Re: [syzbot] KASAN: use-after-free Read in hugetlb_fault

[Mike Kravetz]

On 10/23/22 00:26, syzbot wrote:
> Hello,
> 
> syzbot found the following issue on:
> 
> HEAD commit:    bb1a1146467a Merge tag 'cgroup-for-6.1-rc1-fixes' of git:/..
> git tree:       upstream
> console output: https://syzkaller.appspot.com/x/log.txt?x=15f38cc2880000
> kernel config:  https://syzkaller.appspot.com/x/.config?x=972ee691c428ae39
> dashboard link: https://syzkaller.appspot.com/bug?extid=1b27d7a2722eabc2c5d5
> compiler:       gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2
> 
> Unfortunately, I don't have any reproducer for this issue yet.
> 
> IMPORTANT: if you fix the issue, please add the following tag to the commit:
> Reported-by: syzbot+1b27d7a2722eabc2c5d5@xxxxxxxxxxxxxxxxxxxxxxxxx
> 
> ==================================================================
> BUG: KASAN: use-after-free in instrument_atomic_read include/linux/instrumented.h:72 [inline]
> BUG: KASAN: use-after-free in atomic_long_read include/linux/atomic/atomic-instrumented.h:1265 [inline]
> BUG: KASAN: use-after-free in is_rwsem_reader_owned kernel/locking/rwsem.c:193 [inline]
> BUG: KASAN: use-after-free in __down_read_common kernel/locking/rwsem.c:1262 [inline]
> BUG: KASAN: use-after-free in __down_read_common kernel/locking/rwsem.c:1255 [inline]
> BUG: KASAN: use-after-free in __down_read kernel/locking/rwsem.c:1269 [inline]
> BUG: KASAN: use-after-free in down_read+0x1d3/0x450 kernel/locking/rwsem.c:1511
> Read of size 8 at addr ffff88801eabdc08 by task syz-executor.2/4171
> 
> CPU: 0 PID: 4171 Comm: syz-executor.2 Not tainted 6.1.0-rc1-syzkaller-00010-gbb1a1146467a #0
> Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014
> Call Trace:
>  <TASK>
>  __dump_stack lib/dump_stack.c:88 [inline]
>  dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106
>  print_address_description mm/kasan/report.c:284 [inline]
>  print_report+0x15e/0x461 mm/kasan/report.c:395
>  kasan_report+0xbb/0x1f0 mm/kasan/report.c:495
>  check_region_inline mm/kasan/generic.c:183 [inline]
>  kasan_check_range+0x13d/0x180 mm/kasan/generic.c:189
>  instrument_atomic_read include/linux/instrumented.h:72 [inline]
>  atomic_long_read include/linux/atomic/atomic-instrumented.h:1265 [inline]
>  is_rwsem_reader_owned kernel/locking/rwsem.c:193 [inline]
>  __down_read_common kernel/locking/rwsem.c:1262 [inline]
>  __down_read_common kernel/locking/rwsem.c:1255 [inline]
>  __down_read kernel/locking/rwsem.c:1269 [inline]
>  down_read+0x1d3/0x450 kernel/locking/rwsem.c:1511
>  hugetlb_vma_lock_read mm/hugetlb.c:6797 [inline]
>  hugetlb_fault+0x40a/0x2060 mm/hugetlb.c:5840
>  follow_hugetlb_page+0x3f3/0x1850 mm/hugetlb.c:6282
>  __get_user_pages+0x2cb/0xf10 mm/gup.c:1202
>  populate_vma_page_range+0x23d/0x320 mm/gup.c:1587
>  __mm_populate+0x101/0x3a0 mm/gup.c:1701
>  mm_populate include/linux/mm.h:2744 [inline]
>  vm_mmap_pgoff+0x1fd/0x270 mm/util.c:525
>  ksys_mmap_pgoff+0x1c3/0x5a0 mm/mmap.c:1457
>  do_syscall_x64 arch/x86/entry/common.c:50 [inline]
>  do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
>  entry_SYSCALL_64_after_hwframe+0x63/0xcd
> RIP: 0033:0x7f07e488b5a9
> Code: ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b8 ff ff ff f7 d8 64 89 01 48
> RSP: 002b:00007f07e59e7168 EFLAGS: 00000246 ORIG_RAX: 0000000000000009
> RAX: ffffffffffffffda RBX: 00007f07e49abf80 RCX: 00007f07e488b5a9
> RDX: 0000000000000003 RSI: 0000000000b36000 RDI: 0000000020000000
> RBP: 00007f07e48e6580 R08: ffffffffffffffff R09: 0000000000000000
> R10: 0000000000068831 R11: 0000000000000246 R12: 0000000000000000
> R13: 00007ffc053fba9f R14: 00007f07e59e7300 R15: 0000000000022000
>  </TASK>
> 
> Allocated by task 4171:
>  kasan_save_stack+0x1e/0x40 mm/kasan/common.c:45
>  kasan_set_track+0x21/0x30 mm/kasan/common.c:52
>  ____kasan_kmalloc mm/kasan/common.c:371 [inline]
>  ____kasan_kmalloc mm/kasan/common.c:330 [inline]
>  __kasan_kmalloc+0x9f/0xb0 mm/kasan/common.c:380
>  kmalloc include/linux/slab.h:576 [inline]
>  hugetlb_vma_lock_alloc.part.0+0x3f/0x130 mm/hugetlb.c:6907
>  hugetlb_vma_lock_alloc mm/hugetlb.c:6663 [inline]
>  hugetlb_reserve_pages+0xa3f/0xe80 mm/hugetlb.c:6532
>  hugetlbfs_file_mmap+0x40c/0x5c0 fs/hugetlbfs/inode.c:167
>  call_mmap include/linux/fs.h:2196 [inline]
>  mmap_region+0x6bf/0x1bf0 mm/mmap.c:2624
>  do_mmap+0x825/0xf50 mm/mmap.c:1411
>  vm_mmap_pgoff+0x1ab/0x270 mm/util.c:520
>  ksys_mmap_pgoff+0x1c3/0x5a0 mm/mmap.c:1457
>  do_syscall_x64 arch/x86/entry/common.c:50 [inline]
>  do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
>  entry_SYSCALL_64_after_hwframe+0x63/0xcd
> 
> Freed by task 4172:
>  kasan_save_stack+0x1e/0x40 mm/kasan/common.c:45
>  kasan_set_track+0x21/0x30 mm/kasan/common.c:52
>  kasan_save_free_info+0x27/0x40 mm/kasan/generic.c:511
>  ____kasan_slab_free mm/kasan/common.c:236 [inline]
>  ____kasan_slab_free+0x13b/0x1a0 mm/kasan/common.c:200
>  kasan_slab_free include/linux/kasan.h:177 [inline]
>  __cache_free mm/slab.c:3389 [inline]
>  __do_kmem_cache_free mm/slab.c:3585 [inline]
>  __kmem_cache_free+0xc9/0x3a0 mm/slab.c:3592
>  hugetlb_vma_lock_release mm/hugetlb.c:6852 [inline]
>  kref_put include/linux/kref.h:65 [inline]
>  __hugetlb_vma_unlock_write_put mm/hugetlb.c:6867 [inline]
>  __hugetlb_vma_unlock_write_free mm/hugetlb.c:6875 [inline]
>  __unmap_hugepage_range_final+0x2ad/0x340 mm/hugetlb.c:5196
>  unmap_single_vma+0x23d/0x2a0 mm/memory.c:1690
>  zap_page_range+0x38a/0x520 mm/memory.c:1762
>  madvise_dontneed_single_vma mm/madvise.c:793 [inline]
>  madvise_dontneed_free mm/madvise.c:876 [inline]
>  madvise_vma_behavior+0xee8/0x1cc0 mm/madvise.c:1015
>  madvise_walk_vmas+0x1c7/0x2b0 mm/madvise.c:1240
>  do_madvise.part.0+0x24a/0x340 mm/madvise.c:1419
>  do_madvise mm/madvise.c:1432 [inline]
>  __do_sys_madvise mm/madvise.c:1432 [inline]
>  __se_sys_madvise mm/madvise.c:1430 [inline]
>  __x64_sys_madvise+0x113/0x150 mm/madvise.c:1430
>  do_syscall_x64 arch/x86/entry/common.c:50 [inline]
>  do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
>  entry_SYSCALL_64_after_hwframe+0x63/0xcd

Looks like page fault is racing with the vma_lock freeing in
__unmap_hugepage_range_final as the result of a madvise(MADV_DONTNEED).

It should be addressed with,
https://lore.kernel.org/linux-mm/20221023025047.470646-1-mike.kravetz@xxxxxxxxxx/

-- 
Mike Kravetz