pgtable.c 18.3 KB
Newer Older
1
// SPDX-License-Identifier: GPL-2.0
2
#include <linux/mm.h>
3
#include <linux/gfp.h>
4
#include <linux/hugetlb.h>
5
#include <asm/pgalloc.h>
6
#include <asm/pgtable.h>
7
#include <asm/tlb.h>
I
Ingo Molnar 已提交
8
#include <asm/fixmap.h>
9
#include <asm/mtrr.h>
10

11 12 13 14 15
#ifdef CONFIG_DYNAMIC_PHYSICAL_MASK
phys_addr_t physical_mask __ro_after_init = (1ULL << __PHYSICAL_MASK_SHIFT) - 1;
EXPORT_SYMBOL(physical_mask);
#endif

16
#define PGALLOC_GFP (GFP_KERNEL_ACCOUNT | __GFP_ZERO)
17

18 19 20 21 22 23 24 25
#ifdef CONFIG_HIGHPTE
#define PGALLOC_USER_GFP __GFP_HIGHMEM
#else
#define PGALLOC_USER_GFP 0
#endif

gfp_t __userpte_alloc_gfp = PGALLOC_GFP | PGALLOC_USER_GFP;

26 27
pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address)
{
28
	return (pte_t *)__get_free_page(PGALLOC_GFP & ~__GFP_ACCOUNT);
29 30 31 32 33 34
}

pgtable_t pte_alloc_one(struct mm_struct *mm, unsigned long address)
{
	struct page *pte;

35
	pte = alloc_pages(__userpte_alloc_gfp, 0);
36 37 38 39 40 41
	if (!pte)
		return NULL;
	if (!pgtable_page_ctor(pte)) {
		__free_page(pte);
		return NULL;
	}
42 43 44
	return pte;
}

45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
static int __init setup_userpte(char *arg)
{
	if (!arg)
		return -EINVAL;

	/*
	 * "userpte=nohigh" disables allocation of user pagetables in
	 * high memory.
	 */
	if (strcmp(arg, "nohigh") == 0)
		__userpte_alloc_gfp &= ~__GFP_HIGHMEM;
	else
		return -EINVAL;
	return 0;
}
early_param("userpte", setup_userpte);

62
void ___pte_free_tlb(struct mmu_gather *tlb, struct page *pte)
63 64
{
	pgtable_page_dtor(pte);
65
	paravirt_release_pte(page_to_pfn(pte));
66
	tlb_remove_table(tlb, pte);
67 68
}

69
#if CONFIG_PGTABLE_LEVELS > 2
70
void ___pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd)
71
{
72
	struct page *page = virt_to_page(pmd);
73
	paravirt_release_pmd(__pa(pmd) >> PAGE_SHIFT);
74 75 76 77 78 79 80
	/*
	 * NOTE! For PAE, any changes to the top page-directory-pointer-table
	 * entries need a full cr3 reload to flush.
	 */
#ifdef CONFIG_X86_PAE
	tlb->need_flush_all = 1;
#endif
81
	pgtable_pmd_page_dtor(page);
82
	tlb_remove_table(tlb, page);
83
}
84

85
#if CONFIG_PGTABLE_LEVELS > 3
86
void ___pud_free_tlb(struct mmu_gather *tlb, pud_t *pud)
87
{
88
	paravirt_release_pud(__pa(pud) >> PAGE_SHIFT);
89
	tlb_remove_table(tlb, virt_to_page(pud));
90
}
91 92 93 94 95

#if CONFIG_PGTABLE_LEVELS > 4
void ___p4d_free_tlb(struct mmu_gather *tlb, p4d_t *p4d)
{
	paravirt_release_p4d(__pa(p4d) >> PAGE_SHIFT);
96
	tlb_remove_table(tlb, virt_to_page(p4d));
97 98
}
#endif	/* CONFIG_PGTABLE_LEVELS > 4 */
99 100
#endif	/* CONFIG_PGTABLE_LEVELS > 3 */
#endif	/* CONFIG_PGTABLE_LEVELS > 2 */
101

102 103 104 105 106 107 108 109 110 111 112 113 114 115 116
static inline void pgd_list_add(pgd_t *pgd)
{
	struct page *page = virt_to_page(pgd);

	list_add(&page->lru, &pgd_list);
}

static inline void pgd_list_del(pgd_t *pgd)
{
	struct page *page = virt_to_page(pgd);

	list_del(&page->lru);
}

#define UNSHARED_PTRS_PER_PGD				\
J
Jeremy Fitzhardinge 已提交
117
	(SHARED_KERNEL_PMD ? KERNEL_PGD_BOUNDARY : PTRS_PER_PGD)
118

119 120 121

static void pgd_set_mm(pgd_t *pgd, struct mm_struct *mm)
{
M
Matthew Wilcox 已提交
122
	virt_to_page(pgd)->pt_mm = mm;
123 124 125 126
}

struct mm_struct *pgd_page_get_mm(struct page *page)
{
M
Matthew Wilcox 已提交
127
	return page->pt_mm;
128 129 130
}

static void pgd_ctor(struct mm_struct *mm, pgd_t *pgd)
131 132 133 134
{
	/* If the pgd points to a shared pagetable level (either the
	   ptes in non-PAE, or shared PMD in PAE), then just copy the
	   references from swapper_pg_dir. */
135 136
	if (CONFIG_PGTABLE_LEVELS == 2 ||
	    (CONFIG_PGTABLE_LEVELS == 3 && SHARED_KERNEL_PMD) ||
137
	    CONFIG_PGTABLE_LEVELS >= 4) {
J
Jeremy Fitzhardinge 已提交
138 139
		clone_pgd_range(pgd + KERNEL_PGD_BOUNDARY,
				swapper_pg_dir + KERNEL_PGD_BOUNDARY,
140 141 142 143
				KERNEL_PGD_PTRS);
	}

	/* list required to sync kernel mapping updates */
144 145
	if (!SHARED_KERNEL_PMD) {
		pgd_set_mm(pgd, mm);
146
		pgd_list_add(pgd);
147
	}
148 149
}

J
Jan Beulich 已提交
150
static void pgd_dtor(pgd_t *pgd)
151 152 153 154
{
	if (SHARED_KERNEL_PMD)
		return;

A
Andrea Arcangeli 已提交
155
	spin_lock(&pgd_lock);
156
	pgd_list_del(pgd);
A
Andrea Arcangeli 已提交
157
	spin_unlock(&pgd_lock);
158 159
}

J
Jeremy Fitzhardinge 已提交
160 161 162 163 164 165 166 167
/*
 * List of all pgd's needed for non-PAE so it can invalidate entries
 * in both cached and uncached pgd's; not needed for PAE since the
 * kernel pmd is shared. If PAE were not to share the pmd a similar
 * tactic would be needed. This is essentially codepath-based locking
 * against pageattr.c; it is the unique case in which a valid change
 * of kernel pagetables can't be lazily synchronized by vmalloc faults.
 * vmalloc faults work because attached pagetables are never freed.
168
 * -- nyc
J
Jeremy Fitzhardinge 已提交
169 170
 */

171
#ifdef CONFIG_X86_PAE
172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198
/*
 * In PAE mode, we need to do a cr3 reload (=tlb flush) when
 * updating the top-level pagetable entries to guarantee the
 * processor notices the update.  Since this is expensive, and
 * all 4 top-level entries are used almost immediately in a
 * new process's life, we just pre-populate them here.
 *
 * Also, if we're in a paravirt environment where the kernel pmd is
 * not shared between pagetables (!SHARED_KERNEL_PMDS), we allocate
 * and initialize the kernel pmds here.
 */
#define PREALLOCATED_PMDS	UNSHARED_PTRS_PER_PGD

void pud_populate(struct mm_struct *mm, pud_t *pudp, pmd_t *pmd)
{
	paravirt_alloc_pmd(mm, __pa(pmd) >> PAGE_SHIFT);

	/* Note: almost everything apart from _PAGE_PRESENT is
	   reserved at the pmd (PDPT) level. */
	set_pud(pudp, __pud(__pa(pmd) | _PAGE_PRESENT));

	/*
	 * According to Intel App note "TLBs, Paging-Structure Caches,
	 * and Their Invalidation", April 2007, document 317080-001,
	 * section 8.1: in PAE mode we explicitly have to flush the
	 * TLB via cr3 if the top-level pgd is changed...
	 */
199
	flush_tlb_mm(mm);
200 201 202 203 204 205 206 207
}
#else  /* !CONFIG_X86_PAE */

/* No need to prepopulate any pagetable entries in non-PAE modes. */
#define PREALLOCATED_PMDS	0

#endif	/* CONFIG_X86_PAE */

208
static void free_pmds(struct mm_struct *mm, pmd_t *pmds[])
209 210 211 212
{
	int i;

	for(i = 0; i < PREALLOCATED_PMDS; i++)
213 214
		if (pmds[i]) {
			pgtable_pmd_page_dtor(virt_to_page(pmds[i]));
215
			free_page((unsigned long)pmds[i]);
216
			mm_dec_nr_pmds(mm);
217
		}
218 219
}

220
static int preallocate_pmds(struct mm_struct *mm, pmd_t *pmds[])
221 222 223
{
	int i;
	bool failed = false;
224 225 226 227
	gfp_t gfp = PGALLOC_GFP;

	if (mm == &init_mm)
		gfp &= ~__GFP_ACCOUNT;
228 229

	for(i = 0; i < PREALLOCATED_PMDS; i++) {
230
		pmd_t *pmd = (pmd_t *)__get_free_page(gfp);
231
		if (!pmd)
232
			failed = true;
233
		if (pmd && !pgtable_pmd_page_ctor(virt_to_page(pmd))) {
234
			free_page((unsigned long)pmd);
235 236 237
			pmd = NULL;
			failed = true;
		}
238 239
		if (pmd)
			mm_inc_nr_pmds(mm);
240 241 242 243
		pmds[i] = pmd;
	}

	if (failed) {
244
		free_pmds(mm, pmds);
245 246 247 248 249 250
		return -ENOMEM;
	}

	return 0;
}

251 252 253 254 255 256 257 258 259 260
/*
 * Mop up any pmd pages which may still be attached to the pgd.
 * Normally they will be freed by munmap/exit_mmap, but any pmd we
 * preallocate which never got a corresponding vma will need to be
 * freed manually.
 */
static void pgd_mop_up_pmds(struct mm_struct *mm, pgd_t *pgdp)
{
	int i;

261
	for(i = 0; i < PREALLOCATED_PMDS; i++) {
262 263 264 265 266 267 268
		pgd_t pgd = pgdp[i];

		if (pgd_val(pgd) != 0) {
			pmd_t *pmd = (pmd_t *)pgd_page_vaddr(pgd);

			pgdp[i] = native_make_pgd(0);

269
			paravirt_release_pmd(pgd_val(pgd) >> PAGE_SHIFT);
270
			pmd_free(mm, pmd);
271
			mm_dec_nr_pmds(mm);
272 273 274 275
		}
	}
}

276
static void pgd_prepopulate_pmd(struct mm_struct *mm, pgd_t *pgd, pmd_t *pmds[])
277
{
278
	p4d_t *p4d;
279 280 281
	pud_t *pud;
	int i;

282 283 284
	if (PREALLOCATED_PMDS == 0) /* Work around gcc-3.4.x bug */
		return;

285 286
	p4d = p4d_offset(pgd, 0);
	pud = pud_offset(p4d, 0);
287

288
	for (i = 0; i < PREALLOCATED_PMDS; i++, pud++) {
289
		pmd_t *pmd = pmds[i];
290

J
Jeremy Fitzhardinge 已提交
291
		if (i >= KERNEL_PGD_BOUNDARY)
292 293 294 295 296 297
			memcpy(pmd, (pmd_t *)pgd_page_vaddr(swapper_pg_dir[i]),
			       sizeof(pmd_t) * PTRS_PER_PMD);

		pud_populate(mm, pud, pmd);
	}
}
298

299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362
/*
 * Xen paravirt assumes pgd table should be in one page. 64 bit kernel also
 * assumes that pgd should be in one page.
 *
 * But kernel with PAE paging that is not running as a Xen domain
 * only needs to allocate 32 bytes for pgd instead of one page.
 */
#ifdef CONFIG_X86_PAE

#include <linux/slab.h>

#define PGD_SIZE	(PTRS_PER_PGD * sizeof(pgd_t))
#define PGD_ALIGN	32

static struct kmem_cache *pgd_cache;

static int __init pgd_cache_init(void)
{
	/*
	 * When PAE kernel is running as a Xen domain, it does not use
	 * shared kernel pmd. And this requires a whole page for pgd.
	 */
	if (!SHARED_KERNEL_PMD)
		return 0;

	/*
	 * when PAE kernel is not running as a Xen domain, it uses
	 * shared kernel pmd. Shared kernel pmd does not require a whole
	 * page for pgd. We are able to just allocate a 32-byte for pgd.
	 * During boot time, we create a 32-byte slab for pgd table allocation.
	 */
	pgd_cache = kmem_cache_create("pgd_cache", PGD_SIZE, PGD_ALIGN,
				      SLAB_PANIC, NULL);
	if (!pgd_cache)
		return -ENOMEM;

	return 0;
}
core_initcall(pgd_cache_init);

static inline pgd_t *_pgd_alloc(void)
{
	/*
	 * If no SHARED_KERNEL_PMD, PAE kernel is running as a Xen domain.
	 * We allocate one page for pgd.
	 */
	if (!SHARED_KERNEL_PMD)
		return (pgd_t *)__get_free_page(PGALLOC_GFP);

	/*
	 * Now PAE kernel is not running as a Xen domain. We can allocate
	 * a 32-byte slab for pgd to save memory space.
	 */
	return kmem_cache_alloc(pgd_cache, PGALLOC_GFP);
}

static inline void _pgd_free(pgd_t *pgd)
{
	if (!SHARED_KERNEL_PMD)
		free_page((unsigned long)pgd);
	else
		kmem_cache_free(pgd_cache, pgd);
}
#else
363

364 365
static inline pgd_t *_pgd_alloc(void)
{
366
	return (pgd_t *)__get_free_pages(PGALLOC_GFP, PGD_ALLOCATION_ORDER);
367 368 369 370
}

static inline void _pgd_free(pgd_t *pgd)
{
371
	free_pages((unsigned long)pgd, PGD_ALLOCATION_ORDER);
372 373 374
}
#endif /* CONFIG_X86_PAE */

375
pgd_t *pgd_alloc(struct mm_struct *mm)
376
{
377 378
	pgd_t *pgd;
	pmd_t *pmds[PREALLOCATED_PMDS];
379

380
	pgd = _pgd_alloc();
381 382 383 384 385 386

	if (pgd == NULL)
		goto out;

	mm->pgd = pgd;

387
	if (preallocate_pmds(mm, pmds) != 0)
388 389 390 391
		goto out_free_pgd;

	if (paravirt_pgd_alloc(mm) != 0)
		goto out_free_pmds;
392 393

	/*
394 395 396
	 * Make sure that pre-populating the pmds is atomic with
	 * respect to anything walking the pgd_list, so that they
	 * never see a partially populated pgd.
397
	 */
A
Andrea Arcangeli 已提交
398
	spin_lock(&pgd_lock);
399

400
	pgd_ctor(mm, pgd);
401
	pgd_prepopulate_pmd(mm, pgd, pmds);
402

A
Andrea Arcangeli 已提交
403
	spin_unlock(&pgd_lock);
404 405

	return pgd;
406 407

out_free_pmds:
408
	free_pmds(mm, pmds);
409
out_free_pgd:
410
	_pgd_free(pgd);
411 412
out:
	return NULL;
413 414 415 416 417 418
}

void pgd_free(struct mm_struct *mm, pgd_t *pgd)
{
	pgd_mop_up_pmds(mm, pgd);
	pgd_dtor(pgd);
419
	paravirt_pgd_free(mm, pgd);
420
	_pgd_free(pgd);
421
}
422

423 424 425 426 427 428 429
/*
 * Used to set accessed or dirty bits in the page table entries
 * on other architectures. On x86, the accessed and dirty bits
 * are tracked by hardware. However, do_wp_page calls this function
 * to also make the pte writeable at the same time the dirty bit is
 * set. In that case we do actually need to write the PTE.
 */
430 431 432 433 434 435
int ptep_set_access_flags(struct vm_area_struct *vma,
			  unsigned long address, pte_t *ptep,
			  pte_t entry, int dirty)
{
	int changed = !pte_same(*ptep, entry);

436
	if (changed && dirty)
437 438 439 440
		*ptep = entry;

	return changed;
}
441

442 443 444 445 446 447 448 449 450 451 452
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
int pmdp_set_access_flags(struct vm_area_struct *vma,
			  unsigned long address, pmd_t *pmdp,
			  pmd_t entry, int dirty)
{
	int changed = !pmd_same(*pmdp, entry);

	VM_BUG_ON(address & ~HPAGE_PMD_MASK);

	if (changed && dirty) {
		*pmdp = entry;
453 454 455 456 457 458
		/*
		 * We had a write-protection fault here and changed the pmd
		 * to to more permissive. No need to flush the TLB for that,
		 * #PF is architecturally guaranteed to do that and in the
		 * worst-case we'll generate a spurious fault.
		 */
459 460 461 462
	}

	return changed;
}
463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482

int pudp_set_access_flags(struct vm_area_struct *vma, unsigned long address,
			  pud_t *pudp, pud_t entry, int dirty)
{
	int changed = !pud_same(*pudp, entry);

	VM_BUG_ON(address & ~HPAGE_PUD_MASK);

	if (changed && dirty) {
		*pudp = entry;
		/*
		 * We had a write-protection fault here and changed the pud
		 * to to more permissive. No need to flush the TLB for that,
		 * #PF is architecturally guaranteed to do that and in the
		 * worst-case we'll generate a spurious fault.
		 */
	}

	return changed;
}
483 484
#endif

485 486 487 488 489 490 491
int ptep_test_and_clear_young(struct vm_area_struct *vma,
			      unsigned long addr, pte_t *ptep)
{
	int ret = 0;

	if (pte_young(*ptep))
		ret = test_and_clear_bit(_PAGE_BIT_ACCESSED,
492
					 (unsigned long *) &ptep->pte);
493 494 495

	return ret;
}
496

497 498 499 500 501 502 503 504
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
int pmdp_test_and_clear_young(struct vm_area_struct *vma,
			      unsigned long addr, pmd_t *pmdp)
{
	int ret = 0;

	if (pmd_young(*pmdp))
		ret = test_and_clear_bit(_PAGE_BIT_ACCESSED,
J
Johannes Weiner 已提交
505
					 (unsigned long *)pmdp);
506 507 508

	return ret;
}
509 510 511 512 513 514 515 516 517 518 519
int pudp_test_and_clear_young(struct vm_area_struct *vma,
			      unsigned long addr, pud_t *pudp)
{
	int ret = 0;

	if (pud_young(*pudp))
		ret = test_and_clear_bit(_PAGE_BIT_ACCESSED,
					 (unsigned long *)pudp);

	return ret;
}
520 521
#endif

522 523 524
int ptep_clear_flush_young(struct vm_area_struct *vma,
			   unsigned long address, pte_t *ptep)
{
525 526 527 528 529 530 531 532 533 534 535 536 537 538
	/*
	 * On x86 CPUs, clearing the accessed bit without a TLB flush
	 * doesn't cause data corruption. [ It could cause incorrect
	 * page aging and the (mistaken) reclaim of hot pages, but the
	 * chance of that should be relatively low. ]
	 *
	 * So as a performance optimization don't flush the TLB when
	 * clearing the accessed bit, it will eventually be flushed by
	 * a context switch or a VM operation anyway. [ In the rare
	 * event of it not getting flushed for a long time the delay
	 * shouldn't really matter because there's no real memory
	 * pressure for swapout to react to. ]
	 */
	return ptep_test_and_clear_young(vma, address, ptep);
539
}
J
Jeremy Fitzhardinge 已提交
540

541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
int pmdp_clear_flush_young(struct vm_area_struct *vma,
			   unsigned long address, pmd_t *pmdp)
{
	int young;

	VM_BUG_ON(address & ~HPAGE_PMD_MASK);

	young = pmdp_test_and_clear_young(vma, address, pmdp);
	if (young)
		flush_tlb_range(vma, address, address + HPAGE_PMD_SIZE);

	return young;
}
#endif

557 558 559 560 561 562 563 564 565 566 567
/**
 * reserve_top_address - reserves a hole in the top of kernel address space
 * @reserve - size of hole to reserve
 *
 * Can be used to relocate the fixmap area and poke a hole in the top
 * of kernel address space to make room for a hypervisor.
 */
void __init reserve_top_address(unsigned long reserve)
{
#ifdef CONFIG_X86_32
	BUG_ON(fixmaps_set > 0);
568 569 570
	__FIXADDR_TOP = round_down(-reserve, 1 << PMD_SHIFT) - PAGE_SIZE;
	printk(KERN_INFO "Reserving virtual address space above 0x%08lx (rounded to 0x%08lx)\n",
	       -reserve, __FIXADDR_TOP + PAGE_SIZE);
571 572 573
#endif
}

J
Jeremy Fitzhardinge 已提交
574 575
int fixmaps_set;

576
void __native_set_fixmap(enum fixed_addresses idx, pte_t pte)
J
Jeremy Fitzhardinge 已提交
577 578 579 580 581 582 583
{
	unsigned long address = __fix_to_virt(idx);

	if (idx >= __end_of_fixed_addresses) {
		BUG();
		return;
	}
584
	set_pte_vaddr(address, pte);
J
Jeremy Fitzhardinge 已提交
585 586
	fixmaps_set++;
}
587

588 589
void native_set_fixmap(enum fixed_addresses idx, phys_addr_t phys,
		       pgprot_t flags)
590
{
591 592 593
	/* Sanitize 'prot' against any unsupported bits: */
	pgprot_val(flags) &= __default_kernel_pte_mask;

594 595
	__native_set_fixmap(idx, pfn_pte(phys >> PAGE_SHIFT, flags));
}
596 597

#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619
#ifdef CONFIG_X86_5LEVEL
/**
 * p4d_set_huge - setup kernel P4D mapping
 *
 * No 512GB pages yet -- always return 0
 */
int p4d_set_huge(p4d_t *p4d, phys_addr_t addr, pgprot_t prot)
{
	return 0;
}

/**
 * p4d_clear_huge - clear kernel P4D mapping when it is set
 *
 * No 512GB pages yet -- always return 0
 */
int p4d_clear_huge(p4d_t *p4d)
{
	return 0;
}
#endif

620 621 622
/**
 * pud_set_huge - setup kernel PUD mapping
 *
623 624 625 626 627 628 629 630 631 632 633 634
 * MTRRs can override PAT memory types with 4KiB granularity. Therefore, this
 * function sets up a huge page only if any of the following conditions are met:
 *
 * - MTRRs are disabled, or
 *
 * - MTRRs are enabled and the range is completely covered by a single MTRR, or
 *
 * - MTRRs are enabled and the corresponding MTRR memory type is WB, which
 *   has no effect on the requested PAT memory type.
 *
 * Callers should try to decrease page size (1GB -> 2MB -> 4K) if the bigger
 * page mapping attempt fails.
635 636 637
 *
 * Returns 1 on success and 0 on failure.
 */
638 639
int pud_set_huge(pud_t *pud, phys_addr_t addr, pgprot_t prot)
{
640
	u8 mtrr, uniform;
641

642 643 644
	mtrr = mtrr_type_lookup(addr, addr + PUD_SIZE, &uniform);
	if ((mtrr != MTRR_TYPE_INVALID) && (!uniform) &&
	    (mtrr != MTRR_TYPE_WRBACK))
645 646
		return 0;

647 648 649 650
	/* Bail out if we are we on a populated non-leaf entry: */
	if (pud_present(*pud) && !pud_huge(*pud))
		return 0;

651 652 653 654 655 656 657 658 659
	prot = pgprot_4k_2_large(prot);

	set_pte((pte_t *)pud, pfn_pte(
		(u64)addr >> PAGE_SHIFT,
		__pgprot(pgprot_val(prot) | _PAGE_PSE)));

	return 1;
}

660 661 662
/**
 * pmd_set_huge - setup kernel PMD mapping
 *
663
 * See text over pud_set_huge() above.
664 665 666
 *
 * Returns 1 on success and 0 on failure.
 */
667 668
int pmd_set_huge(pmd_t *pmd, phys_addr_t addr, pgprot_t prot)
{
669
	u8 mtrr, uniform;
670

671 672 673 674 675
	mtrr = mtrr_type_lookup(addr, addr + PMD_SIZE, &uniform);
	if ((mtrr != MTRR_TYPE_INVALID) && (!uniform) &&
	    (mtrr != MTRR_TYPE_WRBACK)) {
		pr_warn_once("%s: Cannot satisfy [mem %#010llx-%#010llx] with a huge-page mapping due to MTRR override.\n",
			     __func__, addr, addr + PMD_SIZE);
676
		return 0;
677
	}
678

679 680 681 682
	/* Bail out if we are we on a populated non-leaf entry: */
	if (pmd_present(*pmd) && !pmd_huge(*pmd))
		return 0;

683 684 685 686 687 688 689 690 691
	prot = pgprot_4k_2_large(prot);

	set_pte((pte_t *)pmd, pfn_pte(
		(u64)addr >> PAGE_SHIFT,
		__pgprot(pgprot_val(prot) | _PAGE_PSE)));

	return 1;
}

692 693 694 695 696
/**
 * pud_clear_huge - clear kernel PUD mapping when it is set
 *
 * Returns 1 on success and 0 on failure (no PUD map is found).
 */
697 698 699 700 701 702 703 704 705 706
int pud_clear_huge(pud_t *pud)
{
	if (pud_large(*pud)) {
		pud_clear(pud);
		return 1;
	}

	return 0;
}

707 708 709 710 711
/**
 * pmd_clear_huge - clear kernel PMD mapping when it is set
 *
 * Returns 1 on success and 0 on failure (no PMD map is found).
 */
712 713 714 715 716 717 718 719 720
int pmd_clear_huge(pmd_t *pmd)
{
	if (pmd_large(*pmd)) {
		pmd_clear(pmd);
		return 1;
	}

	return 0;
}
721

722
#ifdef CONFIG_X86_64
723 724 725
/**
 * pud_free_pmd_page - Clear pud entry and free pmd page.
 * @pud: Pointer to a PUD.
726
 * @addr: Virtual address associated with pud.
727 728 729 730
 *
 * Context: The pud range has been unmaped and TLB purged.
 * Return: 1 if clearing the entry succeeded. 0 otherwise.
 */
731
int pud_free_pmd_page(pud_t *pud, unsigned long addr)
732
{
733 734 735 736 737 738 739 740 741
	pmd_t *pmd;
	int i;

	if (pud_none(*pud))
		return 1;

	pmd = (pmd_t *)pud_page_vaddr(*pud);

	for (i = 0; i < PTRS_PER_PMD; i++)
742
		if (!pmd_free_pte_page(&pmd[i], addr + (i * PMD_SIZE)))
743 744 745 746 747 748
			return 0;

	pud_clear(pud);
	free_page((unsigned long)pmd);

	return 1;
749 750 751 752 753
}

/**
 * pmd_free_pte_page - Clear pmd entry and free pte page.
 * @pmd: Pointer to a PMD.
754
 * @addr: Virtual address associated with pmd.
755 756 757 758
 *
 * Context: The pmd range has been unmaped and TLB purged.
 * Return: 1 if clearing the entry succeeded. 0 otherwise.
 */
759
int pmd_free_pte_page(pmd_t *pmd, unsigned long addr)
760
{
761 762 763 764 765 766 767 768 769 770
	pte_t *pte;

	if (pmd_none(*pmd))
		return 1;

	pte = (pte_t *)pmd_page_vaddr(*pmd);
	pmd_clear(pmd);
	free_page((unsigned long)pte);

	return 1;
771
}
772 773 774

#else /* !CONFIG_X86_64 */

775
int pud_free_pmd_page(pud_t *pud, unsigned long addr)
776 777 778 779 780 781 782 783
{
	return pud_none(*pud);
}

/*
 * Disable free page handling on x86-PAE. This assures that ioremap()
 * does not update sync'd pmd entries. See vmalloc_sync_one().
 */
784
int pmd_free_pte_page(pmd_t *pmd, unsigned long addr)
785 786 787 788 789
{
	return pmd_none(*pmd);
}

#endif /* CONFIG_X86_64 */
790
#endif	/* CONFIG_HAVE_ARCH_HUGE_VMAP */