userfaultfd: shmem: add i_size checks

With MAP_SHARED: recheck the i_size after taking the PT lock, to
serialize against truncate with the PT lock.  Delete the page from the
pagecache if the i_size_read check fails.

With MAP_PRIVATE: check the i_size after the PT lock before mapping
anonymous memory or zeropages into the MAP_PRIVATE shmem mapping.

A mostly irrelevant cleanup: like we do the delete_from_page_cache()
pagecache removal after dropping the PT lock, the PT lock is a spinlock
so drop it before the sleepable page lock.

Link: http://lkml.kernel.org/r/[email protected]
Fixes: 4c27fe4 ("userfaultfd: shmem: add shmem_mcopy_atomic_pte for userfaultfd support")
Signed-off-by: Andrea Arcangeli <[email protected]>
Reviewed-by: Mike Rapoport <[email protected]>
Reviewed-by: Hugh Dickins <[email protected]>
Reported-by: Jann Horn <[email protected]>
Cc: <[email protected]>
Cc: "Dr. David Alan Gilbert" <[email protected]>
Cc: Mike Kravetz <[email protected]>
Cc: Peter Xu <[email protected]>
Cc: [email protected]
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

mm/shmem.c

@@ -2216,6 +2216,7 @@ static int shmem_mfill_atomic_pte(struct mm_struct *dst_mm,
 	struct page *page;
 	pte_t _dst_pte, *dst_pte;
 	int ret;
+	pgoff_t offset, max_off;
 
 	ret = -ENOMEM;
 	if (!shmem_inode_acct_block(inode, 1))
@@ -2253,6 +2254,12 @@ static int shmem_mfill_atomic_pte(struct mm_struct *dst_mm,
 	__SetPageSwapBacked(page);
 	__SetPageUptodate(page);
 
+	ret = -EFAULT;
+	offset = linear_page_index(dst_vma, dst_addr);
+	max_off = DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE);
+	if (unlikely(offset >= max_off))
+		goto out_release;
+
 	ret = mem_cgroup_try_charge_delay(page, dst_mm, gfp, &memcg, false);
 	if (ret)
 		goto out_release;
@@ -2268,8 +2275,14 @@ static int shmem_mfill_atomic_pte(struct mm_struct *dst_mm,
 	if (dst_vma->vm_flags & VM_WRITE)
 		_dst_pte = pte_mkwrite(pte_mkdirty(_dst_pte));
 
-	ret = -EEXIST;
 	dst_pte = pte_offset_map_lock(dst_mm, dst_pmd, dst_addr, &ptl);
+
+	ret = -EFAULT;
+	max_off = DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE);
+	if (unlikely(offset >= max_off))
+		goto out_release_uncharge_unlock;
+
+	ret = -EEXIST;
 	if (!pte_none(*dst_pte))
 		goto out_release_uncharge_unlock;
 
@@ -2287,13 +2300,14 @@ static int shmem_mfill_atomic_pte(struct mm_struct *dst_mm,
 
 	/* No need to invalidate - it was non-present before */
 	update_mmu_cache(dst_vma, dst_addr, dst_pte);
-	unlock_page(page);
 	pte_unmap_unlock(dst_pte, ptl);
+	unlock_page(page);
 	ret = 0;
 out:
 	return ret;
 out_release_uncharge_unlock:
 	pte_unmap_unlock(dst_pte, ptl);
+	delete_from_page_cache(page);
 out_release_uncharge:
 	mem_cgroup_cancel_charge(page, memcg, false);
 out_release:

mm/userfaultfd.c

@@ -33,6 +33,8 @@ static int mcopy_atomic_pte(struct mm_struct *dst_mm,
 	void *page_kaddr;
 	int ret;
 	struct page *page;
+	pgoff_t offset, max_off;
+	struct inode *inode;
 
 	if (!*pagep) {
 		ret = -ENOMEM;
@@ -73,8 +75,17 @@ static int mcopy_atomic_pte(struct mm_struct *dst_mm,
 	if (dst_vma->vm_flags & VM_WRITE)
 		_dst_pte = pte_mkwrite(pte_mkdirty(_dst_pte));
 
-	ret = -EEXIST;
 	dst_pte = pte_offset_map_lock(dst_mm, dst_pmd, dst_addr, &ptl);
+	if (dst_vma->vm_file) {
+		/* the shmem MAP_PRIVATE case requires checking the i_size */
+		inode = dst_vma->vm_file->f_inode;
+		offset = linear_page_index(dst_vma, dst_addr);
+		max_off = DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE);
+		ret = -EFAULT;
+		if (unlikely(offset >= max_off))
+			goto out_release_uncharge_unlock;
+	}
+	ret = -EEXIST;
 	if (!pte_none(*dst_pte))
 		goto out_release_uncharge_unlock;
 
@@ -108,11 +119,22 @@ static int mfill_zeropage_pte(struct mm_struct *dst_mm,
 	pte_t _dst_pte, *dst_pte;
 	spinlock_t *ptl;
 	int ret;
+	pgoff_t offset, max_off;
+	struct inode *inode;
 
 	_dst_pte = pte_mkspecial(pfn_pte(my_zero_pfn(dst_addr),
 					 dst_vma->vm_page_prot));
-	ret = -EEXIST;
 	dst_pte = pte_offset_map_lock(dst_mm, dst_pmd, dst_addr, &ptl);
+	if (dst_vma->vm_file) {
+		/* the shmem MAP_PRIVATE case requires checking the i_size */
+		inode = dst_vma->vm_file->f_inode;
+		offset = linear_page_index(dst_vma, dst_addr);
+		max_off = DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE);
+		ret = -EFAULT;
+		if (unlikely(offset >= max_off))
+			goto out_unlock;
+	}
+	ret = -EEXIST;
 	if (!pte_none(*dst_pte))
 		goto out_unlock;
 	set_pte_at(dst_mm, dst_addr, dst_pte, _dst_pte);