Re: [PATCH RESEND] sched/core: Do not touch SMT siblings that have not joined the core yet

[Peter Zijlstra]

On Thu, Nov 05, 2026 at 10:40:35PM +0900, Seiji Nishikawa wrote:
> On Thu, Nov 05, 2026 at 11:10:18AM +0100, Peter Zijlstra wrote:
> > Of the two, I like your better for being more consistent -- but you're
> > doing the same thing 4 times, which does beg the question if we cannot
> > wrap that nicer.
> >
> > How about so then?
> 
> Thanks, this looks nicer.
> 
> I tested it on v7.4-rc1 with the syzbot C reproducer. The warning did
> not trigger in about 1250 seconds (about 6100 offline/online cycles of
> CPU 1).
> 
> Should I send a v2 with this change, or will you apply it yourself?

I have it; if only 0-day would hurry up with the current batch of
sched/urgent patches :/