Re: [PATCH v6] x86/mm: Don't disable INVLPG if "incomplete Global INVLPG flushes" is fixed by microcode or the kernel is running in a hypervisor

[Dave Hansen <dave.hansen@intel.com>]

On 4/11/24 08:18, Dave Hansen wrote:
> 
>> I'm not a native English speaker, so could you give some hint about how
>> to write this comment clearly?
> Something like this would be fine:
> 
> 	/* Only bare-metal is affected.  PCIDs in guests are OK. */

One more thing...  Instead if cluttering up the main PCID setup path
that everybody has to read, you can put this in the macro:

#define INTEL_MATCH(_model)		    \
	{				    \
	  .vendor  = X86_VENDOR_INTEL,      \
	  .feature = X86_FEATURE_HYPERVISOR \
          .family  = 6,                     \
	  .model   = _model,                \
	}

Then you can supplement the "INVLPG may not properly flush Global
entries..." comment.

/*
  * INVLPG may not properly flush Global entries. But
  * it only misbehaves when PCIDs are in use on bare
  * metal on specific CPU models. It is completely
  * fixed with updated microcode.
  */