[mirror_ubuntu-bionic-kernel.git] / mm / mprotect.c

/*
 *  mm/mprotect.c
 *
 *  (C) Copyright 1994 Linus Torvalds
 *  (C) Copyright 2002 Christoph Hellwig
 *
 *  Address space accounting code	<alan@lxorguk.ukuu.org.uk>
 *  (C) Copyright 2002 Red Hat Inc, All Rights Reserved
 */

#include <linux/mm.h>
#include <linux/hugetlb.h>
#include <linux/shm.h>
#include <linux/mman.h>
#include <linux/fs.h>
#include <linux/highmem.h>
#include <linux/security.h>
#include <linux/mempolicy.h>
#include <linux/personality.h>
#include <linux/syscalls.h>
#include <linux/swap.h>
#include <linux/swapops.h>
#include <linux/mmu_notifier.h>
#include <linux/migrate.h>
#include <linux/perf_event.h>
#include <linux/ksm.h>
#include <asm/uaccess.h>
#include <asm/pgtable.h>
#include <asm/cacheflush.h>
#include <asm/tlbflush.h>

#ifndef pgprot_modify
static inline pgprot_t pgprot_modify(pgprot_t oldprot, pgprot_t newprot)
{
	return newprot;
}
#endif

static unsigned long change_pte_range(struct vm_area_struct *vma, pmd_t *pmd,
		unsigned long addr, unsigned long end, pgprot_t newprot,
		int dirty_accountable, int prot_numa)
{
	struct mm_struct *mm = vma->vm_mm;
	pte_t *pte, oldpte;
	spinlock_t *ptl;
	unsigned long pages = 0;

	pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
	arch_enter_lazy_mmu_mode();
	do {
		oldpte = *pte;
		if (pte_present(oldpte)) {
			pte_t ptent;
			bool updated = false;

			if (!prot_numa) {
				ptent = ptep_modify_prot_start(mm, addr, pte);
				if (pte_numa(ptent))
					ptent = pte_mknonnuma(ptent);
				ptent = pte_modify(ptent, newprot);
				/*
				 * Avoid taking write faults for pages we
				 * know to be dirty.
				 */
				if (dirty_accountable && pte_dirty(ptent))
					ptent = pte_mkwrite(ptent);
				ptep_modify_prot_commit(mm, addr, pte, ptent);
				updated = true;
			} else {
				struct page *page;

				page = vm_normal_page(vma, addr, oldpte);
				if (page && !PageKsm(page)) {
					if (!pte_numa(oldpte)) {
						ptep_set_numa(mm, addr, pte);
						updated = true;
					}
				}
			}
			if (updated)
				pages++;
		} else if (IS_ENABLED(CONFIG_MIGRATION) && !pte_file(oldpte)) {
			swp_entry_t entry = pte_to_swp_entry(oldpte);

			if (is_write_migration_entry(entry)) {
				pte_t newpte;
				/*
				 * A protection check is difficult so
				 * just be safe and disable write
				 */
				make_migration_entry_read(&entry);
				newpte = swp_entry_to_pte(entry);
				if (pte_swp_soft_dirty(oldpte))
					newpte = pte_swp_mksoft_dirty(newpte);
				set_pte_at(mm, addr, pte, newpte);

				pages++;
			}
		}
	} while (pte++, addr += PAGE_SIZE, addr != end);
	arch_leave_lazy_mmu_mode();
	pte_unmap_unlock(pte - 1, ptl);

	return pages;
}

static inline unsigned long change_pmd_range(struct vm_area_struct *vma,
		pud_t *pud, unsigned long addr, unsigned long end,
		pgprot_t newprot, int dirty_accountable, int prot_numa)
{
	pmd_t *pmd;
	unsigned long next;
	unsigned long pages = 0;
	unsigned long nr_huge_updates = 0;

	pmd = pmd_offset(pud, addr);
	do {
		unsigned long this_pages;

		next = pmd_addr_end(addr, end);
		if (pmd_trans_huge(*pmd)) {
			if (next - addr != HPAGE_PMD_SIZE)
				split_huge_page_pmd(vma, addr, pmd);
			else {
				int nr_ptes = change_huge_pmd(vma, pmd, addr,
						newprot, prot_numa);

				if (nr_ptes) {
					if (nr_ptes == HPAGE_PMD_NR) {
						pages += HPAGE_PMD_NR;
						nr_huge_updates++;
					}
					continue;
				}
			}
			/* fall through */
		}
		if (pmd_none_or_clear_bad(pmd))
			continue;
		this_pages = change_pte_range(vma, pmd, addr, next, newprot,
				 dirty_accountable, prot_numa);
		pages += this_pages;
	} while (pmd++, addr = next, addr != end);

	if (nr_huge_updates)
		count_vm_numa_events(NUMA_HUGE_PTE_UPDATES, nr_huge_updates);
	return pages;
}

static inline unsigned long change_pud_range(struct vm_area_struct *vma,
		pgd_t *pgd, unsigned long addr, unsigned long end,
		pgprot_t newprot, int dirty_accountable, int prot_numa)
{
	pud_t *pud;
	unsigned long next;
	unsigned long pages = 0;

	pud = pud_offset(pgd, addr);
	do {
		next = pud_addr_end(addr, end);
		if (pud_none_or_clear_bad(pud))
			continue;
		pages += change_pmd_range(vma, pud, addr, next, newprot,
				 dirty_accountable, prot_numa);
	} while (pud++, addr = next, addr != end);

	return pages;
}

static unsigned long change_protection_range(struct vm_area_struct *vma,
		unsigned long addr, unsigned long end, pgprot_t newprot,
		int dirty_accountable, int prot_numa)
{
	struct mm_struct *mm = vma->vm_mm;
	pgd_t *pgd;
	unsigned long next;
	unsigned long start = addr;
	unsigned long pages = 0;

	BUG_ON(addr >= end);
	pgd = pgd_offset(mm, addr);
	flush_cache_range(vma, addr, end);
	set_tlb_flush_pending(mm);
	do {
		next = pgd_addr_end(addr, end);
		if (pgd_none_or_clear_bad(pgd))
			continue;
		pages += change_pud_range(vma, pgd, addr, next, newprot,
				 dirty_accountable, prot_numa);
	} while (pgd++, addr = next, addr != end);

	/* Only flush the TLB if we actually modified any entries: */
	if (pages)
		flush_tlb_range(vma, start, end);
	clear_tlb_flush_pending(mm);

	return pages;
}

unsigned long change_protection(struct vm_area_struct *vma, unsigned long start,
		       unsigned long end, pgprot_t newprot,
		       int dirty_accountable, int prot_numa)
{
	struct mm_struct *mm = vma->vm_mm;
	unsigned long pages;

	mmu_notifier_invalidate_range_start(mm, start, end);
	if (is_vm_hugetlb_page(vma))
		pages = hugetlb_change_protection(vma, start, end, newprot);
	else
		pages = change_protection_range(vma, start, end, newprot, dirty_accountable, prot_numa);
	mmu_notifier_invalidate_range_end(mm, start, end);

	return pages;
}

int
mprotect_fixup(struct vm_area_struct *vma, struct vm_area_struct **pprev,
	unsigned long start, unsigned long end, unsigned long newflags)
{
	struct mm_struct *mm = vma->vm_mm;
	unsigned long oldflags = vma->vm_flags;
	long nrpages = (end - start) >> PAGE_SHIFT;
	unsigned long charged = 0;
	pgoff_t pgoff;
	int error;
	int dirty_accountable = 0;

	if (newflags == oldflags) {
		*pprev = vma;
		return 0;
	}

	/*
	 * If we make a private mapping writable we increase our commit;
	 * but (without finer accounting) cannot reduce our commit if we
	 * make it unwritable again. hugetlb mapping were accounted for
	 * even if read-only so there is no need to account for them here
	 */
	if (newflags & VM_WRITE) {
		if (!(oldflags & (VM_ACCOUNT|VM_WRITE|VM_HUGETLB|
						VM_SHARED|VM_NORESERVE))) {
			charged = nrpages;
			if (security_vm_enough_memory_mm(mm, charged))
				return -ENOMEM;
			newflags |= VM_ACCOUNT;
		}
	}

	/*
	 * First try to merge with previous and/or next vma.
	 */
	pgoff = vma->vm_pgoff + ((start - vma->vm_start) >> PAGE_SHIFT);
	*pprev = vma_merge(mm, *pprev, start, end, newflags,
			vma->anon_vma, vma->vm_file, pgoff, vma_policy(vma));
	if (*pprev) {
		vma = *pprev;
		goto success;
	}

	*pprev = vma;

	if (start != vma->vm_start) {
		error = split_vma(mm, vma, start, 1);
		if (error)
			goto fail;
	}

	if (end != vma->vm_end) {
		error = split_vma(mm, vma, end, 0);
		if (error)
			goto fail;
	}

success:
	/*
	 * vm_flags and vm_page_prot are protected by the mmap_sem
	 * held in write mode.
	 */
	vma->vm_flags = newflags;
	vma->vm_page_prot = pgprot_modify(vma->vm_page_prot,
					  vm_get_page_prot(newflags));

	if (vma_wants_writenotify(vma)) {
		vma->vm_page_prot = vm_get_page_prot(newflags & ~VM_SHARED);
		dirty_accountable = 1;
	}

	change_protection(vma, start, end, vma->vm_page_prot,
			  dirty_accountable, 0);

	vm_stat_account(mm, oldflags, vma->vm_file, -nrpages);
	vm_stat_account(mm, newflags, vma->vm_file, nrpages);
	perf_event_mmap(vma);
	return 0;

fail:
	vm_unacct_memory(charged);
	return error;
}

SYSCALL_DEFINE3(mprotect, unsigned long, start, size_t, len,
		unsigned long, prot)
{
	unsigned long vm_flags, nstart, end, tmp, reqprot;
	struct vm_area_struct *vma, *prev;
	int error = -EINVAL;
	const int grows = prot & (PROT_GROWSDOWN|PROT_GROWSUP);
	prot &= ~(PROT_GROWSDOWN|PROT_GROWSUP);
	if (grows == (PROT_GROWSDOWN|PROT_GROWSUP)) /* can't be both */
		return -EINVAL;

	if (start & ~PAGE_MASK)
		return -EINVAL;
	if (!len)
		return 0;
	len = PAGE_ALIGN(len);
	end = start + len;
	if (end <= start)
		return -ENOMEM;
	if (!arch_validate_prot(prot))
		return -EINVAL;

	reqprot = prot;
	/*
	 * Does the application expect PROT_READ to imply PROT_EXEC:
	 */
	if ((prot & PROT_READ) && (current->personality & READ_IMPLIES_EXEC))
		prot |= PROT_EXEC;

	vm_flags = calc_vm_prot_bits(prot);

	down_write(&current->mm->mmap_sem);

	vma = find_vma(current->mm, start);
	error = -ENOMEM;
	if (!vma)
		goto out;
	prev = vma->vm_prev;
	if (unlikely(grows & PROT_GROWSDOWN)) {
		if (vma->vm_start >= end)
			goto out;
		start = vma->vm_start;
		error = -EINVAL;
		if (!(vma->vm_flags & VM_GROWSDOWN))
			goto out;
	} else {
		if (vma->vm_start > start)
			goto out;
		if (unlikely(grows & PROT_GROWSUP)) {
			end = vma->vm_end;
			error = -EINVAL;
			if (!(vma->vm_flags & VM_GROWSUP))
				goto out;
		}
	}
	if (start > vma->vm_start)
		prev = vma;

	for (nstart = start ; ; ) {
		unsigned long newflags;

		/* Here we know that vma->vm_start <= nstart < vma->vm_end. */

		newflags = vm_flags;
		newflags |= (vma->vm_flags & ~(VM_READ | VM_WRITE | VM_EXEC));

		/* newflags >> 4 shift VM_MAY% in place of VM_% */
		if ((newflags & ~(newflags >> 4)) & (VM_READ | VM_WRITE | VM_EXEC)) {
			error = -EACCES;
			goto out;
		}

		error = security_file_mprotect(vma, reqprot, prot);
		if (error)
			goto out;

		tmp = vma->vm_end;
		if (tmp > end)
			tmp = end;
		error = mprotect_fixup(vma, &prev, nstart, tmp, newflags);
		if (error)
			goto out;
		nstart = tmp;

		if (nstart < prev->vm_end)
			nstart = prev->vm_end;
		if (nstart >= end)
			goto out;

		vma = prev->vm_next;
		if (!vma || vma->vm_start != nstart) {
			error = -ENOMEM;
			goto out;
		}
	}
out:
	up_write(&current->mm->mmap_sem);
	return error;
}
Commit	Line	Data
1da177e4 LT	1	/*
	2	* mm/mprotect.c
	3	*
	4	* (C) Copyright 1994 Linus Torvalds
	5	* (C) Copyright 2002 Christoph Hellwig
	6	*
046c6884	7	* Address space accounting code <alan@lxorguk.ukuu.org.uk>
1da177e4 LT	8	* (C) Copyright 2002 Red Hat Inc, All Rights Reserved
	9	*/
	10
	11	#include <linux/mm.h>
	12	#include <linux/hugetlb.h>
1da177e4 LT	13	#include <linux/shm.h>
	14	#include <linux/mman.h>
	15	#include <linux/fs.h>
	16	#include <linux/highmem.h>
	17	#include <linux/security.h>
	18	#include <linux/mempolicy.h>
	19	#include <linux/personality.h>
	20	#include <linux/syscalls.h>
0697212a CL	21	#include <linux/swap.h>
0697212a CL	22	#include <linux/swapops.h>
cddb8a5c	23	#include <linux/mmu_notifier.h>
64cdd548	24	#include <linux/migrate.h>
cdd6c482	25	#include <linux/perf_event.h>
64a9a34e	26	#include <linux/ksm.h>
1da177e4 LT	27	#include <asm/uaccess.h>
	28	#include <asm/pgtable.h>
	29	#include <asm/cacheflush.h>
	30	#include <asm/tlbflush.h>
	31
1c12c4cf VP	32	#ifndef pgprot_modify
	33	static inline pgprot_t pgprot_modify(pgprot_t oldprot, pgprot_t newprot)
	34	{
	35	return newprot;
	36	}
	37	#endif
	38
4b10e7d5	39	static unsigned long change_pte_range(struct vm_area_struct vma, pmd_t pmd,
c1e6098b	40	unsigned long addr, unsigned long end, pgprot_t newprot,
0f19c179	41	int dirty_accountable, int prot_numa)
1da177e4	42	{
4b10e7d5	43	struct mm_struct *mm = vma->vm_mm;
0697212a	44	pte_t *pte, oldpte;
705e87c0	45	spinlock_t *ptl;
7da4d641	46	unsigned long pages = 0;
1da177e4	47
705e87c0	48	pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
6606c3e0	49	arch_enter_lazy_mmu_mode();
1da177e4	50	do {
0697212a CL	51	oldpte = *pte;
0697212a CL	52	if (pte_present(oldpte)) {
1da177e4	53	pte_t ptent;
4b10e7d5	54	bool updated = false;
1da177e4	55
4b10e7d5	56	if (!prot_numa) {
0c5f83c2	57	ptent = ptep_modify_prot_start(mm, addr, pte);
1667918b MG	58	if (pte_numa(ptent))
1667918b MG	59	ptent = pte_mknonnuma(ptent);
4b10e7d5	60	ptent = pte_modify(ptent, newprot);
9d85d586 AK	61	/*
	62	* Avoid taking write faults for pages we
	63	* know to be dirty.
	64	*/
	65	if (dirty_accountable && pte_dirty(ptent))
	66	ptent = pte_mkwrite(ptent);
	67	ptep_modify_prot_commit(mm, addr, pte, ptent);
4b10e7d5 MG	68	updated = true;
	69	} else {
	70	struct page *page;
	71
	72	page = vm_normal_page(vma, addr, oldpte);
64a9a34e	73	if (page && !PageKsm(page)) {
1bc115d8	74	if (!pte_numa(oldpte)) {
56eecdb9	75	ptep_set_numa(mm, addr, pte);
4b10e7d5 MG	76	updated = true;
	77	}
	78	}
	79	}
4b10e7d5 MG	80	if (updated)
4b10e7d5 MG	81	pages++;
ce1744f4	82	} else if (IS_ENABLED(CONFIG_MIGRATION) && !pte_file(oldpte)) {
0697212a CL	83	swp_entry_t entry = pte_to_swp_entry(oldpte);
	84
	85	if (is_write_migration_entry(entry)) {
c3d16e16	86	pte_t newpte;
0697212a CL	87	/*
	88	* A protection check is difficult so
	89	* just be safe and disable write
	90	*/
	91	make_migration_entry_read(&entry);
c3d16e16 CG	92	newpte = swp_entry_to_pte(entry);
	93	if (pte_swp_soft_dirty(oldpte))
	94	newpte = pte_swp_mksoft_dirty(newpte);
	95	set_pte_at(mm, addr, pte, newpte);
e920e14c MG	96
e920e14c MG	97	pages++;
0697212a	98	}
1da177e4 LT	99	}
1da177e4 LT	100	} while (pte++, addr += PAGE_SIZE, addr != end);
6606c3e0	101	arch_leave_lazy_mmu_mode();
705e87c0	102	pte_unmap_unlock(pte - 1, ptl);
7da4d641 PZ	103
7da4d641 PZ	104	return pages;
1da177e4 LT	105	}
1da177e4 LT	106
7d12efae AM	107	static inline unsigned long change_pmd_range(struct vm_area_struct *vma,
	108	pud_t *pud, unsigned long addr, unsigned long end,
	109	pgprot_t newprot, int dirty_accountable, int prot_numa)
1da177e4 LT	110	{
	111	pmd_t *pmd;
	112	unsigned long next;
7da4d641	113	unsigned long pages = 0;
72403b4a	114	unsigned long nr_huge_updates = 0;
1da177e4 LT	115
	116	pmd = pmd_offset(pud, addr);
	117	do {
25cbbef1 MG	118	unsigned long this_pages;
25cbbef1 MG	119
1da177e4	120	next = pmd_addr_end(addr, end);
cd7548ab JW	121	if (pmd_trans_huge(*pmd)) {
cd7548ab JW	122	if (next - addr != HPAGE_PMD_SIZE)
e180377f	123	split_huge_page_pmd(vma, addr, pmd);
f123d74a MG	124	else {
	125	int nr_ptes = change_huge_pmd(vma, pmd, addr,
	126	newprot, prot_numa);
	127
	128	if (nr_ptes) {
72403b4a MG	129	if (nr_ptes == HPAGE_PMD_NR) {
	130	pages += HPAGE_PMD_NR;
	131	nr_huge_updates++;
	132	}
f123d74a MG	133	continue;
f123d74a MG	134	}
7da4d641	135	}
cd7548ab JW	136	/* fall through */
cd7548ab JW	137	}
1da177e4 LT	138	if (pmd_none_or_clear_bad(pmd))
1da177e4 LT	139	continue;
25cbbef1	140	this_pages = change_pte_range(vma, pmd, addr, next, newprot,
0f19c179	141	dirty_accountable, prot_numa);
25cbbef1	142	pages += this_pages;
1da177e4	143	} while (pmd++, addr = next, addr != end);
7da4d641	144
72403b4a MG	145	if (nr_huge_updates)
72403b4a MG	146	count_vm_numa_events(NUMA_HUGE_PTE_UPDATES, nr_huge_updates);
7da4d641	147	return pages;
1da177e4 LT	148	}
1da177e4 LT	149
7d12efae AM	150	static inline unsigned long change_pud_range(struct vm_area_struct *vma,
	151	pgd_t *pgd, unsigned long addr, unsigned long end,
	152	pgprot_t newprot, int dirty_accountable, int prot_numa)
1da177e4 LT	153	{
	154	pud_t *pud;
	155	unsigned long next;
7da4d641	156	unsigned long pages = 0;
1da177e4 LT	157
	158	pud = pud_offset(pgd, addr);
	159	do {
	160	next = pud_addr_end(addr, end);
	161	if (pud_none_or_clear_bad(pud))
	162	continue;
7da4d641	163	pages += change_pmd_range(vma, pud, addr, next, newprot,
4b10e7d5	164	dirty_accountable, prot_numa);
1da177e4	165	} while (pud++, addr = next, addr != end);
7da4d641 PZ	166
7da4d641 PZ	167	return pages;
1da177e4 LT	168	}
1da177e4 LT	169
7da4d641	170	static unsigned long change_protection_range(struct vm_area_struct *vma,
c1e6098b	171	unsigned long addr, unsigned long end, pgprot_t newprot,
4b10e7d5	172	int dirty_accountable, int prot_numa)
1da177e4 LT	173	{
	174	struct mm_struct *mm = vma->vm_mm;
	175	pgd_t *pgd;
	176	unsigned long next;
	177	unsigned long start = addr;
7da4d641	178	unsigned long pages = 0;
1da177e4 LT	179
	180	BUG_ON(addr >= end);
	181	pgd = pgd_offset(mm, addr);
	182	flush_cache_range(vma, addr, end);
20841405	183	set_tlb_flush_pending(mm);
1da177e4 LT	184	do {
	185	next = pgd_addr_end(addr, end);
	186	if (pgd_none_or_clear_bad(pgd))
	187	continue;
7da4d641	188	pages += change_pud_range(vma, pgd, addr, next, newprot,
4b10e7d5	189	dirty_accountable, prot_numa);
1da177e4	190	} while (pgd++, addr = next, addr != end);
7da4d641	191
1233d588 IM	192	/* Only flush the TLB if we actually modified any entries: */
	193	if (pages)
	194	flush_tlb_range(vma, start, end);
20841405	195	clear_tlb_flush_pending(mm);
7da4d641 PZ	196
	197	return pages;
	198	}
	199
	200	unsigned long change_protection(struct vm_area_struct *vma, unsigned long start,
	201	unsigned long end, pgprot_t newprot,
4b10e7d5	202	int dirty_accountable, int prot_numa)
7da4d641 PZ	203	{
	204	struct mm_struct *mm = vma->vm_mm;
	205	unsigned long pages;
	206
	207	mmu_notifier_invalidate_range_start(mm, start, end);
	208	if (is_vm_hugetlb_page(vma))
	209	pages = hugetlb_change_protection(vma, start, end, newprot);
	210	else
4b10e7d5	211	pages = change_protection_range(vma, start, end, newprot, dirty_accountable, prot_numa);
7da4d641 PZ	212	mmu_notifier_invalidate_range_end(mm, start, end);
	213
	214	return pages;
1da177e4 LT	215	}
1da177e4 LT	216
b6a2fea3	217	int
1da177e4 LT	218	mprotect_fixup(struct vm_area_struct vma, struct vm_area_struct *pprev,
	219	unsigned long start, unsigned long end, unsigned long newflags)
	220	{
	221	struct mm_struct *mm = vma->vm_mm;
	222	unsigned long oldflags = vma->vm_flags;
	223	long nrpages = (end - start) >> PAGE_SHIFT;
	224	unsigned long charged = 0;
1da177e4 LT	225	pgoff_t pgoff;
1da177e4 LT	226	int error;
c1e6098b	227	int dirty_accountable = 0;
1da177e4 LT	228
	229	if (newflags == oldflags) {
	230	*pprev = vma;
	231	return 0;
	232	}
	233
	234	/*
	235	* If we make a private mapping writable we increase our commit;
	236	* but (without finer accounting) cannot reduce our commit if we
5a6fe125 MG	237	* make it unwritable again. hugetlb mapping were accounted for
5a6fe125 MG	238	* even if read-only so there is no need to account for them here
1da177e4 LT	239	*/
1da177e4 LT	240	if (newflags & VM_WRITE) {
5a6fe125	241	if (!(oldflags & (VM_ACCOUNT\|VM_WRITE\|VM_HUGETLB\|
cdfd4325	242	VM_SHARED\|VM_NORESERVE))) {
1da177e4	243	charged = nrpages;
191c5424	244	if (security_vm_enough_memory_mm(mm, charged))
1da177e4 LT	245	return -ENOMEM;
	246	newflags \|= VM_ACCOUNT;
	247	}
	248	}
	249
1da177e4 LT	250	/*
	251	* First try to merge with previous and/or next vma.
	252	*/
	253	pgoff = vma->vm_pgoff + ((start - vma->vm_start) >> PAGE_SHIFT);
	254	pprev = vma_merge(mm, pprev, start, end, newflags,
	255	vma->anon_vma, vma->vm_file, pgoff, vma_policy(vma));
	256	if (*pprev) {
	257	vma = *pprev;
	258	goto success;
	259	}
	260
	261	*pprev = vma;
	262
	263	if (start != vma->vm_start) {
	264	error = split_vma(mm, vma, start, 1);
	265	if (error)
	266	goto fail;
	267	}
	268
	269	if (end != vma->vm_end) {
	270	error = split_vma(mm, vma, end, 0);
	271	if (error)
	272	goto fail;
	273	}
	274
	275	success:
	276	/*
	277	* vm_flags and vm_page_prot are protected by the mmap_sem
	278	* held in write mode.
	279	*/
	280	vma->vm_flags = newflags;
1c12c4cf VP	281	vma->vm_page_prot = pgprot_modify(vma->vm_page_prot,
	282	vm_get_page_prot(newflags));
	283
c1e6098b	284	if (vma_wants_writenotify(vma)) {
1ddd439e	285	vma->vm_page_prot = vm_get_page_prot(newflags & ~VM_SHARED);
c1e6098b PZ	286	dirty_accountable = 1;
c1e6098b PZ	287	}
d08b3851	288
7d12efae AM	289	change_protection(vma, start, end, vma->vm_page_prot,
7d12efae AM	290	dirty_accountable, 0);
7da4d641	291
ab50b8ed HD	292	vm_stat_account(mm, oldflags, vma->vm_file, -nrpages);
ab50b8ed HD	293	vm_stat_account(mm, newflags, vma->vm_file, nrpages);
63bfd738	294	perf_event_mmap(vma);
1da177e4 LT	295	return 0;
	296
	297	fail:
	298	vm_unacct_memory(charged);
	299	return error;
	300	}
	301
6a6160a7 HC	302	SYSCALL_DEFINE3(mprotect, unsigned long, start, size_t, len,
6a6160a7 HC	303	unsigned long, prot)
1da177e4 LT	304	{
	305	unsigned long vm_flags, nstart, end, tmp, reqprot;
	306	struct vm_area_struct vma, prev;
	307	int error = -EINVAL;
	308	const int grows = prot & (PROT_GROWSDOWN\|PROT_GROWSUP);
	309	prot &= ~(PROT_GROWSDOWN\|PROT_GROWSUP);
	310	if (grows == (PROT_GROWSDOWN\|PROT_GROWSUP)) /* can't be both */
	311	return -EINVAL;
	312
	313	if (start & ~PAGE_MASK)
	314	return -EINVAL;
	315	if (!len)
	316	return 0;
	317	len = PAGE_ALIGN(len);
	318	end = start + len;
	319	if (end <= start)
	320	return -ENOMEM;
b845f313	321	if (!arch_validate_prot(prot))
1da177e4 LT	322	return -EINVAL;
	323
	324	reqprot = prot;
	325	/*
	326	* Does the application expect PROT_READ to imply PROT_EXEC:
	327	*/
b344e05c	328	if ((prot & PROT_READ) && (current->personality & READ_IMPLIES_EXEC))
1da177e4 LT	329	prot \|= PROT_EXEC;
	330
	331	vm_flags = calc_vm_prot_bits(prot);
	332
	333	down_write(&current->mm->mmap_sem);
	334
097d5910	335	vma = find_vma(current->mm, start);
1da177e4 LT	336	error = -ENOMEM;
	337	if (!vma)
	338	goto out;
097d5910	339	prev = vma->vm_prev;
1da177e4 LT	340	if (unlikely(grows & PROT_GROWSDOWN)) {
	341	if (vma->vm_start >= end)
	342	goto out;
	343	start = vma->vm_start;
	344	error = -EINVAL;
	345	if (!(vma->vm_flags & VM_GROWSDOWN))
	346	goto out;
7d12efae	347	} else {
1da177e4 LT	348	if (vma->vm_start > start)
	349	goto out;
	350	if (unlikely(grows & PROT_GROWSUP)) {
	351	end = vma->vm_end;
	352	error = -EINVAL;
	353	if (!(vma->vm_flags & VM_GROWSUP))
	354	goto out;
	355	}
	356	}
	357	if (start > vma->vm_start)
	358	prev = vma;
	359
	360	for (nstart = start ; ; ) {
	361	unsigned long newflags;
	362
7d12efae	363	/* Here we know that vma->vm_start <= nstart < vma->vm_end. */
1da177e4	364
7d12efae AM	365	newflags = vm_flags;
7d12efae AM	366	newflags \|= (vma->vm_flags & ~(VM_READ \| VM_WRITE \| VM_EXEC));
1da177e4	367
7e2cff42 PBG	368	/* newflags >> 4 shift VM_MAY% in place of VM_% */
7e2cff42 PBG	369	if ((newflags & ~(newflags >> 4)) & (VM_READ \| VM_WRITE \| VM_EXEC)) {
1da177e4 LT	370	error = -EACCES;
	371	goto out;
	372	}
	373
	374	error = security_file_mprotect(vma, reqprot, prot);
	375	if (error)
	376	goto out;
	377
	378	tmp = vma->vm_end;
	379	if (tmp > end)
	380	tmp = end;
	381	error = mprotect_fixup(vma, &prev, nstart, tmp, newflags);
	382	if (error)
	383	goto out;
	384	nstart = tmp;
	385
	386	if (nstart < prev->vm_end)
	387	nstart = prev->vm_end;
	388	if (nstart >= end)
	389	goto out;
	390
	391	vma = prev->vm_next;
	392	if (!vma \|\| vma->vm_start != nstart) {
	393	error = -ENOMEM;
	394	goto out;
	395	}
	396	}
	397	out:
	398	up_write(&current->mm->mmap_sem);
	399	return error;
	400	}