[linux-2.6-block.git] / mm / mprotect.c

// SPDX-License-Identifier: GPL-2.0
/*
 *  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 <linux/pkeys.h>
#include <linux/ksm.h>
#include <linux/uaccess.h>
#include <asm/pgtable.h>
#include <asm/cacheflush.h>
#include <asm/mmu_context.h>
#include <asm/tlbflush.h>

#include "internal.h"

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;
	int target_node = NUMA_NO_NODE;

	/*
	 * Can be called with only the mmap_sem for reading by
	 * prot_numa so we must check the pmd isn't constantly
	 * changing from under us from pmd_none to pmd_trans_huge
	 * and/or the other way around.
	 */
	if (pmd_trans_unstable(pmd))
		return 0;

	/*
	 * The pmd points to a regular pte so the pmd can't change
	 * from under us even if the mmap_sem is only hold for
	 * reading.
	 */
	pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl);

	/* Get target node for single threaded private VMAs */
	if (prot_numa && !(vma->vm_flags & VM_SHARED) &&
	    atomic_read(&vma->vm_mm->mm_users) == 1)
		target_node = numa_node_id();

	flush_tlb_batched_pending(vma->vm_mm);
	arch_enter_lazy_mmu_mode();
	do {
		oldpte = *pte;
		if (pte_present(oldpte)) {
			pte_t ptent;
			bool preserve_write = prot_numa && pte_write(oldpte);

			/*
			 * Avoid trapping faults against the zero or KSM
			 * pages. See similar comment in change_huge_pmd.
			 */
			if (prot_numa) {
				struct page *page;

				page = vm_normal_page(vma, addr, oldpte);
				if (!page || PageKsm(page))
					continue;

				/* Also skip shared copy-on-write pages */
				if (is_cow_mapping(vma->vm_flags) &&
				    page_mapcount(page) != 1)
					continue;

				/* Avoid TLB flush if possible */
				if (pte_protnone(oldpte))
					continue;

				/*
				 * Don't mess with PTEs if page is already on the node
				 * a single-threaded process is running on.
				 */
				if (target_node == page_to_nid(page))
					continue;
			}

			ptent = ptep_modify_prot_start(mm, addr, pte);
			ptent = pte_modify(ptent, newprot);
			if (preserve_write)
				ptent = pte_mk_savedwrite(ptent);

			/* Avoid taking write faults for known dirty pages */
			if (dirty_accountable && pte_dirty(ptent) &&
					(pte_soft_dirty(ptent) ||
					 !(vma->vm_flags & VM_SOFTDIRTY))) {
				ptent = pte_mkwrite(ptent);
			}
			ptep_modify_prot_commit(mm, addr, pte, ptent);
			pages++;
		} else if (IS_ENABLED(CONFIG_MIGRATION)) {
			swp_entry_t entry = pte_to_swp_entry(oldpte);

			if (is_write_migration_entry(entry)) {
				pte_t newpte;
				/*
				 * A protection check is difficult so
				 * just be safe and disable write
				 */
				make_migration_entry_read(&entry);
				newpte = swp_entry_to_pte(entry);
				if (pte_swp_soft_dirty(oldpte))
					newpte = pte_swp_mksoft_dirty(newpte);
				set_pte_at(mm, addr, pte, newpte);

				pages++;
			}

			if (is_write_device_private_entry(entry)) {
				pte_t newpte;

				/*
				 * We do not preserve soft-dirtiness. See
				 * copy_one_pte() for explanation.
				 */
				make_device_private_entry_read(&entry);
				newpte = swp_entry_to_pte(entry);
				set_pte_at(mm, addr, pte, newpte);

				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;
	struct mm_struct *mm = vma->vm_mm;
	unsigned long next;
	unsigned long pages = 0;
	unsigned long nr_huge_updates = 0;
	unsigned long mni_start = 0;

	pmd = pmd_offset(pud, addr);
	do {
		unsigned long this_pages;

		next = pmd_addr_end(addr, end);
		if (!is_swap_pmd(*pmd) && !pmd_trans_huge(*pmd) && !pmd_devmap(*pmd)
				&& pmd_none_or_clear_bad(pmd))
			goto next;

		/* invoke the mmu notifier if the pmd is populated */
		if (!mni_start) {
			mni_start = addr;
			mmu_notifier_invalidate_range_start(mm, mni_start, end);
		}

		if (is_swap_pmd(*pmd) || pmd_trans_huge(*pmd) || pmd_devmap(*pmd)) {
			if (next - addr != HPAGE_PMD_SIZE) {
				__split_huge_pmd(vma, pmd, addr, false, NULL);
			} else {
				int nr_ptes = change_huge_pmd(vma, pmd, addr,
						newprot, prot_numa);

				if (nr_ptes) {
					if (nr_ptes == HPAGE_PMD_NR) {
						pages += HPAGE_PMD_NR;
						nr_huge_updates++;
					}

					/* huge pmd was handled */
					goto next;
				}
			}
			/* fall through, the trans huge pmd just split */
		}
		this_pages = change_pte_range(vma, pmd, addr, next, newprot,
				 dirty_accountable, prot_numa);
		pages += this_pages;
next:
		cond_resched();
	} while (pmd++, addr = next, addr != end);

	if (mni_start)
		mmu_notifier_invalidate_range_end(mm, mni_start, end);

	if (nr_huge_updates)
		count_vm_numa_events(NUMA_HUGE_PTE_UPDATES, nr_huge_updates);
	return pages;
}

static inline unsigned long change_pud_range(struct vm_area_struct *vma,
		p4d_t *p4d, 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(p4d, 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 inline unsigned long change_p4d_range(struct vm_area_struct *vma,
		pgd_t *pgd, unsigned long addr, unsigned long end,
		pgprot_t newprot, int dirty_accountable, int prot_numa)
{
	p4d_t *p4d;
	unsigned long next;
	unsigned long pages = 0;

	p4d = p4d_offset(pgd, addr);
	do {
		next = p4d_addr_end(addr, end);
		if (p4d_none_or_clear_bad(p4d))
			continue;
		pages += change_pud_range(vma, p4d, addr, next, newprot,
				 dirty_accountable, prot_numa);
	} while (p4d++, 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);
	inc_tlb_flush_pending(mm);
	do {
		next = pgd_addr_end(addr, end);
		if (pgd_none_or_clear_bad(pgd))
			continue;
		pages += change_p4d_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);
	dec_tlb_flush_pending(mm);

	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)
{
	unsigned long pages;

	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);

	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) {
		/* Check space limits when area turns into data. */
		if (!may_expand_vm(mm, newflags, nrpages) &&
				may_expand_vm(mm, oldflags, nrpages))
			return -ENOMEM;
		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),
			   vma->vm_userfaultfd_ctx);
	if (*pprev) {
		vma = *pprev;
		VM_WARN_ON((vma->vm_flags ^ newflags) & ~VM_SOFTDIRTY);
		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;
	dirty_accountable = vma_wants_writenotify(vma, vma->vm_page_prot);
	vma_set_page_prot(vma);

	change_protection(vma, start, end, vma->vm_page_prot,
			  dirty_accountable, 0);

	/*
	 * Private VM_LOCKED VMA becoming writable: trigger COW to avoid major
	 * fault on access.
	 */
	if ((oldflags & (VM_WRITE | VM_SHARED | VM_LOCKED)) == VM_LOCKED &&
			(newflags & VM_WRITE)) {
		populate_vma_page_range(vma, start, end, NULL);
	}

	vm_stat_account(mm, oldflags, -nrpages);
	vm_stat_account(mm, newflags, nrpages);
	perf_event_mmap(vma);
	return 0;

fail:
	vm_unacct_memory(charged);
	return error;
}

/*
 * pkey==-1 when doing a legacy mprotect()
 */
static int do_mprotect_pkey(unsigned long start, size_t len,
		unsigned long prot, int pkey)
{
	unsigned long nstart, end, tmp, reqprot;
	struct vm_area_struct *vma, *prev;
	int error = -EINVAL;
	const int grows = prot & (PROT_GROWSDOWN|PROT_GROWSUP);
	const bool rier = (current->personality & READ_IMPLIES_EXEC) &&
				(prot & PROT_READ);

	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;

	if (down_write_killable(&current->mm->mmap_sem))
		return -EINTR;

	/*
	 * If userspace did not allocate the pkey, do not let
	 * them use it here.
	 */
	error = -EINVAL;
	if ((pkey != -1) && !mm_pkey_is_allocated(current->mm, pkey))
		goto out;

	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 mask_off_old_flags;
		unsigned long newflags;
		int new_vma_pkey;

		/* Here we know that vma->vm_start <= nstart < vma->vm_end. */

		/* Does the application expect PROT_READ to imply PROT_EXEC */
		if (rier && (vma->vm_flags & VM_MAYEXEC))
			prot |= PROT_EXEC;

		/*
		 * Each mprotect() call explicitly passes r/w/x permissions.
		 * If a permission is not passed to mprotect(), it must be
		 * cleared from the VMA.
		 */
		mask_off_old_flags = VM_READ | VM_WRITE | VM_EXEC |
					ARCH_VM_PKEY_FLAGS;

		new_vma_pkey = arch_override_mprotect_pkey(vma, prot, pkey);
		newflags = calc_vm_prot_bits(prot, new_vma_pkey);
		newflags |= (vma->vm_flags & ~mask_off_old_flags);

		/* 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;
		}
		prot = reqprot;
	}
out:
	up_write(&current->mm->mmap_sem);
	return error;
}

SYSCALL_DEFINE3(mprotect, unsigned long, start, size_t, len,
		unsigned long, prot)
{
	return do_mprotect_pkey(start, len, prot, -1);
}

#ifdef CONFIG_ARCH_HAS_PKEYS

SYSCALL_DEFINE4(pkey_mprotect, unsigned long, start, size_t, len,
		unsigned long, prot, int, pkey)
{
	return do_mprotect_pkey(start, len, prot, pkey);
}

SYSCALL_DEFINE2(pkey_alloc, unsigned long, flags, unsigned long, init_val)
{
	int pkey;
	int ret;

	/* No flags supported yet. */
	if (flags)
		return -EINVAL;
	/* check for unsupported init values */
	if (init_val & ~PKEY_ACCESS_MASK)
		return -EINVAL;

	down_write(&current->mm->mmap_sem);
	pkey = mm_pkey_alloc(current->mm);

	ret = -ENOSPC;
	if (pkey == -1)
		goto out;

	ret = arch_set_user_pkey_access(current, pkey, init_val);
	if (ret) {
		mm_pkey_free(current->mm, pkey);
		goto out;
	}
	ret = pkey;
out:
	up_write(&current->mm->mmap_sem);
	return ret;
}

SYSCALL_DEFINE1(pkey_free, int, pkey)
{
	int ret;

	down_write(&current->mm->mmap_sem);
	ret = mm_pkey_free(current->mm, pkey);
	up_write(&current->mm->mmap_sem);

	/*
	 * We could provie warnings or errors if any VMA still
	 * has the pkey set here.
	 */
	return ret;
}

#endif /* CONFIG_ARCH_HAS_PKEYS */
Commit	Line	Data
b2441318	1	// SPDX-License-Identifier: GPL-2.0
1da177e4 LT	2	/*
	3	* mm/mprotect.c
	4	*
	5	* (C) Copyright 1994 Linus Torvalds
	6	* (C) Copyright 2002 Christoph Hellwig
	7	*
046c6884	8	* Address space accounting code <alan@lxorguk.ukuu.org.uk>
1da177e4 LT	9	* (C) Copyright 2002 Red Hat Inc, All Rights Reserved
	10	*/
	11
	12	#include <linux/mm.h>
	13	#include <linux/hugetlb.h>
1da177e4 LT	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>
e8c24d3a	27	#include <linux/pkeys.h>
64a9a34e	28	#include <linux/ksm.h>
7c0f6ba6	29	#include <linux/uaccess.h>
1da177e4 LT	30	#include <asm/pgtable.h>
1da177e4 LT	31	#include <asm/cacheflush.h>
e8c24d3a	32	#include <asm/mmu_context.h>
1da177e4 LT	33	#include <asm/tlbflush.h>
1da177e4 LT	34
36f88188 KS	35	#include "internal.h"
36f88188 KS	36
4b10e7d5	37	static unsigned long change_pte_range(struct vm_area_struct vma, pmd_t pmd,
c1e6098b	38	unsigned long addr, unsigned long end, pgprot_t newprot,
0f19c179	39	int dirty_accountable, int prot_numa)
1da177e4	40	{
4b10e7d5	41	struct mm_struct *mm = vma->vm_mm;
0697212a	42	pte_t *pte, oldpte;
705e87c0	43	spinlock_t *ptl;
7da4d641	44	unsigned long pages = 0;
3e321587	45	int target_node = NUMA_NO_NODE;
1da177e4	46
175ad4f1 AA	47	/*
	48	* Can be called with only the mmap_sem for reading by
	49	* prot_numa so we must check the pmd isn't constantly
	50	* changing from under us from pmd_none to pmd_trans_huge
	51	* and/or the other way around.
	52	*/
	53	if (pmd_trans_unstable(pmd))
	54	return 0;
	55
	56	/*
	57	* The pmd points to a regular pte so the pmd can't change
	58	* from under us even if the mmap_sem is only hold for
	59	* reading.
	60	*/
	61	pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl);
1ad9f620	62
3e321587 AK	63	/* Get target node for single threaded private VMAs */
	64	if (prot_numa && !(vma->vm_flags & VM_SHARED) &&
	65	atomic_read(&vma->vm_mm->mm_users) == 1)
	66	target_node = numa_node_id();
	67
3ea27719	68	flush_tlb_batched_pending(vma->vm_mm);
6606c3e0	69	arch_enter_lazy_mmu_mode();
1da177e4	70	do {
0697212a CL	71	oldpte = *pte;
0697212a CL	72	if (pte_present(oldpte)) {
1da177e4	73	pte_t ptent;
b191f9b1	74	bool preserve_write = prot_numa && pte_write(oldpte);
1da177e4	75
e944fd67 MG	76	/*
	77	* Avoid trapping faults against the zero or KSM
	78	* pages. See similar comment in change_huge_pmd.
	79	*/
	80	if (prot_numa) {
	81	struct page *page;
	82
	83	page = vm_normal_page(vma, addr, oldpte);
	84	if (!page \|\| PageKsm(page))
	85	continue;
10c1045f	86
859d4adc HW	87	/* Also skip shared copy-on-write pages */
	88	if (is_cow_mapping(vma->vm_flags) &&
	89	page_mapcount(page) != 1)
	90	continue;
	91
10c1045f MG	92	/* Avoid TLB flush if possible */
	93	if (pte_protnone(oldpte))
	94	continue;
3e321587 AK	95
	96	/*
	97	* Don't mess with PTEs if page is already on the node
	98	* a single-threaded process is running on.
	99	*/
	100	if (target_node == page_to_nid(page))
	101	continue;
e944fd67 MG	102	}
e944fd67 MG	103
8a0516ed MG	104	ptent = ptep_modify_prot_start(mm, addr, pte);
8a0516ed MG	105	ptent = pte_modify(ptent, newprot);
b191f9b1	106	if (preserve_write)
288bc549	107	ptent = pte_mk_savedwrite(ptent);
4b10e7d5	108
8a0516ed MG	109	/* Avoid taking write faults for known dirty pages */
	110	if (dirty_accountable && pte_dirty(ptent) &&
	111	(pte_soft_dirty(ptent) \|\|
	112	!(vma->vm_flags & VM_SOFTDIRTY))) {
	113	ptent = pte_mkwrite(ptent);
4b10e7d5	114	}
8a0516ed MG	115	ptep_modify_prot_commit(mm, addr, pte, ptent);
8a0516ed MG	116	pages++;
0661a336	117	} else if (IS_ENABLED(CONFIG_MIGRATION)) {
0697212a CL	118	swp_entry_t entry = pte_to_swp_entry(oldpte);
	119
	120	if (is_write_migration_entry(entry)) {
c3d16e16	121	pte_t newpte;
0697212a CL	122	/*
	123	* A protection check is difficult so
	124	* just be safe and disable write
	125	*/
	126	make_migration_entry_read(&entry);
c3d16e16 CG	127	newpte = swp_entry_to_pte(entry);
	128	if (pte_swp_soft_dirty(oldpte))
	129	newpte = pte_swp_mksoft_dirty(newpte);
	130	set_pte_at(mm, addr, pte, newpte);
e920e14c MG	131
e920e14c MG	132	pages++;
0697212a	133	}
5042db43 JG	134
	135	if (is_write_device_private_entry(entry)) {
	136	pte_t newpte;
	137
	138	/*
	139	* We do not preserve soft-dirtiness. See
	140	* copy_one_pte() for explanation.
	141	*/
	142	make_device_private_entry_read(&entry);
	143	newpte = swp_entry_to_pte(entry);
	144	set_pte_at(mm, addr, pte, newpte);
	145
	146	pages++;
	147	}
1da177e4 LT	148	}
1da177e4 LT	149	} while (pte++, addr += PAGE_SIZE, addr != end);
6606c3e0	150	arch_leave_lazy_mmu_mode();
705e87c0	151	pte_unmap_unlock(pte - 1, ptl);
7da4d641 PZ	152
7da4d641 PZ	153	return pages;
1da177e4 LT	154	}
1da177e4 LT	155
7d12efae AM	156	static inline unsigned long change_pmd_range(struct vm_area_struct *vma,
	157	pud_t *pud, unsigned long addr, unsigned long end,
	158	pgprot_t newprot, int dirty_accountable, int prot_numa)
1da177e4 LT	159	{
1da177e4 LT	160	pmd_t *pmd;
a5338093	161	struct mm_struct *mm = vma->vm_mm;
1da177e4	162	unsigned long next;
7da4d641	163	unsigned long pages = 0;
72403b4a	164	unsigned long nr_huge_updates = 0;
a5338093	165	unsigned long mni_start = 0;
1da177e4 LT	166
	167	pmd = pmd_offset(pud, addr);
	168	do {
25cbbef1 MG	169	unsigned long this_pages;
25cbbef1 MG	170
1da177e4	171	next = pmd_addr_end(addr, end);
84c3fc4e	172	if (!is_swap_pmd(pmd) && !pmd_trans_huge(pmd) && !pmd_devmap(*pmd)
5c7fb56e	173	&& pmd_none_or_clear_bad(pmd))
4991c09c	174	goto next;
a5338093 RR	175
	176	/* invoke the mmu notifier if the pmd is populated */
	177	if (!mni_start) {
	178	mni_start = addr;
	179	mmu_notifier_invalidate_range_start(mm, mni_start, end);
	180	}
	181
84c3fc4e	182	if (is_swap_pmd(pmd) \|\| pmd_trans_huge(pmd) \|\| pmd_devmap(*pmd)) {
6b9116a6	183	if (next - addr != HPAGE_PMD_SIZE) {
fd60775a	184	__split_huge_pmd(vma, pmd, addr, false, NULL);
6b9116a6	185	} else {
f123d74a	186	int nr_ptes = change_huge_pmd(vma, pmd, addr,
e944fd67	187	newprot, prot_numa);
f123d74a MG	188
f123d74a MG	189	if (nr_ptes) {
72403b4a MG	190	if (nr_ptes == HPAGE_PMD_NR) {
	191	pages += HPAGE_PMD_NR;
	192	nr_huge_updates++;
	193	}
1ad9f620 MG	194
1ad9f620 MG	195	/* huge pmd was handled */
4991c09c	196	goto next;
f123d74a	197	}
7da4d641	198	}
88a9ab6e	199	/* fall through, the trans huge pmd just split */
cd7548ab	200	}
25cbbef1	201	this_pages = change_pte_range(vma, pmd, addr, next, newprot,
0f19c179	202	dirty_accountable, prot_numa);
25cbbef1	203	pages += this_pages;
4991c09c AK	204	next:
4991c09c AK	205	cond_resched();
1da177e4	206	} while (pmd++, addr = next, addr != end);
7da4d641	207
a5338093 RR	208	if (mni_start)
	209	mmu_notifier_invalidate_range_end(mm, mni_start, end);
	210
72403b4a MG	211	if (nr_huge_updates)
72403b4a MG	212	count_vm_numa_events(NUMA_HUGE_PTE_UPDATES, nr_huge_updates);
7da4d641	213	return pages;
1da177e4 LT	214	}
1da177e4 LT	215
7d12efae	216	static inline unsigned long change_pud_range(struct vm_area_struct *vma,
c2febafc	217	p4d_t *p4d, unsigned long addr, unsigned long end,
7d12efae	218	pgprot_t newprot, int dirty_accountable, int prot_numa)
1da177e4 LT	219	{
	220	pud_t *pud;
	221	unsigned long next;
7da4d641	222	unsigned long pages = 0;
1da177e4	223
c2febafc	224	pud = pud_offset(p4d, addr);
1da177e4 LT	225	do {
	226	next = pud_addr_end(addr, end);
	227	if (pud_none_or_clear_bad(pud))
	228	continue;
7da4d641	229	pages += change_pmd_range(vma, pud, addr, next, newprot,
4b10e7d5	230	dirty_accountable, prot_numa);
1da177e4	231	} while (pud++, addr = next, addr != end);
7da4d641 PZ	232
7da4d641 PZ	233	return pages;
1da177e4 LT	234	}
1da177e4 LT	235
c2febafc KS	236	static inline unsigned long change_p4d_range(struct vm_area_struct *vma,
	237	pgd_t *pgd, unsigned long addr, unsigned long end,
	238	pgprot_t newprot, int dirty_accountable, int prot_numa)
	239	{
	240	p4d_t *p4d;
	241	unsigned long next;
	242	unsigned long pages = 0;
	243
	244	p4d = p4d_offset(pgd, addr);
	245	do {
	246	next = p4d_addr_end(addr, end);
	247	if (p4d_none_or_clear_bad(p4d))
	248	continue;
	249	pages += change_pud_range(vma, p4d, addr, next, newprot,
	250	dirty_accountable, prot_numa);
	251	} while (p4d++, addr = next, addr != end);
	252
	253	return pages;
	254	}
	255
7da4d641	256	static unsigned long change_protection_range(struct vm_area_struct *vma,
c1e6098b	257	unsigned long addr, unsigned long end, pgprot_t newprot,
4b10e7d5	258	int dirty_accountable, int prot_numa)
1da177e4 LT	259	{
	260	struct mm_struct *mm = vma->vm_mm;
	261	pgd_t *pgd;
	262	unsigned long next;
	263	unsigned long start = addr;
7da4d641	264	unsigned long pages = 0;
1da177e4 LT	265
	266	BUG_ON(addr >= end);
	267	pgd = pgd_offset(mm, addr);
	268	flush_cache_range(vma, addr, end);
16af97dc	269	inc_tlb_flush_pending(mm);
1da177e4 LT	270	do {
	271	next = pgd_addr_end(addr, end);
	272	if (pgd_none_or_clear_bad(pgd))
	273	continue;
c2febafc	274	pages += change_p4d_range(vma, pgd, addr, next, newprot,
4b10e7d5	275	dirty_accountable, prot_numa);
1da177e4	276	} while (pgd++, addr = next, addr != end);
7da4d641	277
1233d588 IM	278	/* Only flush the TLB if we actually modified any entries: */
	279	if (pages)
	280	flush_tlb_range(vma, start, end);
16af97dc	281	dec_tlb_flush_pending(mm);
7da4d641 PZ	282
	283	return pages;
	284	}
	285
	286	unsigned long change_protection(struct vm_area_struct *vma, unsigned long start,
	287	unsigned long end, pgprot_t newprot,
4b10e7d5	288	int dirty_accountable, int prot_numa)
7da4d641	289	{
7da4d641 PZ	290	unsigned long pages;
7da4d641 PZ	291
7da4d641 PZ	292	if (is_vm_hugetlb_page(vma))
	293	pages = hugetlb_change_protection(vma, start, end, newprot);
	294	else
4b10e7d5	295	pages = change_protection_range(vma, start, end, newprot, dirty_accountable, prot_numa);
7da4d641 PZ	296
7da4d641 PZ	297	return pages;
1da177e4 LT	298	}
1da177e4 LT	299
b6a2fea3	300	int
1da177e4 LT	301	mprotect_fixup(struct vm_area_struct vma, struct vm_area_struct *pprev,
	302	unsigned long start, unsigned long end, unsigned long newflags)
	303	{
	304	struct mm_struct *mm = vma->vm_mm;
	305	unsigned long oldflags = vma->vm_flags;
	306	long nrpages = (end - start) >> PAGE_SHIFT;
	307	unsigned long charged = 0;
1da177e4 LT	308	pgoff_t pgoff;
1da177e4 LT	309	int error;
c1e6098b	310	int dirty_accountable = 0;
1da177e4 LT	311
	312	if (newflags == oldflags) {
	313	*pprev = vma;
	314	return 0;
	315	}
	316
	317	/*
	318	* If we make a private mapping writable we increase our commit;
	319	* but (without finer accounting) cannot reduce our commit if we
5a6fe125 MG	320	* make it unwritable again. hugetlb mapping were accounted for
5a6fe125 MG	321	* even if read-only so there is no need to account for them here
1da177e4 LT	322	*/
1da177e4 LT	323	if (newflags & VM_WRITE) {
84638335 KK	324	/* Check space limits when area turns into data. */
	325	if (!may_expand_vm(mm, newflags, nrpages) &&
	326	may_expand_vm(mm, oldflags, nrpages))
	327	return -ENOMEM;
5a6fe125	328	if (!(oldflags & (VM_ACCOUNT\|VM_WRITE\|VM_HUGETLB\|
cdfd4325	329	VM_SHARED\|VM_NORESERVE))) {
1da177e4	330	charged = nrpages;
191c5424	331	if (security_vm_enough_memory_mm(mm, charged))
1da177e4 LT	332	return -ENOMEM;
	333	newflags \|= VM_ACCOUNT;
	334	}
	335	}
	336
1da177e4 LT	337	/*
	338	* First try to merge with previous and/or next vma.
	339	*/
	340	pgoff = vma->vm_pgoff + ((start - vma->vm_start) >> PAGE_SHIFT);
	341	pprev = vma_merge(mm, pprev, start, end, newflags,
19a809af AA	342	vma->anon_vma, vma->vm_file, pgoff, vma_policy(vma),
19a809af AA	343	vma->vm_userfaultfd_ctx);
1da177e4 LT	344	if (*pprev) {
1da177e4 LT	345	vma = *pprev;
e86f15ee	346	VM_WARN_ON((vma->vm_flags ^ newflags) & ~VM_SOFTDIRTY);
1da177e4 LT	347	goto success;
	348	}
	349
	350	*pprev = vma;
	351
	352	if (start != vma->vm_start) {
	353	error = split_vma(mm, vma, start, 1);
	354	if (error)
	355	goto fail;
	356	}
	357
	358	if (end != vma->vm_end) {
	359	error = split_vma(mm, vma, end, 0);
	360	if (error)
	361	goto fail;
	362	}
	363
	364	success:
	365	/*
	366	* vm_flags and vm_page_prot are protected by the mmap_sem
	367	* held in write mode.
	368	*/
	369	vma->vm_flags = newflags;
6d2329f8	370	dirty_accountable = vma_wants_writenotify(vma, vma->vm_page_prot);
64e45507	371	vma_set_page_prot(vma);
d08b3851	372
7d12efae AM	373	change_protection(vma, start, end, vma->vm_page_prot,
7d12efae AM	374	dirty_accountable, 0);
7da4d641	375
36f88188 KS	376	/*
	377	* Private VM_LOCKED VMA becoming writable: trigger COW to avoid major
	378	* fault on access.
	379	*/
	380	if ((oldflags & (VM_WRITE \| VM_SHARED \| VM_LOCKED)) == VM_LOCKED &&
	381	(newflags & VM_WRITE)) {
	382	populate_vma_page_range(vma, start, end, NULL);
	383	}
	384
84638335 KK	385	vm_stat_account(mm, oldflags, -nrpages);
84638335 KK	386	vm_stat_account(mm, newflags, nrpages);
63bfd738	387	perf_event_mmap(vma);
1da177e4 LT	388	return 0;
	389
	390	fail:
	391	vm_unacct_memory(charged);
	392	return error;
	393	}
	394
7d06d9c9 DH	395	/*
	396	* pkey==-1 when doing a legacy mprotect()
	397	*/
	398	static int do_mprotect_pkey(unsigned long start, size_t len,
	399	unsigned long prot, int pkey)
1da177e4	400	{
62b5f7d0	401	unsigned long nstart, end, tmp, reqprot;
1da177e4 LT	402	struct vm_area_struct vma, prev;
	403	int error = -EINVAL;
	404	const int grows = prot & (PROT_GROWSDOWN\|PROT_GROWSUP);
f138556d PK	405	const bool rier = (current->personality & READ_IMPLIES_EXEC) &&
	406	(prot & PROT_READ);
	407
1da177e4 LT	408	prot &= ~(PROT_GROWSDOWN\|PROT_GROWSUP);
	409	if (grows == (PROT_GROWSDOWN\|PROT_GROWSUP)) /* can't be both */
	410	return -EINVAL;
	411
	412	if (start & ~PAGE_MASK)
	413	return -EINVAL;
	414	if (!len)
	415	return 0;
	416	len = PAGE_ALIGN(len);
	417	end = start + len;
	418	if (end <= start)
	419	return -ENOMEM;
b845f313	420	if (!arch_validate_prot(prot))
1da177e4 LT	421	return -EINVAL;
	422
	423	reqprot = prot;
1da177e4	424
dc0ef0df MH	425	if (down_write_killable(&current->mm->mmap_sem))
dc0ef0df MH	426	return -EINTR;
1da177e4	427
e8c24d3a DH	428	/*
	429	* If userspace did not allocate the pkey, do not let
	430	* them use it here.
	431	*/
	432	error = -EINVAL;
	433	if ((pkey != -1) && !mm_pkey_is_allocated(current->mm, pkey))
	434	goto out;
	435
097d5910	436	vma = find_vma(current->mm, start);
1da177e4 LT	437	error = -ENOMEM;
	438	if (!vma)
	439	goto out;
097d5910	440	prev = vma->vm_prev;
1da177e4 LT	441	if (unlikely(grows & PROT_GROWSDOWN)) {
	442	if (vma->vm_start >= end)
	443	goto out;
	444	start = vma->vm_start;
	445	error = -EINVAL;
	446	if (!(vma->vm_flags & VM_GROWSDOWN))
	447	goto out;
7d12efae	448	} else {
1da177e4 LT	449	if (vma->vm_start > start)
	450	goto out;
	451	if (unlikely(grows & PROT_GROWSUP)) {
	452	end = vma->vm_end;
	453	error = -EINVAL;
	454	if (!(vma->vm_flags & VM_GROWSUP))
	455	goto out;
	456	}
	457	}
	458	if (start > vma->vm_start)
	459	prev = vma;
	460
	461	for (nstart = start ; ; ) {
a8502b67	462	unsigned long mask_off_old_flags;
1da177e4	463	unsigned long newflags;
7d06d9c9	464	int new_vma_pkey;
1da177e4	465
7d12efae	466	/* Here we know that vma->vm_start <= nstart < vma->vm_end. */
1da177e4	467
f138556d PK	468	/* Does the application expect PROT_READ to imply PROT_EXEC */
	469	if (rier && (vma->vm_flags & VM_MAYEXEC))
	470	prot \|= PROT_EXEC;
	471
a8502b67 DH	472	/*
	473	* Each mprotect() call explicitly passes r/w/x permissions.
	474	* If a permission is not passed to mprotect(), it must be
	475	* cleared from the VMA.
	476	*/
	477	mask_off_old_flags = VM_READ \| VM_WRITE \| VM_EXEC \|
	478	ARCH_VM_PKEY_FLAGS;
	479
7d06d9c9 DH	480	new_vma_pkey = arch_override_mprotect_pkey(vma, prot, pkey);
7d06d9c9 DH	481	newflags = calc_vm_prot_bits(prot, new_vma_pkey);
a8502b67	482	newflags \|= (vma->vm_flags & ~mask_off_old_flags);
1da177e4	483
7e2cff42 PBG	484	/* newflags >> 4 shift VM_MAY% in place of VM_% */
7e2cff42 PBG	485	if ((newflags & ~(newflags >> 4)) & (VM_READ \| VM_WRITE \| VM_EXEC)) {
1da177e4 LT	486	error = -EACCES;
	487	goto out;
	488	}
	489
	490	error = security_file_mprotect(vma, reqprot, prot);
	491	if (error)
	492	goto out;
	493
	494	tmp = vma->vm_end;
	495	if (tmp > end)
	496	tmp = end;
	497	error = mprotect_fixup(vma, &prev, nstart, tmp, newflags);
	498	if (error)
	499	goto out;
	500	nstart = tmp;
	501
	502	if (nstart < prev->vm_end)
	503	nstart = prev->vm_end;
	504	if (nstart >= end)
	505	goto out;
	506
	507	vma = prev->vm_next;
	508	if (!vma \|\| vma->vm_start != nstart) {
	509	error = -ENOMEM;
	510	goto out;
	511	}
f138556d	512	prot = reqprot;
1da177e4 LT	513	}
	514	out:
	515	up_write(&current->mm->mmap_sem);
	516	return error;
	517	}
7d06d9c9 DH	518
	519	SYSCALL_DEFINE3(mprotect, unsigned long, start, size_t, len,
	520	unsigned long, prot)
	521	{
	522	return do_mprotect_pkey(start, len, prot, -1);
	523	}
	524
c7142aea HC	525	#ifdef CONFIG_ARCH_HAS_PKEYS
c7142aea HC	526
7d06d9c9 DH	527	SYSCALL_DEFINE4(pkey_mprotect, unsigned long, start, size_t, len,
	528	unsigned long, prot, int, pkey)
	529	{
	530	return do_mprotect_pkey(start, len, prot, pkey);
	531	}
e8c24d3a DH	532
	533	SYSCALL_DEFINE2(pkey_alloc, unsigned long, flags, unsigned long, init_val)
	534	{
	535	int pkey;
	536	int ret;
	537
	538	/* No flags supported yet. */
	539	if (flags)
	540	return -EINVAL;
	541	/* check for unsupported init values */
	542	if (init_val & ~PKEY_ACCESS_MASK)
	543	return -EINVAL;
	544
	545	down_write(&current->mm->mmap_sem);
	546	pkey = mm_pkey_alloc(current->mm);
	547
	548	ret = -ENOSPC;
	549	if (pkey == -1)
	550	goto out;
	551
	552	ret = arch_set_user_pkey_access(current, pkey, init_val);
	553	if (ret) {
	554	mm_pkey_free(current->mm, pkey);
	555	goto out;
	556	}
	557	ret = pkey;
	558	out:
	559	up_write(&current->mm->mmap_sem);
	560	return ret;
	561	}
	562
	563	SYSCALL_DEFINE1(pkey_free, int, pkey)
	564	{
	565	int ret;
	566
	567	down_write(&current->mm->mmap_sem);
	568	ret = mm_pkey_free(current->mm, pkey);
	569	up_write(&current->mm->mmap_sem);
	570
	571	/*
	572	* We could provie warnings or errors if any VMA still
	573	* has the pkey set here.
	574	*/
	575	return ret;
	576	}
c7142aea HC	577
c7142aea HC	578	#endif /* CONFIG_ARCH_HAS_PKEYS */