Re: [PATCH v1 0/2] x86/smp: Fix power regression introduced by commit 96040f7273e2

[Peter Zijlstra <peterz@infradead.org>]

On Wed, May 28, 2025 at 04:25:19PM +0200, Rafael J. Wysocki wrote:

> If cpuidle is available and works, it will do the same thing.

Why can't we make it available sooner? But no, cpuidle does not do the
same thing -- it was argued it does the right thing because it has them
tables with C states on and doesn't try and divinate from CPUID.

> > The whole point was that mwait_play_dead did not DTRT because hints are
> > stupid and it could not select the deepest C state in an unambiguous
> > fashion.
> 
> Yes, on some systems.

The 'on some systems' thing is irrelevant. Either it always works, or it
doesn't and we shouldnt be having it.

> > And now you're restoring that -- code you all argued was fundamentally
> > buggered.
> >
> > Yes is 'fixes' things on old platforms, but it is equally broken on the
> > new platforms where you all argued it was broken on. So either way
> > around you're going to need to fix those, and this isn't it.

> The commit reverted by the first patch removed
> mwait_play_dead_cpuid_hint() altogether, so it never runs and the only
> fallback is hlt_play_dead(), but this doesn't work for disabling SMT
> siblings.

It should either be fixed to always work or stay dead.

> > Now, SMT siblings are all AP, by definition. So can't we simply send
> > them INIT instead of doing CLI;HLT, that way they drop into
> > Wait-for-SIPI and the ucode can sort it out?
> 
> No, I don't think so.  I don't think that Wait-for-SIPI is an idle state.
> 
> But we are discussing patch [2/2] here while really the problem is
> that the commit in question is broken, so it needs to be reverted in
> the first place.

No, you all very much argued that mwait_play_dead couldn't be fixed, as
such it must die and stay dead. Sometimes working is worse than never
working.

So no, I very much object to the revert.