===================================================== WARNING: HARDIRQ-safe -> HARDIRQ-unsafe lock order detected 6.1.82-syzkaller #0 Not tainted ----------------------------------------------------- kworker/0:2/151 [HC0[0]:SC0[2]:HE0:SE0] is trying to acquire: ffff88807ba9c8d8 (&htab->buckets[i].lock){+...}-{2:2}, at: sock_hash_delete_elem+0xac/0x2f0 net/core/sock_map.c:932 and this task is already holding: ffff8880b9828358 (&base->lock){-.-.}-{2:2}, at: lock_timer_base+0x120/0x260 kernel/time/timer.c:999 which would create a new lock dependency: (&base->lock){-.-.}-{2:2} -> (&htab->buckets[i].lock){+...}-{2:2} but this new dependency connects a HARDIRQ-irq-safe lock: (&base->lock){-.-.}-{2:2} ... which became HARDIRQ-irq-safe at: lock_acquire+0x1f8/0x5a0 kernel/locking/lockdep.c:5662 __raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:110 [inline] _raw_spin_lock_irqsave+0xd1/0x120 kernel/locking/spinlock.c:162 lock_timer_base+0x120/0x260 kernel/time/timer.c:999 add_timer_on+0x1eb/0x580 kernel/time/timer.c:1222 handle_irq_event_percpu kernel/irq/handle.c:195 [inline] handle_irq_event+0xa9/0x1e0 kernel/irq/handle.c:210 handle_level_irq+0x3ab/0x6c0 kernel/irq/chip.c:650 generic_handle_irq_desc include/linux/irqdesc.h:158 [inline] handle_irq arch/x86/kernel/irq.c:231 [inline] __common_interrupt+0xd7/0x1f0 arch/x86/kernel/irq.c:250 common_interrupt+0x9f/0xc0 arch/x86/kernel/irq.c:240 asm_common_interrupt+0x22/0x40 arch/x86/include/asm/idtentry.h:644 __raw_spin_unlock_irqrestore include/linux/spinlock_api_smp.h:152 [inline] _raw_spin_unlock_irqrestore+0xd4/0x130 kernel/locking/spinlock.c:194 __setup_irq+0x12fa/0x1d80 kernel/irq/manage.c:1809 request_threaded_irq+0x2a7/0x380 kernel/irq/manage.c:2198 request_irq include/linux/interrupt.h:168 [inline] setup_default_timer_irq+0x1f/0x30 arch/x86/kernel/time.c:70 x86_late_time_init+0x51/0x86 arch/x86/kernel/time.c:94 start_kernel+0x414/0x53f init/main.c:1101 secondary_startup_64_no_verify+0xcf/0xdb to a HARDIRQ-irq-unsafe lock: (&htab->buckets[i].lock){+...}-{2:2} ... which became HARDIRQ-irq-unsafe at: ... lock_acquire+0x1f8/0x5a0 kernel/locking/lockdep.c:5662 __raw_spin_lock_bh include/linux/spinlock_api_smp.h:126 [inline] _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:178 sock_hash_free+0x160/0x820 net/core/sock_map.c:1149 process_one_work+0x8a9/0x11d0 kernel/workqueue.c:2292 worker_thread+0xa47/0x1200 kernel/workqueue.c:2439 kthread+0x28d/0x320 kernel/kthread.c:376 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:307 other info that might help us debug this: Possible interrupt unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&htab->buckets[i].lock); local_irq_disable(); lock(&base->lock); lock(&htab->buckets[i].lock); lock(&base->lock); *** DEADLOCK *** 4 locks held by kworker/0:2/151: #0: ffff888012472138 ((wq_completion)rcu_gp){+.+.}-{0:0}, at: process_one_work+0x7a9/0x11d0 kernel/workqueue.c:2267 #1: ffffc90002defd20 ((work_completion)(&rew->rew_work)){+.+.}-{0:0}, at: process_one_work+0x7a9/0x11d0 kernel/workqueue.c:2267 #2: ffff8880b9828358 (&base->lock){-.-.}-{2:2}, at: lock_timer_base+0x120/0x260 kernel/time/timer.c:999 #3: ffffffff8d12a940 (rcu_read_lock){....}-{1:2}, at: rcu_lock_acquire include/linux/rcupdate.h:319 [inline] #3: ffffffff8d12a940 (rcu_read_lock){....}-{1:2}, at: rcu_read_lock include/linux/rcupdate.h:760 [inline] #3: ffffffff8d12a940 (rcu_read_lock){....}-{1:2}, at: __bpf_trace_run kernel/trace/bpf_trace.c:2272 [inline] #3: ffffffff8d12a940 (rcu_read_lock){....}-{1:2}, at: bpf_trace_run3+0x146/0x440 kernel/trace/bpf_trace.c:2313 the dependencies between HARDIRQ-irq-safe lock and the holding lock: -> (&base->lock){-.-.}-{2:2} { IN-HARDIRQ-W at: lock_acquire+0x1f8/0x5a0 kernel/locking/lockdep.c:5662 __raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:110 [inline] _raw_spin_lock_irqsave+0xd1/0x120 kernel/locking/spinlock.c:162 lock_timer_base+0x120/0x260 kernel/time/timer.c:999 add_timer_on+0x1eb/0x580 kernel/time/timer.c:1222 handle_irq_event_percpu kernel/irq/handle.c:195 [inline] handle_irq_event+0xa9/0x1e0 kernel/irq/handle.c:210 handle_level_irq+0x3ab/0x6c0 kernel/irq/chip.c:650 generic_handle_irq_desc include/linux/irqdesc.h:158 [inline] handle_irq arch/x86/kernel/irq.c:231 [inline] __common_interrupt+0xd7/0x1f0 arch/x86/kernel/irq.c:250 common_interrupt+0x9f/0xc0 arch/x86/kernel/irq.c:240 asm_common_interrupt+0x22/0x40 arch/x86/include/asm/idtentry.h:644 __raw_spin_unlock_irqrestore include/linux/spinlock_api_smp.h:152 [inline] _raw_spin_unlock_irqrestore+0xd4/0x130 kernel/locking/spinlock.c:194 __setup_irq+0x12fa/0x1d80 kernel/irq/manage.c:1809 request_threaded_irq+0x2a7/0x380 kernel/irq/manage.c:2198 request_irq include/linux/interrupt.h:168 [inline] setup_default_timer_irq+0x1f/0x30 arch/x86/kernel/time.c:70 x86_late_time_init+0x51/0x86 arch/x86/kernel/time.c:94 start_kernel+0x414/0x53f init/main.c:1101 secondary_startup_64_no_verify+0xcf/0xdb IN-SOFTIRQ-W at: lock_acquire+0x1f8/0x5a0 kernel/locking/lockdep.c:5662 __raw_spin_lock_irq include/linux/spinlock_api_smp.h:119 [inline] _raw_spin_lock_irq+0xcf/0x110 kernel/locking/spinlock.c:170 __run_timers+0x111/0x890 kernel/time/timer.c:1772 run_timer_softirq+0x63/0xf0 kernel/time/timer.c:1803 __do_softirq+0x2e9/0xa4c kernel/softirq.c:571 invoke_softirq kernel/softirq.c:445 [inline] __irq_exit_rcu+0x155/0x240 kernel/softirq.c:650 irq_exit_rcu+0x5/0x20 kernel/softirq.c:662 common_interrupt+0xa4/0xc0 arch/x86/kernel/irq.c:240 asm_common_interrupt+0x22/0x40 arch/x86/include/asm/idtentry.h:644 console_emit_next_record+0xc69/0xea0 kernel/printk/printk.c:2751 console_unlock+0x278/0x7c0 kernel/printk/printk.c:2871 vprintk_emit+0x523/0x740 kernel/printk/printk.c:2268 _printk+0xd1/0x111 kernel/printk/printk.c:2293 calibrate_delay+0x158f/0x16a0 init/calibrate.c:308 start_kernel+0x41e/0x53f init/main.c:1103 secondary_startup_64_no_verify+0xcf/0xdb INITIAL USE at: lock_acquire+0x1f8/0x5a0 kernel/locking/lockdep.c:5662 __raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:110 [inline] _raw_spin_lock_irqsave+0xd1/0x120 kernel/locking/spinlock.c:162 lock_timer_base+0x120/0x260 kernel/time/timer.c:999 add_timer_on+0x1eb/0x580 kernel/time/timer.c:1222 handle_irq_event_percpu kernel/irq/handle.c:195 [inline] handle_irq_event+0xa9/0x1e0 kernel/irq/handle.c:210 handle_level_irq+0x3ab/0x6c0 kernel/irq/chip.c:650 generic_handle_irq_desc include/linux/irqdesc.h:158 [inline] handle_irq arch/x86/kernel/irq.c:231 [inline] __common_interrupt+0xd7/0x1f0 arch/x86/kernel/irq.c:250 common_interrupt+0x9f/0xc0 arch/x86/kernel/irq.c:240 asm_common_interrupt+0x22/0x40 arch/x86/include/asm/idtentry.h:644 __raw_spin_unlock_irqrestore include/linux/spinlock_api_smp.h:152 [inline] _raw_spin_unlock_irqrestore+0xd4/0x130 kernel/locking/spinlock.c:194 __setup_irq+0x12fa/0x1d80 kernel/irq/manage.c:1809 request_threaded_irq+0x2a7/0x380 kernel/irq/manage.c:2198 request_irq include/linux/interrupt.h:168 [inline] setup_default_timer_irq+0x1f/0x30 arch/x86/kernel/time.c:70 x86_late_time_init+0x51/0x86 arch/x86/kernel/time.c:94 start_kernel+0x414/0x53f init/main.c:1101 secondary_startup_64_no_verify+0xcf/0xdb } ... key at: [] init_timer_cpu.__key+0x0/0x20 the dependencies between the lock to be acquired and HARDIRQ-irq-unsafe lock: -> (&htab->buckets[i].lock){+...}-{2:2} { HARDIRQ-ON-W at: lock_acquire+0x1f8/0x5a0 kernel/locking/lockdep.c:5662 __raw_spin_lock_bh include/linux/spinlock_api_smp.h:126 [inline] _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:178 sock_hash_free+0x160/0x820 net/core/sock_map.c:1149 process_one_work+0x8a9/0x11d0 kernel/workqueue.c:2292 worker_thread+0xa47/0x1200 kernel/workqueue.c:2439 kthread+0x28d/0x320 kernel/kthread.c:376 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:307 INITIAL USE at: lock_acquire+0x1f8/0x5a0 kernel/locking/lockdep.c:5662 __raw_spin_lock_bh include/linux/spinlock_api_smp.h:126 [inline] _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:178 sock_hash_free+0x160/0x820 net/core/sock_map.c:1149 process_one_work+0x8a9/0x11d0 kernel/workqueue.c:2292 worker_thread+0xa47/0x1200 kernel/workqueue.c:2439 kthread+0x28d/0x320 kernel/kthread.c:376 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:307 } ... key at: [] sock_hash_alloc.__key+0x0/0x20 ... acquired at: lock_acquire+0x1f8/0x5a0 kernel/locking/lockdep.c:5662 __raw_spin_lock_bh include/linux/spinlock_api_smp.h:126 [inline] _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:178 sock_hash_delete_elem+0xac/0x2f0 net/core/sock_map.c:932 bpf_prog_2c29ac5cdc6b1842+0x3a/0x3e bpf_dispatcher_nop_func include/linux/bpf.h:989 [inline] __bpf_prog_run include/linux/filter.h:600 [inline] bpf_prog_run include/linux/filter.h:607 [inline] __bpf_trace_run kernel/trace/bpf_trace.c:2273 [inline] bpf_trace_run3+0x231/0x440 kernel/trace/bpf_trace.c:2313 trace_timer_start include/trace/events/timer.h:53 [inline] enqueue_timer+0x440/0x600 kernel/time/timer.c:609 __mod_timer+0x92b/0xee0 schedule_timeout+0x1b4/0x300 kernel/time/timer.c:1934 synchronize_rcu_expedited_wait_once kernel/rcu/tree_exp.h:570 [inline] synchronize_rcu_expedited_wait kernel/rcu/tree_exp.h:621 [inline] rcu_exp_wait_wake kernel/rcu/tree_exp.h:689 [inline] rcu_exp_sel_wait_wake+0x764/0x1d50 kernel/rcu/tree_exp.h:723 process_one_work+0x8a9/0x11d0 kernel/workqueue.c:2292 worker_thread+0xa47/0x1200 kernel/workqueue.c:2439 kthread+0x28d/0x320 kernel/kthread.c:376 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:307 stack backtrace: CPU: 0 PID: 151 Comm: kworker/0:2 Not tainted 6.1.82-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 02/29/2024 Workqueue: rcu_gp wait_rcu_exp_gp Call Trace: __dump_stack lib/dump_stack.c:88 [inline] dump_stack_lvl+0x1e3/0x2cb lib/dump_stack.c:106 print_bad_irq_dependency kernel/locking/lockdep.c:2604 [inline] check_irq_usage kernel/locking/lockdep.c:2843 [inline] check_prev_add kernel/locking/lockdep.c:3094 [inline] check_prevs_add kernel/locking/lockdep.c:3209 [inline] validate_chain+0x4d16/0x5950 kernel/locking/lockdep.c:3825 __lock_acquire+0x125b/0x1f80 kernel/locking/lockdep.c:5049 lock_acquire+0x1f8/0x5a0 kernel/locking/lockdep.c:5662 __raw_spin_lock_bh include/linux/spinlock_api_smp.h:126 [inline] _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:178 sock_hash_delete_elem+0xac/0x2f0 net/core/sock_map.c:932 bpf_prog_2c29ac5cdc6b1842+0x3a/0x3e bpf_dispatcher_nop_func include/linux/bpf.h:989 [inline] __bpf_prog_run include/linux/filter.h:600 [inline] bpf_prog_run include/linux/filter.h:607 [inline] __bpf_trace_run kernel/trace/bpf_trace.c:2273 [inline] bpf_trace_run3+0x231/0x440 kernel/trace/bpf_trace.c:2313 trace_timer_start include/trace/events/timer.h:53 [inline] enqueue_timer+0x440/0x600 kernel/time/timer.c:609 __mod_timer+0x92b/0xee0 schedule_timeout+0x1b4/0x300 kernel/time/timer.c:1934 synchronize_rcu_expedited_wait_once kernel/rcu/tree_exp.h:570 [inline] synchronize_rcu_expedited_wait kernel/rcu/tree_exp.h:621 [inline] rcu_exp_wait_wake kernel/rcu/tree_exp.h:689 [inline] rcu_exp_sel_wait_wake+0x764/0x1d50 kernel/rcu/tree_exp.h:723 process_one_work+0x8a9/0x11d0 kernel/workqueue.c:2292 worker_thread+0xa47/0x1200 kernel/workqueue.c:2439 kthread+0x28d/0x320 kernel/kthread.c:376 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:307