[linux-2.6-block.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 <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);
				ptent = pte_modify(ptent, newprot);
				updated = true;
			} else {
				struct page *page;

				ptent = *pte;
				page = vm_normal_page(vma, addr, oldpte);
				if (page) {
					if (!pte_numa(oldpte)) {
						ptent = pte_mknuma(ptent);
						set_pte_at(mm, addr, pte, ptent);
						updated = true;
					}
				}
			}

			/*
			 * Avoid taking write faults for pages we know to be
			 * dirty.
			 */
			if (dirty_accountable && pte_dirty(ptent)) {
				ptent = pte_mkwrite(ptent);
				updated = true;
			}

			if (updated)
				pages++;

			/* Only !prot_numa always clears the pte */
			if (!prot_numa)
				ptep_modify_prot_commit(mm, addr, pte, ptent);
		} 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);
	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);

	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>
1da177e4 LT	26	#include <asm/uaccess.h>
	27	#include <asm/pgtable.h>
	28	#include <asm/cacheflush.h>
	29	#include <asm/tlbflush.h>
	30
1c12c4cf VP	31	#ifndef pgprot_modify
	32	static inline pgprot_t pgprot_modify(pgprot_t oldprot, pgprot_t newprot)
	33	{
	34	return newprot;
	35	}
	36	#endif
	37
4b10e7d5	38	static unsigned long change_pte_range(struct vm_area_struct vma, pmd_t pmd,
c1e6098b	39	unsigned long addr, unsigned long end, pgprot_t newprot,
0f19c179	40	int dirty_accountable, int prot_numa)
1da177e4	41	{
4b10e7d5	42	struct mm_struct *mm = vma->vm_mm;
0697212a	43	pte_t *pte, oldpte;
705e87c0	44	spinlock_t *ptl;
7da4d641	45	unsigned long pages = 0;
1da177e4	46
705e87c0	47	pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
6606c3e0	48	arch_enter_lazy_mmu_mode();
1da177e4	49	do {
0697212a CL	50	oldpte = *pte;
0697212a CL	51	if (pte_present(oldpte)) {
1da177e4	52	pte_t ptent;
4b10e7d5	53	bool updated = false;
1da177e4	54
4b10e7d5	55	if (!prot_numa) {
0c5f83c2	56	ptent = ptep_modify_prot_start(mm, addr, pte);
4b10e7d5 MG	57	ptent = pte_modify(ptent, newprot);
	58	updated = true;
	59	} else {
	60	struct page *page;
	61
0c5f83c2	62	ptent = *pte;
4b10e7d5 MG	63	page = vm_normal_page(vma, addr, oldpte);
4b10e7d5 MG	64	if (page) {
1bc115d8	65	if (!pte_numa(oldpte)) {
4b10e7d5	66	ptent = pte_mknuma(ptent);
0c5f83c2	67	set_pte_at(mm, addr, pte, ptent);
4b10e7d5 MG	68	updated = true;
	69	}
	70	}
	71	}
1ea0704e	72
c1e6098b PZ	73	/*
	74	* Avoid taking write faults for pages we know to be
	75	* dirty.
	76	*/
4b10e7d5	77	if (dirty_accountable && pte_dirty(ptent)) {
c1e6098b	78	ptent = pte_mkwrite(ptent);
4b10e7d5 MG	79	updated = true;
4b10e7d5 MG	80	}
1ea0704e	81
4b10e7d5 MG	82	if (updated)
4b10e7d5 MG	83	pages++;
0c5f83c2 MG	84
	85	/* Only !prot_numa always clears the pte */
	86	if (!prot_numa)
	87	ptep_modify_prot_commit(mm, addr, pte, ptent);
ce1744f4	88	} else if (IS_ENABLED(CONFIG_MIGRATION) && !pte_file(oldpte)) {
0697212a CL	89	swp_entry_t entry = pte_to_swp_entry(oldpte);
	90
	91	if (is_write_migration_entry(entry)) {
c3d16e16	92	pte_t newpte;
0697212a CL	93	/*
	94	* A protection check is difficult so
	95	* just be safe and disable write
	96	*/
	97	make_migration_entry_read(&entry);
c3d16e16 CG	98	newpte = swp_entry_to_pte(entry);
	99	if (pte_swp_soft_dirty(oldpte))
	100	newpte = pte_swp_mksoft_dirty(newpte);
	101	set_pte_at(mm, addr, pte, newpte);
e920e14c MG	102
e920e14c MG	103	pages++;
0697212a	104	}
1da177e4 LT	105	}
1da177e4 LT	106	} while (pte++, addr += PAGE_SIZE, addr != end);
6606c3e0	107	arch_leave_lazy_mmu_mode();
705e87c0	108	pte_unmap_unlock(pte - 1, ptl);
7da4d641 PZ	109
7da4d641 PZ	110	return pages;
1da177e4 LT	111	}
1da177e4 LT	112
7d12efae AM	113	static inline unsigned long change_pmd_range(struct vm_area_struct *vma,
	114	pud_t *pud, unsigned long addr, unsigned long end,
	115	pgprot_t newprot, int dirty_accountable, int prot_numa)
1da177e4 LT	116	{
	117	pmd_t *pmd;
	118	unsigned long next;
7da4d641	119	unsigned long pages = 0;
72403b4a	120	unsigned long nr_huge_updates = 0;
1da177e4 LT	121
	122	pmd = pmd_offset(pud, addr);
	123	do {
25cbbef1 MG	124	unsigned long this_pages;
25cbbef1 MG	125
1da177e4	126	next = pmd_addr_end(addr, end);
cd7548ab JW	127	if (pmd_trans_huge(*pmd)) {
cd7548ab JW	128	if (next - addr != HPAGE_PMD_SIZE)
e180377f	129	split_huge_page_pmd(vma, addr, pmd);
f123d74a MG	130	else {
	131	int nr_ptes = change_huge_pmd(vma, pmd, addr,
	132	newprot, prot_numa);
	133
	134	if (nr_ptes) {
72403b4a MG	135	if (nr_ptes == HPAGE_PMD_NR) {
	136	pages += HPAGE_PMD_NR;
	137	nr_huge_updates++;
	138	}
f123d74a MG	139	continue;
f123d74a MG	140	}
7da4d641	141	}
cd7548ab JW	142	/* fall through */
cd7548ab JW	143	}
1da177e4 LT	144	if (pmd_none_or_clear_bad(pmd))
1da177e4 LT	145	continue;
25cbbef1	146	this_pages = change_pte_range(vma, pmd, addr, next, newprot,
0f19c179	147	dirty_accountable, prot_numa);
25cbbef1	148	pages += this_pages;
1da177e4	149	} while (pmd++, addr = next, addr != end);
7da4d641	150
72403b4a MG	151	if (nr_huge_updates)
72403b4a MG	152	count_vm_numa_events(NUMA_HUGE_PTE_UPDATES, nr_huge_updates);
7da4d641	153	return pages;
1da177e4 LT	154	}
1da177e4 LT	155
7d12efae AM	156	static inline unsigned long change_pud_range(struct vm_area_struct *vma,
	157	pgd_t *pgd, unsigned long addr, unsigned long end,
	158	pgprot_t newprot, int dirty_accountable, int prot_numa)
1da177e4 LT	159	{
	160	pud_t *pud;
	161	unsigned long next;
7da4d641	162	unsigned long pages = 0;
1da177e4 LT	163
	164	pud = pud_offset(pgd, addr);
	165	do {
	166	next = pud_addr_end(addr, end);
	167	if (pud_none_or_clear_bad(pud))
	168	continue;
7da4d641	169	pages += change_pmd_range(vma, pud, addr, next, newprot,
4b10e7d5	170	dirty_accountable, prot_numa);
1da177e4	171	} while (pud++, addr = next, addr != end);
7da4d641 PZ	172
7da4d641 PZ	173	return pages;
1da177e4 LT	174	}
1da177e4 LT	175
7da4d641	176	static unsigned long change_protection_range(struct vm_area_struct *vma,
c1e6098b	177	unsigned long addr, unsigned long end, pgprot_t newprot,
4b10e7d5	178	int dirty_accountable, int prot_numa)
1da177e4 LT	179	{
	180	struct mm_struct *mm = vma->vm_mm;
	181	pgd_t *pgd;
	182	unsigned long next;
	183	unsigned long start = addr;
7da4d641	184	unsigned long pages = 0;
1da177e4 LT	185
	186	BUG_ON(addr >= end);
	187	pgd = pgd_offset(mm, addr);
	188	flush_cache_range(vma, addr, end);
1da177e4 LT	189	do {
	190	next = pgd_addr_end(addr, end);
	191	if (pgd_none_or_clear_bad(pgd))
	192	continue;
7da4d641	193	pages += change_pud_range(vma, pgd, addr, next, newprot,
4b10e7d5	194	dirty_accountable, prot_numa);
1da177e4	195	} while (pgd++, addr = next, addr != end);
7da4d641	196
1233d588 IM	197	/* Only flush the TLB if we actually modified any entries: */
	198	if (pages)
	199	flush_tlb_range(vma, start, end);
7da4d641 PZ	200
	201	return pages;
	202	}
	203
	204	unsigned long change_protection(struct vm_area_struct *vma, unsigned long start,
	205	unsigned long end, pgprot_t newprot,
4b10e7d5	206	int dirty_accountable, int prot_numa)
7da4d641 PZ	207	{
	208	struct mm_struct *mm = vma->vm_mm;
	209	unsigned long pages;
	210
	211	mmu_notifier_invalidate_range_start(mm, start, end);
	212	if (is_vm_hugetlb_page(vma))
	213	pages = hugetlb_change_protection(vma, start, end, newprot);
	214	else
4b10e7d5	215	pages = change_protection_range(vma, start, end, newprot, dirty_accountable, prot_numa);
7da4d641 PZ	216	mmu_notifier_invalidate_range_end(mm, start, end);
	217
	218	return pages;
1da177e4 LT	219	}
1da177e4 LT	220
b6a2fea3	221	int
1da177e4 LT	222	mprotect_fixup(struct vm_area_struct vma, struct vm_area_struct *pprev,
	223	unsigned long start, unsigned long end, unsigned long newflags)
	224	{
	225	struct mm_struct *mm = vma->vm_mm;
	226	unsigned long oldflags = vma->vm_flags;
	227	long nrpages = (end - start) >> PAGE_SHIFT;
	228	unsigned long charged = 0;
1da177e4 LT	229	pgoff_t pgoff;
1da177e4 LT	230	int error;
c1e6098b	231	int dirty_accountable = 0;
1da177e4 LT	232
	233	if (newflags == oldflags) {
	234	*pprev = vma;
	235	return 0;
	236	}
	237
	238	/*
	239	* If we make a private mapping writable we increase our commit;
	240	* but (without finer accounting) cannot reduce our commit if we
5a6fe125 MG	241	* make it unwritable again. hugetlb mapping were accounted for
5a6fe125 MG	242	* even if read-only so there is no need to account for them here
1da177e4 LT	243	*/
1da177e4 LT	244	if (newflags & VM_WRITE) {
5a6fe125	245	if (!(oldflags & (VM_ACCOUNT\|VM_WRITE\|VM_HUGETLB\|
cdfd4325	246	VM_SHARED\|VM_NORESERVE))) {
1da177e4	247	charged = nrpages;
191c5424	248	if (security_vm_enough_memory_mm(mm, charged))
1da177e4 LT	249	return -ENOMEM;
	250	newflags \|= VM_ACCOUNT;
	251	}
	252	}
	253
1da177e4 LT	254	/*
	255	* First try to merge with previous and/or next vma.
	256	*/
	257	pgoff = vma->vm_pgoff + ((start - vma->vm_start) >> PAGE_SHIFT);
	258	pprev = vma_merge(mm, pprev, start, end, newflags,
	259	vma->anon_vma, vma->vm_file, pgoff, vma_policy(vma));
	260	if (*pprev) {
	261	vma = *pprev;
	262	goto success;
	263	}
	264
	265	*pprev = vma;
	266
	267	if (start != vma->vm_start) {
	268	error = split_vma(mm, vma, start, 1);
	269	if (error)
	270	goto fail;
	271	}
	272
	273	if (end != vma->vm_end) {
	274	error = split_vma(mm, vma, end, 0);
	275	if (error)
	276	goto fail;
	277	}
	278
	279	success:
	280	/*
	281	* vm_flags and vm_page_prot are protected by the mmap_sem
	282	* held in write mode.
	283	*/
	284	vma->vm_flags = newflags;
1c12c4cf VP	285	vma->vm_page_prot = pgprot_modify(vma->vm_page_prot,
	286	vm_get_page_prot(newflags));
	287
c1e6098b	288	if (vma_wants_writenotify(vma)) {
1ddd439e	289	vma->vm_page_prot = vm_get_page_prot(newflags & ~VM_SHARED);
c1e6098b PZ	290	dirty_accountable = 1;
c1e6098b PZ	291	}
d08b3851	292
7d12efae AM	293	change_protection(vma, start, end, vma->vm_page_prot,
7d12efae AM	294	dirty_accountable, 0);
7da4d641	295
ab50b8ed HD	296	vm_stat_account(mm, oldflags, vma->vm_file, -nrpages);
ab50b8ed HD	297	vm_stat_account(mm, newflags, vma->vm_file, nrpages);
63bfd738	298	perf_event_mmap(vma);
1da177e4 LT	299	return 0;
	300
	301	fail:
	302	vm_unacct_memory(charged);
	303	return error;
	304	}
	305
6a6160a7 HC	306	SYSCALL_DEFINE3(mprotect, unsigned long, start, size_t, len,
6a6160a7 HC	307	unsigned long, prot)
1da177e4 LT	308	{
	309	unsigned long vm_flags, nstart, end, tmp, reqprot;
	310	struct vm_area_struct vma, prev;
	311	int error = -EINVAL;
	312	const int grows = prot & (PROT_GROWSDOWN\|PROT_GROWSUP);
	313	prot &= ~(PROT_GROWSDOWN\|PROT_GROWSUP);
	314	if (grows == (PROT_GROWSDOWN\|PROT_GROWSUP)) /* can't be both */
	315	return -EINVAL;
	316
	317	if (start & ~PAGE_MASK)
	318	return -EINVAL;
	319	if (!len)
	320	return 0;
	321	len = PAGE_ALIGN(len);
	322	end = start + len;
	323	if (end <= start)
	324	return -ENOMEM;
b845f313	325	if (!arch_validate_prot(prot))
1da177e4 LT	326	return -EINVAL;
	327
	328	reqprot = prot;
	329	/*
	330	* Does the application expect PROT_READ to imply PROT_EXEC:
	331	*/
b344e05c	332	if ((prot & PROT_READ) && (current->personality & READ_IMPLIES_EXEC))
1da177e4 LT	333	prot \|= PROT_EXEC;
	334
	335	vm_flags = calc_vm_prot_bits(prot);
	336
	337	down_write(&current->mm->mmap_sem);
	338
097d5910	339	vma = find_vma(current->mm, start);
1da177e4 LT	340	error = -ENOMEM;
	341	if (!vma)
	342	goto out;
097d5910	343	prev = vma->vm_prev;
1da177e4 LT	344	if (unlikely(grows & PROT_GROWSDOWN)) {
	345	if (vma->vm_start >= end)
	346	goto out;
	347	start = vma->vm_start;
	348	error = -EINVAL;
	349	if (!(vma->vm_flags & VM_GROWSDOWN))
	350	goto out;
7d12efae	351	} else {
1da177e4 LT	352	if (vma->vm_start > start)
	353	goto out;
	354	if (unlikely(grows & PROT_GROWSUP)) {
	355	end = vma->vm_end;
	356	error = -EINVAL;
	357	if (!(vma->vm_flags & VM_GROWSUP))
	358	goto out;
	359	}
	360	}
	361	if (start > vma->vm_start)
	362	prev = vma;
	363
	364	for (nstart = start ; ; ) {
	365	unsigned long newflags;
	366
7d12efae	367	/* Here we know that vma->vm_start <= nstart < vma->vm_end. */
1da177e4	368
7d12efae AM	369	newflags = vm_flags;
7d12efae AM	370	newflags \|= (vma->vm_flags & ~(VM_READ \| VM_WRITE \| VM_EXEC));
1da177e4	371
7e2cff42 PBG	372	/* newflags >> 4 shift VM_MAY% in place of VM_% */
7e2cff42 PBG	373	if ((newflags & ~(newflags >> 4)) & (VM_READ \| VM_WRITE \| VM_EXEC)) {
1da177e4 LT	374	error = -EACCES;
	375	goto out;
	376	}
	377
	378	error = security_file_mprotect(vma, reqprot, prot);
	379	if (error)
	380	goto out;
	381
	382	tmp = vma->vm_end;
	383	if (tmp > end)
	384	tmp = end;
	385	error = mprotect_fixup(vma, &prev, nstart, tmp, newflags);
	386	if (error)
	387	goto out;
	388	nstart = tmp;
	389
	390	if (nstart < prev->vm_end)
	391	nstart = prev->vm_end;
	392	if (nstart >= end)
	393	goto out;
	394
	395	vma = prev->vm_next;
	396	if (!vma \|\| vma->vm_start != nstart) {
	397	error = -ENOMEM;
	398	goto out;
	399	}
	400	}
	401	out:
	402	up_write(&current->mm->mmap_sem);
	403	return error;
	404	}