Re: KASAN: use-after-free Read in __do_page_fault

[Vlastimil Babka <vbabka@suse.cz>]

On 10/31/2017 08:15 PM, Andrea Arcangeli wrote:
> On Tue, Oct 31, 2017 at 03:28:26PM +0100, Vlastimil Babka wrote:
>> Hmm that could indeed work, Dmitry can you try the patch below?
>> But it still seems rather fragile so I'd hope Andrea can do it more
>> robust, or at least make sure that we don't reintroduce this kind of
>> problem in the future (explicitly set vma to NULL with a comment?).
> 
> Reviewed-by: Andrea Arcangeli <aarcange@redhat.com>

Thanks. OK so here's the full patch for the immediate issue, unless we
decide to do something more general.

----8<----
>From a5f887fcac65372f4e76a290ed59855de0b08e2e Mon Sep 17 00:00:00 2001
From: Vlastimil Babka <vbabka@suse.cz>
Date: Wed, 1 Nov 2017 08:21:25 +0100
Subject: [PATCH] x86/mm: fix use-after-free of vma during userfaultfd fault

Syzkaller with KASAN has reported a use-after-free of vma->vm_flags in
__do_page_fault() with the following reproducer:

#{Threaded:true Collide:true Repeat:true Procs:8 Sandbox:none Fault:false FaultCall:-1 FaultNth:0 EnableTun:true UseTmpDir:true HandleSegv:true WaitRepeat:true Debug:false Repro:false}
mmap(&(0x7f0000000000/0xfff000)=nil, 0xfff000, 0x3, 0x32, 0xffffffffffffffff, 0x0)
mmap(&(0x7f0000011000/0x3000)=nil, 0x3000, 0x1, 0x32, 0xffffffffffffffff, 0x0)
r0 = userfaultfd(0x0)
ioctl$UFFDIO_API(r0, 0xc018aa3f, &(0x7f0000002000-0x18)={0xaa, 0x0, 0x0})
ioctl$UFFDIO_REGISTER(r0, 0xc020aa00, &(0x7f0000019000)={{&(0x7f0000012000/0x2000)=nil, 0x2000}, 0x1, 0x0})
r1 = gettid()
syz_open_dev$evdev(&(0x7f0000013000-0x12)="2f6465762f696e7075742f6576656e742300", 0x0, 0x0)
tkill(r1, 0x7)

The vma should be pinned by mmap_sem, but handle_userfault() will in some
scenarios release it and then acquire again, so when we return to
__do_page_fault() with other result than VM_FAULT_RETRY, the vma might be gone.
However, since a3c4fb7c9c2e ("x86/mm: Fix fault error path using unsafe vma
pointer") there is a vma_pkey() read of vma->vm_flags after that point, which
can thus become use-after-free. Fix this by moving the read before calling
handle_mm_fault().

Reported-by: syzbot <bot+6a5269ce759a7bb12754ed9622076dc93f65a1f6@syzkaller.appspotmail.com>
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Suggested-by: Kirill A. Shutemov <kirill@shutemov.name>
Fixes: 3c4fb7c9c2e ("x86/mm: Fix fault error path using unsafe vma pointer")
Reviewed-by: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
---
 arch/x86/mm/fault.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c
index e2baeaa053a5..2f45a959aec2 100644
--- a/arch/x86/mm/fault.c
+++ b/arch/x86/mm/fault.c
@@ -1440,7 +1440,13 @@ __do_page_fault(struct pt_regs *regs, unsigned long error_code,
 	 * make sure we exit gracefully rather than endlessly redo
 	 * the fault.  Since we never set FAULT_FLAG_RETRY_NOWAIT, if
 	 * we get VM_FAULT_RETRY back, the mmap_sem has been unlocked.
+	 *
+	 * Since handle_userfault() may also release and reacquire mmap_sem
+	 * in some scenario (and not return VM_FAULT_RETRY), we have to be
+	 * careful about not touching vma after handling the fault. So we
+	 * read the pkey beforehand.
 	 */
+	pkey = vma_pkey(vma);
 	fault = handle_mm_fault(vma, address, flags);
 	major |= fault & VM_FAULT_MAJOR;
 
@@ -1467,7 +1473,6 @@ __do_page_fault(struct pt_regs *regs, unsigned long error_code,
 		return;
 	}
 
-	pkey = vma_pkey(vma);
 	up_read(&mm->mmap_sem);
 	if (unlikely(fault & VM_FAULT_ERROR)) {
 		mm_fault_error(regs, error_code, address, &pkey, fault);
-- 
2.14.3