[deliverable/linux.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;

			ptent = ptep_modify_prot_start(mm, addr, pte);
			if (!prot_numa) {
				ptent = pte_modify(ptent, newprot);
				updated = true;
			} else {
				struct page *page;

				page = vm_normal_page(vma, addr, oldpte);
				if (page) {
					if (!pte_numa(oldpte)) {
						ptent = pte_mknuma(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++;
			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)) {
				/*
				 * A protection check is difficult so
				 * just be safe and disable write
				 */
				make_migration_entry_read(&entry);
				set_pte_at(mm, addr, pte,
					swp_entry_to_pte(entry));

				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;

	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++;

					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);

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