Re: [PATCH] mm/vmalloc: fix vmap_purge_lock livelock under memory pressure

[Ye Liu <ye.liu@linux.dev>]

在 2026/8/26 16:19, Uladzislau Rezki 写道:
> On Wed, Aug 26, 2026 at 11:09:21AM +0530, Dev Jain wrote:
>>
>>
>> On 25/08/26 11:10 pm, Uladzislau Rezki wrote:
>>> On Tue, Aug 25, 2026 at 09:04:24AM +0800, Ye Liu wrote:
>>>>
>>>>
>>>> 在 2026/8/24 21:19, Uladzislau Rezki 写道:
>>>>> On Mon, Aug 24, 2026 at 05:50:19PM +0800, Ye Liu wrote:
>>>>>> From: Ye Liu <liuye@kylinos.cn>
>>>>>>
>>>>>> The vmap_purge_lock mutex can be held for an extended period by
>>>>>> __purge_vmap_area_lazy() which calls flush_work() to wait for
>>>>>> purge_vmap_node workers while holding the lock.  Under memory
>>>>>> pressure, those workers may themselves be blocked in direct
>>>>>> reclaim trying to acquire the same lock via the
>>>>>> vmap_node_shrink_scan() shrinker callback, creating a circular
>>>>>> dependency that deadlocks the entire system.
>>>>>>
>>>>>> Two changes:
>>>>>>
>>>>>> 1. vmap_node_shrink_scan(): replace blocking guard(mutex) with
>>>>>>    mutex_trylock().  This is a shrinker that only decays the vmap
>>>>>>    pool and returns SHRINK_STOP without freeing memory; skipping a
>>>>>>    decay cycle when the lock is contended is harmless and prevents
>>>>>>    tasks from piling up on the mutex in the direct reclaim path.
>>>>>>
>>>>>> 2. __purge_vmap_area_lazy(): purge all vmap nodes inline instead of
>>>>>>    scheduling purge_vmap_node via workqueue and calling flush_work()
>>>>>>    while holding vmap_purge_lock.  The workqueue dispatch + flush
>>>>>>    pattern under a mutex is the deadlock trigger: if no free worker
>>>>>>    threads are available (all blocked on the same lock), flush_work()
>>>>>>    never returns and the lock is held indefinitely.
>>>>>>
>>>>>> Fixes: 7679ba6b36db ("mm: vmalloc: add a shrinker to drain vmap pools")
>>>>>> Signed-off-by: Ye Liu <liuye@kylinos.cn>
>>>>>> ---
>>>>>>  mm/vmalloc.c | 51 +++++++++++++++++++++------------------------------
>>>>>>  1 file changed, 21 insertions(+), 30 deletions(-)
>>>>>>
>>>>>> diff --git a/mm/vmalloc.c b/mm/vmalloc.c
>>>>>> index bea9f76ed7e7..41443708d22d 100644
>>>>>> --- a/mm/vmalloc.c
>>>>>> +++ b/mm/vmalloc.c
>>>>>> @@ -2358,7 +2358,6 @@ static bool __purge_vmap_area_lazy(unsigned long start, unsigned long end,
>>>>>>  		bool full_pool_decay)
>>>>>>  {
>>>>>>  	unsigned long nr_purged_areas = 0;
>>>>>> -	unsigned int nr_purge_helpers;
>>>>>>  	static cpumask_t purge_nodes;
>>>>>>  	unsigned int nr_purge_nodes;
>>>>>>  	struct vmap_node *vn;
>>>>>> @@ -2397,36 +2396,17 @@ static bool __purge_vmap_area_lazy(unsigned long start, unsigned long end,
>>>>>>  	if (nr_purge_nodes > 0) {
>>>>>>  		flush_tlb_kernel_range(start, end);
>>>>>>  
>>>>>> -		/* One extra worker is per a lazy_max_pages() full set minus one. */
>>>>>> -		nr_purge_helpers = atomic_long_read(&vmap_lazy_nr) / lazy_max_pages();
>>>>>> -		nr_purge_helpers = clamp(nr_purge_helpers, 1U, nr_purge_nodes) - 1;
>>>>>> -
>>>>>> -		for_each_cpu(i, &purge_nodes) {
>>>>>> -			vn = &vmap_nodes[i];
>>>>>> -
>>>>>> -			if (nr_purge_helpers > 0) {
>>>>>> -				INIT_WORK(&vn->purge_work, purge_vmap_node);
>>>>>> -
>>>>>> -				if (cpumask_test_cpu(i, cpu_online_mask))
>>>>>> -					schedule_work_on(i, &vn->purge_work);
>>>>>> -				else
>>>>>> -					schedule_work(&vn->purge_work);
>>>>>> -
>>>>>> -				nr_purge_helpers--;
>>>>>> -			} else {
>>>>>> -				vn->purge_work.func = NULL;
>>>>>> -				purge_vmap_node(&vn->purge_work);
>>>>>> -				nr_purged_areas += vn->nr_purged;
>>>>>> -			}
>>>>>> -		}
>>>>>> -
>>>>>> +		/*
>>>>>> +		 * Purge all nodes inline.  Do not schedule_work() and
>>>>>> +		 * flush_work() here: flush_work() while holding
>>>>>> +		 * vmap_purge_lock can deadlock if the worker pool is
>>>>>> +		 * starved (e.g. all workers blocked on this same lock
>>>>>> +		 * in the direct reclaim path via vmap_node_shrink_scan).
>>>>>> +		 */
>>>>>>  		for_each_cpu(i, &purge_nodes) {
>>>>>>  			vn = &vmap_nodes[i];
>>>>>> -
>>>>>> -			if (vn->purge_work.func) {
>>>>>> -				flush_work(&vn->purge_work);
>>>>>> -				nr_purged_areas += vn->nr_purged;
>>>>>> -			}
>>>>>> +			purge_vmap_node(&vn->purge_work);
>>>>>> +			nr_purged_areas += vn->nr_purged;
>>>>>>  		}
>>>>>>  	}
>>>>>>  
>>>>>> @@ -5519,10 +5499,21 @@ vmap_node_shrink_scan(struct shrinker *shrink, struct shrink_control *sc)
>>>>>>  {
>>>>>>  	struct vmap_node *vn;
>>>>>>  
>>>>>> -	guard(mutex)(&vmap_purge_lock);
>>>>>> +	/*
>>>>>> +	 * This shrinker is invoked from direct reclaim path where memory
>>>>>> +	 * pressure is already high.  Blocking on vmap_purge_lock here can
>>>>>> +	 * cause a pile-up of tasks all waiting for the same mutex while
>>>>>> +	 * the lock holder may itself be blocked in flush_work() waiting for
>>>>>> +	 * a worker that is stuck in the same reclaim path.  Use trylock to
>>>>>> +	 * avoid this; skipping a pool decay cycle is harmless.
>>>>>> +	 */
>>>>>> +	if (!mutex_trylock(&vmap_purge_lock))
>>>>>> +		return SHRINK_STOP;
>>>>>> +
>>>>>>  	for_each_vmap_node(vn)
>>>>>>  		decay_va_pool_node(vn, true);
>>>>>>  
>>>>>> +	mutex_unlock(&vmap_purge_lock);
>>>>>>  	return SHRINK_STOP;
>>>>>>  }
>>>>>>  
>>>>>> -- 
>>>>>> 2.25.1
>>>>>>
>>>>> Have you seen any report about the problem you described?
>>>>>
>>>> Our testing team discovered the issue during testing with syzkaller 
>>>> on kernel version 6.6.89. The relevant warnings are shown below.
>>>>
>>>> INFO: task kworker/u80:6:141 blocked for more than 143 seconds.
>>>> Tainted: G S ------- ---- 6.6.89+ #931
>>>> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>> task:kworker/u80:6 state:D stack:0 pid:141 tgid:141 ppid:2 flags:0x00004000
>>>> Workqueue: events_unbound call_usermodehelper_exec_work
>>>> Call Trace:
>>>>
>>>> context_switch root/klinux/kernel/sched/core.c:5346 [inline]
>>>> __schedule+0x7bc/0x1310 root/klinux/kernel/sched/core.c:6720
>>>> __schedule_loop root/klinux/kernel/sched/core.c:6797 [inline]
>>>> schedule+0xe2/0x290 root/klinux/kernel/sched/core.c:6812
>>>> schedule_preempt_disabled+0x18/0x30 root/klinux/kernel/sched/core.c:6869
>>>> __mutex_lock_common root/klinux/kernel/locking/mutex.c:679 [inline]
>>>> __mutex_lock+0x914/0x13a0 root/klinux/kernel/locking/mutex.c:747
>>>> class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> do_shrink_slab+0x3c9/0xf30 root/klinux/mm/shrinker.c:435
>>>> shrink_slab+0x175/0x5e0 root/klinux/mm/shrinker.c:662
>>>> shrink_node_memcgs+0x4d6/0x700 root/klinux/mm/vmscan.c:5900
>>>> shrink_node+0x247/0x12c0 root/klinux/mm/vmscan.c:5933
>>>> shrink_zones root/klinux/mm/vmscan.c:6175 [inline]
>>>> do_try_to_free_pages+0x314/0x1580 root/klinux/mm/vmscan.c:6237
>>>> try_to_free_pages+0x293/0x740 root/klinux/mm/vmscan.c:6472
>>>> __perform_reclaim root/klinux/mm/page_alloc.c:3965 [inline]
>>>> __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> __alloc_pages_slowpath.constprop.0+0x800/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> __alloc_pages_noprof+0x490/0x570 root/klinux/mm/page_alloc.c:4792
>>>> __alloc_pages_node_noprof root/klinux/./include/linux/gfp.h:244 [inline]
>>>> alloc_pages_node_noprof root/klinux/./include/linux/gfp.h:271 [inline]
>>>> alloc_thread_stack_node root/klinux/kernel/fork.c:367 [inline]
>>>> dup_task_struct+0xa4/0x6c0 root/klinux/kernel/fork.c:1161
>>>> copy_process+0x4eb/0x57d0 root/klinux/kernel/fork.c:2261
>>>> kernel_clone+0xdd/0x910 root/klinux/kernel/fork.c:2849
>>>> user_mode_thread+0xcb/0x110 root/klinux/kernel/fork.c:2927
>>>> call_usermodehelper_exec_sync root/klinux/kernel/umh.c:133 [inline]
>>>> call_usermodehelper_exec_work+0x69/0x170 root/klinux/kernel/umh.c:164
>>>> process_one_work+0x90d/0x1aa0 root/klinux/kernel/workqueue.c:3235
>>>> process_scheduled_works root/klinux/kernel/workqueue.c:3315 [inline]
>>>> worker_thread+0x633/0xe80 root/klinux/kernel/workqueue.c:3396
>>>> kthread+0x317/0x420 root/klinux/kernel/kthread.c:388
>>>> ret_from_fork+0x4ae/0x5e0 root/klinux/arch/x86/kernel/process.c:152
>>>> ret_from_fork_asm+0x1a/0x30 root/klinux/arch/x86/entry/entry_64.S:293
>>>>
>>>> INFO: task syz-executor.0:37820 blocked for more than 143 seconds.
>>>> Tainted: G S ------- ---- 6.6.89+ #931
>>>> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>> task:syz-executor.0 state:D stack:0 pid:37820 tgid:37820 ppid:1 flags:0x00000004
>>>> Call Trace:
>>>>
>>>> context_switch root/klinux/kernel/sched/core.c:5346 [inline]
>>>> __schedule+0x7bc/0x1310 root/klinux/kernel/sched/core.c:6720
>>>> __schedule_loop root/klinux/kernel/sched/core.c:6797 [inline]
>>>> schedule+0xe2/0x290 root/klinux/kernel/sched/core.c:6812
>>>> schedule_preempt_disabled+0x18/0x30 root/klinux/kernel/sched/core.c:6869
>>>> __mutex_lock_common root/klinux/kernel/locking/mutex.c:679 [inline]
>>>> __mutex_lock+0x914/0x13a0 root/klinux/kernel/locking/mutex.c:747
>>>> class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> do_shrink_slab+0x3c9/0xf30 root/klinux/mm/shrinker.c:435
>>>> shrink_slab+0x175/0x5e0 root/klinux/mm/shrinker.c:662
>>>> shrink_node_memcgs+0x4d6/0x700 root/klinux/mm/vmscan.c:5900
>>>> shrink_node+0x247/0x12c0 root/klinux/mm/vmscan.c:5933
>>>> shrink_zones root/klinux/mm/vmscan.c:6175 [inline]
>>>> do_try_to_free_pages+0x314/0x1580 root/klinux/mm/vmscan.c:6237
>>>> try_to_free_pages+0x293/0x740 root/klinux/mm/vmscan.c:6472
>>>> __perform_reclaim root/klinux/mm/page_alloc.c:3965 [inline]
>>>> __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> __alloc_pages_slowpath.constprop.0+0x800/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> __alloc_pages_noprof+0x490/0x570 root/klinux/mm/page_alloc.c:4792
>>>> alloc_pages_mpol_noprof+0x1ee/0x530 root/klinux/mm/mempolicy.c:2270
>>>> pagetable_alloc_noprof root/klinux/./include/linux/mm.h:2982 [inline]
>>>> __pud_alloc_one_noprof root/klinux/./include/asm-generic/pgalloc.h:174 [inline]
>>>> pud_alloc_one_noprof root/klinux/./include/asm-generic/pgalloc.h:195 [inline]
>>>> __pud_alloc+0x40/0x650 root/klinux/mm/memory.c:6222
>>>> pud_alloc root/klinux/./include/linux/mm.h:2939 [inline]
>>>> copy_pud_range root/klinux/mm/memory.c:1270 [inline]
>>>> copy_p4d_range+0x95e/0xb70 root/klinux/mm/memory.c:1314
>>>> copy_page_range+0x432/0xa30 root/klinux/mm/memory.c:1408
>>>> dup_mmap+0x1127/0x1ef0 root/klinux/kernel/fork.c:773
>>>> dup_mm+0x93/0x4e0 root/klinux/kernel/fork.c:1722
>>>> copy_mm root/klinux/kernel/fork.c:1781 [inline]
>>>> copy_process+0x2c89/0x57d0 root/klinux/kernel/fork.c:2436
>>>> kernel_clone+0xdd/0x910 root/klinux/kernel/fork.c:2849
>>>> __do_sys_clone+0xcf/0x120 root/klinux/kernel/fork.c:2992
>>>> do_syscall_x64 root/klinux/arch/x86/entry/common.c:51 [inline]
>>>> do_syscall_64+0x74/0x360 root/klinux/arch/x86/entry/common.c:81
>>>> entry_SYSCALL_64_after_hwframe+0x76/0xe0
>>>> RIP: 0033:0x45d778
>>>> RSP: 002b:00007ffc4d846860 EFLAGS: 00000246 ORIG_RAX: 0000000000000038
>>>> RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 000000000045d778
>>>> RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000001200011
>>>> RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000013782980
>>>> R10: 0000000013782c50 R11: 0000000000000246 R12: 0000000000000000
>>>> R13: 0000000000000032 R14: 0000000000000000 R15: 0000000000000000
>>>>
>>>> INFO: task syz-executor.10:56553 blocked for more than 143 seconds.
>>>> Tainted: G S ------- ---- 6.6.89+ #931
>>>> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>> task:syz-executor.10 state:D stack:0 pid:56553 tgid:56553 ppid:1 flags:0x00000004
>>>> Call Trace:
>>>>
>>>> context_switch root/klinux/kernel/sched/core.c:5346 [inline]
>>>> __schedule+0x7bc/0x1310 root/klinux/kernel/sched/core.c:6720
>>>> __schedule_loop root/klinux/kernel/sched/core.c:6797 [inline]
>>>> schedule+0xe2/0x290 root/klinux/kernel/sched/core.c:6812
>>>> schedule_preempt_disabled+0x18/0x30 root/klinux/kernel/sched/core.c:6869
>>>> __mutex_lock_common root/klinux/kernel/locking/mutex.c:679 [inline]
>>>> __mutex_lock+0x914/0x13a0 root/klinux/kernel/locking/mutex.c:747
>>>> class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> do_shrink_slab+0x3c9/0xf30 root/klinux/mm/shrinker.c:435
>>>> shrink_slab+0x175/0x5e0 root/klinux/mm/shrinker.c:662
>>>> shrink_node_memcgs+0x4d6/0x700 root/klinux/mm/vmscan.c:5900
>>>> shrink_node+0x247/0x12c0 root/klinux/mm/vmscan.c:5933
>>>> shrink_zones root/klinux/mm/vmscan.c:6175 [inline]
>>>> do_try_to_free_pages+0x314/0x1580 root/klinux/mm/vmscan.c:6237
>>>> try_to_free_pages+0x293/0x740 root/klinux/mm/vmscan.c:6472
>>>> __perform_reclaim root/klinux/mm/page_alloc.c:3965 [inline]
>>>> __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> __alloc_pages_slowpath.constprop.0+0x800/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> __alloc_pages_noprof+0x490/0x570 root/klinux/mm/page_alloc.c:4792
>>>> __alloc_pages_node_noprof root/klinux/./include/linux/gfp.h:244 [inline]
>>>> alloc_pages_node_noprof root/klinux/./include/linux/gfp.h:271 [inline]
>>>> alloc_thread_stack_node root/klinux/kernel/fork.c:367 [inline]
>>>> dup_task_struct+0xa4/0x6c0 root/klinux/kernel/fork.c:1161
>>>> copy_process+0x4eb/0x57d0 root/klinux/kernel/fork.c:2261
>>>> kernel_clone+0xdd/0x910 root/klinux/kernel/fork.c:2849
>>>> __do_sys_clone+0xcf/0x120 root/klinux/kernel/fork.c:2992
>>>> do_syscall_x64 root/klinux/arch/x86/entry/common.c:51 [inline]
>>>> do_syscall_64+0x74/0x360 root/klinux/arch/x86/entry/common.c:81
>>>> entry_SYSCALL_64_after_hwframe+0x76/0xe0
>>>> RIP: 0033:0x45d778
>>>> RSP: 002b:00007ffc6a4dad70 EFLAGS: 00000246 ORIG_RAX: 0000000000000038
>>>> RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 000000000045d778
>>>> RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000001200011
>>>> RBP: 0000000000000000 R08: 0000000000000000 R09: 000000002d906980
>>>> R10: 000000002d906c50 R11: 0000000000000246 R12: 0000000000000000
>>>> R13: 0000000000000032 R14: 0000000000000001 R15: 0000000000000000
>>>>
>>>> INFO: task kworker/12:13:56977 blocked for more than 144 seconds.
>>>> Tainted: G S ------- ---- 6.6.89+ #931
>>>> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>> task:kworker/12:13 state:D stack:0 pid:56977 tgid:56977 ppid:2 flags:0x00004000
>>>> Workqueue: events drain_vmap_area_work
>>>> Call Trace:
>>>>
>>>> context_switch root/klinux/kernel/sched/core.c:5346 [inline]
>>>> __schedule+0x7bc/0x1310 root/klinux/kernel/sched/core.c:6720
>>>> __schedule_loop root/klinux/kernel/sched/core.c:6797 [inline]
>>>> schedule+0xe2/0x290 root/klinux/kernel/sched/core.c:6812
>>>> schedule_timeout+0x268/0x2b0 root/klinux/kernel/time/timer.c:2142
>>>> do_wait_for_common root/klinux/kernel/sched/completion.c:95 [inline]
>>>> __wait_for_common+0x3ea/0x600 root/klinux/kernel/sched/completion.c:116
>>>> __flush_work+0xf4/0x180 root/klinux/kernel/workqueue.c:4125
>>>> __purge_vmap_area_lazy+0x8e5/0xcf0 root/klinux/mm/vmalloc.c:2324
>>>> drain_vmap_area_work+0x2c/0x40 root/klinux/mm/vmalloc.c:2349
>>>> process_one_work+0x90d/0x1aa0 root/klinux/kernel/workqueue.c:3235
>>>> process_scheduled_works root/klinux/kernel/workqueue.c:3315 [inline]
>>>> worker_thread+0x633/0xe80 root/klinux/kernel/workqueue.c:3396
>>>> kthread+0x317/0x420 root/klinux/kernel/kthread.c:388
>>>> ret_from_fork+0x4ae/0x5e0 root/klinux/arch/x86/kernel/process.c:152
>>>> ret_from_fork_asm+0x1a/0x30 root/klinux/arch/x86/entry/entry_64.S:293
>>>>
>>>> INFO: task kdamond.0:66163 blocked for more than 144 seconds.
>>>> Tainted: G S ------- ---- 6.6.89+ #931
>>>> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>> task:kdamond.0 state:D stack:0 pid:66163 tgid:66163 ppid:2 flags:0x00004000
>>>> Call Trace:
>>>>
>>>> context_switch root/klinux/kernel/sched/core.c:5346 [inline]
>>>> __schedule+0x7bc/0x1310 root/klinux/kernel/sched/core.c:6720
>>>> __schedule_loop root/klinux/kernel/sched/core.c:6797 [inline]
>>>> schedule+0xe2/0x290 root/klinux/kernel/sched/core.c:6812
>>>> schedule_preempt_disabled+0x18/0x30 root/klinux/kernel/sched/core.c:6869
>>>> __mutex_lock_common root/klinux/kernel/locking/mutex.c:679 [inline]
>>>> __mutex_lock+0x914/0x13a0 root/klinux/kernel/locking/mutex.c:747
>>>> class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> do_shrink_slab+0x3c9/0xf30 root/klinux/mm/shrinker.c:435
>>>> shrink_slab+0x175/0x5e0 root/klinux/mm/shrinker.c:662
>>>> shrink_node_memcgs+0x4d6/0x700 root/klinux/mm/vmscan.c:5900
>>>> shrink_node+0x247/0x12c0 root/klinux/mm/vmscan.c:5933
>>>> shrink_zones root/klinux/mm/vmscan.c:6175 [inline]
>>>> do_try_to_free_pages+0x314/0x1580 root/klinux/mm/vmscan.c:6237
>>>> try_to_free_pages+0x293/0x740 root/klinux/mm/vmscan.c:6472
>>>> __perform_reclaim root/klinux/mm/page_alloc.c:3965 [inline]
>>>> __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> __alloc_pages_slowpath.constprop.0+0x800/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> __alloc_pages_noprof+0x490/0x570 root/klinux/mm/page_alloc.c:4792
>>>> alloc_pages_mpol_noprof+0x1ee/0x530 root/klinux/mm/mempolicy.c:2270
>>>> alloc_slab_page root/klinux/mm/slub.c:2315 [inline]
>>>> allocate_slab+0x27a/0x3f0 root/klinux/mm/slub.c:2485
>>>> ___slab_alloc+0x889/0x1060 root/klinux/mm/slub.c:3727
>>>> __slab_alloc.constprop.0+0x52/0xa0 root/klinux/mm/slub.c:3817
>>>> __slab_alloc_node root/klinux/mm/slub.c:3870 [inline]
>>>> slab_alloc_node root/klinux/mm/slub.c:4031 [inline]
>>>> kmem_cache_alloc_noprof+0x385/0x3e0 root/klinux/mm/slub.c:4050
>>>> damon_new_region+0x29/0x1a0 root/klinux/mm/damon/core.c:124
>>>> damon_va_evenly_split_region+0x1f8/0x2c0 root/klinux/mm/damon/vaddr.c:85
>>>> __damon_va_init_regions root/klinux/mm/damon/vaddr.c:271 [inline]
>>>> damon_va_init+0x2b5/0x470 root/klinux/mm/damon/vaddr.c:283
>>>> kdamond_fn+0x289/0x1240 root/klinux/mm/damon/core.c:1432
>>>> kthread+0x317/0x420 root/klinux/kernel/kthread.c:388
>>>> ret_from_fork+0x4ae/0x5e0 root/klinux/arch/x86/kernel/process.c:152
>>>> ret_from_fork_asm+0x1a/0x30 root/klinux/arch/x86/entry/entry_64.S:293
>>>>
>>>> INFO: task kworker/u80:5:67724 blocked for more than 144 seconds.
>>>> Tainted: G S ------- ---- 6.6.89+ #931
>>>> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>> task:kworker/u80:5 state:D stack:0 pid:67724 tgid:67724 ppid:2 flags:0x00004000
>>>> Workqueue: events_unbound call_usermodehelper_exec_work
>>>> Call Trace:
>>>>
>>>> context_switch root/klinux/kernel/sched/core.c:5346 [inline]
>>>> __schedule+0x7bc/0x1310 root/klinux/kernel/sched/core.c:6720
>>>> __schedule_loop root/klinux/kernel/sched/core.c:6797 [inline]
>>>> schedule+0xe2/0x290 root/klinux/kernel/sched/core.c:6812
>>>> schedule_preempt_disabled+0x18/0x30 root/klinux/kernel/sched/core.c:6869
>>>> __mutex_lock_common root/klinux/kernel/locking/mutex.c:679 [inline]
>>>> __mutex_lock+0x914/0x13a0 root/klinux/kernel/locking/mutex.c:747
>>>> class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> do_shrink_slab+0x3c9/0xf30 root/klinux/mm/shrinker.c:435
>>>> shrink_slab+0x175/0x5e0 root/klinux/mm/shrinker.c:662
>>>> shrink_node_memcgs+0x4d6/0x700 root/klinux/mm/vmscan.c:5900
>>>> shrink_node+0x247/0x12c0 root/klinux/mm/vmscan.c:5933
>>>> shrink_zones root/klinux/mm/vmscan.c:6175 [inline]
>>>> do_try_to_free_pages+0x314/0x1580 root/klinux/mm/vmscan.c:6237
>>>> try_to_free_pages+0x293/0x740 root/klinux/mm/vmscan.c:6472
>>>> __perform_reclaim root/klinux/mm/page_alloc.c:3965 [inline]
>>>> __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> __alloc_pages_slowpath.constprop.0+0x800/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> __alloc_pages_noprof+0x490/0x570 root/klinux/mm/page_alloc.c:4792
>>>> __alloc_pages_node_noprof root/klinux/./include/linux/gfp.h:244 [inline]
>>>> alloc_pages_node_noprof root/klinux/./include/linux/gfp.h:271 [inline]
>>>> alloc_thread_stack_node root/klinux/kernel/fork.c:367 [inline]
>>>> dup_task_struct+0xa4/0x6c0 root/klinux/kernel/fork.c:1161
>>>> copy_process+0x4eb/0x57d0 root/klinux/kernel/fork.c:2261
>>>> kernel_clone+0xdd/0x910 root/klinux/kernel/fork.c:2849
>>>> user_mode_thread+0xcb/0x110 root/klinux/kernel/fork.c:2927
>>>> call_usermodehelper_exec_sync root/klinux/kernel/umh.c:133 [inline]
>>>> call_usermodehelper_exec_work+0x69/0x170 root/klinux/kernel/umh.c:164
>>>> process_one_work+0x90d/0x1aa0 root/klinux/kernel/workqueue.c:3235
>>>> process_scheduled_works root/klinux/kernel/workqueue.c:3315 [inline]
>>>> worker_thread+0x633/0xe80 root/klinux/kernel/workqueue.c:3396
>>>> kthread+0x317/0x420 root/klinux/kernel/kthread.c:388
>>>> ret_from_fork+0x4ae/0x5e0 root/klinux/arch/x86/kernel/process.c:152
>>>> ret_from_fork_asm+0x1a/0x30 root/klinux/arch/x86/entry/entry_64.S:293
>>>>
>>>> INFO: task kworker/u80:9:69412 blocked for more than 144 seconds.
>>>> Tainted: G S ------- ---- 6.6.89+ #931
>>>> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>> task:kworker/u80:9 state:D stack:0 pid:69412 tgid:69412 ppid:2 flags:0x00004000
>>>> Workqueue: events_unbound call_usermodehelper_exec_work
>>>> Call Trace:
>>>>
>>>> context_switch root/klinux/kernel/sched/core.c:5346 [inline]
>>>> __schedule+0x7bc/0x1310 root/klinux/kernel/sched/core.c:6720
>>>> __schedule_loop root/klinux/kernel/sched/core.c:6797 [inline]
>>>> schedule+0xe2/0x290 root/klinux/kernel/sched/core.c:6812
>>>> schedule_preempt_disabled+0x18/0x30 root/klinux/kernel/sched/core.c:6869
>>>> __mutex_lock_common root/klinux/kernel/locking/mutex.c:679 [inline]
>>>> __mutex_lock+0x914/0x13a0 root/klinux/kernel/locking/mutex.c:747
>>>> class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> do_shrink_slab+0x3c9/0xf30 root/klinux/mm/shrinker.c:435
>>>> shrink_slab+0x175/0x5e0 root/klinux/mm/shrinker.c:662
>>>> shrink_node_memcgs+0x4d6/0x700 root/klinux/mm/vmscan.c:5900
>>>> shrink_node+0x247/0x12c0 root/klinux/mm/vmscan.c:5933
>>>> shrink_zones root/klinux/mm/vmscan.c:6175 [inline]
>>>> do_try_to_free_pages+0x314/0x1580 root/klinux/mm/vmscan.c:6237
>>>> try_to_free_pages+0x293/0x740 root/klinux/mm/vmscan.c:6472
>>>> __perform_reclaim root/klinux/mm/page_alloc.c:3965 [inline]
>>>> __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> __alloc_pages_slowpath.constprop.0+0x800/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> __alloc_pages_noprof+0x490/0x570 root/klinux/mm/page_alloc.c:4792
>>>> __alloc_pages_node_noprof root/klinux/./include/linux/gfp.h:244 [inline]
>>>> alloc_pages_node_noprof root/klinux/./include/linux/gfp.h:271 [inline]
>>>> alloc_thread_stack_node root/klinux/kernel/fork.c:367 [inline]
>>>> dup_task_struct+0xa4/0x6c0 root/klinux/kernel/fork.c:1161
>>>> copy_process+0x4eb/0x57d0 root/klinux/kernel/fork.c:2261
>>>> kernel_clone+0xdd/0x910 root/klinux/kernel/fork.c:2849
>>>> user_mode_thread+0xcb/0x110 root/klinux/kernel/fork.c:2927
>>>> call_usermodehelper_exec_sync root/klinux/kernel/umh.c:133 [inline]
>>>> call_usermodehelper_exec_work+0x69/0x170 root/klinux/kernel/umh.c:164
>>>> process_one_work+0x90d/0x1aa0 root/klinux/kernel/workqueue.c:3235
>>>> process_scheduled_works root/klinux/kernel/workqueue.c:3315 [inline]
>>>> worker_thread+0x633/0xe80 root/klinux/kernel/workqueue.c:3396
>>>> kthread+0x317/0x420 root/klinux/kernel/kthread.c:388
>>>> ret_from_fork+0x4ae/0x5e0 root/klinux/arch/x86/kernel/process.c:152
>>>> ret_from_fork_asm+0x1a/0x30 root/klinux/arch/x86/entry/entry_64.S:293
>>>>
>>>> INFO: task syz-executor.18:73039 blocked for more than 144 seconds.
>>>> Tainted: G S ------- ---- 6.6.89+ #931
>>>> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>> task:syz-executor.18 state:D stack:0 pid:73039 tgid:73039 ppid:1 flags:0x00004004
>>>> Call Trace:
>>>>
>>>> context_switch root/klinux/kernel/sched/core.c:5346 [inline]
>>>> __schedule+0x7bc/0x1310 root/klinux/kernel/sched/core.c:6720
>>>> __schedule_loop root/klinux/kernel/sched/core.c:6797 [inline]
>>>> schedule+0xe2/0x290 root/klinux/kernel/sched/core.c:6812
>>>> schedule_preempt_disabled+0x18/0x30 root/klinux/kernel/sched/core.c:6869
>>>> __mutex_lock_common root/klinux/kernel/locking/mutex.c:679 [inline]
>>>> __mutex_lock+0x914/0x13a0 root/klinux/kernel/locking/mutex.c:747
>>>> class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> do_shrink_slab+0x3c9/0xf30 root/klinux/mm/shrinker.c:435
>>>> shrink_slab+0x175/0x5e0 root/klinux/mm/shrinker.c:662
>>>> shrink_node_memcgs+0x4d6/0x700 root/klinux/mm/vmscan.c:5900
>>>> shrink_node+0x247/0x12c0 root/klinux/mm/vmscan.c:5933
>>>> shrink_zones root/klinux/mm/vmscan.c:6175 [inline]
>>>> do_try_to_free_pages+0x314/0x1580 root/klinux/mm/vmscan.c:6237
>>>> try_to_free_pages+0x293/0x740 root/klinux/mm/vmscan.c:6472
>>>> __perform_reclaim root/klinux/mm/page_alloc.c:3965 [inline]
>>>> __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> __alloc_pages_slowpath.constprop.0+0x800/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> __alloc_pages_noprof+0x490/0x570 root/klinux/mm/page_alloc.c:4792
>>>> alloc_pages_mpol_noprof+0x1ee/0x530 root/klinux/mm/mempolicy.c:2270
>>>> folio_alloc_mpol_noprof+0x3b/0xd0 root/klinux/mm/mempolicy.c:2288
>>>> vma_alloc_folio_noprof+0xaf/0x140 root/klinux/mm/mempolicy.c:2316
>>>> folio_prealloc root/klinux/mm/memory.c:1062 [inline]
>>>> wp_page_copy+0x21a/0x11c0 root/klinux/mm/memory.c:3308
>>>> do_wp_page+0x762/0x14e0 root/klinux/mm/memory.c:3700
>>>> handle_pte_fault+0x573/0x730 root/klinux/mm/memory.c:5717
>>>> __handle_mm_fault+0xae4/0x10d0 root/klinux/mm/memory.c:5844
>>>> handle_mm_fault+0x376/0x650 root/klinux/mm/memory.c:6009
>>>> do_user_addr_fault+0x2be/0xc40 root/klinux/arch/x86/mm/fault.c:1330
>>>> handle_page_fault root/klinux/arch/x86/mm/fault.c:1471 [inline]
>>>> exc_page_fault+0x5c/0xe0 root/klinux/arch/x86/mm/fault.c:1527
>>>> asm_exc_page_fault+0x26/0x30 root/klinux/./arch/x86/include/asm/idtentry.h:608
>>>> RIP: 0033:0x45d8e1
>>>> RSP: 002b:00007ffd015d5e10 EFLAGS: 00010246
>>>> RAX: 000000000000003b RBX: 0000000000000000 RCX: 000000000045d778
>>>> RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000002
>>>> RBP: 0000000000000000 R08: 0000000000000000 R09: 000000000701d980
>>>> R10: 000000000701dc50 R11: 0000000000000246 R12: 000000000000003b
>>>> R13: 0000000000000032 R14: 0000000000000000 R15: 0000000000000000
>>>>
>>>> INFO: task syz-executor.13:74790 blocked for more than 144 seconds.
>>>> Tainted: G S ------- ---- 6.6.89+ #931
>>>> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>> task:syz-executor.13 state:D stack:0 pid:74790 tgid:74790 ppid:1 flags:0x00000004
>>>> Call Trace:
>>>>
>>>> context_switch root/klinux/kernel/sched/core.c:5346 [inline]
>>>> __schedule+0x7bc/0x1310 root/klinux/kernel/sched/core.c:6720
>>>> __schedule_loop root/klinux/kernel/sched/core.c:6797 [inline]
>>>> schedule+0xe2/0x290 root/klinux/kernel/sched/core.c:6812
>>>> schedule_preempt_disabled+0x18/0x30 root/klinux/kernel/sched/core.c:6869
>>>> __mutex_lock_common root/klinux/kernel/locking/mutex.c:679 [inline]
>>>> __mutex_lock+0x914/0x13a0 root/klinux/kernel/locking/mutex.c:747
>>>> class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> do_shrink_slab+0x3c9/0xf30 root/klinux/mm/shrinker.c:435
>>>> shrink_slab+0x175/0x5e0 root/klinux/mm/shrinker.c:662
>>>> shrink_node_memcgs+0x4d6/0x700 root/klinux/mm/vmscan.c:5900
>>>> shrink_node+0x247/0x12c0 root/klinux/mm/vmscan.c:5933
>>>> shrink_zones root/klinux/mm/vmscan.c:6175 [inline]
>>>> do_try_to_free_pages+0x314/0x1580 root/klinux/mm/vmscan.c:6237
>>>> try_to_free_pages+0x293/0x740 root/klinux/mm/vmscan.c:6472
>>>> __perform_reclaim root/klinux/mm/page_alloc.c:3965 [inline]
>>>> __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> __alloc_pages_slowpath.constprop.0+0x800/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> __alloc_pages_noprof+0x490/0x570 root/klinux/mm/page_alloc.c:4792
>>>> alloc_pages_mpol_noprof+0x1ee/0x530 root/klinux/mm/mempolicy.c:2270
>>>> folio_alloc_mpol_noprof+0x3b/0xd0 root/klinux/mm/mempolicy.c:2288
>>>> vma_alloc_folio_noprof+0xaf/0x140 root/klinux/mm/mempolicy.c:2316
>>>> folio_prealloc root/klinux/mm/memory.c:1062 [inline]
>>>> wp_page_copy+0x21a/0x11c0 root/klinux/mm/memory.c:3308
>>>> do_wp_page+0x762/0x14e0 root/klinux/mm/memory.c:3700
>>>> handle_pte_fault+0x573/0x730 root/klinux/mm/memory.c:5717
>>>> __handle_mm_fault+0xae4/0x10d0 root/klinux/mm/memory.c:5844
>>>> handle_mm_fault+0x376/0x650 root/klinux/mm/memory.c:6009
>>>> do_user_addr_fault+0x2be/0xc40 root/klinux/arch/x86/mm/fault.c:1330
>>>> handle_page_fault root/klinux/arch/x86/mm/fault.c:1471 [inline]
>>>> exc_page_fault+0x5c/0xe0 root/klinux/arch/x86/mm/fault.c:1527
>>>> asm_exc_page_fault+0x26/0x30 root/klinux/./arch/x86/include/asm/idtentry.h:608
>>>> RIP: 0033:0x45d8e1
>>>> RSP: 002b:00007ffe18ac7000 EFLAGS: 00010246
>>>> RAX: 0000000000000038 RBX: 0000000000000000 RCX: 000000000045d778
>>>> RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000002
>>>> RBP: 0000000000000000 R08: 0000000000000000 R09: 000000000eb5f980
>>>> R10: 000000000eb5fc50 R11: 0000000000000246 R12: 0000000000000038
>>>> R13: 0000000000000032 R14: 0000000000000001 R15: 0000000000000000
>>>>
>>>> INFO: task syz-executor.5:78550 blocked for more than 144 seconds.
>>>> Tainted: G S ------- ---- 6.6.89+ #931
>>>> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>>> task:syz-executor.5 state:D stack:0 pid:78550 tgid:78550 ppid:1 flags:0x00000004
>>>> Call Trace:
>>>>
>>>> context_switch root/klinux/kernel/sched/core.c:5346 [inline]
>>>> __schedule+0x7bc/0x1310 root/klinux/kernel/sched/core.c:6720
>>>> __schedule_loop root/klinux/kernel/sched/core.c:6797 [inline]
>>>> schedule+0xe2/0x290 root/klinux/kernel/sched/core.c:6812
>>>> schedule_preempt_disabled+0x18/0x30 root/klinux/kernel/sched/core.c:6869
>>>> __mutex_lock_common root/klinux/kernel/locking/mutex.c:679 [inline]
>>>> __mutex_lock+0x914/0x13a0 root/klinux/kernel/locking/mutex.c:747
>>>> class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> do_shrink_slab+0x3c9/0xf30 root/klinux/mm/shrinker.c:435
>>>> shrink_slab+0x175/0x5e0 root/klinux/mm/shrinker.c:662
>>>> shrink_node_memcgs+0x4d6/0x700 root/klinux/mm/vmscan.c:5900
>>>> shrink_node+0x247/0x12c0 root/klinux/mm/vmscan.c:5933
>>>> shrink_zones root/klinux/mm/vmscan.c:6175 [inline]
>>>> do_try_to_free_pages+0x314/0x1580 root/klinux/mm/vmscan.c:6237
>>>> try_to_free_pages+0x293/0x740 root/klinux/mm/vmscan.c:6472
>>>> __perform_reclaim root/klinux/mm/page_alloc.c:3965 [inline]
>>>> __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> __alloc_pages_slowpath.constprop.0+0x800/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> __alloc_pages_noprof+0x490/0x570 root/klinux/mm/page_alloc.c:4792
>>>> alloc_pages_mpol_noprof+0x1ee/0x530 root/klinux/mm/mempolicy.c:2270
>>>> folio_alloc_mpol_noprof+0x3b/0xd0 root/klinux/mm/mempolicy.c:2288
>>>> vma_alloc_folio_noprof+0xaf/0x140 root/klinux/mm/mempolicy.c:2316
>>>> folio_prealloc root/klinux/mm/memory.c:1062 [inline]
>>>> wp_page_copy+0x21a/0x11c0 root/klinux/mm/memory.c:3308
>>>> do_wp_page+0x762/0x14e0 root/klinux/mm/memory.c:3700
>>>> handle_pte_fault+0x573/0x730 root/klinux/mm/memory.c:5717
>>>> __handle_mm_fault+0xae4/0x10d0 root/klinux/mm/memory.c:5844
>>>> handle_mm_fault+0x376/0x650 root/klinux/mm/memory.c:6009
>>>> do_user_addr_fault+0x2be/0xc40 root/klinux/arch/x86/mm/fault.c:1330
>>>> handle_page_fault root/klinux/arch/x86/mm/fault.c:1471 [inline]
>>>> exc_page_fault+0x5c/0xe0 root/klinux/arch/x86/mm/fault.c:1527
>>>> asm_exc_page_fault+0x26/0x30 root/klinux/./arch/x86/include/asm/idtentry.h:608
>>>> RIP: 0033:0x45d8e1
>>>> RSP: 002b:00007fff9b912230 EFLAGS: 00010246
>>>> RAX: 0000000000000019 RBX: 0000000000000000 RCX: 000000000045d778
>>>> RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000002
>>>> RBP: 0000000000000000 R08: 0000000000000000 R09: 000000000cd3c980
>>>> R10: 000000000cd3cc50 R11: 0000000000000246 R12: 0000000000000019
>>>> R13: 0000000000000032 R14: 0000000000000001 R15: 0000000000000000
>>>>
>>>> Future hung task reports are suppressed, see sysctl kernel.hung_task_warnings
>>>>
>>>> Showing all locks held in the system:
>>>> 2 locks held by kthreadd/2:
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 4 locks held by kworker/u80:6/141:
>>>> #0: ffff8881041a4d58 ((wq_completion)events_unbound){+.+.}-{0:0}, at: process_one_work+0x803/0x1aa0 root/klinux/kernel/workqueue.c:3210
>>>> #1: ffff8881053cfd80 ((work_completion)(&sub_info->work)){+.+.}-{0:0}, at: process_one_work+0x859/0x1aa0 root/klinux/kernel/workqueue.c:3211
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #3: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #3: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 1 lock held by khungtaskd/155:
>>>> #0: ffffffff8b33eb00 (rcu_read_lock){....}-{1:2}, at: rcu_lock_acquire root/klinux/./include/linux/rcupdate.h:334 [inline]
>>>> #0: ffffffff8b33eb00 (rcu_read_lock){....}-{1:2}, at: rcu_read_lock root/klinux/./include/linux/rcupdate.h:786 [inline]
>>>> #0: ffffffff8b33eb00 (rcu_read_lock){....}-{1:2}, at: debug_show_all_locks+0x75/0x340 root/klinux/kernel/locking/lockdep.c:6721
>>>> 2 locks held by bash/957:
>>>> #0: ffff88800310d0a8 (&tty->ldisc_sem){++++}-{0:0}, at: tty_ldisc_ref_wait+0x29/0x80 root/klinux/drivers/tty/tty_ldisc.c:243
>>>> #1: ffffc900039852f8 (&ldata->atomic_read_lock){+.+.}-{3:3}, at: n_tty_read+0xd55/0x1230 root/klinux/drivers/tty/n_tty.c:2217
>>>> 5 locks held by syz-executor.0/37820:
>>>> #0: ffffffff8b511780 (dup_mmap_sem){++++}-{0:0}, at: dup_mmap+0xbe/0x1ef0 root/klinux/kernel/fork.c:668
>>>> #1: ffff888106b396e8 (&mm->mmap_lock){++++}-{3:3}, at: mmap_write_lock_killable root/klinux/./include/linux/mmap_lock.h:124 [inline]
>>>> #1: ffff888106b396e8 (&mm->mmap_lock){++++}-{3:3}, at: dup_mmap+0xdc/0x1ef0 root/klinux/kernel/fork.c:669
>>>> #2: ffff88801ff62be8 (&mm->mmap_lock/1){+.+.}-{3:3}, at: mmap_write_lock_nested root/klinux/./include/linux/mmap_lock.h:115 [inline]
>>>> #2: ffff88801ff62be8 (&mm->mmap_lock/1){+.+.}-{3:3}, at: dup_mmap+0x125/0x1ef0 root/klinux/kernel/fork.c:678
>>>> #3: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #3: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #3: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #4: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #4: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 2 locks held by bash/41651:
>>>> #0: ffff88801f8cb8a8 (&tty->ldisc_sem){++++}-{0:0}, at: tty_ldisc_ref_wait+0x29/0x80 root/klinux/drivers/tty/tty_ldisc.c:243
>>>> #1: ffffc900001fd2f8 (&ldata->atomic_read_lock){+.+.}-{3:3}, at: n_tty_read+0xd55/0x1230 root/klinux/drivers/tty/n_tty.c:2217
>>>> 2 locks held by syz-executor.10/56553:
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by kworker/12:13/56977:
>>>> #0: ffff88810419cb58 ((wq_completion)events){+.+.}-{0:0}, at: process_one_work+0x803/0x1aa0 root/klinux/kernel/workqueue.c:3210
>>>> #1: ffff88804264fd80 (drain_vmap_work){+.+.}-{0:0}, at: process_one_work+0x859/0x1aa0 root/klinux/kernel/workqueue.c:3211
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: drain_vmap_area_work+0x1c/0x40 root/klinux/mm/vmalloc.c:2348
>>>> 2 locks held by kdamond.0/66163:
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 4 locks held by kworker/u80:5/67724:
>>>> #0: ffff8881041a4d58 ((wq_completion)events_unbound){+.+.}-{0:0}, at: process_one_work+0x803/0x1aa0 root/klinux/kernel/workqueue.c:3210
>>>> #1: ffff88804664fd80 ((work_completion)(&sub_info->work)){+.+.}-{0:0}, at: process_one_work+0x859/0x1aa0 root/klinux/kernel/workqueue.c:3211
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #3: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #3: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 4 locks held by kworker/u80:9/69412:
>>>> #0: ffff8881041a4d58 ((wq_completion)events_unbound){+.+.}-{0:0}, at: process_one_work+0x803/0x1aa0 root/klinux/kernel/workqueue.c:3210
>>>> #1: ffff8880346afd80 ((work_completion)(&sub_info->work)){+.+.}-{0:0}, at: process_one_work+0x859/0x1aa0 root/klinux/kernel/workqueue.c:3211
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #3: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #3: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by syz-executor.18/73039:
>>>> #0: ffff888048652940 (&vma->vm_lock->lock){++++}-{3:3}, at: vma_start_read root/klinux/./include/linux/mm.h:694 [inline]
>>>> #0: ffff888048652940 (&vma->vm_lock->lock){++++}-{3:3}, at: lock_vma_under_rcu+0x1c7/0x6f0 root/klinux/mm/memory.c:6165
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by syz-executor.13/74790:
>>>> #0: ffff88803e37eeb8 (&vma->vm_lock->lock){++++}-{3:3}, at: vma_start_read root/klinux/./include/linux/mm.h:694 [inline]
>>>> #0: ffff88803e37eeb8 (&vma->vm_lock->lock){++++}-{3:3}, at: lock_vma_under_rcu+0x1c7/0x6f0 root/klinux/mm/memory.c:6165
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by syz-executor.5/78550:
>>>> #0: ffff88801f4310e8 (&vma->vm_lock->lock){++++}-{3:3}, at: vma_start_read root/klinux/./include/linux/mm.h:694 [inline]
>>>> #0: ffff88801f4310e8 (&vma->vm_lock->lock){++++}-{3:3}, at: lock_vma_under_rcu+0x1c7/0x6f0 root/klinux/mm/memory.c:6165
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by syz-executor.26/80991:
>>>> #0: ffff88801b3c6068 (&mm->mmap_lock){++++}-{3:3}, at: mmap_read_lock root/klinux/./include/linux/mmap_lock.h:146 [inline]
>>>> #0: ffff88801b3c6068 (&mm->mmap_lock){++++}-{3:3}, at: __mm_populate+0x220/0x360 root/klinux/mm/gup.c:2170
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 4 locks held by syz-executor.5/80992:
>>>> #0: ffff888104871430 (sb_writers#2){.+.+}-{0:0}, at: get_signal+0x18d9/0x1a60 root/klinux/kernel/signal.c:2915
>>>> #1: ffff888001a29818 (&sb->s_type->i_mutex_key){++++}-{3:3}, at: inode_lock root/klinux/./include/linux/fs.h:828 [inline]
>>>> #1: ffff888001a29818 (&sb->s_type->i_mutex_key){++++}-{3:3}, at: generic_file_write_iter+0x91/0x2f0 root/klinux/mm/filemap.c:4181
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #3: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #3: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 2 locks held by syz-executor.22/81067:
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 2 locks held by syz-executor.31/81012:
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 2 locks held by syz-executor.31/81059:
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by syz-executor.23/81024:
>>>> #0: ffff88804a3aa080 (&vma->vm_lock->lock){++++}-{3:3}, at: vma_start_read root/klinux/./include/linux/mm.h:694 [inline]
>>>> #0: ffff88804a3aa080 (&vma->vm_lock->lock){++++}-{3:3}, at: lock_vma_under_rcu+0x1c7/0x6f0 root/klinux/mm/memory.c:6165
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 2 locks held by syz-executor.23/81030:
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by syz-executor.27/81031:
>>>> #0: ffff88802008cb70 (&vma->vm_lock->lock){++++}-{3:3}, at: vma_start_read root/klinux/./include/linux/mm.h:694 [inline]
>>>> #0: ffff88802008cb70 (&vma->vm_lock->lock){++++}-{3:3}, at: lock_vma_under_rcu+0x1c7/0x6f0 root/klinux/mm/memory.c:6165
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 2 locks held by syz-executor.30/81070:
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 4 locks held by syz-executor.28/81036:
>>>> #0: ffff888104871430 (sb_writers#2){.+.+}-{0:0}, at: filename_create+0xf9/0x4c0 root/klinux/fs/namei.c:3928
>>>> #1: ffff88817b041270 (&sb->s_type->i_mutex_key/1){+.+.}-{3:3}, at: inode_lock_nested root/klinux/./include/linux/fs.h:863 [inline]
>>>> #1: ffff88817b041270 (&sb->s_type->i_mutex_key/1){+.+.}-{3:3}, at: filename_create+0x1a1/0x4c0 root/klinux/fs/namei.c:3935
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #3: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #3: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by syz-executor.12/81086:
>>>> #0: ffff88801f2a45f8 (&vma->vm_lock->lock){++++}-{3:3}, at: vma_start_read root/klinux/./include/linux/mm.h:694 [inline]
>>>> #0: ffff88801f2a45f8 (&vma->vm_lock->lock){++++}-{3:3}, at: lock_vma_under_rcu+0x1c7/0x6f0 root/klinux/mm/memory.c:6165
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 4 locks held by syz-executor.22/81082:
>>>> #0: ffff888104871430 (sb_writers#2){.+.+}-{0:0}, at: get_signal+0x18d9/0x1a60 root/klinux/kernel/signal.c:2915
>>>> #1: ffff88805c122910 (&sb->s_type->i_mutex_key){++++}-{3:3}, at: inode_lock root/klinux/./include/linux/fs.h:828 [inline]
>>>> #1: ffff88805c122910 (&sb->s_type->i_mutex_key){++++}-{3:3}, at: generic_file_write_iter+0x91/0x2f0 root/klinux/mm/filemap.c:4181
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #2: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #3: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #3: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by syz-executor.17/81084:
>>>> #0: ffff888084e01430 (&vma->vm_lock->lock){++++}-{3:3}, at: vma_start_read root/klinux/./include/linux/mm.h:694 [inline]
>>>> #0: ffff888084e01430 (&vma->vm_lock->lock){++++}-{3:3}, at: lock_vma_under_rcu+0x1c7/0x6f0 root/klinux/mm/memory.c:6165
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by syz-executor.21/81088:
>>>> #0: ffff88801cde23c8 (&vma->vm_lock->lock){++++}-{3:3}, at: vma_start_read root/klinux/./include/linux/mm.h:694 [inline]
>>>> #0: ffff88801cde23c8 (&vma->vm_lock->lock){++++}-{3:3}, at: lock_vma_under_rcu+0x1c7/0x6f0 root/klinux/mm/memory.c:6165
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by syz-executor.5/81089:
>>>> #0: ffff888004d1abe8 (&mm->mmap_lock){++++}-{3:3}, at: mmap_read_trylock root/klinux/./include/linux/mmap_lock.h:165 [inline]
>>>> #0: ffff888004d1abe8 (&mm->mmap_lock){++++}-{3:3}, at: get_mmap_lock_carefully root/klinux/mm/memory.c:6036 [inline]
>>>> #0: ffff888004d1abe8 (&mm->mmap_lock){++++}-{3:3}, at: lock_mm_and_find_vma+0x35/0x5c0 root/klinux/mm/memory.c:6096
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by syz-executor.13/81090:
>>>> #0: ffff888004d1f568 (&mm->mmap_lock){++++}-{3:3}, at: mmap_read_trylock root/klinux/./include/linux/mmap_lock.h:165 [inline]
>>>> #0: ffff888004d1f568 (&mm->mmap_lock){++++}-{3:3}, at: get_mmap_lock_carefully root/klinux/mm/memory.c:6036 [inline]
>>>> #0: ffff888004d1f568 (&mm->mmap_lock){++++}-{3:3}, at: lock_mm_and_find_vma+0x35/0x5c0 root/klinux/mm/memory.c:6096
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by syz-executor.18/81091:
>>>> #0: ffff888004d1d5e8 (&mm->mmap_lock){++++}-{3:3}, at: mmap_read_trylock root/klinux/./include/linux/mmap_lock.h:165 [inline]
>>>> #0: ffff888004d1d5e8 (&mm->mmap_lock){++++}-{3:3}, at: get_mmap_lock_carefully root/klinux/mm/memory.c:6036 [inline]
>>>> #0: ffff888004d1d5e8 (&mm->mmap_lock){++++}-{3:3}, at: lock_mm_and_find_vma+0x35/0x5c0 root/klinux/mm/memory.c:6096
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 2 locks held by syz-executor.7/81096:
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #0: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #1: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>> 3 locks held by syz-executor.16/81095:
>>>> #0: ffff88801de15e08 (&vma->vm_lock->lock){++++}-{3:3}, at: vma_start_read root/klinux/./include/linux/mm.h:694 [inline]
>>>> #0: ffff88801de15e08 (&vma->vm_lock->lock){++++}-{3:3}, at: lock_vma_under_rcu+0x1c7/0x6f0 root/klinux/mm/memory.c:6165
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __perform_reclaim root/klinux/mm/page_alloc.c:3962 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_direct_reclaim root/klinux/mm/page_alloc.c:3987 [inline]
>>>> #1: ffffffff8b56b4c0 (fs_reclaim){+.+.}-{0:0}, at: __alloc_pages_slowpath.constprop.0+0x74d/0x1fb0 root/klinux/mm/page_alloc.c:4394
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: class_mutex_constructor root/klinux/./include/linux/mutex.h:200 [inline]
>>>> #2: ffffffff8b567390 (vmap_purge_lock){+.+.}-{3:3}, at: vmap_node_shrink_scan+0x1d/0x90 root/klinux/mm/vmalloc.c:5108
>>>>
>>>> =============================================
>>>>
>>>>
>>>>> --
>>>>> Uladzislau Rezki
>>>>
>>>> -- 
>>>> Thanks,
>>>> Ye Liu
>>>>
>>> Thank you!
>>>
>>> I can trigger below:
>>>
>>> <snip>
>>> Aug 25 13:56:52 pc638 kernel: Future hung task reports are suppressed, see sysctl kernel.hung_task_warnings
>>> Aug 25 13:56:52 pc638 kernel:
>>>                               Showing all locks held in the system:
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by khungtaskd/488:
>>> Aug 25 13:56:52 pc638 kernel:  #0: ffffffffa3306840 (rcu_read_lock){....}-{1:3}, at: debug_show_all_locks+0x47/0x19b
>>> Aug 25 13:56:52 pc638 kernel: 3 locks held by kworker/52:1/694:
>>> Aug 25 13:56:52 pc638 kernel:  #0: ffff8ac4401a4140 ((wq_completion)events){+.+.}-{0:0}, at: process_one_work+0x57b/0x600
>>> Aug 25 13:56:52 pc638 kernel:  #1: ffffce26435ffe30 (drain_vmap_work){+.+.}-{0:0}, at: process_one_work+0x1fc/0x600
>>> Aug 25 13:56:52 pc638 kernel:  #2: ffffffffa3406300 (vmap_purge_lock){+.+.}-{4:4}, at: drain_vmap_area_work+0x1b/0x40
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by in:imklog/3831:
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by vmalloc_test/2/5009:
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by vmalloc_test/4/5011:
>>> Aug 25 13:56:52 pc638 kernel:  #0: ffffffffa3406300 (vmap_purge_lock){+.+.}-{4:4}, at: alloc_vmap_area+0x8a8/0xd20
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by vmalloc_test/5/5012:
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by vmalloc_test/8/5015:
>>> Aug 25 13:56:52 pc638 kernel:  #0: ffffffffa3406300 (vmap_purge_lock){+.+.}-{4:4}, at: alloc_vmap_area+0x8a8/0xd20
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by vmalloc_test/30/5037:
>>> Aug 25 13:56:52 pc638 kernel:  #0: ffffffffa3406300 (vmap_purge_lock){+.+.}-{4:4}, at: alloc_vmap_area+0x8a8/0xd20
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by vmalloc_test/31/5038:
>>> Aug 25 13:56:52 pc638 kernel:  #0: ffffffffa3406300 (vmap_purge_lock){+.+.}-{4:4}, at: alloc_vmap_area+0x8a8/0xd20
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by vmalloc_test/38/5045:
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by vmalloc_test/39/5046:
>>> Aug 25 13:56:52 pc638 kernel:  #0: ffffffffa3406300 (vmap_purge_lock){+.+.}-{4:4}, at: alloc_vmap_area+0x8a8/0xd20
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by vmalloc_test/41/5048:
>>> Aug 25 13:56:52 pc638 kernel:  #0: ffffffffa3406300 (vmap_purge_lock){+.+.}-{4:4}, at: alloc_vmap_area+0x8a8/0xd20
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by vmalloc_test/49/5056:
>>> Aug 25 13:56:52 pc638 kernel:  #0: ffffffffa3406300 (vmap_purge_lock){+.+.}-{4:4}, at: alloc_vmap_area+0x8a8/0xd20
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by vmalloc_test/51/5058:
>>> Aug 25 13:56:52 pc638 kernel:  #0: ffffffffa3406300 (vmap_purge_lock){+.+.}-{4:4}, at: alloc_vmap_area+0x8a8/0xd20
>>> Aug 25 13:56:52 pc638 kernel: 1 lock held by vmalloc_test/57/5064:
>>> ...
>>> <snip>
>>>
>>> I think this is with your patch. But i can check also without.
>>> One potential issue to inline vnode-drain - is a time. At least
>>> we need to inject cond_resched() after each vnode flush +
>>>
>>> /*
>>>  * Reclaim vmap areas by purging fragmented blocks and purge_vmap_area_list.
>>>  */
>>> static void reclaim_and_purge_vmap_areas(void)
>>>
>>> {
>>> 	mutex_lock(&vmap_purge_lock);
>>> 	purge_fragmented_blocks_allcpus();
>>> 	__purge_vmap_area_lazy(ULONG_MAX, 0, true);
>>> 	mutex_unlock(&vmap_purge_lock);
>>> }
>>>
>>> here use try-lock as well due to potentially increased flush time.
>>>
>>> Any thoughts?
>>>
>>> --
>>> Uladzislau Rezki
>>
>>
>> IIUC we won't have to do inline purge if we switch both places to use mutex_trylock?
>>
> Looks like that. One concern about inline is a CPU hogging by one thread
> due to slow progress.
> 

So, should we fix it using the method below?

diff --git a/mm/vmalloc.c b/mm/vmalloc.c                                                         
index bea9f76ed7e7..e5c68b795a3e 100644                                                          
--- a/mm/vmalloc.c                                                                               
+++ b/mm/vmalloc.c                                                                               
@@ -2440,7 +2440,8 @@ static bool __purge_vmap_area_lazy(unsigned long start, unsigned long end, 
 static void reclaim_and_purge_vmap_areas(void)                                                  
                                                                                                 
 {                                                                                               
-       mutex_lock(&vmap_purge_lock);                                                            
+       if (!mutex_trylock(&vmap_purge_lock))                                                    
+               return;                                                                          
        purge_fragmented_blocks_allcpus();                                                       
        __purge_vmap_area_lazy(ULONG_MAX, 0, true);                                              
        mutex_unlock(&vmap_purge_lock);                                                          
@@ -5519,10 +5520,20 @@ vmap_node_shrink_scan(struct shrinker *shrink, struct shrink_control *sc)
 {                                                                                               
        struct vmap_node *vn;                                                                    
                                                                                                 
-       guard(mutex)(&vmap_purge_lock);                                                          
+       /*                                                                                       
+        * This shrinker is invoked from direct reclaim where memory                             
+        * pressure is already high.  Blocking on vmap_purge_lock here                           
+        * can deadlock the system: the lock holder may be blocked in                            
+        * flush_work() waiting for a worker that is stuck in this same                          
+        * reclaim path trying to acquire the same lock.  Use trylock                            
+        * to avoid this; skipping a pool decay cycle is harmless.                               
+        */                                                                                      
+       if (!mutex_trylock(&vmap_purge_lock))                                                    
+               return SHRINK_STOP;                                                              
        for_each_vmap_node(vn)                                                                   
                decay_va_pool_node(vn, true);                                                    
                                                                                                 
+       mutex_unlock(&vmap_purge_lock);                                                          
        return SHRINK_STOP;                                                                      
 }                                                                                               


> --
> Uladzislau Rezki

-- 
Thanks,
Ye Liu