Re: [syzbot] [mm?] possible deadlock in move_pages

[David Hildenbrand <david@redhat.com>]

On 19.03.24 10:52, syzbot wrote:
> Hello,
> 
> syzbot found the following issue on:
> 
> HEAD commit:    e5eb28f6d1af Merge tag 'mm-nonmm-stable-2024-03-14-09-36' ..
> git tree:       upstream
> console output: https://syzkaller.appspot.com/x/log.txt?x=160dc26e180000
> kernel config:  https://syzkaller.appspot.com/x/.config?x=4ffb854606e658d
> dashboard link: https://syzkaller.appspot.com/bug?extid=49056626fe41e01f2ba7
> compiler:       gcc (Debian 12.2.0-14) 12.2.0, GNU ld (GNU Binutils for Debian) 2.40
> syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=10f467b9180000
> C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=173b7ac9180000
> 
> Downloadable assets:
> disk image (non-bootable): https://storage.googleapis.com/syzbot-assets/7bc7510fe41f/non_bootable_disk-e5eb28f6.raw.xz
> vmlinux: https://storage.googleapis.com/syzbot-assets/a5c7ad05d6b2/vmlinux-e5eb28f6.xz
> kernel image: https://storage.googleapis.com/syzbot-assets/531cb1917612/bzImage-e5eb28f6.xz
> 
> IMPORTANT: if you fix the issue, please add the following tag to the commit:
> Reported-by: syzbot+49056626fe41e01f2ba7@syzkaller.appspotmail.com
> 
> ============================================
> WARNING: possible recursive locking detected
> 6.8.0-syzkaller-09791-ge5eb28f6d1af #0 Not tainted
> --------------------------------------------
> syz-executor258/5169 is trying to acquire lock:
> ffff88802a6d23d0 (&vma->vm_lock->lock){++++}-{3:3}, at: uffd_move_lock mm/userfaultfd.c:1447 [inline]
> ffff88802a6d23d0 (&vma->vm_lock->lock){++++}-{3:3}, at: move_pages+0xbab/0x4970 mm/userfaultfd.c:1583
> 
> but task is already holding lock:
> ffff88802a6d2580 (&vma->vm_lock->lock){++++}-{3:3}, at: uffd_move_lock mm/userfaultfd.c:1445 [inline]
> ffff88802a6d2580 (&vma->vm_lock->lock){++++}-{3:3}, at: move_pages+0xb6f/0x4970 mm/userfaultfd.c:1583
> 
> other info that might help us debug this:
>   Possible unsafe locking scenario:
> 
>         CPU0
>         ----
>    lock(&vma->vm_lock->lock);
>    lock(&vma->vm_lock->lock);
> 
>   *** DEADLOCK ***
> 
>   May be due to missing lock nesting notation
> 
> 2 locks held by syz-executor258/5169:
>   #0: ffff888015086a20 (&mm->mmap_lock){++++}-{3:3}, at: mmap_read_lock include/linux/mmap_lock.h:146 [inline]
>   #0: ffff888015086a20 (&mm->mmap_lock){++++}-{3:3}, at: uffd_move_lock mm/userfaultfd.c:1438 [inline]
>   #0: ffff888015086a20 (&mm->mmap_lock){++++}-{3:3}, at: move_pages+0x8df/0x4970 mm/userfaultfd.c:1583
>   #1: ffff88802a6d2580 (&vma->vm_lock->lock){++++}-{3:3}, at: uffd_move_lock mm/userfaultfd.c:1445 [inline]
>   #1: ffff88802a6d2580 (&vma->vm_lock->lock){++++}-{3:3}, at: move_pages+0xb6f/0x4970 mm/userfaultfd.c:1583
> 
> stack backtrace:
> CPU: 2 PID: 5169 Comm: syz-executor258 Not tainted 6.8.0-syzkaller-09791-ge5eb28f6d1af #0
> Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.2-debian-1.16.2-1 04/01/2014
> Call Trace:
>   <TASK>
>   __dump_stack lib/dump_stack.c:88 [inline]
>   dump_stack_lvl+0x116/0x1f0 lib/dump_stack.c:114
>   check_deadlock kernel/locking/lockdep.c:3062 [inline]
>   validate_chain kernel/locking/lockdep.c:3856 [inline]
>   __lock_acquire+0x20e6/0x3b30 kernel/locking/lockdep.c:5137
>   lock_acquire kernel/locking/lockdep.c:5754 [inline]
>   lock_acquire+0x1b1/0x540 kernel/locking/lockdep.c:5719
>   down_read+0x9a/0x330 kernel/locking/rwsem.c:1526
>   uffd_move_lock mm/userfaultfd.c:1447 [inline]
>   move_pages+0xbab/0x4970 mm/userfaultfd.c:1583
>   userfaultfd_move fs/userfaultfd.c:2008 [inline]
>   userfaultfd_ioctl+0x5e1/0x60e0 fs/userfaultfd.c:2126
>   vfs_ioctl fs/ioctl.c:51 [inline]
>   __do_sys_ioctl fs/ioctl.c:904 [inline]
>   __se_sys_ioctl fs/ioctl.c:890 [inline]
>   __x64_sys_ioctl+0x193/0x220 fs/ioctl.c:890
>   do_syscall_x64 arch/x86/entry/common.c:52 [inline]
>   do_syscall_64+0xd2/0x260 arch/x86/entry/common.c:83
>   entry_SYSCALL_64_after_hwframe+0x6d/0x75
> RIP: 0033:0x7fd48da20329
> Code: 48 83 c4 28 c3 e8 37 17 00 00 0f 1f 80 00 00 00 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:00007ffd1244f8e8 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
> RAX: ffffffffffffffda RBX: 00007ffd1244fab8 RCX: 00007fd48da20329
> RDX: 00000000200000c0 RSI: 00000000c028aa05 RDI: 0000000000000003
> RBP: 00007fd48da93610 R08: 00007ffd1244fab8 R09: 00007ffd1244fab8
> R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000001
> R13: 00007ffd1244faa8 R14: 0000000000000001 R15: 0000000000000001
>   </TASK>
> 
> 
> ---
> This report is generated by a bot. It may contain errors.
> See https://goo.gl/tpsmEJ for more information about syzbot.
> syzbot engineers can be reached at syzkaller@googlegroups.com.
> 
> syzbot will keep track of this issue. See:
> https://goo.gl/tpsmEJ#status for how to communicate with syzbot.
> 
> If the report is already addressed, let syzbot know by replying with:
> #syz fix: exact-commit-title
> 
> If you want syzbot to run the reproducer, reply with:
> #syz test: git://repo/address.git branch-or-commit-hash
> If you attach or paste a git patch, syzbot will apply it before testing.
> 
> If you want to overwrite report's subsystems, reply with:
> #syz set subsystems: new-subsystem
> (See the list of subsystem names on the web dashboard)
> 
> If the report is a duplicate of another one, reply with:
> #syz dup: exact-subject-of-another-report
> 
> If you want to undo deduplication, reply with:
> #syz undup
> 

Possibly

commit 867a43a34ff8a38772212045262b2c9b77807ea3
Author: Lokesh Gidra <lokeshgidra@google.com>
Date:   Thu Feb 15 10:27:56 2024 -0800

     userfaultfd: use per-vma locks in userfaultfd operations
     
     All userfaultfd operations, except write-protect, opportunistically use
     per-vma locks to lock vmas.  On failure, attempt again inside mmap_lock
     critical section.
     
     Write-protect operation requires mmap_lock as it iterates over multiple
     vmas.

and

commit 5e4c24a57b0c126686534b5b159a406c5dd02400
Author: Lokesh Gidra <lokeshgidra@google.com>
Date:   Thu Feb 15 10:27:54 2024 -0800

     userfaultfd: protect mmap_changing with rw_sem in userfaulfd_ctx
     
     Increments and loads to mmap_changing are always in mmap_lock critical
     section.  This ensures that if userspace requests event notification for
     non-cooperative operations (e.g.  mremap), userfaultfd operations don't
     occur concurrently.
     
     This can be achieved by using a separate read-write semaphore in
     userfaultfd_ctx such that increments are done in write-mode and loads in
     read-mode, thereby eliminating the dependency on mmap_lock for this
     purpose.
     
     This is a preparatory step before we replace mmap_lock usage with per-vma
     locks in fill/move ioctls.

might responsible.

CCin Lokesh

-- 
Cheers,

David / dhildenb