Re: [syzbot] [mm?] KASAN: slab-use-after-free Read in move_pages_pte

From: Qi Zheng
Date: Mon Dec 09 2024 - 02:00:20 EST

Next message: Stephen Rothwell: "linux-next: Signed-off-by missing for commit in the bcachefs tree"
Previous message: Ryusuke Konishi: "[PATCH] nilfs2: prevent use of deleted inode"
In reply to: syzbot: "Re: [syzbot] [mm?] KASAN: slab-use-after-free Read in move_pages_pte"
Next in thread: syzbot: "Re: [syzbot] [mm?] KASAN: slab-use-after-free Read in move_pages_pte"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On 2024/12/9 14:25, Qi Zheng wrote:

On 2024/12/8 01:14, syzbot wrote:

Hello,

syzbot found the following issue on:

HEAD commit:    af2ea8ab7a54 Add linux-next specific files for 20241205
git tree:       linux-next
console output: https://syzkaller.appspot.com/x/log.txt?x=13c4e8df980000
kernel config: https://syzkaller.appspot.com/x/.config?x=76f158395f6f15fd
dashboard link: https://syzkaller.appspot.com/bug?extid=1c58afed1cfd2f57efee
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=133850f8580000
C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=17be9330580000

Downloadable assets:
disk image: https://storage.googleapis.com/syzbot-assets/8af0861258fa/disk-af2ea8ab.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/ffb38cf7a344/vmlinux-af2ea8ab.xz
kernel image: https://storage.googleapis.com/syzbot-assets/6fbd2e50358a/bzImage-af2ea8ab.xz

The issue was bisected to:

commit 5b29c4156f5801fced2ec504b44ab98f60c480bf
Author: Qi Zheng <zhengqi.arch@xxxxxxxxxxxxx>
Date:   Wed Dec 4 11:09:51 2024 +0000

     x86: select ARCH_SUPPORTS_PT_RECLAIM if X86_64

bisection log: https://syzkaller.appspot.com/x/bisect.txt?x=16d344df980000
final oops: https://syzkaller.appspot.com/x/report.txt?x=15d344df980000
console output: https://syzkaller.appspot.com/x/log.txt?x=11d344df980000

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+1c58afed1cfd2f57efee@xxxxxxxxxxxxxxxxxxxxxxxxx
Fixes: 5b29c4156f58 ("x86: select ARCH_SUPPORTS_PT_RECLAIM if X86_64")

==================================================================
BUG: KASAN: slab-use-after-free in __lock_acquire+0x78/0x2100 kernel/locking/lockdep.c:5089
Read of size 8 at addr ffff888034718978 by task syz-executor352/6070

CPU: 0 UID: 0 PID: 6070 Comm: syz-executor352 Not tainted 6.13.0-rc1-next-20241205-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 09/13/2024
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:378 [inline]
print_report+0x169/0x550 mm/kasan/report.c:489
kasan_report+0x143/0x180 mm/kasan/report.c:602
__lock_acquire+0x78/0x2100 kernel/locking/lockdep.c:5089
lock_acquire+0x1ed/0x550 kernel/locking/lockdep.c:5849
__raw_spin_lock include/linux/spinlock_api_smp.h:133 [inline]
_raw_spin_lock+0x2e/0x40 kernel/locking/spinlock.c:154
spin_lock include/linux/spinlock.h:351 [inline]
move_pages_pte+0x8aa/0x3400 mm/userfaultfd.c:1248
move_pages+0xe75/0x16a0 mm/userfaultfd.c:1754
userfaultfd_move fs/userfaultfd.c:1899 [inline]
userfaultfd_ioctl+0x5221/0x6840 fs/userfaultfd.c:2022
vfs_ioctl fs/ioctl.c:51 [inline]
__do_sys_ioctl fs/ioctl.c:906 [inline]
__se_sys_ioctl+0xf5/0x170 fs/ioctl.c:892
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: 0033:0x7fed8de85af9
Code: 28 00 00 00 75 05 48 83 c4 28 c3 e8 e1 18 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:00007fed8de40238 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
RAX: ffffffffffffffda RBX: 00007fed8df10328 RCX: 00007fed8de85af9
RDX: 0000000020000080 RSI: 00000000c028aa05 RDI: 0000000000000003
RBP: 00007fed8df10320 R08: 00007fed8de406c0 R09: 00007fed8de406c0
R10: 00007fed8de406c0 R11: 0000000000000246 R12: 00007fed8dedd334
R13: 0000000000000010 R14: 00007ffc241241e0 R15: 00007ffc241242c8
</TASK>

Allocated by task 6070:
kasan_save_stack mm/kasan/common.c:47 [inline]
kasan_save_track+0x3f/0x80 mm/kasan/common.c:68
unpoison_slab_object mm/kasan/common.c:319 [inline]
__kasan_slab_alloc+0x66/0x80 mm/kasan/common.c:345
kasan_slab_alloc include/linux/kasan.h:250 [inline]
slab_post_alloc_hook mm/slub.c:4104 [inline]
slab_alloc_node mm/slub.c:4153 [inline]
kmem_cache_alloc_noprof+0x1d9/0x380 mm/slub.c:4160
ptlock_alloc+0x20/0x70 mm/memory.c:7026
ptlock_init include/linux/mm.h:2971 [inline]
pagetable_pte_ctor include/linux/mm.h:2998 [inline]
__pte_alloc_one_noprof include/asm-generic/pgalloc.h:73 [inline]
pte_alloc_one+0xd3/0x510 arch/x86/mm/pgtable.c:41
__do_huge_pmd_anonymous_page mm/huge_memory.c:1229 [inline]
do_huge_pmd_anonymous_page+0x2fb/0xb30 mm/huge_memory.c:1374
create_huge_pmd mm/memory.c:5737 [inline]
__handle_mm_fault mm/memory.c:5986 [inline]
handle_mm_fault+0x15a7/0x1bb0 mm/memory.c:6183
do_user_addr_fault arch/x86/mm/fault.c:1338 [inline]
handle_page_fault arch/x86/mm/fault.c:1481 [inline]
exc_page_fault+0x459/0x8b0 arch/x86/mm/fault.c:1539
asm_exc_page_fault+0x26/0x30 arch/x86/include/asm/idtentry.h:623

Freed by task 6071:
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:576
poison_slab_object mm/kasan/common.c:247 [inline]
__kasan_slab_free+0x59/0x70 mm/kasan/common.c:264
kasan_slab_free include/linux/kasan.h:233 [inline]
slab_free_hook mm/slub.c:2338 [inline]
slab_free mm/slub.c:4598 [inline]
kmem_cache_free+0x195/0x410 mm/slub.c:4700
pagetable_pte_dtor include/linux/mm.h:3009 [inline]

OK, so the problem is that ptdesc->ptl is not freed via RCU:

___pte_free_tlb
--> pagetable_pte_dtor
    --> ptlock_free
        --> kmem_cache_free (free immediately!)
    paravirt_tlb_remove_table
    --> free PTE page via RCU

In retract_page_tables(), it calls pte_free_defer() to free
ptdesc->ptl and PTE page via RCU, so there is no problem.

To fix it, will also free ptdesc->ptl in ptlock_free() via RCU.

___pte_free_tlb+0x2b/0x140 arch/x86/mm/pgtable.c:63
__pte_free_tlb arch/x86/include/asm/pgalloc.h:61 [inline]
free_pte+0x142/0x190 mm/pt_reclaim.c:31
zap_pte_range mm/memory.c:1780 [inline]
zap_pmd_range mm/memory.c:1822 [inline]
zap_pud_range mm/memory.c:1851 [inline]
zap_p4d_range mm/memory.c:1872 [inline]
unmap_page_range+0x4062/0x48d0 mm/memory.c:1893
zap_page_range_single+0x45c/0x630 mm/memory.c:2018
madvise_dontneed_single_vma mm/madvise.c:859 [inline]
madvise_dontneed_free mm/madvise.c:940 [inline]
madvise_vma_behavior mm/madvise.c:1270 [inline]
madvise_walk_vmas mm/madvise.c:1502 [inline]
do_madvise+0x2774/0x4d90 mm/madvise.c:1689
__do_sys_madvise mm/madvise.c:1705 [inline]
__se_sys_madvise mm/madvise.c:1703 [inline]
__x64_sys_madvise+0xa6/0xc0 mm/madvise.c:1703
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 ffff888034718960
which belongs to the cache page->ptl of size 64
The buggy address is located 24 bytes inside of
freed 64-byte region [ffff888034718960, ffff8880347189a0)

The buggy address belongs to the physical page:
page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x34718
flags: 0xfff00000000000(node=0|zone=1|lastcpupid=0x7ff)
page_type: f5(slab)
raw: 00fff00000000000 ffff88801ac4f780 dead000000000122 0000000000000000
raw: 0000000000000000 00000000802a002a 00000000f5000000 0000000000000000
page dumped because: kasan: bad access detected
page_owner tracks the page as allocated
page last allocated via order 0, migratetype Unmovable, gfp_mask 0x52cc0(GFP_KERNEL|__GFP_NOWARN|__GFP_NORETRY|__GFP_COMP), pid 5823, tgid 5823 (syz-executor352), ts 65548803787, free_ts 65433386693
set_page_owner include/linux/page_owner.h:32 [inline]
post_alloc_hook+0x1f4/0x240 mm/page_alloc.c:1549
prep_new_page mm/page_alloc.c:1557 [inline]
get_page_from_freelist+0x365c/0x37a0 mm/page_alloc.c:3475
__alloc_frozen_pages_noprof+0x292/0x710 mm/page_alloc.c:4752
alloc_pages_mpol+0x30e/0x550 mm/mempolicy.c:2270
alloc_slab_page mm/slub.c:2408 [inline]
allocate_slab+0x8f/0x3a0 mm/slub.c:2574
new_slab mm/slub.c:2627 [inline]
___slab_alloc+0xc27/0x14a0 mm/slub.c:3815
__slab_alloc+0x58/0xa0 mm/slub.c:3905
__slab_alloc_node mm/slub.c:3980 [inline]
slab_alloc_node mm/slub.c:4141 [inline]
kmem_cache_alloc_noprof+0x268/0x380 mm/slub.c:4160
ptlock_alloc mm/memory.c:7026 [inline]
ptlock_init include/linux/mm.h:2971 [inline]
pmd_ptlock_init include/linux/mm.h:3078 [inline]
pagetable_pmd_ctor include/linux/mm.h:3116 [inline]
pmd_alloc_one_noprof include/asm-generic/pgalloc.h:141 [inline]
__pmd_alloc+0x10b/0x670 mm/memory.c:6436
pmd_alloc include/linux/mm.h:2862 [inline]
copy_pmd_range+0x7352/0x77a0 mm/memory.c:1241
copy_pud_range mm/memory.c:1298 [inline]
copy_p4d_range mm/memory.c:1322 [inline]
copy_page_range+0x99f/0xe90 mm/memory.c:1420
dup_mmap kernel/fork.c:751 [inline]
dup_mm kernel/fork.c:1693 [inline]
copy_mm+0x12d2/0x2060 kernel/fork.c:1742
copy_process+0x1845/0x3d80 kernel/fork.c:2393
kernel_clone+0x226/0x8e0 kernel/fork.c:2805
__do_sys_clone kernel/fork.c:2948 [inline]
__se_sys_clone kernel/fork.c:2932 [inline]
__x64_sys_clone+0x258/0x2a0 kernel/fork.c:2932
do_syscall_x64 arch/x86/entry/common.c:52 [inline]
do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83
page last free pid 6052 tgid 6051 stack trace:
reset_page_owner include/linux/page_owner.h:25 [inline]
free_pages_prepare mm/page_alloc.c:1127 [inline]
free_frozen_pages+0xe0d/0x10e0 mm/page_alloc.c:2658
__folio_put+0x2b3/0x360 mm/swap.c:112
__tlb_remove_table arch/x86/include/asm/tlb.h:34 [inline]
__tlb_remove_table_free mm/mmu_gather.c:227 [inline]
tlb_remove_table_rcu+0x76/0xf0 mm/mmu_gather.c:282
rcu_do_batch kernel/rcu/tree.c:2567 [inline]
rcu_core+0xaaa/0x17a0 kernel/rcu/tree.c:2823
handle_softirqs+0x2d4/0x9b0 kernel/softirq.c:561
__do_softirq kernel/softirq.c:595 [inline]
invoke_softirq kernel/softirq.c:435 [inline]
__irq_exit_rcu+0xf7/0x220 kernel/softirq.c:662
irq_exit_rcu+0x9/0x30 kernel/softirq.c:678
instr_sysvec_apic_timer_interrupt arch/x86/kernel/apic/apic.c:1049 [inline]
sysvec_apic_timer_interrupt+0xa6/0xc0 arch/x86/kernel/apic/apic.c:1049
asm_sysvec_apic_timer_interrupt+0x1a/0x20 arch/x86/include/asm/idtentry.h:702

Memory state around the buggy address:
ffff888034718800: 00 00 00 00 fc fc fc fc 00 00 00 00 00 00 00 00
ffff888034718880: fc fc fc fc 00 00 00 00 00 00 00 00 fc fc fc fc

ffff888034718900: 00 00 00 00 00 00 00 00 fc fc fc fc fa fb fb fb

                                                                 ^
ffff888034718980: fb fb fb fb fc fc fc fc fa fb fb fb fb fb fb fb
ffff888034718a00: fc fc fc fc 00 00 00 00 00 00 00 00 fc fc fc fc
==================================================================

---
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@xxxxxxxxxxxxxxxx.

syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.
For information about bisection process see: https://goo.gl/tpsmEJ#bisection

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.

#syz test: git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm.git mm-unstable

diff --git a/mm/memory.c b/mm/memory.c
index 83fd35c034d7a..28526a4205d1b 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -7023,7 +7023,7 @@ static struct kmem_cache *page_ptl_cachep;
void __init ptlock_cache_init(void)
{
page_ptl_cachep = kmem_cache_create("page->ptl", sizeof(spinlock_t), 0,
- SLAB_PANIC, NULL);
+ SLAB_PANIC|SLAB_TYPESAFE_BY_RCU, NULL);
}

bool ptlock_alloc(struct ptdesc *ptdesc)

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

Next message: Stephen Rothwell: "linux-next: Signed-off-by missing for commit in the bcachefs tree"
Previous message: Ryusuke Konishi: "[PATCH] nilfs2: prevent use of deleted inode"
In reply to: syzbot: "Re: [syzbot] [mm?] KASAN: slab-use-after-free Read in move_pages_pte"
Next in thread: syzbot: "Re: [syzbot] [mm?] KASAN: slab-use-after-free Read in move_pages_pte"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]