linux-kernel - [RFC PATCH 4/5] userfaultfd: mcopy_atomic: return -ENOENT when no compatible VMA found

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <1485265923-20256-5-git-send-email-rppt@linux.vnet.ibm.com>
Date:   Tue, 24 Jan 2017 15:52:02 +0200
From:   Mike Rapoport <rppt@...ux.vnet.ibm.com>
To:     Linux-MM <linux-mm@...ck.org>
Cc:     Andrea Arcangeli <aarcange@...hat.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        "Dr. David Alan Gilbert" <dgilbert@...hat.com>,
        Hillf Danton <hillf.zj@...baba-inc.com>,
        Mike Kravetz <mike.kravetz@...cle.com>,
        Pavel Emelyanov <xemul@...tuozzo.com>,
        LKML <linux-kernel@...r.kernel.org>,
        Mike Rapoport <rppt@...ux.vnet.ibm.com>
Subject: [RFC PATCH 4/5] userfaultfd: mcopy_atomic: return -ENOENT when no compatible VMA found

The memory mapping of a process may change between #PF event and the call
to mcopy_atomic that comes to resolve the page fault. In such case, there
will be no VMA covering the range passed to mcopy_atomic or the VMA will
not have userfaultfd context.
To allow uffd monitor to distinguish those case from other errors, let's
return -ENOENT instead of -EINVAL.

Note, that despite availability of UFFD_EVENT_UNMAP there still might be
race between the processing of UFFD_EVENT_UNMAP and outstanding
mcopy_atomic in case of non-cooperative uffd usage.

Signed-off-by: Mike Rapoport <rppt@...ux.vnet.ibm.com>
---
 mm/userfaultfd.c | 42 ++++++++++++++++++++++--------------------
 1 file changed, 22 insertions(+), 20 deletions(-)

diff --git a/mm/userfaultfd.c b/mm/userfaultfd.c
index a0817cc..b861cf9 100644
--- a/mm/userfaultfd.c
+++ b/mm/userfaultfd.c
@@ -195,11 +195,18 @@ static __always_inline ssize_t __mcopy_atomic_hugetlb(struct mm_struct *dst_mm,
 	 * retry, dst_vma will be set to NULL and we must lookup again.
 	 */
 	if (!dst_vma) {
-		err = -EINVAL;
+		err = -ENOENT;
 		dst_vma = find_vma(dst_mm, dst_start);
 		if (!dst_vma || !is_vm_hugetlb_page(dst_vma))
 			goto out_unlock;
+		/*
+		 * Only allow __mcopy_atomic_hugetlb on userfaultfd
+		 * registered ranges.
+		 */
+		if (!dst_vma->vm_userfaultfd_ctx.ctx)
+			goto out_unlock;
 
+		err = -EINVAL;
 		if (vma_hpagesize != vma_kernel_pagesize(dst_vma))
 			goto out_unlock;
 
@@ -219,12 +226,6 @@ static __always_inline ssize_t __mcopy_atomic_hugetlb(struct mm_struct *dst_mm,
 		goto out_unlock;
 
 	/*
-	 * Only allow __mcopy_atomic_hugetlb on userfaultfd registered ranges.
-	 */
-	if (!dst_vma->vm_userfaultfd_ctx.ctx)
-		goto out_unlock;
-
-	/*
 	 * Ensure the dst_vma has a anon_vma.
 	 */
 	err = -ENOMEM;
@@ -368,10 +369,23 @@ static __always_inline ssize_t __mcopy_atomic(struct mm_struct *dst_mm,
 	 * Make sure the vma is not shared, that the dst range is
 	 * both valid and fully within a single existing vma.
 	 */
-	err = -EINVAL;
+	err = -ENOENT;
 	dst_vma = find_vma(dst_mm, dst_start);
 	if (!dst_vma)
 		goto out_unlock;
+	/*
+	 * Be strict and only allow __mcopy_atomic on userfaultfd
+	 * registered ranges to prevent userland errors going
+	 * unnoticed. As far as the VM consistency is concerned, it
+	 * would be perfectly safe to remove this check, but there's
+	 * no useful usage for __mcopy_atomic ouside of userfaultfd
+	 * registered ranges. This is after all why these are ioctls
+	 * belonging to the userfaultfd and not syscalls.
+	 */
+	if (!dst_vma->vm_userfaultfd_ctx.ctx)
+		goto out_unlock;
+
+	err = -EINVAL;
 	if (!vma_is_shmem(dst_vma) && dst_vma->vm_flags & VM_SHARED)
 		goto out_unlock;
 	if (dst_start < dst_vma->vm_start ||
@@ -385,18 +399,6 @@ static __always_inline ssize_t __mcopy_atomic(struct mm_struct *dst_mm,
 		return  __mcopy_atomic_hugetlb(dst_mm, dst_vma, dst_start,
 						src_start, len, zeropage);
 
-	/*
-	 * Be strict and only allow __mcopy_atomic on userfaultfd
-	 * registered ranges to prevent userland errors going
-	 * unnoticed. As far as the VM consistency is concerned, it
-	 * would be perfectly safe to remove this check, but there's
-	 * no useful usage for __mcopy_atomic ouside of userfaultfd
-	 * registered ranges. This is after all why these are ioctls
-	 * belonging to the userfaultfd and not syscalls.
-	 */
-	if (!dst_vma->vm_userfaultfd_ctx.ctx)
-		goto out_unlock;
-
 	if (!vma_is_anonymous(dst_vma) && !vma_is_shmem(dst_vma))
 		goto out_unlock;
 
-- 
1.9.1