[linux.git] / mm / mprotect.c

/*
 *  mm/mprotect.c
 *
 *  (C) Copyright 1994 Linus Torvalds
 *  (C) Copyright 2002 Christoph Hellwig
 *
 *  Address space accounting code	<[email protected]>
 *  (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 void change_pte_range(struct mm_struct *mm, pmd_t *pmd,
		unsigned long addr, unsigned long end, pgprot_t newprot,
		int dirty_accountable)
{
	pte_t *pte, oldpte;
	spinlock_t *ptl;

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

			ptent = ptep_modify_prot_start(mm, addr, pte);
			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);
		} else if (PAGE_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));
			}
		}
	} while (pte++, addr += PAGE_SIZE, addr != end);
	arch_leave_lazy_mmu_mode();
	pte_unmap_unlock(pte - 1, ptl);
}

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

	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->vm_mm, pmd);
			else if (change_huge_pmd(vma, pmd, addr, newprot))
				continue;
			/* fall through */
		}
		if (pmd_none_or_clear_bad(pmd))
			continue;
		change_pte_range(vma->vm_mm, pmd, addr, next, newprot,
				 dirty_accountable);
	} while (pmd++, addr = next, addr != end);
}

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

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

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

	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;
		change_pud_range(vma, pgd, addr, next, newprot,
				 dirty_accountable);
	} while (pgd++, addr = next, addr != end);
	flush_tlb_range(vma, start, end);
}

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(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;
	}

	mmu_notifier_invalidate_range_start(mm, start, end);
	if (is_vm_hugetlb_page(vma))
		hugetlb_change_protection(vma, start, end, vma->vm_page_prot);
	else
		change_protection(vma, start, end, vma->vm_page_prot, dirty_accountable);
	mmu_notifier_invalidate_range_end(mm, start, end);
	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_prev(current->mm, start, &prev);
	error = -ENOMEM;
	if (!vma)
		goto out;
	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 | (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 <[email protected]>
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
1da177e4	38	static void change_pte_range(struct mm_struct mm, pmd_t pmd,
c1e6098b PZ	39	unsigned long addr, unsigned long end, pgprot_t newprot,
c1e6098b PZ	40	int dirty_accountable)
1da177e4	41	{
0697212a	42	pte_t *pte, oldpte;
705e87c0	43	spinlock_t *ptl;
1da177e4	44
705e87c0	45	pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
6606c3e0	46	arch_enter_lazy_mmu_mode();
1da177e4	47	do {
0697212a CL	48	oldpte = *pte;
0697212a CL	49	if (pte_present(oldpte)) {
1da177e4 LT	50	pte_t ptent;
1da177e4 LT	51
1ea0704e	52	ptent = ptep_modify_prot_start(mm, addr, pte);
c1e6098b	53	ptent = pte_modify(ptent, newprot);
1ea0704e	54
c1e6098b PZ	55	/*
	56	* Avoid taking write faults for pages we know to be
	57	* dirty.
	58	*/
	59	if (dirty_accountable && pte_dirty(ptent))
	60	ptent = pte_mkwrite(ptent);
1ea0704e JF	61
1ea0704e JF	62	ptep_modify_prot_commit(mm, addr, pte, ptent);
64cdd548	63	} else if (PAGE_MIGRATION && !pte_file(oldpte)) {
0697212a CL	64	swp_entry_t entry = pte_to_swp_entry(oldpte);
	65
	66	if (is_write_migration_entry(entry)) {
	67	/*
	68	* A protection check is difficult so
	69	* just be safe and disable write
	70	*/
	71	make_migration_entry_read(&entry);
	72	set_pte_at(mm, addr, pte,
	73	swp_entry_to_pte(entry));
	74	}
1da177e4 LT	75	}
1da177e4 LT	76	} while (pte++, addr += PAGE_SIZE, addr != end);
6606c3e0	77	arch_leave_lazy_mmu_mode();
705e87c0	78	pte_unmap_unlock(pte - 1, ptl);
1da177e4 LT	79	}
1da177e4 LT	80
b36f5b07	81	static inline void change_pmd_range(struct vm_area_struct vma, pud_t pud,
c1e6098b PZ	82	unsigned long addr, unsigned long end, pgprot_t newprot,
c1e6098b PZ	83	int dirty_accountable)
1da177e4 LT	84	{
	85	pmd_t *pmd;
	86	unsigned long next;
	87
	88	pmd = pmd_offset(pud, addr);
	89	do {
	90	next = pmd_addr_end(addr, end);
cd7548ab JW	91	if (pmd_trans_huge(*pmd)) {
	92	if (next - addr != HPAGE_PMD_SIZE)
	93	split_huge_page_pmd(vma->vm_mm, pmd);
	94	else if (change_huge_pmd(vma, pmd, addr, newprot))
	95	continue;
	96	/* fall through */
	97	}
1da177e4 LT	98	if (pmd_none_or_clear_bad(pmd))
1da177e4 LT	99	continue;
b36f5b07 JW	100	change_pte_range(vma->vm_mm, pmd, addr, next, newprot,
b36f5b07 JW	101	dirty_accountable);
1da177e4 LT	102	} while (pmd++, addr = next, addr != end);
	103	}
	104
b36f5b07	105	static inline void change_pud_range(struct vm_area_struct vma, pgd_t pgd,
c1e6098b PZ	106	unsigned long addr, unsigned long end, pgprot_t newprot,
c1e6098b PZ	107	int dirty_accountable)
1da177e4 LT	108	{
	109	pud_t *pud;
	110	unsigned long next;
	111
	112	pud = pud_offset(pgd, addr);
	113	do {
	114	next = pud_addr_end(addr, end);
	115	if (pud_none_or_clear_bad(pud))
	116	continue;
b36f5b07 JW	117	change_pmd_range(vma, pud, addr, next, newprot,
b36f5b07 JW	118	dirty_accountable);
1da177e4 LT	119	} while (pud++, addr = next, addr != end);
	120	}
	121
	122	static void change_protection(struct vm_area_struct *vma,
c1e6098b PZ	123	unsigned long addr, unsigned long end, pgprot_t newprot,
c1e6098b PZ	124	int dirty_accountable)
1da177e4 LT	125	{
	126	struct mm_struct *mm = vma->vm_mm;
	127	pgd_t *pgd;
	128	unsigned long next;
	129	unsigned long start = addr;
	130
	131	BUG_ON(addr >= end);
	132	pgd = pgd_offset(mm, addr);
	133	flush_cache_range(vma, addr, end);
1da177e4 LT	134	do {
	135	next = pgd_addr_end(addr, end);
	136	if (pgd_none_or_clear_bad(pgd))
	137	continue;
b36f5b07 JW	138	change_pud_range(vma, pgd, addr, next, newprot,
b36f5b07 JW	139	dirty_accountable);
1da177e4 LT	140	} while (pgd++, addr = next, addr != end);
1da177e4 LT	141	flush_tlb_range(vma, start, end);
1da177e4 LT	142	}
1da177e4 LT	143
b6a2fea3	144	int
1da177e4 LT	145	mprotect_fixup(struct vm_area_struct vma, struct vm_area_struct *pprev,
	146	unsigned long start, unsigned long end, unsigned long newflags)
	147	{
	148	struct mm_struct *mm = vma->vm_mm;
	149	unsigned long oldflags = vma->vm_flags;
	150	long nrpages = (end - start) >> PAGE_SHIFT;
	151	unsigned long charged = 0;
1da177e4 LT	152	pgoff_t pgoff;
1da177e4 LT	153	int error;
c1e6098b	154	int dirty_accountable = 0;
1da177e4 LT	155
	156	if (newflags == oldflags) {
	157	*pprev = vma;
	158	return 0;
	159	}
	160
	161	/*
	162	* If we make a private mapping writable we increase our commit;
	163	* but (without finer accounting) cannot reduce our commit if we
5a6fe125 MG	164	* make it unwritable again. hugetlb mapping were accounted for
5a6fe125 MG	165	* even if read-only so there is no need to account for them here
1da177e4 LT	166	*/
1da177e4 LT	167	if (newflags & VM_WRITE) {
5a6fe125	168	if (!(oldflags & (VM_ACCOUNT\|VM_WRITE\|VM_HUGETLB\|
cdfd4325	169	VM_SHARED\|VM_NORESERVE))) {
1da177e4 LT	170	charged = nrpages;
	171	if (security_vm_enough_memory(charged))
	172	return -ENOMEM;
	173	newflags \|= VM_ACCOUNT;
	174	}
	175	}
	176
1da177e4 LT	177	/*
	178	* First try to merge with previous and/or next vma.
	179	*/
	180	pgoff = vma->vm_pgoff + ((start - vma->vm_start) >> PAGE_SHIFT);
	181	pprev = vma_merge(mm, pprev, start, end, newflags,
	182	vma->anon_vma, vma->vm_file, pgoff, vma_policy(vma));
	183	if (*pprev) {
	184	vma = *pprev;
	185	goto success;
	186	}
	187
	188	*pprev = vma;
	189
	190	if (start != vma->vm_start) {
	191	error = split_vma(mm, vma, start, 1);
	192	if (error)
	193	goto fail;
	194	}
	195
	196	if (end != vma->vm_end) {
	197	error = split_vma(mm, vma, end, 0);
	198	if (error)
	199	goto fail;
	200	}
	201
	202	success:
	203	/*
	204	* vm_flags and vm_page_prot are protected by the mmap_sem
	205	* held in write mode.
	206	*/
	207	vma->vm_flags = newflags;
1c12c4cf VP	208	vma->vm_page_prot = pgprot_modify(vma->vm_page_prot,
	209	vm_get_page_prot(newflags));
	210
c1e6098b	211	if (vma_wants_writenotify(vma)) {
1ddd439e	212	vma->vm_page_prot = vm_get_page_prot(newflags & ~VM_SHARED);
c1e6098b PZ	213	dirty_accountable = 1;
c1e6098b PZ	214	}
d08b3851	215
cddb8a5c	216	mmu_notifier_invalidate_range_start(mm, start, end);
8f860591	217	if (is_vm_hugetlb_page(vma))
d08b3851	218	hugetlb_change_protection(vma, start, end, vma->vm_page_prot);
8f860591	219	else
c1e6098b	220	change_protection(vma, start, end, vma->vm_page_prot, dirty_accountable);
cddb8a5c	221	mmu_notifier_invalidate_range_end(mm, start, end);
ab50b8ed HD	222	vm_stat_account(mm, oldflags, vma->vm_file, -nrpages);
ab50b8ed HD	223	vm_stat_account(mm, newflags, vma->vm_file, nrpages);
63bfd738	224	perf_event_mmap(vma);
1da177e4 LT	225	return 0;
	226
	227	fail:
	228	vm_unacct_memory(charged);
	229	return error;
	230	}
	231
6a6160a7 HC	232	SYSCALL_DEFINE3(mprotect, unsigned long, start, size_t, len,
6a6160a7 HC	233	unsigned long, prot)
1da177e4 LT	234	{
	235	unsigned long vm_flags, nstart, end, tmp, reqprot;
	236	struct vm_area_struct vma, prev;
	237	int error = -EINVAL;
	238	const int grows = prot & (PROT_GROWSDOWN\|PROT_GROWSUP);
	239	prot &= ~(PROT_GROWSDOWN\|PROT_GROWSUP);
	240	if (grows == (PROT_GROWSDOWN\|PROT_GROWSUP)) /* can't be both */
	241	return -EINVAL;
	242
	243	if (start & ~PAGE_MASK)
	244	return -EINVAL;
	245	if (!len)
	246	return 0;
	247	len = PAGE_ALIGN(len);
	248	end = start + len;
	249	if (end <= start)
	250	return -ENOMEM;
b845f313	251	if (!arch_validate_prot(prot))
1da177e4 LT	252	return -EINVAL;
	253
	254	reqprot = prot;
	255	/*
	256	* Does the application expect PROT_READ to imply PROT_EXEC:
	257	*/
b344e05c	258	if ((prot & PROT_READ) && (current->personality & READ_IMPLIES_EXEC))
1da177e4 LT	259	prot \|= PROT_EXEC;
	260
	261	vm_flags = calc_vm_prot_bits(prot);
	262
	263	down_write(&current->mm->mmap_sem);
	264
	265	vma = find_vma_prev(current->mm, start, &prev);
	266	error = -ENOMEM;
	267	if (!vma)
	268	goto out;
	269	if (unlikely(grows & PROT_GROWSDOWN)) {
	270	if (vma->vm_start >= end)
	271	goto out;
	272	start = vma->vm_start;
	273	error = -EINVAL;
	274	if (!(vma->vm_flags & VM_GROWSDOWN))
	275	goto out;
	276	}
	277	else {
	278	if (vma->vm_start > start)
	279	goto out;
	280	if (unlikely(grows & PROT_GROWSUP)) {
	281	end = vma->vm_end;
	282	error = -EINVAL;
	283	if (!(vma->vm_flags & VM_GROWSUP))
	284	goto out;
	285	}
	286	}
	287	if (start > vma->vm_start)
	288	prev = vma;
	289
	290	for (nstart = start ; ; ) {
	291	unsigned long newflags;
	292
	293	/* Here we know that vma->vm_start <= nstart < vma->vm_end. */
	294
1da177e4 LT	295	newflags = vm_flags \| (vma->vm_flags & ~(VM_READ \| VM_WRITE \| VM_EXEC));
1da177e4 LT	296
7e2cff42 PBG	297	/* newflags >> 4 shift VM_MAY% in place of VM_% */
7e2cff42 PBG	298	if ((newflags & ~(newflags >> 4)) & (VM_READ \| VM_WRITE \| VM_EXEC)) {
1da177e4 LT	299	error = -EACCES;
	300	goto out;
	301	}
	302
	303	error = security_file_mprotect(vma, reqprot, prot);
	304	if (error)
	305	goto out;
	306
	307	tmp = vma->vm_end;
	308	if (tmp > end)
	309	tmp = end;
	310	error = mprotect_fixup(vma, &prev, nstart, tmp, newflags);
	311	if (error)
	312	goto out;
	313	nstart = tmp;
	314
	315	if (nstart < prev->vm_end)
	316	nstart = prev->vm_end;
	317	if (nstart >= end)
	318	goto out;
	319
	320	vma = prev->vm_next;
	321	if (!vma \|\| vma->vm_start != nstart) {
	322	error = -ENOMEM;
	323	goto out;
	324	}
	325	}
	326	out:
	327	up_write(&current->mm->mmap_sem);
	328	return error;
	329	}