Re: [PATCH v11 01/13] x86/resctrl: Replace smp_call_function_many() with on_each_cpu_mask()

[Ashok Raj]

On Wed, Jan 11, 2023 at 11:05:48AM -0800, Tony Luck wrote:
> > I wasn't sure if the few ms difference is going to make much material
> > difference for that process. IPI's does shake things up and introduces
> > other overheads not related to this process.
> 
> Is it just a few milli-seconds? What is the scheduler priority of the kernel
> thread you wake to perform this action? How does that compare to the
> priority of a RT thread?  I may be wrong here, but I think an RT thread can
> block a kernel thread from running indefinitely.

That's a good point. I don't have an idea about how the kernel thread
priority is compared to user space RT threads.

I assumed that since these threads have some work to be done from kernel,
user space shouldn't get higher priority than kernel tasks.

Uncharted territory for me ... Just a hunch :)

Maybe @peterz can screw my head in the right direction :-)