Re: [PATCH 7/9] x86/hyperv: Initialize cpu and memory for SEV-SNP enlightened guest

[Tianyu Lan]

On 6/8/2023 10:09 PM, Michael Kelley (LINUX) wrote:
> > +static __init void hv_snp_get_smp_config(unsigned int early)
> > +{
> > +	/*
> > +	 * The "early" is only to be true when there is AMD
> > +	 * numa support. Hyper-V AMD SEV-SNP guest may not
> > +	 * have numa support. To make sure smp config is
> > +	 * always initialized, do that when early is false.
> > +	 */
> I didn't really understand this comment.  After doing a little research, let
> me suggest this wording:
>
> 	/*
> 	 * The "early" parameter can be true only if old-style AMD
> 	 * Opteron NUMA detection is enabled, which should never be
> 	 * the case for an SEV-SNP guest.  See CONFIG_AMD_NUMA.
> 	 * For safety, just do nothing if "early" is true.
> 	 */
>
> Let me know if this new wording makes sense based on your understanding.

Yes, it makes sense. Will update. Thanks.