Re: [syzbot] [crypto?] KASAN: slab-use-after-free Read in handle_mm_fault

[Liam R. Howlett]

* Vlastimil Babka (SUSE) <vbabka@xxxxxxxxxx> [240718 07:00]:
> On 7/16/24 10:29 AM, syzbot wrote:
> > Hello,
> 
> dunno about the [crypto?] parts, sounds rather something for Suren or Liam
> or maybe it's due to some changes to gup?

Yes, that crypto part is very odd.

> 
> > syzbot found the following issue on:
> > 
> > HEAD commit:    3fe121b62282 Add linux-next specific files for 20240712
> > git tree:       linux-next
> > console output: https://syzkaller.appspot.com/x/log.txt?x=1097ebed980000
> > kernel config:  https://syzkaller.appspot.com/x/.config?x=98dd8c4bab5cdce
> > dashboard link: https://syzkaller.appspot.com/bug?extid=4c882a4a0697c4a25364
> > compiler:       Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40
> > syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=11d611a5980000
> > C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=13ce3259980000
> > 
> > Downloadable assets:
> > disk image: https://storage.googleapis.com/syzbot-assets/8c6fbf69718d/disk-3fe121b6.raw.xz
> > vmlinux: https://storage.googleapis.com/syzbot-assets/39fc7e43dfc1/vmlinux-3fe121b6.xz
> > kernel image: https://storage.googleapis.com/syzbot-assets/0a78e70e4b4e/bzImage-3fe121b6.xz
> > mounted in repro: https://storage.googleapis.com/syzbot-assets/66cfe5a679f2/mount_0.gz
> > 
> > IMPORTANT: if you fix the issue, please add the following tag to the commit:
> > Reported-by: syzbot+4c882a4a0697c4a25364@xxxxxxxxxxxxxxxxxxxxxxxxx
> > 
> > ==================================================================
> > BUG: KASAN: slab-use-after-free in handle_mm_fault+0x14f0/0x19a0 mm/memory.c:5842
> > Read of size 8 at addr ffff88802c4719d0 by task syz-executor125/5235
> > 
> > CPU: 1 UID: 0 PID: 5235 Comm: syz-executor125 Not tainted 6.10.0-rc7-next-20240712-syzkaller #0
> > Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 06/07/2024
> > Call Trace:
> >  <TASK>
> >  __dump_stack lib/dump_stack.c:94 [inline]
> >  dump_stack_lvl+0x241/0x360 lib/dump_stack.c:120
> >  print_address_description mm/kasan/report.c:377 [inline]
> >  print_report+0x169/0x550 mm/kasan/report.c:488
> >  kasan_report+0x143/0x180 mm/kasan/report.c:601
> >  handle_mm_fault+0x14f0/0x19a0 mm/memory.c:5842

/*
 * By the time we get here, we already hold the mm semaphore
 *
 * The mmap_lock may have been released depending on flags and our
 * return value.  See filemap_fault() and __folio_lock_or_retry().
 */

Somehow we are here without an RCU or mmap_lock held?

> >  faultin_page mm/gup.c:1194 [inline]

/*
 * mmap_lock must be held on entry.  If @flags has FOLL_UNLOCKABLE but not
 * FOLL_NOWAIT, the mmap_lock may be released.  If it is, *@locked will be set
 * to 0 and -EBUSY returned.
 */

We should probably have a lockdep check there then?

> >  __get_user_pages+0x6ec/0x16a0 mm/gup.c:1493
> >  populate_vma_page_range+0x264/0x330 mm/gup.c:1932
> >  __mm_populate+0x27a/0x460 mm/gup.c:2035

/*
 * __mm_populate - populate and/or mlock pages within a range of address space.
 *
 * This is used to implement mlock() and the MAP_POPULATE / MAP_LOCKED mmap
 * flags. VMAs must be already marked with the desired vm_flags, and
 * mmap_lock must not be held.
 */

What ensures the vma doesn't go away then? - I guess nothing, because it
went away.

I don't get it.. __mm_populate() must NOT have the mmap_lock, but
faultin_page() must hold the mmap_lock...

> >  mm_populate include/linux/mm.h:3429 [inline]
> >  vm_mmap_pgoff+0x2c3/0x3d0 mm/util.c:593
> >  do_syscall_x64 arch/x86/entry/common.c:52 [inline]
> >  do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83
> >  entry_SYSCALL_64_after_hwframe+0x77/0x7f
> > RIP: 0033:0x7f093ce17fe9
> > Code: 28 00 00 00 75 05 48 83 c4 28 c3 e8 e1 1d 00 00 90 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 b0 ff ff ff f7 d8 64 89 01 48
> > RSP: 002b:00007f093cd9e158 EFLAGS: 00000246 ORIG_RAX: 0000000000000009
> > RAX: ffffffffffffffda RBX: 00007f093ce9f4b8 RCX: 00007f093ce17fe9
> > RDX: 0000000000000002 RSI: 0000000000b36000 RDI: 0000000020000000
> > RBP: 00007f093ce9f4b0 R08: 00000000ffffffff R09: 0000000000000000
> > R10: 0000000000008031 R11: 0000000000000246 R12: 00007f093ce9f4bc
> > R13: 000000000000006e R14: 00007ffe8008cc30 R15: 00007ffe8008cd18
> >  </TASK>
> > 
> > Allocated by task 5235:
...

> > 
> > Freed by task 5237:
> >  kasan_save_stack mm/kasan/common.c:47 [inline]
> >  kasan_save_track+0x3f/0x80 mm/kasan/common.c:68
> >  kasan_save_free_info+0x40/0x50 mm/kasan/generic.c:579
> >  poison_slab_object+0xe0/0x150 mm/kasan/common.c:240
> >  __kasan_slab_free+0x37/0x60 mm/kasan/common.c:256
> >  kasan_slab_free include/linux/kasan.h:184 [inline]
> >  slab_free_hook mm/slub.c:2252 [inline]
> >  slab_free mm/slub.c:4473 [inline]
> >  kmem_cache_free+0x145/0x350 mm/slub.c:4548
> >  rcu_do_batch kernel/rcu/tree.c:2569 [inline]
> >  rcu_core+0xafd/0x1830 kernel/rcu/tree.c:2843

This seems right.  RCU freeing of a vma here, so that's okay.

> >  handle_softirqs+0x2c4/0x970 kernel/softirq.c:554
> >  __do_softirq kernel/softirq.c:588 [inline]
> >  invoke_softirq kernel/softirq.c:428 [inline]
> >  __irq_exit_rcu+0xf4/0x1c0 kernel/softirq.c:637
> >  irq_exit_rcu+0x9/0x30 kernel/softirq.c:649
> >  instr_sysvec_apic_timer_interrupt arch/x86/kernel/apic/apic.c:1043 [inline]
> >  sysvec_apic_timer_interrupt+0xa6/0xc0 arch/x86/kernel/apic/apic.c:1043
> >  asm_sysvec_apic_timer_interrupt+0x1a/0x20 arch/x86/include/asm/idtentry.h:702
> > 
> > Last potentially related work creation:

Also fine.

> >  kasan_save_stack+0x3f/0x60 mm/kasan/common.c:47
> >  __kasan_record_aux_stack+0xac/0xc0 mm/kasan/generic.c:541
> >  __call_rcu_common kernel/rcu/tree.c:3106 [inline]
> >  call_rcu+0x167/0xa70 kernel/rcu/tree.c:3210
> >  remove_vma mm/mmap.c:189 [inline]
> >  remove_mt mm/mmap.c:2415 [inline]
> >  do_vmi_align_munmap+0x155c/0x18c0 mm/mmap.c:2758
> >  do_vmi_munmap+0x261/0x2f0 mm/mmap.c:2830
> >  mmap_region+0x72f/0x2090 mm/mmap.c:2881
> >  do_mmap+0x8f9/0x1010 mm/mmap.c:1468
> >  vm_mmap_pgoff+0x1dd/0x3d0 mm/util.c:588
> >  ksys_mmap_pgoff+0x544/0x720 mm/mmap.c:1514
> >  do_syscall_x64 arch/x86/entry/common.c:52 [inline]
> >  do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83
> >  entry_SYSCALL_64_after_hwframe+0x77/0x7f
> > 
> > The buggy address belongs to the object at ffff88802c4719b0
> >  which belongs to the cache vm_area_struct of size 184

...