Re: [PATCH V2 03/14] x86/set_memory: Add x86_set_memory_enc static call support

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

On 8/4/21 11:44 AM, Tianyu Lan wrote:
> +static int default_set_memory_enc(unsigned long addr, int numpages, bool enc);
> +DEFINE_STATIC_CALL(x86_set_memory_enc, default_set_memory_enc);
> +
>  #define CPA_FLUSHTLB 1
>  #define CPA_ARRAY 2
>  #define CPA_PAGES_ARRAY 4
> @@ -1981,6 +1985,11 @@ int set_memory_global(unsigned long addr, int numpages)
>  }
>  
>  static int __set_memory_enc_dec(unsigned long addr, int numpages, bool enc)
> +{
> +	return static_call(x86_set_memory_enc)(addr, numpages, enc);
> +}
> +
> +static int default_set_memory_enc(unsigned long addr, int numpages, bool enc)
>  {
>  	struct cpa_data cpa;
>  	int ret;

It doesn't make a lot of difference to add this infrastructure and then
ignore it for the existing in-tree user:

> static int __set_memory_enc_dec(unsigned long addr, int numpages, bool enc)
> {
>         struct cpa_data cpa;
>         int ret;
> 
>         /* Nothing to do if memory encryption is not active */
>         if (!mem_encrypt_active())
>                 return 0;

Shouldn't the default be to just "return 0"?  Then on
mem_encrypt_active() systems, do the bulk of what is in
__set_memory_enc_dec() today.