Re: [PATCH 1/7] KVM: TDX: Return -EBUSY when tdh_mem_page_add() encounters TDX_OPERAND_BUSY

[Edgecombe, Rick P]

On Mon, 2025-01-13 at 10:10 +0800, Yan Zhao wrote:
> diff --git a/arch/x86/kvm/vmx/tdx.c b/arch/x86/kvm/vmx/tdx.c
> index d0dc3200fa37..1cf3ef0faff7 100644
> --- a/arch/x86/kvm/vmx/tdx.c
> +++ b/arch/x86/kvm/vmx/tdx.c
> @@ -3024,13 +3024,11 @@ static int tdx_gmem_post_populate(struct kvm *kvm, gfn_t gfn, kvm_pfn_t pfn,
>  	}
>  
>  	ret = 0;
> -	do {
> -		err = tdh_mem_page_add(kvm_tdx->tdr_pa, gpa, pfn_to_hpa(pfn),
> -				       pfn_to_hpa(page_to_pfn(page)),
> -				       &entry, &level_state);
> -	} while (err == TDX_ERROR_SEPT_BUSY);
> +	err = tdh_mem_page_add(kvm_tdx->tdr_pa, gpa, pfn_to_hpa(pfn),
> +			       pfn_to_hpa(page_to_pfn(page)),
> +			       &entry, &level_state);
>  	if (err) {
> -		ret = -EIO;
> +		ret = unlikely(err & TDX_OPERAND_BUSY) ? -EBUSY : -EIO;
>  		goto out;
>  	}

Should we just squash this into "KVM: TDX: Add an ioctl to create initial guest
memory"? I guess we get a little more specific log history on this corner as a
separate patch, but seems strange to add and remove a loop before it even can
get exercised.