Re: [PATCH] rcu: Fix use __this_cpu_read() warning in preemptible code

[Joel Fernandes <joel@joelfernandes.org>]

On Thu, Oct 13, 2022 at 12:41:48PM +0800, Zqiang wrote:
> BUG: using __this_cpu_read() in preemptible [00000000]
> code: rcu_torture_fqs/398
> caller is __this_cpu_preempt_check+0x13/0x20
> CPU: 3 PID: 398 Comm: rcu_torture_fqs Not tainted 6.0.0-rc1-yoctodev-standard+
> Call Trace:
> <TASK>
> dump_stack_lvl+0x5b/0x86
> dump_stack+0x10/0x16
> check_preemption_disabled+0xe5/0xf0
> __this_cpu_preempt_check+0x13/0x20
> rcu_force_quiescent_state.part.0+0x1c/0x170
> rcu_force_quiescent_state+0x1e/0x30
> rcu_torture_fqs+0xca/0x160
> ? rcu_torture_boost+0x430/0x430
> kthread+0x192/0x1d0
> ? kthread_complete_and_exit+0x30/0x30
> ret_from_fork+0x22/0x30
> </TASK>
> 
> When enable rcutorture.fqs_duration, the rcu_force_quiescent_state() be
> invoked in rcu_torture_fqs task context, invoke __this_cpu_read() in
> preemptible code section will trigger the above calltrace.

Makes sense, so in theory the GP thread might get migrated after sampling
rcu_data.mynode but that's Ok since it wont change the behavior of the code
following it AFAICS.

Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org>

thanks,

 - Joel

> This commit convert __this_cpu_read() to raw_cpu_read() to avoid
> this warning.
> 
> Signed-off-by: Zqiang <qiang1.zhang@intel.com>
> ---
>  kernel/rcu/tree.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c
> index 96d678c9cfb6..282002e62cf3 100644
> --- a/kernel/rcu/tree.c
> +++ b/kernel/rcu/tree.c
> @@ -2436,7 +2436,7 @@ void rcu_force_quiescent_state(void)
>  	struct rcu_node *rnp_old = NULL;
>  
>  	/* Funnel through hierarchy to reduce memory contention. */
> -	rnp = __this_cpu_read(rcu_data.mynode);
> +	rnp = raw_cpu_read(rcu_data.mynode);
>  	for (; rnp != NULL; rnp = rnp->parent) {
>  		ret = (READ_ONCE(rcu_state.gp_flags) & RCU_GP_FLAG_FQS) ||
>  		       !raw_spin_trylock(&rnp->fqslock);
> -- 
> 2.25.1
>