BUG: "spinlock bad magic"

BUG: "spinlock bad magic"

[Chuck Lever]

Hi-

I'm testing linux-next and am seeing this BUG pop soon after the
system boots. Afterwards the system seems otherwise healthy.

BUG: spinlock bad magic on CPU#3, kworker/3:1/41
 lock: zswap+0xd0/0x140, .magic: 00000000, .owner: <none>/-1, .owner_cpu: 0
CPU: 3 PID: 41 Comm: kworker/3:1 Not tainted 6.8.0-rc7-next-20240304 #1
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.2-1.fc38 04/01/2014
Workqueue: cgroup_destroy css_killed_work_fn
Call Trace:
 <TASK>
 dump_stack_lvl+0x8d/0xb0
 do_raw_spin_lock+0x15e/0x1b0
 ? __pfx_do_raw_spin_lock+0x10/0x10
 ? mark_held_locks+0x5d/0x90
 zswap_memcg_offline_cleanup+0x15/0x50
 mem_cgroup_css_offline+0x14a/0x1b0
 ? css_killed_work_fn+0x21/0x2c0
 css_killed_work_fn+0xb6/0x2c0
 process_one_work+0x4ac/0xb30
 ? __pfx_lock_acquire+0x10/0x10
 ? __pfx_process_one_work+0x10/0x10
 ? __list_add_valid_or_report+0x33/0xc0
 worker_thread+0x392/0x680
 ? __kthread_parkme+0xdc/0x100
 ? __pfx_worker_thread+0x10/0x10
 kthread+0x1ad/0x1f0
 ? kthread+0xf0/0x1f0
 ? __pfx_kthread+0x10/0x10
 ret_from_fork+0x30/0x50
 ? __pfx_kthread+0x10/0x10
 ret_from_fork_asm+0x1a/0x30
 </TASK>

-- 
Chuck Lever

Re: BUG: "spinlock bad magic"

[Johannes Weiner]

On Tue, Mar 05, 2024 at 12:49:20PM -0500, Chuck Lever wrote:
> Hi-
> 
> I'm testing linux-next and am seeing this BUG pop soon after the
> system boots. Afterwards the system seems otherwise healthy.
> 
> BUG: spinlock bad magic on CPU#3, kworker/3:1/41
>  lock: zswap+0xd0/0x140, .magic: 00000000, .owner: <none>/-1, .owner_cpu: 0
> CPU: 3 PID: 41 Comm: kworker/3:1 Not tainted 6.8.0-rc7-next-20240304 #1
> Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.2-1.fc38 04/01/2014
> Workqueue: cgroup_destroy css_killed_work_fn
> Call Trace:
>  <TASK>
>  dump_stack_lvl+0x8d/0xb0
>  do_raw_spin_lock+0x15e/0x1b0
>  ? __pfx_do_raw_spin_lock+0x10/0x10
>  ? mark_held_locks+0x5d/0x90
>  zswap_memcg_offline_cleanup+0x15/0x50
>  mem_cgroup_css_offline+0x14a/0x1b0
>  ? css_killed_work_fn+0x21/0x2c0
>  css_killed_work_fn+0xb6/0x2c0
>  process_one_work+0x4ac/0xb30
>  ? __pfx_lock_acquire+0x10/0x10
>  ? __pfx_process_one_work+0x10/0x10
>  ? __list_add_valid_or_report+0x33/0xc0
>  worker_thread+0x392/0x680
>  ? __kthread_parkme+0xdc/0x100
>  ? __pfx_worker_thread+0x10/0x10
>  kthread+0x1ad/0x1f0
>  ? kthread+0xf0/0x1f0
>  ? __pfx_kthread+0x10/0x10
>  ret_from_fork+0x30/0x50
>  ? __pfx_kthread+0x10/0x10
>  ret_from_fork_asm+0x1a/0x30
>  </TASK>

Thanks Chuck, this should be addressed with a fix that Andrew just
picked up:

https://lore.kernel.org/mm-commits/20240305161123.CA4BCC433C7@smtp.kernel.org/T/#u

Re: BUG: "spinlock bad magic"

[Chuck Lever III]

> On Mar 5, 2024, at 1:00 PM, Johannes Weiner <hannes@cmpxchg.org> wrote:
> 
> On Tue, Mar 05, 2024 at 12:49:20PM -0500, Chuck Lever wrote:
>> Hi-
>> 
>> I'm testing linux-next and am seeing this BUG pop soon after the
>> system boots. Afterwards the system seems otherwise healthy.
>> 
>> BUG: spinlock bad magic on CPU#3, kworker/3:1/41
>> lock: zswap+0xd0/0x140, .magic: 00000000, .owner: <none>/-1, .owner_cpu: 0
>> CPU: 3 PID: 41 Comm: kworker/3:1 Not tainted 6.8.0-rc7-next-20240304 #1
>> Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.2-1.fc38 04/01/2014
>> Workqueue: cgroup_destroy css_killed_work_fn
>> Call Trace:
>> <TASK>
>> dump_stack_lvl+0x8d/0xb0
>> do_raw_spin_lock+0x15e/0x1b0
>> ? __pfx_do_raw_spin_lock+0x10/0x10
>> ? mark_held_locks+0x5d/0x90
>> zswap_memcg_offline_cleanup+0x15/0x50
>> mem_cgroup_css_offline+0x14a/0x1b0
>> ? css_killed_work_fn+0x21/0x2c0
>> css_killed_work_fn+0xb6/0x2c0
>> process_one_work+0x4ac/0xb30
>> ? __pfx_lock_acquire+0x10/0x10
>> ? __pfx_process_one_work+0x10/0x10
>> ? __list_add_valid_or_report+0x33/0xc0
>> worker_thread+0x392/0x680
>> ? __kthread_parkme+0xdc/0x100
>> ? __pfx_worker_thread+0x10/0x10
>> kthread+0x1ad/0x1f0
>> ? kthread+0xf0/0x1f0
>> ? __pfx_kthread+0x10/0x10
>> ret_from_fork+0x30/0x50
>> ? __pfx_kthread+0x10/0x10
>> ret_from_fork_asm+0x1a/0x30
>> </TASK>
> 
> Thanks Chuck, this should be addressed with a fix that Andrew just
> picked up:
> 
> https://lore.kernel.org/mm-commits/20240305161123.CA4BCC433C7@smtp.kernel.org/T/#u

Thank you, Johannes, I'll watch for it.


--
Chuck Lever