fs: use-after-free in userfaultfd_exit

[Dmitry Vyukov]

Hello,

I've got the following use-after-free report while running syzkaller
fuzzer on 86292b33d4b79ee03e2f43ea0381ef85f077c760:

==================================================================
BUG: KASAN: use-after-free in userfaultfd_exit+0x251/0x270
fs/userfaultfd.c:803 at addr ffff88004cb91910
Read of size 8 by task syz-executor4/6218
CPU: 0 PID: 6218 Comm: syz-executor4 Not tainted 4.10.0+ #234
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
Call Trace:
__dump_stack lib/dump_stack.c:15 [inline]
dump_stack+0x2ee/0x3ef lib/dump_stack.c:51
kasan_object_err+0x1c/0x70 mm/kasan/report.c:166
print_address_description mm/kasan/report.c:204 [inline]
kasan_report_error mm/kasan/report.c:288 [inline]
kasan_report.part.2+0x198/0x440 mm/kasan/report.c:310
kasan_report mm/kasan/report.c:331 [inline]
__asan_report_load8_noabort+0x29/0x30 mm/kasan/report.c:331
userfaultfd_exit+0x251/0x270 fs/userfaultfd.c:803
exit_mm kernel/exit.c:551 [inline]
do_exit+0xa41/0x2900 kernel/exit.c:860
do_group_exit+0x149/0x420 kernel/exit.c:977
SYSC_exit_group kernel/exit.c:988 [inline]
SyS_exit_group+0x1d/0x20 kernel/exit.c:986
entry_SYSCALL_64_fastpath+0x1f/0xc2
RIP: 0033:0x4458d9
RSP: 002b:0000000000a5fe20 EFLAGS: 00000216 ORIG_RAX: 00000000000000e7
RAX: ffffffffffffffda RBX: 0000000000002dbb RCX: 00000000004458d9
RDX: 00000000004458d9 RSI: 00000000007541a0 RDI: 0000000000000000
RBP: 0000000000000000 R08: 0000001ddbc26dec R09: 0000000000000000
R10: 0000000000754198 R11: 0000000000000216 R12: 00000000000f70cd
R13: 00000000007080cc R14: 0000000020311000 R15: 00000000000f6fee
Object at ffff88004cb91900, in cache vm_area_struct size: 192
Allocated:
PID = 6219
save_stack_trace+0x16/0x20 arch/x86/kernel/stacktrace.c:57
save_stack+0x43/0xd0 mm/kasan/kasan.c:502
set_track mm/kasan/kasan.c:514 [inline]
kasan_kmalloc+0xaa/0xd0 mm/kasan/kasan.c:605
9pnet_virtio: no channels available for device ./bus
kasan_slab_alloc+0x12/0x20 mm/kasan/kasan.c:544
kmem_cache_alloc+0x102/0x680 mm/slab.c:3571
__split_vma+0x1b7/0x880 mm/mmap.c:2515
split_vma+0x8f/0xc0 mm/mmap.c:2578
userfaultfd_register fs/userfaultfd.c:1360 [inline]
userfaultfd_ioctl+0x3413/0x42f0 fs/userfaultfd.c:1739
vfs_ioctl fs/ioctl.c:43 [inline]
do_vfs_ioctl+0x1bf/0x1790 fs/ioctl.c:683
SYSC_ioctl fs/ioctl.c:698 [inline]
SyS_ioctl+0x8f/0xc0 fs/ioctl.c:689
entry_SYSCALL_64_fastpath+0x1f/0xc2
Freed:
PID = 6222
save_stack_trace+0x16/0x20 arch/x86/kernel/stacktrace.c:57
save_stack+0x43/0xd0 mm/kasan/kasan.c:502
set_track mm/kasan/kasan.c:514 [inline]
kasan_slab_free+0x6f/0xb0 mm/kasan/kasan.c:578
__cache_free mm/slab.c:3513 [inline]
kmem_cache_free+0x71/0x240 mm/slab.c:3773
__vma_adjust+0x954/0x1d40 mm/mmap.c:890
vma_merge+0xcc2/0x10c0 mm/mmap.c:1135
userfaultfd_release+0x3cf/0x6d0 fs/userfaultfd.c:840
__fput+0x332/0x7f0 fs/file_table.c:208
____fput+0x15/0x20 fs/file_table.c:244
task_work_run+0x18a/0x260 kernel/task_work.c:116
exit_task_work include/linux/task_work.h:21 [inline]
do_exit+0xafa/0x2900 kernel/exit.c:873
do_group_exit+0x149/0x420 kernel/exit.c:977
get_signal+0x7e0/0x1820 kernel/signal.c:2313
do_signal+0xd2/0x2190 arch/x86/kernel/signal.c:807
exit_to_usermode_loop+0x200/0x2a0 arch/x86/entry/common.c:156
prepare_exit_to_usermode arch/x86/entry/common.c:190 [inline]
syscall_return_slowpath+0x4d3/0x570 arch/x86/entry/common.c:259
entry_SYSCALL_64_fastpath+0xc0/0xc2
Memory state around the buggy address:
ffff88004cb91800: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
ffff88004cb91880: fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc fc
>ffff88004cb91900: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
^
ffff88004cb91980: fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc fc
ffff88004cb91a00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
==================================================================

[Andrea Arcangeli]

On Wed, Mar 01, 2017 at 07:48:00PM +0100, Dmitry Vyukov wrote:
> Hello,
>
> I've got the following use-after-free report while running syzkaller
> fuzzer on 86292b33d4b79ee03e2f43ea0381ef85f077c760:

Yes, I posted the fix for this one last Friday, I found it during
stress testing, it triggered the first time post-upstream merging
despite I was running the same stress testing with SLUB poisoning
enabled before.

This affects all apps, also the ones that don't use userfaultfd, it's
a locking issue. Furthermore the cost of userfaultfd_exit was not
acceptable, if something it had to be activated by a flag in mm->flags
(such an optimization would have been absolutely trivial though).

Thankfully I realized another feature (UFFDIO_COPY -ENOSPC retval) can
provide the same information at zero cost so I could drop
userfaultfd_exit as a whole.

https://marc.info/?l=linux-mm&m=148796041217814&w=2

The fix is already included in -mm along with the other fix for
VM_FAULT_NOPAGE.

Thanks,
Andrea