[linux-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/slab.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 mm_struct *mm, 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_none_or_clear_bad(pmd))
			continue;
		change_pte_range(mm, pmd, addr, next, newprot, dirty_accountable);
	} while (pmd++, addr = next, addr != end);
}

static inline void change_pud_range(struct mm_struct *mm, 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(mm, 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(mm, 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);
	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;
		perf_event_mmap(vma);
		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>
	13	#include <linux/slab.h>
	14	#include <linux/shm.h>
	15	#include <linux/mman.h>
	16	#include <linux/fs.h>
	17	#include <linux/highmem.h>
	18	#include <linux/security.h>
	19	#include <linux/mempolicy.h>
	20	#include <linux/personality.h>
	21	#include <linux/syscalls.h>
0697212a CL	22	#include <linux/swap.h>
0697212a CL	23	#include <linux/swapops.h>
cddb8a5c	24	#include <linux/mmu_notifier.h>
64cdd548	25	#include <linux/migrate.h>
cdd6c482	26	#include <linux/perf_event.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
1da177e4	39	static void change_pte_range(struct mm_struct mm, pmd_t pmd,
c1e6098b PZ	40	unsigned long addr, unsigned long end, pgprot_t newprot,
c1e6098b PZ	41	int dirty_accountable)
1da177e4	42	{
0697212a	43	pte_t *pte, oldpte;
705e87c0	44	spinlock_t *ptl;
1da177e4	45
705e87c0	46	pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
6606c3e0	47	arch_enter_lazy_mmu_mode();
1da177e4	48	do {
0697212a CL	49	oldpte = *pte;
0697212a CL	50	if (pte_present(oldpte)) {
1da177e4 LT	51	pte_t ptent;
1da177e4 LT	52
1ea0704e	53	ptent = ptep_modify_prot_start(mm, addr, pte);
c1e6098b	54	ptent = pte_modify(ptent, newprot);
1ea0704e	55
c1e6098b PZ	56	/*
	57	* Avoid taking write faults for pages we know to be
	58	* dirty.
	59	*/
	60	if (dirty_accountable && pte_dirty(ptent))
	61	ptent = pte_mkwrite(ptent);
1ea0704e JF	62
1ea0704e JF	63	ptep_modify_prot_commit(mm, addr, pte, ptent);
64cdd548	64	} else if (PAGE_MIGRATION && !pte_file(oldpte)) {
0697212a CL	65	swp_entry_t entry = pte_to_swp_entry(oldpte);
	66
	67	if (is_write_migration_entry(entry)) {
	68	/*
	69	* A protection check is difficult so
	70	* just be safe and disable write
	71	*/
	72	make_migration_entry_read(&entry);
	73	set_pte_at(mm, addr, pte,
	74	swp_entry_to_pte(entry));
	75	}
1da177e4 LT	76	}
1da177e4 LT	77	} while (pte++, addr += PAGE_SIZE, addr != end);
6606c3e0	78	arch_leave_lazy_mmu_mode();
705e87c0	79	pte_unmap_unlock(pte - 1, ptl);
1da177e4 LT	80	}
	81
	82	static inline void change_pmd_range(struct mm_struct mm, pud_t pud,
c1e6098b PZ	83	unsigned long addr, unsigned long end, pgprot_t newprot,
c1e6098b PZ	84	int dirty_accountable)
1da177e4 LT	85	{
	86	pmd_t *pmd;
	87	unsigned long next;
	88
	89	pmd = pmd_offset(pud, addr);
	90	do {
	91	next = pmd_addr_end(addr, end);
	92	if (pmd_none_or_clear_bad(pmd))
	93	continue;
c1e6098b	94	change_pte_range(mm, pmd, addr, next, newprot, dirty_accountable);
1da177e4 LT	95	} while (pmd++, addr = next, addr != end);
	96	}
	97
	98	static inline void change_pud_range(struct mm_struct mm, pgd_t pgd,
c1e6098b PZ	99	unsigned long addr, unsigned long end, pgprot_t newprot,
c1e6098b PZ	100	int dirty_accountable)
1da177e4 LT	101	{
	102	pud_t *pud;
	103	unsigned long next;
	104
	105	pud = pud_offset(pgd, addr);
	106	do {
	107	next = pud_addr_end(addr, end);
	108	if (pud_none_or_clear_bad(pud))
	109	continue;
c1e6098b	110	change_pmd_range(mm, pud, addr, next, newprot, dirty_accountable);
1da177e4 LT	111	} while (pud++, addr = next, addr != end);
	112	}
	113
	114	static void change_protection(struct vm_area_struct *vma,
c1e6098b PZ	115	unsigned long addr, unsigned long end, pgprot_t newprot,
c1e6098b PZ	116	int dirty_accountable)
1da177e4 LT	117	{
	118	struct mm_struct *mm = vma->vm_mm;
	119	pgd_t *pgd;
	120	unsigned long next;
	121	unsigned long start = addr;
	122
	123	BUG_ON(addr >= end);
	124	pgd = pgd_offset(mm, addr);
	125	flush_cache_range(vma, addr, end);
1da177e4 LT	126	do {
	127	next = pgd_addr_end(addr, end);
	128	if (pgd_none_or_clear_bad(pgd))
	129	continue;
c1e6098b	130	change_pud_range(mm, pgd, addr, next, newprot, dirty_accountable);
1da177e4 LT	131	} while (pgd++, addr = next, addr != end);
1da177e4 LT	132	flush_tlb_range(vma, start, end);
1da177e4 LT	133	}
1da177e4 LT	134
b6a2fea3	135	int
1da177e4 LT	136	mprotect_fixup(struct vm_area_struct vma, struct vm_area_struct *pprev,
	137	unsigned long start, unsigned long end, unsigned long newflags)
	138	{
	139	struct mm_struct *mm = vma->vm_mm;
	140	unsigned long oldflags = vma->vm_flags;
	141	long nrpages = (end - start) >> PAGE_SHIFT;
	142	unsigned long charged = 0;
1da177e4 LT	143	pgoff_t pgoff;
1da177e4 LT	144	int error;
c1e6098b	145	int dirty_accountable = 0;
1da177e4 LT	146
	147	if (newflags == oldflags) {
	148	*pprev = vma;
	149	return 0;
	150	}
	151
	152	/*
	153	* If we make a private mapping writable we increase our commit;
	154	* but (without finer accounting) cannot reduce our commit if we
5a6fe125 MG	155	* make it unwritable again. hugetlb mapping were accounted for
5a6fe125 MG	156	* even if read-only so there is no need to account for them here
1da177e4 LT	157	*/
1da177e4 LT	158	if (newflags & VM_WRITE) {
5a6fe125	159	if (!(oldflags & (VM_ACCOUNT\|VM_WRITE\|VM_HUGETLB\|
cdfd4325	160	VM_SHARED\|VM_NORESERVE))) {
1da177e4 LT	161	charged = nrpages;
	162	if (security_vm_enough_memory(charged))
	163	return -ENOMEM;
	164	newflags \|= VM_ACCOUNT;
	165	}
	166	}
	167
1da177e4 LT	168	/*
	169	* First try to merge with previous and/or next vma.
	170	*/
	171	pgoff = vma->vm_pgoff + ((start - vma->vm_start) >> PAGE_SHIFT);
	172	pprev = vma_merge(mm, pprev, start, end, newflags,
	173	vma->anon_vma, vma->vm_file, pgoff, vma_policy(vma));
	174	if (*pprev) {
	175	vma = *pprev;
	176	goto success;
	177	}
	178
	179	*pprev = vma;
	180
	181	if (start != vma->vm_start) {
	182	error = split_vma(mm, vma, start, 1);
	183	if (error)
	184	goto fail;
	185	}
	186
	187	if (end != vma->vm_end) {
	188	error = split_vma(mm, vma, end, 0);
	189	if (error)
	190	goto fail;
	191	}
	192
	193	success:
	194	/*
	195	* vm_flags and vm_page_prot are protected by the mmap_sem
	196	* held in write mode.
	197	*/
	198	vma->vm_flags = newflags;
1c12c4cf VP	199	vma->vm_page_prot = pgprot_modify(vma->vm_page_prot,
	200	vm_get_page_prot(newflags));
	201
c1e6098b	202	if (vma_wants_writenotify(vma)) {
1ddd439e	203	vma->vm_page_prot = vm_get_page_prot(newflags & ~VM_SHARED);
c1e6098b PZ	204	dirty_accountable = 1;
c1e6098b PZ	205	}
d08b3851	206
cddb8a5c	207	mmu_notifier_invalidate_range_start(mm, start, end);
8f860591	208	if (is_vm_hugetlb_page(vma))
d08b3851	209	hugetlb_change_protection(vma, start, end, vma->vm_page_prot);
8f860591	210	else
c1e6098b	211	change_protection(vma, start, end, vma->vm_page_prot, dirty_accountable);
cddb8a5c	212	mmu_notifier_invalidate_range_end(mm, start, end);
ab50b8ed HD	213	vm_stat_account(mm, oldflags, vma->vm_file, -nrpages);
ab50b8ed HD	214	vm_stat_account(mm, newflags, vma->vm_file, nrpages);
1da177e4 LT	215	return 0;
	216
	217	fail:
	218	vm_unacct_memory(charged);
	219	return error;
	220	}
	221
6a6160a7 HC	222	SYSCALL_DEFINE3(mprotect, unsigned long, start, size_t, len,
6a6160a7 HC	223	unsigned long, prot)
1da177e4 LT	224	{
	225	unsigned long vm_flags, nstart, end, tmp, reqprot;
	226	struct vm_area_struct vma, prev;
	227	int error = -EINVAL;
	228	const int grows = prot & (PROT_GROWSDOWN\|PROT_GROWSUP);
	229	prot &= ~(PROT_GROWSDOWN\|PROT_GROWSUP);
	230	if (grows == (PROT_GROWSDOWN\|PROT_GROWSUP)) /* can't be both */
	231	return -EINVAL;
	232
	233	if (start & ~PAGE_MASK)
	234	return -EINVAL;
	235	if (!len)
	236	return 0;
	237	len = PAGE_ALIGN(len);
	238	end = start + len;
	239	if (end <= start)
	240	return -ENOMEM;
b845f313	241	if (!arch_validate_prot(prot))
1da177e4 LT	242	return -EINVAL;
	243
	244	reqprot = prot;
	245	/*
	246	* Does the application expect PROT_READ to imply PROT_EXEC:
	247	*/
b344e05c	248	if ((prot & PROT_READ) && (current->personality & READ_IMPLIES_EXEC))
1da177e4 LT	249	prot \|= PROT_EXEC;
	250
	251	vm_flags = calc_vm_prot_bits(prot);
	252
	253	down_write(&current->mm->mmap_sem);
	254
	255	vma = find_vma_prev(current->mm, start, &prev);
	256	error = -ENOMEM;
	257	if (!vma)
	258	goto out;
	259	if (unlikely(grows & PROT_GROWSDOWN)) {
	260	if (vma->vm_start >= end)
	261	goto out;
	262	start = vma->vm_start;
	263	error = -EINVAL;
	264	if (!(vma->vm_flags & VM_GROWSDOWN))
	265	goto out;
	266	}
	267	else {
	268	if (vma->vm_start > start)
	269	goto out;
	270	if (unlikely(grows & PROT_GROWSUP)) {
	271	end = vma->vm_end;
	272	error = -EINVAL;
	273	if (!(vma->vm_flags & VM_GROWSUP))
	274	goto out;
	275	}
	276	}
	277	if (start > vma->vm_start)
	278	prev = vma;
	279
	280	for (nstart = start ; ; ) {
	281	unsigned long newflags;
	282
	283	/* Here we know that vma->vm_start <= nstart < vma->vm_end. */
	284
1da177e4 LT	285	newflags = vm_flags \| (vma->vm_flags & ~(VM_READ \| VM_WRITE \| VM_EXEC));
1da177e4 LT	286
7e2cff42 PBG	287	/* newflags >> 4 shift VM_MAY% in place of VM_% */
7e2cff42 PBG	288	if ((newflags & ~(newflags >> 4)) & (VM_READ \| VM_WRITE \| VM_EXEC)) {
1da177e4 LT	289	error = -EACCES;
	290	goto out;
	291	}
	292
	293	error = security_file_mprotect(vma, reqprot, prot);
	294	if (error)
	295	goto out;
	296
	297	tmp = vma->vm_end;
	298	if (tmp > end)
	299	tmp = end;
	300	error = mprotect_fixup(vma, &prev, nstart, tmp, newflags);
	301	if (error)
	302	goto out;
cdd6c482	303	perf_event_mmap(vma);
1da177e4 LT	304	nstart = tmp;
	305
	306	if (nstart < prev->vm_end)
	307	nstart = prev->vm_end;
	308	if (nstart >= end)
	309	goto out;
	310
	311	vma = prev->vm_next;
	312	if (!vma \|\| vma->vm_start != nstart) {
	313	error = -ENOMEM;
	314	goto out;
	315	}
	316	}
	317	out:
	318	up_write(&current->mm->mmap_sem);
	319	return error;
	320	}