mm/hugetlb: fix possible deadlock in __hugetlb_zap_begin
Subsystems (2)
PATCH 1/1 mm/hugetlb: fix possible deadlock in __hugetlb_zap_begin
From: Lance Yang <[email protected]>
The hugetlb VMA unmap path contains a potential deadlock, as reported by syzbot. In __hugetlb_zap_begin(), vma_lock is acquired before i_mmap_lock. This lock ordering conflicts with the page fault path in hugetlb_fault(), which acquires i_mmap_lock first, establishing the correct dependency as i_mmap_lock -> vma_lock.
Chain exists of:
&hugetlbfs_i_mmap_rwsem_key
--> &hugetlb_fault_mutex_table[i]
--> &vma_lock->rw_sema
Possible unsafe locking scenario:CPU0 CPU1
---- ----
lock(&vma_lock->rw_sema);
lock(&hugetlb_fault_mutex_table[i]);
lock(&vma_lock->rw_sema);lock(&hugetlbfs_i_mmap_rwsem_key);
Resolve the deadlock by reordering the locks in __hugetlb_zap_begin() to follow the established i_mmap_lock -> vma_lock order.
--- mm/hugetlb.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
mm/hugetlb.c +1 -1
@@ -5327,9 +5327,9 @@ void __hugetlb_zap_begin(struct vm_area_struct *vma,return ;adjust_range_if_pmd_sharing_possible (vma, start, end); -hugetlb_vma_lock_write (vma);if (vma->vm_file)i_mmap_lock_write (vma->vm_file->f_mapping); +hugetlb_vma_lock_write (vma); } void __hugetlb_zap_end(struct vm_area_struct *vma,
Signature · 1 line
2.49.0
syzbot ci has tested the following series
syzbot ci has tested the following series
[v1] mm/hugetlb: fix possible deadlock in __hugetlb_zap_begin
https://lore.kernel.org/all/[email protected]
* [PATCH 1/1] mm/hugetlb: fix possible deadlock in __hugetlb_zap_begin
and found the following issues:
* possible deadlock in move_hugetlb_page_tables
* possible deadlock in remove_inode_hugepages
* possible deadlock in unmap_vmas
Full report is available here:
https://ci.syzbot.org/series/4162a577-196d-48cb-908e-f31fe5f186e2
***
possible deadlock in move_hugetlb_page_tables
tree: mm-new URL: https://kernel.googlesource.com/pub/scm/linux/kernel/git/akpm/mm.git base: 02dafa01ec9a00c3758c1c6478d82fe601f5f1ba arch: amd64 compiler: Debian clang version 20.1.8 (++20250708063551+0c9f909b7976-1~exp1~20250708183702.136), Debian LLD 20.1.8 config: https://ci.syzbot.org/builds/b444c811-3577-4885-8522-093253539e5d/config C repro: https://ci.syzbot.org/findings/118b63ad-708b-4ada-abd9-4d6e18a8384f/c_repro syz repro: https://ci.syzbot.org/findings/118b63ad-708b-4ada-abd9-4d6e18a8384f/syz_repro
======================================================
WARNING: possible circular locking dependency detected
syzkaller #0 Not tainted
------------------------------------------------------
syz.0.17/5959 is trying to acquire lock:ffff888167d5c918 (&hugetlbfs_i_mmap_rwsem_key){+.+.}-{4:4}, at: i_mmap_lock_write include/linux/fs.h:548 [inline] ffff888167d5c918 (&hugetlbfs_i_mmap_rwsem_key){+.+.}-{4:4}, at: move_hugetlb_page_tables+0x5a5/0x1560 mm/hugetlb.c:5114
but task is already holding lock:ffff8881b5f35e78 (&vma_lock->rw_sema){+.+.}-{4:4}, at: hugetlb_vma_lock_write mm/hugetlb.c:-1 [inline] ffff8881b5f35e78 (&vma_lock->rw_sema){+.+.}-{4:4}, at: move_hugetlb_page_tables+0x589/0x1560 mm/hugetlb.c:5113
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (&vma_lock->rw_sema){+.+.}-{4:4}:
lock_acquire+0x120/0x360 kernel/locking/lockdep.c:5868
down_write+0x96/0x1f0 kernel/locking/rwsem.c:1590
hugetlb_zap_begin include/linux/hugetlb.h:257 [inline]
unmap_vmas+0x23d/0x580 mm/memory.c:2098
vms_clear_ptes+0x353/0x530 mm/vma.c:1231
vms_complete_munmap_vmas+0x206/0x8a0 mm/vma.c:1280
do_vmi_align_munmap+0x364/0x440 mm/vma.c:1539
do_vmi_munmap+0x253/0x2e0 mm/vma.c:1587
shrink_vma mm/mremap.c:1340 [inline]
mremap_to+0x21e/0x790 mm/mremap.c:1394
do_mremap mm/mremap.c:1940 [inline]
__do_sys_mremap mm/mremap.c:1996 [inline]
__se_sys_mremap+0xd45/0xfb0 mm/mremap.c:1964
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xfa/0xfa0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f-> #0 (&hugetlbfs_i_mmap_rwsem_key){+.+.}-{4:4}:
check_prev_add kernel/locking/lockdep.c:3165 [inline]
check_prevs_add kernel/locking/lockdep.c:3284 [inline]
validate_chain+0xb9b/0x2140 kernel/locking/lockdep.c:3908
__lock_acquire+0xab9/0xd20 kernel/locking/lockdep.c:5237
lock_acquire+0x120/0x360 kernel/locking/lockdep.c:5868
down_write+0x96/0x1f0 kernel/locking/rwsem.c:1590
i_mmap_lock_write include/linux/fs.h:548 [inline]
move_hugetlb_page_tables+0x5a5/0x1560 mm/hugetlb.c:5114
move_page_tables+0x35c/0x1c80 mm/mremap.c:806
copy_vma_and_data mm/mremap.c:1211 [inline]
move_vma+0x942/0x1820 mm/mremap.c:1291
mremap_to+0x6ce/0x790 mm/mremap.c:1415
do_mremap mm/mremap.c:1940 [inline]
__do_sys_mremap mm/mremap.c:1996 [inline]
__se_sys_mremap+0xd45/0xfb0 mm/mremap.c:1964
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xfa/0xfa0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7fother info that might help us debug this:
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(&vma_lock->rw_sema);
lock(&hugetlbfs_i_mmap_rwsem_key);
lock(&vma_lock->rw_sema); lock(&hugetlbfs_i_mmap_rwsem_key);
*** DEADLOCK ***
2 locks held by syz.0.17/5959:
#0: ffff888172d8d7e0 (&mm->mmap_lock){++++}-{4:4}, at: mmap_write_lock_killable include/linux/mmap_lock.h:327 [inline] #0: ffff888172d8d7e0 (&mm->mmap_lock){++++}-{4:4}, at: do_mremap mm/mremap.c:1927 [inline]
#0: ffff888172d8d7e0 (&mm->mmap_lock){++++}-{4:4}, at: __do_sys_mremap mm/mremap.c:1996 [inline] #0: ffff888172d8d7e0 (&mm->mmap_lock){++++}-{4:4}, at: __se_sys_mremap+0x42e/0xfb0 mm/mremap.c:1964 #1: ffff8881b5f35e78 (&vma_lock->rw_sema){+.+.}-{4:4}, at: hugetlb_vma_lock_write mm/hugetlb.c:-1 [inline] #1: ffff8881b5f35e78 (&vma_lock->rw_sema){+.+.}-{4:4}, at: move_hugetlb_page_tables+0x589/0x1560 mm/hugetlb.c:5113
stack backtrace: CPU: 1 UID: 0 PID: 5959 Comm: syz.0.17 Not tainted syzkaller #0 PREEMPT(full) 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_lvl+0x189/0x250 lib/dump_stack.c:120 print_circular_bug+0x2ee/0x310 kernel/locking/lockdep.c:2043 check_noncircular+0x134/0x160 kernel/locking/lockdep.c:2175 check_prev_add kernel/locking/lockdep.c:3165 [inline] check_prevs_add kernel/locking/lockdep.c:3284 [inline] validate_chain+0xb9b/0x2140 kernel/locking/lockdep.c:3908 __lock_acquire+0xab9/0xd20 kernel/locking/lockdep.c:5237 lock_acquire+0x120/0x360 kernel/locking/lockdep.c:5868 down_write+0x96/0x1f0 kernel/locking/rwsem.c:1590 i_mmap_lock_write include/linux/fs.h:548 [inline] move_hugetlb_page_tables+0x5a5/0x1560 mm/hugetlb.c:5114 move_page_tables+0x35c/0x1c80 mm/mremap.c:806 copy_vma_and_data mm/mremap.c:1211 [inline] move_vma+0x942/0x1820 mm/mremap.c:1291 mremap_to+0x6ce/0x790 mm/mremap.c:1415 do_mremap mm/mremap.c:1940 [inline] __do_sys_mremap mm/mremap.c:1996 [inline] __se_sys_mremap+0xd45/0xfb0 mm/mremap.c:1964 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0xfa/0xfa0 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f RIP: 0033:0x7f043598efc9 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 a8 ff ff ff f7 d8 64 89 01 48 RSP: 002b:00007ffe808cfc08 EFLAGS: 00000246 ORIG_RAX: 0000000000000019 RAX: ffffffffffffffda RBX: 00007f0435be5fa0 RCX: 00007f043598efc9 RDX: 0000000000400000 RSI: 0000000000c00000 RDI: 0000200000000000 RBP: 00007f0435a11f91 R08: 0000200000c00000 R09: 0000000000000000 R10: 0000000000000003 R11: 0000000000000246 R12: 0000000000000000 R13: 00007f0435be5fa0 R14: 00007f0435be5fa0 R15: 0000000000000005 </TASK>
***
possible deadlock in remove_inode_hugepages
tree: mm-new URL: https://kernel.googlesource.com/pub/scm/linux/kernel/git/akpm/mm.git base: 02dafa01ec9a00c3758c1c6478d82fe601f5f1ba arch: amd64 compiler: Debian clang version 20.1.8 (++20250708063551+0c9f909b7976-1~exp1~20250708183702.136), Debian LLD 20.1.8 config: https://ci.syzbot.org/builds/b444c811-3577-4885-8522-093253539e5d/config syz repro: https://ci.syzbot.org/findings/e3555fcc-3852-4e88-aea4-f3445e8a6ffb/syz_repro
======================================================
WARNING: possible circular locking dependency detected
syzkaller #0 Not tainted
------------------------------------------------------
syz.2.22/5989 is trying to acquire lock:ffff8881bc18d318 (&hugetlbfs_i_mmap_rwsem_key){+.+.}-{4:4}, at: i_mmap_lock_write include/linux/fs.h:548 [inline] ffff8881bc18d318 (&hugetlbfs_i_mmap_rwsem_key){+.+.}-{4:4}, at: hugetlb_unmap_file_folio fs/hugetlbfs/inode.c:451 [inline] ffff8881bc18d318 (&hugetlbfs_i_mmap_rwsem_key){+.+.}-{4:4}, at: remove_inode_single_folio fs/hugetlbfs/inode.c:535 [inline] ffff8881bc18d318 (&hugetlbfs_i_mmap_rwsem_key){+.+.}-{4:4}, at: remove_inode_hugepages+0x972/0x1110 fs/hugetlbfs/inode.c:602
but task is already holding lock:ffff8881ae823178 (&vma_lock->rw_sema){++++}-{4:4}, at: hugetlb_unmap_file_folio fs/hugetlbfs/inode.c:450 [inline] ffff8881ae823178 (&vma_lock->rw_sema){++++}-{4:4}, at: remove_inode_single_folio fs/hugetlbfs/inode.c:535 [inline] ffff8881ae823178 (&vma_lock->rw_sema){++++}-{4:4}, at: remove_inode_hugepages+0x96a/0x1110 fs/hugetlbfs/inode.c:602
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (&vma_lock->rw_sema){++++}-{4:4}:
lock_acquire+0x120/0x360 kernel/locking/lockdep.c:5868
down_write+0x96/0x1f0 kernel/locking/rwsem.c:1590
hugetlb_zap_begin include/linux/hugetlb.h:257 [inline]
unmap_vmas+0x23d/0x580 mm/memory.c:2098
exit_mmap+0x23e/0xb30 mm/mmap.c:1277
__mmput+0x118/0x430 kernel/fork.c:1133
exit_mm+0x1da/0x2c0 kernel/exit.c:582
do_exit+0x648/0x2300 kernel/exit.c:954
do_group_exit+0x21c/0x2d0 kernel/exit.c:1107
get_signal+0x1285/0x1340 kernel/signal.c:3034
arch_do_signal_or_restart+0xa0/0x790 arch/x86/kernel/signal.c:337
exit_to_user_mode_loop+0x72/0x130 kernel/entry/common.c:40
exit_to_user_mode_prepare include/linux/irq-entry-common.h:225 [inline]
syscall_exit_to_user_mode_work include/linux/entry-common.h:175 [inline]
syscall_exit_to_user_mode include/linux/entry-common.h:210 [inline]
do_syscall_64+0x2bd/0xfa0 arch/x86/entry/syscall_64.c:100
entry_SYSCALL_64_after_hwframe+0x77/0x7f-> #0 (&hugetlbfs_i_mmap_rwsem_key){+.+.}-{4:4}:
check_prev_add kernel/locking/lockdep.c:3165 [inline]
check_prevs_add kernel/locking/lockdep.c:3284 [inline]
validate_chain+0xb9b/0x2140 kernel/locking/lockdep.c:3908
__lock_acquire+0xab9/0xd20 kernel/locking/lockdep.c:5237
lock_acquire+0x120/0x360 kernel/locking/lockdep.c:5868
down_write+0x96/0x1f0 kernel/locking/rwsem.c:1590
i_mmap_lock_write include/linux/fs.h:548 [inline]
hugetlb_unmap_file_folio fs/hugetlbfs/inode.c:451 [inline]
remove_inode_single_folio fs/hugetlbfs/inode.c:535 [inline]
remove_inode_hugepages+0x972/0x1110 fs/hugetlbfs/inode.c:602
hugetlbfs_punch_hole fs/hugetlbfs/inode.c:725 [inline]
hugetlbfs_fallocate+0xbc2/0x1100 fs/hugetlbfs/inode.c:751
vfs_fallocate+0x669/0x7e0 fs/open.c:342
madvise_remove mm/madvise.c:1046 [inline]
madvise_vma_behavior+0x341a/0x3d70 mm/madvise.c:1350
madvise_walk_vmas+0x51c/0xa30 mm/madvise.c:1673
madvise_do_behavior+0x38e/0x550 mm/madvise.c:1889
do_madvise+0x1bc/0x270 mm/madvise.c:1982
__do_sys_madvise mm/madvise.c:1991 [inline]
__se_sys_madvise mm/madvise.c:1989 [inline]
__x64_sys_madvise+0xa7/0xc0 mm/madvise.c:1989
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xfa/0xfa0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7fother info that might help us debug this:
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(&vma_lock->rw_sema);
lock(&hugetlbfs_i_mmap_rwsem_key);
lock(&vma_lock->rw_sema); lock(&hugetlbfs_i_mmap_rwsem_key);
*** DEADLOCK ***
4 locks held by syz.2.22/5989:
#0: ffff8881056ac420 (sb_writers#12){.+.+}-{0:0}, at: file_start_write include/linux/fs.h:3111 [inline]
#0: ffff8881056ac420 (sb_writers#12){.+.+}-{0:0}, at: vfs_fallocate+0x5f0/0x7e0 fs/open.c:341
#1: ffff8881bc18d048 (&sb->s_type->i_mutex_key#20){+.+.}-{4:4}, at: inode_lock include/linux/fs.h:980 [inline] #1: ffff8881bc18d048 (&sb->s_type->i_mutex_key#20){+.+.}-{4:4}, at: hugetlbfs_punch_hole fs/hugetlbfs/inode.c:693 [inline] #1: ffff8881bc18d048 (&sb->s_type->i_mutex_key#20){+.+.}-{4:4}, at: hugetlbfs_fallocate+0x3cc/0x1100 fs/hugetlbfs/inode.c:751 #2: ffff888103e8e4e8 (&hugetlb_fault_mutex_table[i]){+.+.}-{4:4}, at: remove_inode_hugepages+0x332/0x1110 fs/hugetlbfs/inode.c:597 #3: ffff8881ae823178 (&vma_lock->rw_sema){++++}-{4:4}, at: hugetlb_unmap_file_folio fs/hugetlbfs/inode.c:450 [inline] #3: ffff8881ae823178 (&vma_lock->rw_sema){++++}-{4:4}, at: remove_inode_single_folio fs/hugetlbfs/inode.c:535 [inline] #3: ffff8881ae823178 (&vma_lock->rw_sema){++++}-{4:4}, at: remove_inode_hugepages+0x96a/0x1110 fs/hugetlbfs/inode.c:602
stack backtrace: CPU: 1 UID: 0 PID: 5989 Comm: syz.2.22 Not tainted syzkaller #0 PREEMPT(full) 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_lvl+0x189/0x250 lib/dump_stack.c:120 print_circular_bug+0x2ee/0x310 kernel/locking/lockdep.c:2043 check_noncircular+0x134/0x160 kernel/locking/lockdep.c:2175 check_prev_add kernel/locking/lockdep.c:3165 [inline] check_prevs_add kernel/locking/lockdep.c:3284 [inline] validate_chain+0xb9b/0x2140 kernel/locking/lockdep.c:3908 __lock_acquire+0xab9/0xd20 kernel/locking/lockdep.c:5237 lock_acquire+0x120/0x360 kernel/locking/lockdep.c:5868 down_write+0x96/0x1f0 kernel/locking/rwsem.c:1590 i_mmap_lock_write include/linux/fs.h:548 [inline] hugetlb_unmap_file_folio fs/hugetlbfs/inode.c:451 [inline] remove_inode_single_folio fs/hugetlbfs/inode.c:535 [inline] remove_inode_hugepages+0x972/0x1110 fs/hugetlbfs/inode.c:602 hugetlbfs_punch_hole fs/hugetlbfs/inode.c:725 [inline] hugetlbfs_fallocate+0xbc2/0x1100 fs/hugetlbfs/inode.c:751 vfs_fallocate+0x669/0x7e0 fs/open.c:342 madvise_remove mm/madvise.c:1046 [inline] madvise_vma_behavior+0x341a/0x3d70 mm/madvise.c:1350 madvise_walk_vmas+0x51c/0xa30 mm/madvise.c:1673 madvise_do_behavior+0x38e/0x550 mm/madvise.c:1889 do_madvise+0x1bc/0x270 mm/madvise.c:1982 __do_sys_madvise mm/madvise.c:1991 [inline] __se_sys_madvise mm/madvise.c:1989 [inline] __x64_sys_madvise+0xa7/0xc0 mm/madvise.c:1989 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0xfa/0xfa0 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f RIP: 0033:0x7fca36b8efc9 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 a8 ff ff ff f7 d8 64 89 01 48 RSP: 002b:00007fca37a80038 EFLAGS: 00000246 ORIG_RAX: 000000000000001c RAX: ffffffffffffffda RBX: 00007fca36de6090 RCX: 00007fca36b8efc9 RDX: 0000000000000009 RSI: 0000000000600000 RDI: 0000200000bc0000 RBP: 00007fca36c11f91 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000 R13: 00007fca36de6128 R14: 00007fca36de6090 R15: 00007ffeef05aaf8 </TASK>
***
possible deadlock in unmap_vmas
tree: mm-new URL: https://kernel.googlesource.com/pub/scm/linux/kernel/git/akpm/mm.git base: 02dafa01ec9a00c3758c1c6478d82fe601f5f1ba arch: amd64 compiler: Debian clang version 20.1.8 (++20250708063551+0c9f909b7976-1~exp1~20250708183702.136), Debian LLD 20.1.8 config: https://ci.syzbot.org/builds/b444c811-3577-4885-8522-093253539e5d/config C repro: https://ci.syzbot.org/findings/0f0a7c70-e594-4359-9971-50a723bbbd7a/c_repro syz repro: https://ci.syzbot.org/findings/0f0a7c70-e594-4359-9971-50a723bbbd7a/syz_repro
======================================================
WARNING: possible circular locking dependency detected
syzkaller #0 Not tainted
------------------------------------------------------
syz.0.17/5961 is trying to acquire lock:ffff88810be5c178 (&vma_lock->rw_sema){+.+.}-{4:4}, at: hugetlb_zap_begin include/linux/hugetlb.h:257 [inline] ffff88810be5c178 (&vma_lock->rw_sema){+.+.}-{4:4}, at: unmap_vmas+0x23d/0x580 mm/memory.c:2098
but task is already holding lock:ffff88816a110418 (&hugetlbfs_i_mmap_rwsem_key){+.+.}-{4:4}, at: i_mmap_lock_write include/linux/fs.h:548 [inline] ffff88816a110418 (&hugetlbfs_i_mmap_rwsem_key){+.+.}-{4:4}, at: __hugetlb_zap_begin+0x2aa/0x3a0 mm/hugetlb.c:5331
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (&hugetlbfs_i_mmap_rwsem_key){+.+.}-{4:4}:
lock_acquire+0x120/0x360 kernel/locking/lockdep.c:5868
down_write+0x96/0x1f0 kernel/locking/rwsem.c:1590
i_mmap_lock_write include/linux/fs.h:548 [inline]
hugetlb_change_protection+0x56a/0x18c0 mm/hugetlb.c:6439
change_protection+0x386/0x38c0 mm/mprotect.c:655
mprotect_fixup+0x6fe/0x9c0 mm/mprotect.c:774
do_mprotect_pkey+0x8c5/0xcd0 mm/mprotect.c:930
__do_sys_mprotect mm/mprotect.c:951 [inline]
__se_sys_mprotect mm/mprotect.c:948 [inline]
__x64_sys_mprotect+0x80/0x90 mm/mprotect.c:948
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xfa/0xfa0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f-> #0 (&vma_lock->rw_sema){+.+.}-{4:4}:
check_prev_add kernel/locking/lockdep.c:3165 [inline]
check_prevs_add kernel/locking/lockdep.c:3284 [inline]
validate_chain+0xb9b/0x2140 kernel/locking/lockdep.c:3908
__lock_acquire+0xab9/0xd20 kernel/locking/lockdep.c:5237
lock_acquire+0x120/0x360 kernel/locking/lockdep.c:5868
down_write+0x96/0x1f0 kernel/locking/rwsem.c:1590
hugetlb_zap_begin include/linux/hugetlb.h:257 [inline]
unmap_vmas+0x23d/0x580 mm/memory.c:2098
exit_mmap+0x23e/0xb30 mm/mmap.c:1277
__mmput+0x118/0x430 kernel/fork.c:1133
exit_mm+0x1da/0x2c0 kernel/exit.c:582
do_exit+0x648/0x2300 kernel/exit.c:954
do_group_exit+0x21c/0x2d0 kernel/exit.c:1107
__do_sys_exit_group kernel/exit.c:1118 [inline]
__se_sys_exit_group kernel/exit.c:1116 [inline]
__x64_sys_exit_group+0x3f/0x40 kernel/exit.c:1116
x64_sys_call+0x21f7/0x2200 arch/x86/include/generated/asm/syscalls_64.h:232
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xfa/0xfa0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7fother info that might help us debug this:
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(&hugetlbfs_i_mmap_rwsem_key);
lock(&vma_lock->rw_sema);
lock(&hugetlbfs_i_mmap_rwsem_key); lock(&vma_lock->rw_sema);
*** DEADLOCK ***
2 locks held by syz.0.17/5961:
#0: ffff8881151362a0 (&mm->mmap_lock){++++}-{4:4}, at: mmap_read_lock include/linux/mmap_lock.h:365 [inline] #0: ffff8881151362a0 (&mm->mmap_lock){++++}-{4:4}, at: exit_mmap+0x126/0xb30 mm/mmap.c:1262
#1: ffff88816a110418 (&hugetlbfs_i_mmap_rwsem_key){+.+.}-{4:4}, at: i_mmap_lock_write include/linux/fs.h:548 [inline] #1: ffff88816a110418 (&hugetlbfs_i_mmap_rwsem_key){+.+.}-{4:4}, at: __hugetlb_zap_begin+0x2aa/0x3a0 mm/hugetlb.c:5331
stack backtrace: CPU: 1 UID: 0 PID: 5961 Comm: syz.0.17 Not tainted syzkaller #0 PREEMPT(full) 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_lvl+0x189/0x250 lib/dump_stack.c:120 print_circular_bug+0x2ee/0x310 kernel/locking/lockdep.c:2043 check_noncircular+0x134/0x160 kernel/locking/lockdep.c:2175 check_prev_add kernel/locking/lockdep.c:3165 [inline] check_prevs_add kernel/locking/lockdep.c:3284 [inline] validate_chain+0xb9b/0x2140 kernel/locking/lockdep.c:3908 __lock_acquire+0xab9/0xd20 kernel/locking/lockdep.c:5237 lock_acquire+0x120/0x360 kernel/locking/lockdep.c:5868 down_write+0x96/0x1f0 kernel/locking/rwsem.c:1590 hugetlb_zap_begin include/linux/hugetlb.h:257 [inline] unmap_vmas+0x23d/0x580 mm/memory.c:2098 exit_mmap+0x23e/0xb30 mm/mmap.c:1277 __mmput+0x118/0x430 kernel/fork.c:1133 exit_mm+0x1da/0x2c0 kernel/exit.c:582 do_exit+0x648/0x2300 kernel/exit.c:954 do_group_exit+0x21c/0x2d0 kernel/exit.c:1107 __do_sys_exit_group kernel/exit.c:1118 [inline] __se_sys_exit_group kernel/exit.c:1116 [inline] __x64_sys_exit_group+0x3f/0x40 kernel/exit.c:1116 x64_sys_call+0x21f7/0x2200 arch/x86/include/generated/asm/syscalls_64.h:232 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0xfa/0xfa0 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f RIP: 0033:0x7f470758efc9 Code: Unable to access opcode bytes at 0x7f470758ef9f. RSP: 002b:00007ffed0f73318 EFLAGS: 00000246 ORIG_RAX: 00000000000000e7 RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f470758efc9 RDX: 0000000000000064 RSI: 0000000000000000 RDI: 0000000000000000 RBP: 0000000000000003 R08: 00000002d0f7340f R09: 00007f47077b1280 R10: 0000000000000001 R11: 0000000000000246 R12: 0000000000000000 R13: 00007f47077b1280 R14: 0000000000000003 R15: 00007ffed0f733d0 </TASK>
***
If these findings have caused you to resend the series or submit a separate fix, please add the following tag to your commit message:
---
This report is generated by a bot. It may contain errors. syzbot ci engineers can be reached at [email protected].
Quoted 10 lines · syzbot ci
syzbot ci has tested the following series
[v1] mm/hugetlb: fix possible deadlock in
__hugetlb_zap_begin
https://lore.kernel.org/all/[email protected]
* [PATCH 1/1] mm/hugetlb: fix possible deadlock in__hugetlb_zap_beginand found the following issues:
* possible deadlock inmove_hugetlb_page_tables
* possible deadlock inremove_inode_hugepages
* possible deadlock inunmap_vmas
Oops, this fix was incomplete, as other code paths with the reverse locking order still exist, leading to the deadlocks reported by syzbot ...
Let me take a closer look ...