Re: [PATCH v10 2/6] x86/sev: Initialize RMPOPT configuration MSRs

Hello Boris,

On 7/20/2026 8:48 PM, Borislav Petkov wrote:
> On Mon, Jul 20, 2026 at 05:38:05PM -0500, Kalra, Ashish wrote:
>> Agreed on dropping the bool — I'll derive it in a local helper.
>>
>> One issue though: rmp_segment_table != NULL isn't segmented-only. setup_contiguous_rmptable() also allocates rmp_segment_table
>> (the contiguous RMP is stored as a single segment in the same table), so it's non-NULL for the contiguous case too.
>>
>> To keep it segmented-only, I'll also need to gate on the segmented-RMP mode, something like: 
>>
>>   static bool rmpopt_capable(void)
>>   {
>>         return cpu_feature_enabled(X86_FEATURE_RMPOPT) &&
>>                cc_platform_has(CC_ATTR_HOST_SEV_SNP) &&
>>                (rmp_cfg & MSR_AMD64_SEG_RMP_ENABLED) &&
>>                rmp_segment_table;
>>   }
>>
>> The CC_ATTR_HOST_SEV_SNP check also handles SNP being disabled at runtime, so snp_clear_rmpopt_capable() and its caller go away as well.
> 
 Ok, you can simply do:
> 
> 	setup_clear_cpu_cap(X86_FEATURE_RMPOPT)
> 
> and clear our internal representation of RMPOPT support and make it a lot
> simpler.
> 

We did exactly that in v4, but had to drop it in v7 (flagged during Sashiko review) because it's unreliable here — we only
find out whether RMPOPT is usable after alternatives are patched — too late for setup_clear_cpu_cap() to take effect.

Segmented RMP (which RMPOPT requires) is only determined in setup_rmptable(), reached via:

  rootfs_initcall(pci_iommu_init)
    -> amd_iommu_init() 
	-> state_next() 
		-> iommu_snp_enable()
      			-> snp_rmptable_init() 
				-> setup_rmptable()

That's a rootfs_initcall, which runs after alternative_instructions(). By then static_cpu_has(X86_FEATURE_RMPOPT) is already
patched, so a setup_clear_cpu_cap() at that point doesn't take effect and cpu_feature_enabled(X86_FEATURE_RMPOPT) keeps returning true.

That's why v7 moved to a runtime check. In v11 (as i mentioned) this will be a small local helper,  it will also handles two things
the cap-clear can't:
  - a contiguous (non-segmented) RMP, where RMPOPT isn't usable even with X86_FEATURE_RMPOPT set.
  - SNP being disabled at runtime (CC_ATTR_HOST_SEV_SNP).

  static bool rmpopt_capable(void)
  {
        return cpu_feature_enabled(X86_FEATURE_RMPOPT) &&
               cc_platform_has(CC_ATTR_HOST_SEV_SNP) &&
               (rmp_cfg & MSR_AMD64_SEG_RMP_ENABLED) &&
               rmp_segment_table;
  }

Thanks,
Ashish