[net-next-2.6.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 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>
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	}
	80
	81	static inline void change_pmd_range(struct mm_struct mm, 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);
	91	if (pmd_none_or_clear_bad(pmd))
	92	continue;
c1e6098b	93	change_pte_range(mm, pmd, addr, next, newprot, dirty_accountable);
1da177e4 LT	94	} while (pmd++, addr = next, addr != end);
	95	}
	96
	97	static inline void change_pud_range(struct mm_struct mm, pgd_t pgd,
c1e6098b PZ	98	unsigned long addr, unsigned long end, pgprot_t newprot,
c1e6098b PZ	99	int dirty_accountable)
1da177e4 LT	100	{
	101	pud_t *pud;
	102	unsigned long next;
	103
	104	pud = pud_offset(pgd, addr);
	105	do {
	106	next = pud_addr_end(addr, end);
	107	if (pud_none_or_clear_bad(pud))
	108	continue;
c1e6098b	109	change_pmd_range(mm, pud, addr, next, newprot, dirty_accountable);
1da177e4 LT	110	} while (pud++, addr = next, addr != end);
	111	}
	112
	113	static void change_protection(struct vm_area_struct *vma,
c1e6098b PZ	114	unsigned long addr, unsigned long end, pgprot_t newprot,
c1e6098b PZ	115	int dirty_accountable)
1da177e4 LT	116	{
	117	struct mm_struct *mm = vma->vm_mm;
	118	pgd_t *pgd;
	119	unsigned long next;
	120	unsigned long start = addr;
	121
	122	BUG_ON(addr >= end);
	123	pgd = pgd_offset(mm, addr);
	124	flush_cache_range(vma, addr, end);
1da177e4 LT	125	do {
	126	next = pgd_addr_end(addr, end);
	127	if (pgd_none_or_clear_bad(pgd))
	128	continue;
c1e6098b	129	change_pud_range(mm, pgd, addr, next, newprot, dirty_accountable);
1da177e4 LT	130	} while (pgd++, addr = next, addr != end);
1da177e4 LT	131	flush_tlb_range(vma, start, end);
1da177e4 LT	132	}
1da177e4 LT	133
b6a2fea3	134	int
1da177e4 LT	135	mprotect_fixup(struct vm_area_struct vma, struct vm_area_struct *pprev,
	136	unsigned long start, unsigned long end, unsigned long newflags)
	137	{
	138	struct mm_struct *mm = vma->vm_mm;
	139	unsigned long oldflags = vma->vm_flags;
	140	long nrpages = (end - start) >> PAGE_SHIFT;
	141	unsigned long charged = 0;
1da177e4 LT	142	pgoff_t pgoff;
1da177e4 LT	143	int error;
c1e6098b	144	int dirty_accountable = 0;
1da177e4 LT	145
	146	if (newflags == oldflags) {
	147	*pprev = vma;
	148	return 0;
	149	}
	150
	151	/*
	152	* If we make a private mapping writable we increase our commit;
	153	* but (without finer accounting) cannot reduce our commit if we
5a6fe125 MG	154	* make it unwritable again. hugetlb mapping were accounted for
5a6fe125 MG	155	* even if read-only so there is no need to account for them here
1da177e4 LT	156	*/
1da177e4 LT	157	if (newflags & VM_WRITE) {
5a6fe125	158	if (!(oldflags & (VM_ACCOUNT\|VM_WRITE\|VM_HUGETLB\|
cdfd4325	159	VM_SHARED\|VM_NORESERVE))) {
1da177e4 LT	160	charged = nrpages;
	161	if (security_vm_enough_memory(charged))
	162	return -ENOMEM;
	163	newflags \|= VM_ACCOUNT;
	164	}
	165	}
	166
1da177e4 LT	167	/*
	168	* First try to merge with previous and/or next vma.
	169	*/
	170	pgoff = vma->vm_pgoff + ((start - vma->vm_start) >> PAGE_SHIFT);
	171	pprev = vma_merge(mm, pprev, start, end, newflags,
	172	vma->anon_vma, vma->vm_file, pgoff, vma_policy(vma));
	173	if (*pprev) {
	174	vma = *pprev;
	175	goto success;
	176	}
	177
	178	*pprev = vma;
	179
	180	if (start != vma->vm_start) {
	181	error = split_vma(mm, vma, start, 1);
	182	if (error)
	183	goto fail;
	184	}
	185
	186	if (end != vma->vm_end) {
	187	error = split_vma(mm, vma, end, 0);
	188	if (error)
	189	goto fail;
	190	}
	191
	192	success:
	193	/*
	194	* vm_flags and vm_page_prot are protected by the mmap_sem
	195	* held in write mode.
	196	*/
	197	vma->vm_flags = newflags;
1c12c4cf VP	198	vma->vm_page_prot = pgprot_modify(vma->vm_page_prot,
	199	vm_get_page_prot(newflags));
	200
c1e6098b	201	if (vma_wants_writenotify(vma)) {
1ddd439e	202	vma->vm_page_prot = vm_get_page_prot(newflags & ~VM_SHARED);
c1e6098b PZ	203	dirty_accountable = 1;
c1e6098b PZ	204	}
d08b3851	205
cddb8a5c	206	mmu_notifier_invalidate_range_start(mm, start, end);
8f860591	207	if (is_vm_hugetlb_page(vma))
d08b3851	208	hugetlb_change_protection(vma, start, end, vma->vm_page_prot);
8f860591	209	else
c1e6098b	210	change_protection(vma, start, end, vma->vm_page_prot, dirty_accountable);
cddb8a5c	211	mmu_notifier_invalidate_range_end(mm, start, end);
ab50b8ed HD	212	vm_stat_account(mm, oldflags, vma->vm_file, -nrpages);
ab50b8ed HD	213	vm_stat_account(mm, newflags, vma->vm_file, nrpages);
1da177e4 LT	214	return 0;
	215
	216	fail:
	217	vm_unacct_memory(charged);
	218	return error;
	219	}
	220
6a6160a7 HC	221	SYSCALL_DEFINE3(mprotect, unsigned long, start, size_t, len,
6a6160a7 HC	222	unsigned long, prot)
1da177e4 LT	223	{
	224	unsigned long vm_flags, nstart, end, tmp, reqprot;
	225	struct vm_area_struct vma, prev;
	226	int error = -EINVAL;
	227	const int grows = prot & (PROT_GROWSDOWN\|PROT_GROWSUP);
	228	prot &= ~(PROT_GROWSDOWN\|PROT_GROWSUP);
	229	if (grows == (PROT_GROWSDOWN\|PROT_GROWSUP)) /* can't be both */
	230	return -EINVAL;
	231
	232	if (start & ~PAGE_MASK)
	233	return -EINVAL;
	234	if (!len)
	235	return 0;
	236	len = PAGE_ALIGN(len);
	237	end = start + len;
	238	if (end <= start)
	239	return -ENOMEM;
b845f313	240	if (!arch_validate_prot(prot))
1da177e4 LT	241	return -EINVAL;
	242
	243	reqprot = prot;
	244	/*
	245	* Does the application expect PROT_READ to imply PROT_EXEC:
	246	*/
b344e05c	247	if ((prot & PROT_READ) && (current->personality & READ_IMPLIES_EXEC))
1da177e4 LT	248	prot \|= PROT_EXEC;
	249
	250	vm_flags = calc_vm_prot_bits(prot);
	251
	252	down_write(&current->mm->mmap_sem);
	253
	254	vma = find_vma_prev(current->mm, start, &prev);
	255	error = -ENOMEM;
	256	if (!vma)
	257	goto out;
	258	if (unlikely(grows & PROT_GROWSDOWN)) {
	259	if (vma->vm_start >= end)
	260	goto out;
	261	start = vma->vm_start;
	262	error = -EINVAL;
	263	if (!(vma->vm_flags & VM_GROWSDOWN))
	264	goto out;
	265	}
	266	else {
	267	if (vma->vm_start > start)
	268	goto out;
	269	if (unlikely(grows & PROT_GROWSUP)) {
	270	end = vma->vm_end;
	271	error = -EINVAL;
	272	if (!(vma->vm_flags & VM_GROWSUP))
	273	goto out;
	274	}
	275	}
	276	if (start > vma->vm_start)
	277	prev = vma;
	278
	279	for (nstart = start ; ; ) {
	280	unsigned long newflags;
	281
	282	/* Here we know that vma->vm_start <= nstart < vma->vm_end. */
	283
1da177e4 LT	284	newflags = vm_flags \| (vma->vm_flags & ~(VM_READ \| VM_WRITE \| VM_EXEC));
1da177e4 LT	285
7e2cff42 PBG	286	/* newflags >> 4 shift VM_MAY% in place of VM_% */
7e2cff42 PBG	287	if ((newflags & ~(newflags >> 4)) & (VM_READ \| VM_WRITE \| VM_EXEC)) {
1da177e4 LT	288	error = -EACCES;
	289	goto out;
	290	}
	291
	292	error = security_file_mprotect(vma, reqprot, prot);
	293	if (error)
	294	goto out;
	295
	296	tmp = vma->vm_end;
	297	if (tmp > end)
	298	tmp = end;
	299	error = mprotect_fixup(vma, &prev, nstart, tmp, newflags);
	300	if (error)
	301	goto out;
cdd6c482	302	perf_event_mmap(vma);
1da177e4 LT	303	nstart = tmp;
	304
	305	if (nstart < prev->vm_end)
	306	nstart = prev->vm_end;
	307	if (nstart >= end)
	308	goto out;
	309
	310	vma = prev->vm_next;
	311	if (!vma \|\| vma->vm_start != nstart) {
	312	error = -ENOMEM;
	313	goto out;
	314	}
	315	}
	316	out:
	317	up_write(&current->mm->mmap_sem);
	318	return error;
	319	}