
2451
	if (!head)
2452 2453 2454
		return 0;

	if (unlikely(pud_val(orig) != pud_val(*pudp))) {
2455
		put_compound_head(head, refs, flags);
2456 2457 2458
		return 0;
	}

2459
	*nr += refs;
2460
	SetPageReferenced(head);
2461 2462 2463
	return 1;
}

2464
static int gup_huge_pgd(pgd_t orig, pgd_t *pgdp, unsigned long addr,
2465
			unsigned long end, unsigned int flags,
2466 2467 2468
			struct page **pages, int *nr)
{
	int refs;
2469
	struct page *head, *page;
2470

2471
	if (!pgd_access_permitted(orig, flags & FOLL_WRITE))
2472 2473
		return 0;

2474
	BUILD_BUG_ON(pgd_devmap(orig));
2475

2476
	page = pgd_page(orig) + ((addr & ~PGDIR_MASK) >> PAGE_SHIFT);
2477
	refs = record_subpages(page, addr, end, pages + *nr);
2478

J
John Hubbard 已提交
2479
	head = try_grab_compound_head(pgd_page(orig), refs, flags);
2480
	if (!head)
2481 2482 2483
		return 0;

	if (unlikely(pgd_val(orig) != pgd_val(*pgdp))) {
2484
		put_compound_head(head, refs, flags);
2485 2486 2487
		return 0;
	}

2488
	*nr += refs;
2489
	SetPageReferenced(head);
2490 2491 2492
	return 1;
}

2493
static int gup_pmd_range(pud_t pud, unsigned long addr, unsigned long end,
2494
		unsigned int flags, struct page **pages, int *nr)
2495 2496 2497 2498 2499 2500
{
	unsigned long next;
	pmd_t *pmdp;

	pmdp = pmd_offset(&pud, addr);
	do {
2501
		pmd_t pmd = READ_ONCE(*pmdp);
2502 2503

		next = pmd_addr_end(addr, end);
2504
		if (!pmd_present(pmd))
2505 2506
			return 0;

Y
Yu Zhao 已提交
2507 2508
		if (unlikely(pmd_trans_huge(pmd) || pmd_huge(pmd) ||
			     pmd_devmap(pmd))) {
2509 2510 2511 2512 2513
			/*
			 * NUMA hinting faults need to be handled in the GUP
			 * slowpath for accounting purposes and so that they
			 * can be serialised against THP migration.
			 */
2514
			if (pmd_protnone(pmd))
2515 2516
				return 0;

2517
			if (!gup_huge_pmd(pmd, pmdp, addr, next, flags,
2518 2519 2520
				pages, nr))
				return 0;

2521 2522 2523 2524 2525 2526
		} 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,
2527
					 PMD_SHIFT, next, flags, pages, nr))
2528
				return 0;
2529
		} else if (!gup_pte_range(pmd, addr, next, flags, pages, nr))
2530
			return 0;
2531 2532 2533 2534 2535
	} while (pmdp++, addr = next, addr != end);

	return 1;
}

2536
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
2537
			 unsigned int flags, struct page **pages, int *nr)
2538 2539 2540 2541
{
	unsigned long next;
	pud_t *pudp;

2542
	pudp = pud_offset(&p4d, addr);
2543
	do {
2544
		pud_t pud = READ_ONCE(*pudp);
2545 2546

		next = pud_addr_end(addr, end);
Q
Qiujun Huang 已提交
2547
		if (unlikely(!pud_present(pud)))
2548
			return 0;
2549
		if (unlikely(pud_huge(pud))) {
2550
			if (!gup_huge_pud(pud, pudp, addr, next, flags,
2551 2552 2553 2554
					  pages, nr))
				return 0;
		} else if (unlikely(is_hugepd(__hugepd(pud_val(pud))))) {
			if (!gup_huge_pd(__hugepd(pud_val(pud)), addr,
2555
					 PUD_SHIFT, next, flags, pages, nr))
2556
				return 0;
2557
		} else if (!gup_pmd_range(pud, addr, next, flags, pages, nr))
2558 2559 2560 2561 2562 2563
			return 0;
	} while (pudp++, addr = next, addr != end);

	return 1;
}

2564
static int gup_p4d_range(pgd_t pgd, unsigned long addr, unsigned long end,
2565
			 unsigned int flags, struct page **pages, int *nr)
2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579
{
	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,
2580
					 P4D_SHIFT, next, flags, pages, nr))
2581
				return 0;
2582
		} else if (!gup_pud_range(p4d, addr, next, flags, pages, nr))
2583 2584 2585 2586 2587 2588
			return 0;
	} while (p4dp++, addr = next, addr != end);

	return 1;
}

2589
static void gup_pgd_range(unsigned long addr, unsigned long end,
2590
		unsigned int flags, struct page **pages, int *nr)
2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602
{
	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))) {
2603
			if (!gup_huge_pgd(pgd, pgdp, addr, next, flags,
2604 2605 2606 2607
					  pages, nr))
				return;
		} else if (unlikely(is_hugepd(__hugepd(pgd_val(pgd))))) {
			if (!gup_huge_pd(__hugepd(pgd_val(pgd)), addr,
2608
					 PGDIR_SHIFT, next, flags, pages, nr))
2609
				return;
2610
		} else if (!gup_p4d_range(pgd, addr, next, flags, pages, nr))
2611 2612 2613
			return;
	} while (pgdp++, addr = next, addr != end);
}
2614 2615 2616 2617 2618 2619
#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 */
2620 2621 2622 2623 2624 2625

#ifndef gup_fast_permitted
/*
 * Check if it's allowed to use __get_user_pages_fast() for the range, or
 * we need to fall back to the slow version:
 */
2626
static bool gup_fast_permitted(unsigned long start, unsigned long end)
2627
{
2628
	return true;
2629 2630 2631
}
#endif

2632 2633
/*
 * Like get_user_pages_fast() except it's IRQ-safe in that it won't fall back to
2634 2635 2636
 * 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.
2637 2638 2639
 *
 * If the architecture does not support this function, simply return with no
 * pages pinned.
2640 2641 2642 2643
 */
int __get_user_pages_fast(unsigned long start, int nr_pages, int write,
			  struct page **pages)
{
2644
	unsigned long len, end;
2645
	unsigned long flags;
2646
	int nr_pinned = 0;
2647 2648 2649 2650 2651 2652 2653 2654
	/*
	 * 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.
	 */
	unsigned int gup_flags = FOLL_GET;

	if (write)
		gup_flags |= FOLL_WRITE;
2655

2656
	start = untagged_addr(start) & PAGE_MASK;
2657 2658 2659
	len = (unsigned long) nr_pages << PAGE_SHIFT;
	end = start + len;

2660 2661
	if (end <= start)
		return 0;
2662
	if (unlikely(!access_ok((void __user *)start, len)))
2663 2664 2665 2666 2667 2668 2669
		return 0;

	/*
	 * Disable interrupts.  We use the nested form as we can already have
	 * interrupts disabled by get_futex_key.
	 *
	 * With interrupts disabled, we block page table pages from being
2670 2671
	 * freed from under us. See struct mmu_table_batch comments in
	 * include/asm-generic/tlb.h for more details.
2672 2673 2674 2675 2676
	 *
	 * We do not adopt an rcu_read_lock(.) here as we also want to
	 * block IPIs that come from THPs splitting.
	 */

2677 2678
	if (IS_ENABLED(CONFIG_HAVE_FAST_GUP) &&
	    gup_fast_permitted(start, end)) {
2679
		local_irq_save(flags);
2680
		gup_pgd_range(start, end, gup_flags, pages, &nr_pinned);
2681 2682
		local_irq_restore(flags);
	}
2683

2684
	return nr_pinned;
2685
}
2686
EXPORT_SYMBOL_GPL(__get_user_pages_fast);
2687

2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710
static int __gup_longterm_unlocked(unsigned long start, int nr_pages,
				   unsigned int gup_flags, struct page **pages)
{
	int ret;

	/*
	 * FIXME: FOLL_LONGTERM does not work with
	 * get_user_pages_unlocked() (see comments in that function)
	 */
	if (gup_flags & FOLL_LONGTERM) {
		down_read(&current->mm->mmap_sem);
		ret = __gup_longterm_locked(current, current->mm,
					    start, nr_pages,
					    pages, NULL, gup_flags);
		up_read(&current->mm->mmap_sem);
	} else {
		ret = get_user_pages_unlocked(start, nr_pages,
					      pages, gup_flags);
	}

	return ret;
}

2711 2712 2713
static int internal_get_user_pages_fast(unsigned long start, int nr_pages,
					unsigned int gup_flags,
					struct page **pages)
2714
{
2715
	unsigned long addr, len, end;
2716
	int nr_pinned = 0, ret = 0;
2717

2718
	if (WARN_ON_ONCE(gup_flags & ~(FOLL_WRITE | FOLL_LONGTERM |
2719
				       FOLL_FORCE | FOLL_PIN | FOLL_GET)))
2720 2721
		return -EINVAL;

2722
	start = untagged_addr(start) & PAGE_MASK;
2723 2724 2725 2726
	addr = start;
	len = (unsigned long) nr_pages << PAGE_SHIFT;
	end = start + len;

2727
	if (end <= start)
2728
		return 0;
2729
	if (unlikely(!access_ok((void __user *)start, len)))
2730
		return -EFAULT;
2731

2732 2733
	if (IS_ENABLED(CONFIG_HAVE_FAST_GUP) &&
	    gup_fast_permitted(start, end)) {
2734
		local_irq_disable();
2735
		gup_pgd_range(addr, end, gup_flags, pages, &nr_pinned);
2736
		local_irq_enable();
2737
		ret = nr_pinned;
2738
	}
2739

2740
	if (nr_pinned < nr_pages) {
2741
		/* Try to get the remaining pages with get_user_pages */
2742 2743
		start += nr_pinned << PAGE_SHIFT;
		pages += nr_pinned;
2744

2745
		ret = __gup_longterm_unlocked(start, nr_pages - nr_pinned,
2746
					      gup_flags, pages);
2747 2748

		/* Have to be a bit careful with return values */
2749
		if (nr_pinned > 0) {
2750
			if (ret < 0)
2751
				ret = nr_pinned;
2752
			else
2753
				ret += nr_pinned;
2754 2755 2756 2757 2758
		}
	}

	return ret;
}
2759 2760 2761

/**
 * get_user_pages_fast() - pin user pages in memory
J
John Hubbard 已提交
2762 2763 2764 2765 2766
 * @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.
2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785
 *
 * 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;

2786 2787 2788 2789 2790 2791 2792
	/*
	 * 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;
2793 2794
	return internal_get_user_pages_fast(start, nr_pages, gup_flags, pages);
}
2795
EXPORT_SYMBOL_GPL(get_user_pages_fast);
2796 2797 2798 2799

/**
 * pin_user_pages_fast() - pin user pages in memory without taking locks
 *
J
John Hubbard 已提交
2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811
 * @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
 * see Documentation/vm/pin_user_pages.rst for further details.
2812 2813 2814 2815 2816 2817 2818
 *
 * This is intended for Case 1 (DIO) in Documentation/vm/pin_user_pages.rst. It
 * is NOT intended for Case 2 (RDMA: long-term pins).
 */
int pin_user_pages_fast(unsigned long start, int nr_pages,
			unsigned int gup_flags, struct page **pages)
{
J
John Hubbard 已提交
2819 2820 2821 2822 2823 2824
	/* 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);
2825 2826 2827 2828 2829 2830
}
EXPORT_SYMBOL_GPL(pin_user_pages_fast);

/**
 * pin_user_pages_remote() - pin pages of a remote process (task != current)
 *
J
John Hubbard 已提交
2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851
 * @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
 * see Documentation/vm/pin_user_pages.rst for details.
2852 2853 2854 2855 2856 2857 2858 2859 2860
 *
 * This is intended for Case 1 (DIO) in Documentation/vm/pin_user_pages.rst. It
 * is NOT intended for Case 2 (RDMA: long-term pins).
 */
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
2861 2862 2863 2864 2865 2866 2867
	/* 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);
2868 2869 2870 2871 2872 2873
}
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
2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887
 * @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
 * see Documentation/vm/pin_user_pages.rst for details.
2888 2889 2890 2891 2892 2893 2894 2895
 *
 * This is intended for Case 1 (DIO) in Documentation/vm/pin_user_pages.rst. It
 * is NOT intended for Case 2 (RDMA: long-term pins).
 */
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
2896 2897 2898 2899 2900 2901 2902
	/* 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);
2903 2904
}
EXPORT_SYMBOL(pin_user_pages);