Re: fs: use-after-free in path_lookupat

[Dmitry Vyukov]

On Sun, Mar 5, 2017 at 12:20 PM, Dmitry Vyukov <dvyukov@xxxxxxxxxx> wrote:
> On Sun, Mar 5, 2017 at 12:15 PM, Dmitry Vyukov <dvyukov@xxxxxxxxxx> wrote:
>> On Sat, Mar 4, 2017 at 8:39 PM, Al Viro <viro@xxxxxxxxxxxxxxxxxx> wrote:
>>> On Sat, Mar 04, 2017 at 03:59:36PM +0100, Dmitry Vyukov wrote:
>>>
>>>> I am getting the following use-after-free reports while running
>>>> syzkaller fuzzer on 86292b33d4b79ee03e2f43ea0381ef85f077c760 (but also
>>>> happened on 6dc39c50e4aeb769c8ae06edf2b1a732f3490913 and
>>>> c82be9d2244aacea9851c86f4fb74694c99cd874).
>>>
>>> IOW, it's not fs/namei.c patches from this window...
>>>
>>>>  unlazy_walk+0xf2/0x4b0 fs/namei.c:692
>>>
>>> Could you post disassembly (e.g. from objdump -d) of your unlazy_walk()?
>>> For the kernel the trace is from...
>>>
>>>> r4 = memfd_create(&(0x7f0000013000)="2f6465762f6877726e6700", 0x0)
>>>> name_to_handle_at(r4, &(0x7f0000003000-0x6)="2e2f62757300",
>>>> &(0x7f0000003000-0xd)={0xc, 0x0, "cd21"}, &(0x7f0000002000)=0x0,
>>>> 0x1000)
>>>>
>>>> What's strange is that dirfd passed to name_to_handle_at is memfd
>>>> handle (sic). And path lookup somehow does not fail early on this.
>>>> Does it make any sense?
>>>
>>> It doesn't, but is that the triggering call of name_to_handle_at(), or do you
>>> have it called elsewhere?
>>>
>>> FWIW, no LOOKUP_ROOT in filename_lookup() flags + NULL root + dfd not
>>> equal to AT_FDCWD + non-empty name should've ended up in
>>>                         if (!d_can_lookup(dentry)) {
>>>                                 fdput(f);
>>>                                 return ERR_PTR(-ENOTDIR);
>>>                         }
>>> in path_init() and it shouldn't have progressed any further.  And in case
>>> of name_to_handle_at() we have user_path_at(dfd, name, lookup_flags, &path),
>>> i.e. user_path_at_empty(dfd, name, lookup_flags, &path, NULL), i.e.
>>> filename_lookup(dfd, getname_flags(name, lookup_flags, NULL), lookup_flags,
>>> &path, NULL).  IOW, filename_lookup() is called with root equal to NULL,
>>> dfd and name coming straight from userland and lookup_flags containing
>>> nothing beyond LOOKUP_EMPTY and LOOKUP_FOLLOW...

Yes, but still it somehow happens...

I don't know if it's related or not, but in all cases the path passed
to memfd_create is used in openat (that "2f6465762f6877726e6700" which
stands for "/dev/hwrng"):

r3 = openat$hwrng(0xffffffffffffff9c,
&(0x7f000000f000)="2f6465762f6877726e6700", 0x0, 0x0)
mmap(&(0x7f0000013000/0x1000)=nil, (0x1000), 0x3, 0x32, 0xffffffffffffffff, 0x0)
r4 = memfd_create(&(0x7f0000013000)="2f6465762f6877726e6700", 0x0)
name_to_handle_at(r4, &(0x7f0000003000-0x6)="2e2f62757300",
&(0x7f0000003000-0xd)={0xc, 0x0, "cd21"}, &(0x7f0000002000)=0x0,
0x1000)



>> I probably messed something. Here is a new report on
>> 0710f3ff91ecc4a715db6e4d0690472b13c4dac6. Line numbers seem to match
>> now.
>>
>> BUG: KASAN: use-after-free in debug_spin_lock_before
>> kernel/locking/spinlock_debug.c:83 [inline] at addr ffff880059c2ace4
>> BUG: KASAN: use-after-free in do_raw_spin_lock+0x1bb/0x1f0
>> kernel/locking/spinlock_debug.c:112 at addr ffff880059c2ace4
>> Read of size 4 by task syz-executor/21853
>> CPU: 0 PID: 21853 Comm: syz-executor Not tainted 4.10.0+ #293
>> Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
>> Call Trace:
>>  __dump_stack lib/dump_stack.c:16 [inline]
>>  dump_stack+0x2fb/0x3fd lib/dump_stack.c:52
>>  kasan_object_err+0x1c/0x90 mm/kasan/report.c:166
>>  print_address_description mm/kasan/report.c:208 [inline]
>>  kasan_report_error mm/kasan/report.c:292 [inline]
>>  kasan_report.part.2+0x1b0/0x460 mm/kasan/report.c:314
>>  kasan_report mm/kasan/report.c:334 [inline]
>>  __asan_report_load4_noabort+0x29/0x30 mm/kasan/report.c:334
>>  debug_spin_lock_before kernel/locking/spinlock_debug.c:83 [inline]
>>  do_raw_spin_lock+0x1bb/0x1f0 kernel/locking/spinlock_debug.c:112
>>  __raw_spin_lock include/linux/spinlock_api_smp.h:143 [inline]
>>  _raw_spin_lock+0x3b/0x50 kernel/locking/spinlock.c:151
>>  spin_lock include/linux/spinlock.h:299 [inline]
>>  lockref_get_not_dead+0x19/0x80 lib/lockref.c:179
>>  legitimize_path.isra.36+0x7d/0x1a0 fs/namei.c:640
>>  unlazy_walk+0xf2/0x4b0 fs/namei.c:692
>>  complete_walk+0xb2/0x1f0 fs/namei.c:805
>>  path_lookupat+0x1c1/0x400 fs/namei.c:2275
>>  filename_lookup+0x282/0x540 fs/namei.c:2301
>>  user_path_at_empty+0x40/0x50 fs/namei.c:2555
>>  user_path_at include/linux/namei.h:55 [inline]
>>  SYSC_name_to_handle_at fs/fhandle.c:106 [inline]
>>  SyS_name_to_handle_at+0xff/0x720 fs/fhandle.c:92
>>  entry_SYSCALL_64_fastpath+0x1f/0xc2
>> RIP: 0033:0x4458d9
>> RSP: 002b:00007f462e243b58 EFLAGS: 00000286 ORIG_RAX: 000000000000012f
>> RAX: ffffffffffffffda RBX: 0000000000000051 RCX: 00000000004458d9
>> RDX: 0000000020002ff3 RSI: 0000000020002ffa RDI: 0000000000000051
>> RBP: 00000000006e11b0 R08: 0000000000001000 R09: 0000000000000000
>> R10: 0000000020002000 R11: 0000000000000286 R12: 0000000000708000
>> R13: 0000000020000000 R14: 0000000000013000 R15: 0000000000000003
>> Object at ffff880059c2ac60, in cache dentry size: 288
>> Allocated:
>> PID = 21878
>>  save_stack_trace+0x16/0x20 arch/x86/kernel/stacktrace.c:59
>>  save_stack+0x43/0xd0 mm/kasan/kasan.c:513
>>  set_track mm/kasan/kasan.c:525 [inline]
>>  kasan_kmalloc+0xaa/0xd0 mm/kasan/kasan.c:616
>>  kasan_slab_alloc+0x12/0x20 mm/kasan/kasan.c:555
>>  kmem_cache_alloc+0x102/0x6e0 mm/slab.c:3572
>>  __d_alloc+0xb3/0xbb0 fs/dcache.c:1571
>>  d_alloc_pseudo+0x1d/0x30 fs/dcache.c:1692
>>  __shmem_file_setup+0x20c/0x5a0 mm/shmem.c:4157
>>  shmem_file_setup mm/shmem.c:4212 [inline]
>>  SYSC_memfd_create mm/shmem.c:3672 [inline]
>>  SyS_memfd_create+0x172/0x2c0 mm/shmem.c:3630
>>  entry_SYSCALL_64_fastpath+0x1f/0xc2
>> Freed:
>> PID = 21878
>>  save_stack_trace+0x16/0x20 arch/x86/kernel/stacktrace.c:59
>>  save_stack+0x43/0xd0 mm/kasan/kasan.c:513
>>  set_track mm/kasan/kasan.c:525 [inline]
>>  kasan_slab_free+0x6f/0xb0 mm/kasan/kasan.c:589
>>  __cache_free mm/slab.c:3514 [inline]
>>  kmem_cache_free+0x71/0x240 mm/slab.c:3774
>>  __d_free fs/dcache.c:265 [inline]
>>  dentry_free+0xd5/0x160 fs/dcache.c:314
>>  __dentry_kill+0x471/0x6d0 fs/dcache.c:552
>>  dentry_kill fs/dcache.c:579 [inline]
>>  dput.part.25+0x5ce/0x7c0 fs/dcache.c:791
>>  dput+0x1f/0x30 fs/dcache.c:753
>>  __fput+0x538/0x800 fs/file_table.c:227
>>  ____fput+0x15/0x20 fs/file_table.c:245
>>  task_work_run+0x197/0x260 kernel/task_work.c:116
>>  tracehook_notify_resume include/linux/tracehook.h:191 [inline]
>>  exit_to_usermode_loop+0x23b/0x2a0 arch/x86/entry/common.c:161
>>  prepare_exit_to_usermode arch/x86/entry/common.c:191 [inline]
>>  syscall_return_slowpath+0x4d3/0x570 arch/x86/entry/common.c:260
>>  entry_SYSCALL_64_fastpath+0xc0/0xc2
>> Memory state around the buggy address:
>>  ffff880059c2ab80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>>  ffff880059c2ac00: 00 00 00 00 fc fc fc fc fc fc fc fc fb fb fb fb
>>>ffff880059c2ac80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
>>                                                        ^
>>  ffff880059c2ad00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
>>  ffff880059c2ad80: fc fc fc fc fc fc fc fc 00 00 00 00 00 00 00 00
>> ==================================================================
>
>
> Here is unlazy_walk if you still need it:
> https://gist.githubusercontent.com/dvyukov/03e3a5eef174628c6985e24c46870c69/raw/7d8600b364edf8508a716a389a6c69eb72942c28/gistfile1.txt
>
> complete_walk:
> https://gist.githubusercontent.com/dvyukov/e5f31d3539d9f332e6946ea32b6749f5/raw/95cd520d8ca0f6a18622da35bbc92f9e53120e56/gistfile1.txt
>
> legitimize_path:
> https://gist.githubusercontent.com/dvyukov/91e5c9bfc188124c16e3e0c15c597c96/raw/cd12ec81b0ff20320c216495c7e9189a86be145a/gistfile1.txt