Re: [PATCH 3/3] mm/mmu_gather: send tlb_remove_table_smp_sync IPI only to CPUs in kernel mode

[David Hildenbrand]

On 05.04.23 13:41, Peter Zijlstra wrote:
> On Wed, Apr 05, 2023 at 01:10:07PM +0200, Frederic Weisbecker wrote:
> > On Wed, Apr 05, 2023 at 12:44:04PM +0200, Frederic Weisbecker wrote:
> > > On Tue, Apr 04, 2023 at 04:42:24PM +0300, Yair Podemsky wrote:
> > > > +	int state = atomic_read(&ct->state);
> > > > +	/* will return true only for cpus in kernel space */
> > > > +	return state & CT_STATE_MASK == CONTEXT_KERNEL;
> > > > +}
> > >
> > > Also note that this doesn't stricly prevent userspace from being interrupted.
> > > You may well observe the CPU in kernel but it may receive the IPI later after
> > > switching to userspace.
> > >
> > > We could arrange for avoiding that with marking ct->state with a pending work bit
> > > to flush upon user entry/exit but that's a bit more overhead so I first need to
> > > know about your expectations here, ie: can you tolerate such an occasional
> > > interruption or not?
> >
> > Bah, actually what can we do to prevent from that racy IPI? Not much I fear...
>
> Yeah, so I don't think that's actually a problem. The premise is that
> *IFF* NOHZ_FULL stays in userspace, then it will never observe the IPI.
>
> If it violates this by doing syscalls or other kernel entries; it gets
> to keep the pieces.

Yair is currently on vacation, so I'm replying on his behalf.

Indeed, RT userspace is supposed to not call into the kernel, that's the 
premise.

--
Thanks,

David / dhildenb