Re: [PATCH] sched/debug: avoid executing show_state and causing rcu stall warning

[Ingo Molnar]

* Liu Song <liusong@xxxxxxxxxxxxxxxxx> wrote:

> From: Liu Song <liusong@xxxxxxxxxxxxxxxxx>
> 
> If the number of CPUs is large, "sysrq_sched_debug_show" will execute for
> a long time. Every time I execute "echo t > /proc/sysrq-trigger" on my
> 128-core machine, the rcu stall warning will be triggered. Moreover,
> sysrq_sched_debug_show does not need to be protected by rcu_read_lock,
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> and no rcu stall warning will appear after adjustment.
> 

That doesn't mean it doesn't have to be protected by *any* lock - which 
your patch implements AFAICS.

There's a couple of lines such as:

        for_each_online_cpu(cpu) {

... which need to be protected against CPU hotplug events.

I haven't checked any of the deeper code to see what RCU or other 
protection it may need, but clearly you didn't either ...

Thanks,

	Ingo