Re: [PATCHv2 3/3] x86/64/mm: Make 5-level paging support unconditional

[Ingo Molnar]

* Borislav Petkov <bp@xxxxxxxxx> wrote:

> On Fri, May 16, 2025 at 05:46:51PM +0200, Ingo Molnar wrote:
> > 
> > * Borislav Petkov <bp@xxxxxxxxx> wrote:
> > 
> > > On Fri, May 16, 2025 at 12:15:33PM +0300, Kirill A. Shutemov wrote:
> > > > @@ -173,10 +173,10 @@ For example, when an old kernel is running on new hardware.
> > > >  The kernel disabled support for it at compile-time
> > > >  --------------------------------------------------
> > > >  
> > > > -For example, if 5-level-paging is not enabled when building (i.e.,
> > > > -CONFIG_X86_5LEVEL is not selected) the flag "la57" will not show up [#f1]_.
> > > > +For example, if Linear Address Masking (LAM) is not enabled when building (i.e.,
> > > > +CONFIG_ADDRESS_MASKING is not selected) the flag "lam" will not show up.
> > > >  Even though the feature will still be detected via CPUID, the kernel disables
> > > > -it by clearing via setup_clear_cpu_cap(X86_FEATURE_LA57).
> > > > +it by clearing via setup_clear_cpu_cap(X86_FEATURE_LAM).
> > > 
> > > LOL, good one.
> > > 
> > > The rest looks nice and good to me. And FWIW, it boots fine on my Zen5 with
> > > 5lvl enabled.
> > > 
> > > Reviewed-by: Borislav Petkov (AMD) <bp@xxxxxxxxx>
> 
> Bah, and I thought I'm replying to v3. :-\
> 
> Anyway...
> 
> > What's your preference on timing? v6.17 or v6.16?
> 
> Right, here's what I'm thinking:
> 
> * Kirill's patches would simplify Ard's cleanup a bit

Yeah.

> * The 4th one: Kirill A. Shutemov ( :  85|) ├─>[PATCHv3 4/4] x86/paravirt: Restrict PARAVIRT_XXL to 64-bit only
> 
> looks ok too.

Yeah, and now has an Ack from Jürgen too.

> So, I don't see anything speaking against queueing them *now* for the 
> upcoming merge window, I am testing the tip lineup on a daily basis 
> this and next week and if it all looks good, we could probably send 
> them.

Cool!

> If not, we delay.
> 
> And if there's other issues which get detected later, during the 
> 6.16-rc phase, we revert.
> 
> So we have an exit route from each scenario.
> 
> So I guess let's...
> 
> Unless I'm missing an aspect.

I think that's a good plan. I've queued up Kirill's latest in 
tip:x86/core, with tags updated, and it boots fine on my
testsystems as well. Knock on wood. :)

Thanks,

	Ingo