Re: [patch 14/19] softirq: Make softirq control and processing RT aware

From: Thomas Gleixner
Date: Fri Nov 20 2020 - 08:27:37 EST


On Fri, Nov 20 2020 at 01:26, Frederic Weisbecker wrote:
> On Fri, Nov 13, 2020 at 03:02:21PM +0100, Thomas Gleixner wrote:
>> +void __local_bh_disable_ip(unsigned long ip, unsigned int cnt)
>> +{
>> + unsigned long flags;
>> + int newcnt;
>> +
>> + WARN_ON_ONCE(in_hardirq());
>> +
>> + /* First entry of a task into a BH disabled section? */
>> + if (!current->softirq_disable_cnt) {
>> + if (preemptible()) {
>> + local_lock(&softirq_ctrl.lock);
>> + rcu_read_lock();
>
> Ah you lock RCU because local_bh_disable() implies it and
> since it doesn't disable preemption anymore, you must do it
> explicitly?
>
> Perhaps local_lock() should itself imply rcu_read_lock() ?

It's really only required for local_bh_disable(). Lemme add a comment.

>> + } else {
>> + DEBUG_LOCKS_WARN_ON(this_cpu_read(softirq_ctrl.cnt));
>> + }
>> + }
>> +
>> + preempt_disable();
>
> Do you really need to disable preemption here? Migration is disabled by local_lock()
> and I can't figure out a scenario where the below can conflict with a
> preempting task.

Indeed it's pointless.

>> + /*
>> + * Track the per CPU softirq disabled state. On RT this is per CPU
>> + * state to allow preemption of bottom half disabled sections.
>> + */
>> + newcnt = this_cpu_add_return(softirq_ctrl.cnt, cnt);
>
> __this_cpu_add_return() ?

Yep.

Thanks,

tglx