syzbot


possible deadlock in try_charge_memcg

Status: auto-obsoleted due to no activity on 2024/10/25 13:08
Subsystems: cgroups mm
[Documentation on labels]
Reported-by: syzbot+4825929f4608b51ff236@syzkaller.appspotmail.com
First crash: 156d, last: 156d
Discussions (1)
Title Replies (including bot) Last reply
[syzbot] [cgroups?] [mm?] possible deadlock in try_charge_memcg 0 (1) 2024/07/21 13:14

Sample crash report:
======================================================
WARNING: possible circular locking dependency detected
6.10.0-syzkaller-04472-g51835949dda3 #0 Not tainted
------------------------------------------------------
kworker/1:6/23181 is trying to acquire lock:
ffff8880b933dbd8 (&pool->lock){-.-.}-{2:2}, at: __queue_work+0x26b/0x1070 kernel/workqueue.c:2301

but task is already holding lock:
ffff8880b9339680 (stock_lock){-.-.}-{2:2}, at: local_lock_acquire+0x10/0x150 include/linux/local_lock_internal.h:29

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #4 (stock_lock){-.-.}-{2:2}:
       local_lock_acquire+0x23/0x150 include/linux/local_lock_internal.h:29
       consume_stock mm/memcontrol.c:2476 [inline]
       try_charge_memcg+0x15c/0x10b0 mm/memcontrol.c:2905
       obj_cgroup_charge_pages mm/memcontrol.c:3416 [inline]
       __memcg_kmem_charge_page+0xda/0x420 mm/memcontrol.c:3442
       __alloc_pages_noprof+0x25b/0x2460 mm/page_alloc.c:4700
       __alloc_pages_node_noprof include/linux/gfp.h:269 [inline]
       alloc_pages_node_noprof include/linux/gfp.h:296 [inline]
       __kmalloc_large_node+0x7f/0x1a0 mm/slub.c:4068
       __do_kmalloc_node mm/slub.c:4111 [inline]
       __kmalloc_node_noprof.cold+0x5/0x5f mm/slub.c:4130
       kmalloc_node_noprof include/linux/slab.h:681 [inline]
       bpf_map_kmalloc_node+0x98/0x4a0 kernel/bpf/syscall.c:422
       lpm_trie_node_alloc kernel/bpf/lpm_trie.c:299 [inline]
       trie_update_elem+0x1ef/0xe00 kernel/bpf/lpm_trie.c:342
       bpf_map_update_value+0x2c4/0x6c0 kernel/bpf/syscall.c:203
       map_update_elem+0x623/0x910 kernel/bpf/syscall.c:1654
       __sys_bpf+0x90c/0x4a30 kernel/bpf/syscall.c:5698
       __do_sys_bpf kernel/bpf/syscall.c:5817 [inline]
       __se_sys_bpf kernel/bpf/syscall.c:5815 [inline]
       __x64_sys_bpf+0x78/0xc0 kernel/bpf/syscall.c:5815
       do_syscall_x64 arch/x86/entry/common.c:52 [inline]
       do_syscall_64+0xcd/0x250 arch/x86/entry/common.c:83
       entry_SYSCALL_64_after_hwframe+0x77/0x7f

-> #3 (&trie->lock){-.-.}-{2:2}:
       __raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:110 [inline]
       _raw_spin_lock_irqsave+0x3a/0x60 kernel/locking/spinlock.c:162
       trie_delete_elem+0xb0/0x820 kernel/bpf/lpm_trie.c:462
       0xffffffffa00008ee
       bpf_dispatcher_nop_func include/linux/bpf.h:1243 [inline]
       __bpf_prog_run include/linux/filter.h:691 [inline]
       bpf_prog_run include/linux/filter.h:698 [inline]
       __bpf_trace_run kernel/trace/bpf_trace.c:2406 [inline]
       bpf_trace_run4+0x248/0x5a0 kernel/trace/bpf_trace.c:2449
       __bpf_trace_sched_switch+0x13e/0x190 include/trace/events/sched.h:222
       __traceiter_sched_switch+0x6f/0xc0 include/trace/events/sched.h:222
       trace_sched_switch include/trace/events/sched.h:222 [inline]
       __schedule+0x17cf/0x5490 kernel/sched/core.c:6526
       preempt_schedule_common+0x44/0xc0 kernel/sched/core.c:6708
       preempt_schedule_thunk+0x1a/0x30 arch/x86/entry/thunk.S:12
       class_preempt_destructor include/linux/preempt.h:480 [inline]
       class_preempt_destructor include/linux/preempt.h:480 [inline]
       try_to_wake_up+0xc08/0x13e0 kernel/sched/core.c:4022
       wake_up_process kernel/sched/core.c:4299 [inline]
       wake_up_q+0x91/0x140 kernel/sched/core.c:1029
       futex_wake+0x43e/0x4e0 kernel/futex/waitwake.c:199
       do_futex+0x1e5/0x350 kernel/futex/syscalls.c:107
       __do_sys_futex kernel/futex/syscalls.c:179 [inline]
       __se_sys_futex kernel/futex/syscalls.c:160 [inline]
       __x64_sys_futex+0x1e1/0x4c0 kernel/futex/syscalls.c:160
       do_syscall_x64 arch/x86/entry/common.c:52 [inline]
       do_syscall_64+0xcd/0x250 arch/x86/entry/common.c:83
       entry_SYSCALL_64_after_hwframe+0x77/0x7f

-> #2 (&rq->__lock){-.-.}-{2:2}:
       _raw_spin_lock_nested+0x31/0x40 kernel/locking/spinlock.c:378
       raw_spin_rq_lock_nested+0x29/0x130 kernel/sched/core.c:560
       raw_spin_rq_lock kernel/sched/sched.h:1415 [inline]
       rq_lock kernel/sched/sched.h:1714 [inline]
       task_fork_fair+0x73/0x250 kernel/sched/fair.c:12710
       sched_cgroup_fork+0x3d2/0x510 kernel/sched/core.c:4633
       copy_process+0x43a1/0x8de0 kernel/fork.c:2482
       kernel_clone+0xfd/0x980 kernel/fork.c:2780
       user_mode_thread+0xb4/0xf0 kernel/fork.c:2858
       rest_init+0x23/0x2b0 init/main.c:712
       start_kernel+0x3df/0x4c0 init/main.c:1103
       x86_64_start_reservations+0x18/0x30 arch/x86/kernel/head64.c:507
       x86_64_start_kernel+0xb2/0xc0 arch/x86/kernel/head64.c:488
       common_startup_64+0x13e/0x148

-> #1 (&p->pi_lock){-.-.}-{2:2}:
       __raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:110 [inline]
       _raw_spin_lock_irqsave+0x3a/0x60 kernel/locking/spinlock.c:162
       class_raw_spinlock_irqsave_constructor include/linux/spinlock.h:551 [inline]
       try_to_wake_up+0x9a/0x13e0 kernel/sched/core.c:4051
       create_worker+0x333/0x800 kernel/workqueue.c:2827
       workqueue_init+0x4b4/0xb70 kernel/workqueue.c:7837
       kernel_init_freeable+0x32f/0xca0 init/main.c:1562
       kernel_init+0x1c/0x2b0 init/main.c:1467
       ret_from_fork+0x48/0x80 arch/x86/kernel/process.c:147
       ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244

-> #0 (&pool->lock){-.-.}-{2:2}:
       check_prev_add kernel/locking/lockdep.c:3133 [inline]
       check_prevs_add kernel/locking/lockdep.c:3252 [inline]
       validate_chain kernel/locking/lockdep.c:3868 [inline]
       __lock_acquire+0x2478/0x3b30 kernel/locking/lockdep.c:5136
       lock_acquire kernel/locking/lockdep.c:5753 [inline]
       lock_acquire+0x1b1/0x560 kernel/locking/lockdep.c:5718
       __raw_spin_lock include/linux/spinlock_api_smp.h:133 [inline]
       _raw_spin_lock+0x2e/0x40 kernel/locking/spinlock.c:154
       __queue_work+0x26b/0x1070 kernel/workqueue.c:2301
       queue_work_on+0x11a/0x140 kernel/workqueue.c:2392
       percpu_ref_put_many include/linux/percpu-refcount.h:335 [inline]
       percpu_ref_put include/linux/percpu-refcount.h:351 [inline]
       css_put include/linux/cgroup_refcnt.h:79 [inline]
       css_put include/linux/cgroup_refcnt.h:76 [inline]
       drain_stock+0x287/0x2e0 mm/memcontrol.c:2509
       __refill_stock+0x80/0x270 mm/memcontrol.c:2546
       drain_obj_stock+0x293/0x3a0 mm/memcontrol.c:3574
       drain_local_stock+0xb0/0x330 mm/memcontrol.c:2527
       process_one_work+0x9c8/0x1b40 kernel/workqueue.c:3231
       process_scheduled_works kernel/workqueue.c:3312 [inline]
       worker_thread+0x6c8/0xf20 kernel/workqueue.c:3390
       kthread+0x2c4/0x3a0 kernel/kthread.c:389
       ret_from_fork+0x48/0x80 arch/x86/kernel/process.c:147
       ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244

other info that might help us debug this:

Chain exists of:
  &pool->lock --> &trie->lock --> stock_lock

 Possible unsafe locking scenario:

       CPU0                    CPU1
       ----                    ----
  lock(stock_lock);
                               lock(&trie->lock);
                               lock(stock_lock);
  lock(&pool->lock);

 *** DEADLOCK ***

5 locks held by kworker/1:6/23181:
 #0: ffff888015478948 ((wq_completion)events){+.+.}-{0:0}, at: process_one_work+0x1277/0x1b40 kernel/workqueue.c:3206
 #1: ffffc90004a2fd80 ((work_completion)(&({ do { const void *__vpp_verify = (typeof((&memcg_stock) + 0))((void *)0); (void)__vpp_verify; } while (0); ({ unsigned long __ptr; __asm__ ("" : "=r"(__ptr) : "0"((typeof(*((&memcg_stock))) *)((&memcg_stock)))); (typeof((typeof(*((&memcg_stock))) *)((&memcg_stock)))) (__ptr + (((__per_cpu_offset[(cpu)])))); }); })->work)){+.+.}-{0:0}, at: process_one_work+0x921/0x1b40 kernel/workqueue.c:3207
 #2: ffff8880b9339680 (stock_lock){-.-.}-{2:2}, at: local_lock_acquire+0x10/0x150 include/linux/local_lock_internal.h:29
 #3: ffffffff8dbb58a0 (rcu_read_lock){....}-{1:2}, at: rcu_lock_acquire include/linux/rcupdate.h:327 [inline]
 #3: ffffffff8dbb58a0 (rcu_read_lock){....}-{1:2}, at: rcu_read_lock include/linux/rcupdate.h:839 [inline]
 #3: ffffffff8dbb58a0 (rcu_read_lock){....}-{1:2}, at: percpu_ref_put_many include/linux/percpu-refcount.h:330 [inline]
 #3: ffffffff8dbb58a0 (rcu_read_lock){....}-{1:2}, at: percpu_ref_put include/linux/percpu-refcount.h:351 [inline]
 #3: ffffffff8dbb58a0 (rcu_read_lock){....}-{1:2}, at: css_put include/linux/cgroup_refcnt.h:79 [inline]
 #3: ffffffff8dbb58a0 (rcu_read_lock){....}-{1:2}, at: css_put include/linux/cgroup_refcnt.h:76 [inline]
 #3: ffffffff8dbb58a0 (rcu_read_lock){....}-{1:2}, at: drain_stock+0x13c/0x2e0 mm/memcontrol.c:2509
 #4: ffffffff8dbb58a0 (rcu_read_lock){....}-{1:2}, at: rcu_lock_acquire include/linux/rcupdate.h:327 [inline]
 #4: ffffffff8dbb58a0 (rcu_read_lock){....}-{1:2}, at: rcu_read_lock include/linux/rcupdate.h:839 [inline]
 #4: ffffffff8dbb58a0 (rcu_read_lock){....}-{1:2}, at: __queue_work+0xf2/0x1070 kernel/workqueue.c:2261

stack backtrace:
CPU: 1 PID: 23181 Comm: kworker/1:6 Not tainted 6.10.0-syzkaller-04472-g51835949dda3 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 06/07/2024
Workqueue: events drain_local_stock
Call Trace:
 <TASK>
 __dump_stack lib/dump_stack.c:88 [inline]
 dump_stack_lvl+0x116/0x1f0 lib/dump_stack.c:114
 check_noncircular+0x31a/0x400 kernel/locking/lockdep.c:2186
 check_prev_add kernel/locking/lockdep.c:3133 [inline]
 check_prevs_add kernel/locking/lockdep.c:3252 [inline]
 validate_chain kernel/locking/lockdep.c:3868 [inline]
 __lock_acquire+0x2478/0x3b30 kernel/locking/lockdep.c:5136
 lock_acquire kernel/locking/lockdep.c:5753 [inline]
 lock_acquire+0x1b1/0x560 kernel/locking/lockdep.c:5718
 __raw_spin_lock include/linux/spinlock_api_smp.h:133 [inline]
 _raw_spin_lock+0x2e/0x40 kernel/locking/spinlock.c:154
 __queue_work+0x26b/0x1070 kernel/workqueue.c:2301
 queue_work_on+0x11a/0x140 kernel/workqueue.c:2392
 percpu_ref_put_many include/linux/percpu-refcount.h:335 [inline]
 percpu_ref_put include/linux/percpu-refcount.h:351 [inline]
 css_put include/linux/cgroup_refcnt.h:79 [inline]
 css_put include/linux/cgroup_refcnt.h:76 [inline]
 drain_stock+0x287/0x2e0 mm/memcontrol.c:2509
 __refill_stock+0x80/0x270 mm/memcontrol.c:2546
 drain_obj_stock+0x293/0x3a0 mm/memcontrol.c:3574
 drain_local_stock+0xb0/0x330 mm/memcontrol.c:2527
 process_one_work+0x9c8/0x1b40 kernel/workqueue.c:3231
 process_scheduled_works kernel/workqueue.c:3312 [inline]
 worker_thread+0x6c8/0xf20 kernel/workqueue.c:3390
 kthread+0x2c4/0x3a0 kernel/kthread.c:389
 ret_from_fork+0x48/0x80 arch/x86/kernel/process.c:147
 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244
 </TASK>

Crashes (1):
Time Kernel Commit Syzkaller Config Log Report Syz repro C repro VM info Assets (help?) Manager Title
2024/07/17 13:08 upstream 51835949dda3 215bec2d .config console log report info [disk image] [vmlinux] [kernel image] ci-upstream-kasan-badwrites-root possible deadlock in try_charge_memcg
* Struck through repros no longer work on HEAD.