Re: [PATCH] net: raise RCU qs after each threaded NAPI poll

From: Paul E. McKenney
Date: Thu Mar 07 2024 - 13:58:48 EST


On Thu, Mar 07, 2024 at 01:52:15PM -0500, Steven Rostedt wrote:
> On Thu, 7 Mar 2024 16:57:33 +0000
> Mark Rutland <mark.rutland@xxxxxxx> wrote:
>
> > * Use rcu_tasks_trace to synchronize updates?
>
> Yes. I think I wanted both. The above to make sure it covers all cases
> where something could be preempted, and a case for those covered when RCU
> isn't watching (which always has preemption disabled).
>
> My mistake was I thought synchronize_rcu_tasks_rude() did both. But I just
> found out recently that it is not a superset of synchronize_rcu_tasks().
>
> But it really needs it in every location that synchronize_rcu_tasks_rude()
> is called.

Should any RCU Tasks Rude grace-period request also wait for an RCU Tasks
grace period?

I would feel better about proposing this, especially for
call_rcu_tasks_rude(), if RCU Tasks Rude was not supposed to be going
away after all the noinstr tags are in place.

Thanx, Paul