[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, bool *ret_all_same_node)
{
	struct mm_struct *mm = vma->vm_mm;
	pte_t *pte, oldpte;
	spinlock_t *ptl;
	unsigned long pages = 0;
	bool all_same_node = true;
	int last_nid = -1;

	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) {
					int this_nid = page_to_nid(page);
					if (last_nid == -1)
						last_nid = this_nid;
					if (last_nid != this_nid)
						all_same_node = false;

					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);

	*ret_all_same_node = all_same_node;
	return pages;
}

#ifdef CONFIG_NUMA_BALANCING
static inline void change_pmd_protnuma(struct mm_struct *mm, unsigned long addr,
				       pmd_t *pmd)
{
	spin_lock(&mm->page_table_lock);
	set_pmd_at(mm, addr & PMD_MASK, pmd, pmd_mknuma(*pmd));
	spin_unlock(&mm->page_table_lock);
}
#else
static inline void change_pmd_protnuma(struct mm_struct *mm, unsigned long addr,
				       pmd_t *pmd)
{
	BUG();
}
#endif /* CONFIG_NUMA_BALANCING */

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;
	bool all_same_node;

	pmd = pmd_offset(pud, addr);
	do {
		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;
		pages += change_pte_range(vma, pmd, addr, next, newprot,
				 dirty_accountable, prot_numa, &all_same_node);

		/*
		 * If we are changing protections for NUMA hinting faults then
		 * set pmd_numa if the examined pages were all on the same
		 * node. This allows a regular PMD to be handled as one fault
		 * and effectively batches the taking of the PTL
		 */
		if (prot_numa && all_same_node)
			change_pmd_protnuma(vma->vm_mm, addr, pmd);
	} 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,
9532fec1	40	int dirty_accountable, int prot_numa, bool *ret_all_same_node)
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;
9532fec1 MG	46	bool all_same_node = true;
9532fec1 MG	47	int last_nid = -1;
1da177e4	48
705e87c0	49	pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
6606c3e0	50	arch_enter_lazy_mmu_mode();
1da177e4	51	do {
0697212a CL	52	oldpte = *pte;
0697212a CL	53	if (pte_present(oldpte)) {
1da177e4	54	pte_t ptent;
4b10e7d5	55	bool updated = false;
1da177e4	56
1ea0704e	57	ptent = ptep_modify_prot_start(mm, addr, pte);
4b10e7d5 MG	58	if (!prot_numa) {
	59	ptent = pte_modify(ptent, newprot);
	60	updated = true;
	61	} else {
	62	struct page *page;
	63
	64	page = vm_normal_page(vma, addr, oldpte);
	65	if (page) {
9532fec1 MG	66	int this_nid = page_to_nid(page);
	67	if (last_nid == -1)
	68	last_nid = this_nid;
	69	if (last_nid != this_nid)
	70	all_same_node = false;
	71
1bc115d8	72	if (!pte_numa(oldpte)) {
4b10e7d5 MG	73	ptent = pte_mknuma(ptent);
	74	updated = true;
	75	}
	76	}
	77	}
1ea0704e	78
c1e6098b PZ	79	/*
	80	* Avoid taking write faults for pages we know to be
	81	* dirty.
	82	*/
4b10e7d5	83	if (dirty_accountable && pte_dirty(ptent)) {
c1e6098b	84	ptent = pte_mkwrite(ptent);
4b10e7d5 MG	85	updated = true;
4b10e7d5 MG	86	}
1ea0704e	87
4b10e7d5 MG	88	if (updated)
4b10e7d5 MG	89	pages++;
1ea0704e	90	ptep_modify_prot_commit(mm, addr, pte, ptent);
ce1744f4	91	} else if (IS_ENABLED(CONFIG_MIGRATION) && !pte_file(oldpte)) {
0697212a CL	92	swp_entry_t entry = pte_to_swp_entry(oldpte);
	93
	94	if (is_write_migration_entry(entry)) {
	95	/*
	96	* A protection check is difficult so
	97	* just be safe and disable write
	98	*/
	99	make_migration_entry_read(&entry);
	100	set_pte_at(mm, addr, pte,
	101	swp_entry_to_pte(entry));
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	109
9532fec1	110	*ret_all_same_node = all_same_node;
7da4d641	111	return pages;
1da177e4 LT	112	}
1da177e4 LT	113
4b10e7d5 MG	114	#ifdef CONFIG_NUMA_BALANCING
4b10e7d5 MG	115	static inline void change_pmd_protnuma(struct mm_struct *mm, unsigned long addr,
7d12efae	116	pmd_t *pmd)
4b10e7d5 MG	117	{
	118	spin_lock(&mm->page_table_lock);
	119	set_pmd_at(mm, addr & PMD_MASK, pmd, pmd_mknuma(*pmd));
	120	spin_unlock(&mm->page_table_lock);
	121	}
	122	#else
	123	static inline void change_pmd_protnuma(struct mm_struct *mm, unsigned long addr,
7d12efae	124	pmd_t *pmd)
4b10e7d5 MG	125	{
4b10e7d5 MG	126	BUG();
1da177e4	127	}
4b10e7d5	128	#endif /* CONFIG_NUMA_BALANCING */
1da177e4	129
7d12efae AM	130	static inline unsigned long change_pmd_range(struct vm_area_struct *vma,
	131	pud_t *pud, unsigned long addr, unsigned long end,
	132	pgprot_t newprot, int dirty_accountable, int prot_numa)
1da177e4 LT	133	{
	134	pmd_t *pmd;
	135	unsigned long next;
7da4d641	136	unsigned long pages = 0;
9532fec1	137	bool all_same_node;
1da177e4 LT	138
	139	pmd = pmd_offset(pud, addr);
	140	do {
	141	next = pmd_addr_end(addr, end);
cd7548ab JW	142	if (pmd_trans_huge(*pmd)) {
cd7548ab JW	143	if (next - addr != HPAGE_PMD_SIZE)
e180377f	144	split_huge_page_pmd(vma, addr, pmd);
f123d74a MG	145	else {
	146	int nr_ptes = change_huge_pmd(vma, pmd, addr,
	147	newprot, prot_numa);
	148
	149	if (nr_ptes) {
	150	if (nr_ptes == HPAGE_PMD_NR)
	151	pages++;
	152
	153	continue;
	154	}
7da4d641	155	}
cd7548ab JW	156	/* fall through */
cd7548ab JW	157	}
1da177e4 LT	158	if (pmd_none_or_clear_bad(pmd))
1da177e4 LT	159	continue;
4b10e7d5	160	pages += change_pte_range(vma, pmd, addr, next, newprot,
9532fec1	161	dirty_accountable, prot_numa, &all_same_node);
4b10e7d5	162
9532fec1 MG	163	/*
	164	* If we are changing protections for NUMA hinting faults then
	165	* set pmd_numa if the examined pages were all on the same
	166	* node. This allows a regular PMD to be handled as one fault
	167	* and effectively batches the taking of the PTL
	168	*/
	169	if (prot_numa && all_same_node)
4b10e7d5	170	change_pmd_protnuma(vma->vm_mm, addr, pmd);
1da177e4	171	} while (pmd++, addr = next, addr != end);
7da4d641 PZ	172
7da4d641 PZ	173	return pages;
1da177e4 LT	174	}
1da177e4 LT	175
7d12efae AM	176	static inline unsigned long change_pud_range(struct vm_area_struct *vma,
	177	pgd_t *pgd, unsigned long addr, unsigned long end,
	178	pgprot_t newprot, int dirty_accountable, int prot_numa)
1da177e4 LT	179	{
	180	pud_t *pud;
	181	unsigned long next;
7da4d641	182	unsigned long pages = 0;
1da177e4 LT	183
	184	pud = pud_offset(pgd, addr);
	185	do {
	186	next = pud_addr_end(addr, end);
	187	if (pud_none_or_clear_bad(pud))
	188	continue;
7da4d641	189	pages += change_pmd_range(vma, pud, addr, next, newprot,
4b10e7d5	190	dirty_accountable, prot_numa);
1da177e4	191	} while (pud++, addr = next, addr != end);
7da4d641 PZ	192
7da4d641 PZ	193	return pages;
1da177e4 LT	194	}
1da177e4 LT	195
7da4d641	196	static unsigned long change_protection_range(struct vm_area_struct *vma,
c1e6098b	197	unsigned long addr, unsigned long end, pgprot_t newprot,
4b10e7d5	198	int dirty_accountable, int prot_numa)
1da177e4 LT	199	{
	200	struct mm_struct *mm = vma->vm_mm;
	201	pgd_t *pgd;
	202	unsigned long next;
	203	unsigned long start = addr;
7da4d641	204	unsigned long pages = 0;
1da177e4 LT	205
	206	BUG_ON(addr >= end);
	207	pgd = pgd_offset(mm, addr);
	208	flush_cache_range(vma, addr, end);
1da177e4 LT	209	do {
	210	next = pgd_addr_end(addr, end);
	211	if (pgd_none_or_clear_bad(pgd))
	212	continue;
7da4d641	213	pages += change_pud_range(vma, pgd, addr, next, newprot,
4b10e7d5	214	dirty_accountable, prot_numa);
1da177e4	215	} while (pgd++, addr = next, addr != end);
7da4d641	216
1233d588 IM	217	/* Only flush the TLB if we actually modified any entries: */
	218	if (pages)
	219	flush_tlb_range(vma, start, end);
7da4d641 PZ	220
	221	return pages;
	222	}
	223
	224	unsigned long change_protection(struct vm_area_struct *vma, unsigned long start,
	225	unsigned long end, pgprot_t newprot,
4b10e7d5	226	int dirty_accountable, int prot_numa)
7da4d641 PZ	227	{
	228	struct mm_struct *mm = vma->vm_mm;
	229	unsigned long pages;
	230
	231	mmu_notifier_invalidate_range_start(mm, start, end);
	232	if (is_vm_hugetlb_page(vma))
	233	pages = hugetlb_change_protection(vma, start, end, newprot);
	234	else
4b10e7d5	235	pages = change_protection_range(vma, start, end, newprot, dirty_accountable, prot_numa);
7da4d641 PZ	236	mmu_notifier_invalidate_range_end(mm, start, end);
	237
	238	return pages;
1da177e4 LT	239	}
1da177e4 LT	240
b6a2fea3	241	int
1da177e4 LT	242	mprotect_fixup(struct vm_area_struct vma, struct vm_area_struct *pprev,
	243	unsigned long start, unsigned long end, unsigned long newflags)
	244	{
	245	struct mm_struct *mm = vma->vm_mm;
	246	unsigned long oldflags = vma->vm_flags;
	247	long nrpages = (end - start) >> PAGE_SHIFT;
	248	unsigned long charged = 0;
1da177e4 LT	249	pgoff_t pgoff;
1da177e4 LT	250	int error;
c1e6098b	251	int dirty_accountable = 0;
1da177e4 LT	252
	253	if (newflags == oldflags) {
	254	*pprev = vma;
	255	return 0;
	256	}
	257
	258	/*
	259	* If we make a private mapping writable we increase our commit;
	260	* but (without finer accounting) cannot reduce our commit if we
5a6fe125 MG	261	* make it unwritable again. hugetlb mapping were accounted for
5a6fe125 MG	262	* even if read-only so there is no need to account for them here
1da177e4 LT	263	*/
1da177e4 LT	264	if (newflags & VM_WRITE) {
5a6fe125	265	if (!(oldflags & (VM_ACCOUNT\|VM_WRITE\|VM_HUGETLB\|
cdfd4325	266	VM_SHARED\|VM_NORESERVE))) {
1da177e4	267	charged = nrpages;
191c5424	268	if (security_vm_enough_memory_mm(mm, charged))
1da177e4 LT	269	return -ENOMEM;
	270	newflags \|= VM_ACCOUNT;
	271	}
	272	}
	273
1da177e4 LT	274	/*
	275	* First try to merge with previous and/or next vma.
	276	*/
	277	pgoff = vma->vm_pgoff + ((start - vma->vm_start) >> PAGE_SHIFT);
	278	pprev = vma_merge(mm, pprev, start, end, newflags,
	279	vma->anon_vma, vma->vm_file, pgoff, vma_policy(vma));
	280	if (*pprev) {
	281	vma = *pprev;
	282	goto success;
	283	}
	284
	285	*pprev = vma;
	286
	287	if (start != vma->vm_start) {
	288	error = split_vma(mm, vma, start, 1);
	289	if (error)
	290	goto fail;
	291	}
	292
	293	if (end != vma->vm_end) {
	294	error = split_vma(mm, vma, end, 0);
	295	if (error)
	296	goto fail;
	297	}
	298
	299	success:
	300	/*
	301	* vm_flags and vm_page_prot are protected by the mmap_sem
	302	* held in write mode.
	303	*/
	304	vma->vm_flags = newflags;
1c12c4cf VP	305	vma->vm_page_prot = pgprot_modify(vma->vm_page_prot,
	306	vm_get_page_prot(newflags));
	307
c1e6098b	308	if (vma_wants_writenotify(vma)) {
1ddd439e	309	vma->vm_page_prot = vm_get_page_prot(newflags & ~VM_SHARED);
c1e6098b PZ	310	dirty_accountable = 1;
c1e6098b PZ	311	}
d08b3851	312
7d12efae AM	313	change_protection(vma, start, end, vma->vm_page_prot,
7d12efae AM	314	dirty_accountable, 0);
7da4d641	315
ab50b8ed HD	316	vm_stat_account(mm, oldflags, vma->vm_file, -nrpages);
ab50b8ed HD	317	vm_stat_account(mm, newflags, vma->vm_file, nrpages);
63bfd738	318	perf_event_mmap(vma);
1da177e4 LT	319	return 0;
	320
	321	fail:
	322	vm_unacct_memory(charged);
	323	return error;
	324	}
	325
6a6160a7 HC	326	SYSCALL_DEFINE3(mprotect, unsigned long, start, size_t, len,
6a6160a7 HC	327	unsigned long, prot)
1da177e4 LT	328	{
	329	unsigned long vm_flags, nstart, end, tmp, reqprot;
	330	struct vm_area_struct vma, prev;
	331	int error = -EINVAL;
	332	const int grows = prot & (PROT_GROWSDOWN\|PROT_GROWSUP);
	333	prot &= ~(PROT_GROWSDOWN\|PROT_GROWSUP);
	334	if (grows == (PROT_GROWSDOWN\|PROT_GROWSUP)) /* can't be both */
	335	return -EINVAL;
	336
	337	if (start & ~PAGE_MASK)
	338	return -EINVAL;
	339	if (!len)
	340	return 0;
	341	len = PAGE_ALIGN(len);
	342	end = start + len;
	343	if (end <= start)
	344	return -ENOMEM;
b845f313	345	if (!arch_validate_prot(prot))
1da177e4 LT	346	return -EINVAL;
	347
	348	reqprot = prot;
	349	/*
	350	* Does the application expect PROT_READ to imply PROT_EXEC:
	351	*/
b344e05c	352	if ((prot & PROT_READ) && (current->personality & READ_IMPLIES_EXEC))
1da177e4 LT	353	prot \|= PROT_EXEC;
	354
	355	vm_flags = calc_vm_prot_bits(prot);
	356
	357	down_write(&current->mm->mmap_sem);
	358
097d5910	359	vma = find_vma(current->mm, start);
1da177e4 LT	360	error = -ENOMEM;
	361	if (!vma)
	362	goto out;
097d5910	363	prev = vma->vm_prev;
1da177e4 LT	364	if (unlikely(grows & PROT_GROWSDOWN)) {
	365	if (vma->vm_start >= end)
	366	goto out;
	367	start = vma->vm_start;
	368	error = -EINVAL;
	369	if (!(vma->vm_flags & VM_GROWSDOWN))
	370	goto out;
7d12efae	371	} else {
1da177e4 LT	372	if (vma->vm_start > start)
	373	goto out;
	374	if (unlikely(grows & PROT_GROWSUP)) {
	375	end = vma->vm_end;
	376	error = -EINVAL;
	377	if (!(vma->vm_flags & VM_GROWSUP))
	378	goto out;
	379	}
	380	}
	381	if (start > vma->vm_start)
	382	prev = vma;
	383
	384	for (nstart = start ; ; ) {
	385	unsigned long newflags;
	386
7d12efae	387	/* Here we know that vma->vm_start <= nstart < vma->vm_end. */
1da177e4	388
7d12efae AM	389	newflags = vm_flags;
7d12efae AM	390	newflags \|= (vma->vm_flags & ~(VM_READ \| VM_WRITE \| VM_EXEC));
1da177e4	391
7e2cff42 PBG	392	/* newflags >> 4 shift VM_MAY% in place of VM_% */
7e2cff42 PBG	393	if ((newflags & ~(newflags >> 4)) & (VM_READ \| VM_WRITE \| VM_EXEC)) {
1da177e4 LT	394	error = -EACCES;
	395	goto out;
	396	}
	397
	398	error = security_file_mprotect(vma, reqprot, prot);
	399	if (error)
	400	goto out;
	401
	402	tmp = vma->vm_end;
	403	if (tmp > end)
	404	tmp = end;
	405	error = mprotect_fixup(vma, &prev, nstart, tmp, newflags);
	406	if (error)
	407	goto out;
	408	nstart = tmp;
	409
	410	if (nstart < prev->vm_end)
	411	nstart = prev->vm_end;
	412	if (nstart >= end)
	413	goto out;
	414
	415	vma = prev->vm_next;
	416	if (!vma \|\| vma->vm_start != nstart) {
	417	error = -ENOMEM;
	418	goto out;
	419	}
	420	}
	421	out:
	422	up_write(&current->mm->mmap_sem);
	423	return error;
	424	}