Re: [Xen-devel] [PATCH v2 10/14] x86/hvm: Replace architecture TSC scaling by a common function

[Boris Ostrovsky <boris.ostrovsky@xxxxxxxxxx>]

On 12/06/2015 03:58 PM, Haozhong Zhang wrote:
> This patch implements a common function hvm_scale_tsc() to scale TSC by
> using TSC scaling information collected by architecture code.
>
> Signed-off-by: Haozhong Zhang <haozhong.zhang@xxxxxxxxx>
> ---
>   xen/arch/x86/hvm/hvm.c        | 18 +++++++++--
>   xen/arch/x86/hvm/svm/svm.c    | 12 --------
>   xen/arch/x86/time.c           |  2 +-
>   xen/include/asm-x86/hvm/hvm.h |  3 +-
>   xen/include/asm-x86/math64.h  | 70 +++++++++++++++++++++++++++++++++++++++++++
>   5 files changed, 88 insertions(+), 17 deletions(-)
>
> diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c
> index 52a0ef8..1e3d89f 100644
> --- a/xen/arch/x86/hvm/hvm.c
> +++ b/xen/arch/x86/hvm/hvm.c
> @@ -302,6 +302,20 @@ int hvm_set_guest_pat(struct vcpu *v, u64 guest_pat)
>       return 1;
>   }
>   
> +u64 hvm_scale_tsc(const struct vcpu *v, u64 tsc)
> +{
> +    u64 ratio = v->arch.hvm_vcpu.tsc_scaling_ratio;
> +
> +    if ( !hvm_funcs.tsc_scaling_supported ||
> +         ratio == hvm_funcs.default_tsc_scaling_ratio )
> +        return tsc;
> +
> +    BUG_ON(hvm_funcs.tsc_scaling_ratio_frac_bits >= 64 ||
> +           hvm_funcs.tsc_scaling_ratio_frac_bits == 0);
Since these values never change, I am not sure it's worth testing this. 
And if you think it is then I'd do it somewhere in initialization code.
> +
> +    return mul_u64_u64_shr(tsc, ratio, hvm_funcs.tsc_scaling_ratio_frac_bits);
> +}
> +


> diff --git a/xen/include/asm-x86/math64.h b/xen/include/asm-x86/math64.h
> index 9af6aee..c6a472b 100644
> --- a/xen/include/asm-x86/math64.h
> +++ b/xen/include/asm-x86/math64.h
> @@ -27,4 +27,74 @@ static inline u64 mul_u64_u32_div(u64 a, u32 mul, u32 
> divisor)
>       return rl.ll;
>   }
>   
> +/*
> + * Multiply two 64-bit unsigned integers a and b. The most and least
> + * significant 64 bits of the 128-bit result are returned through hi
> + * and lo respectively.
> + */
> +static inline void mul64(u64 *lo, u64 *hi, u64 a, u64 b)
Please move these routines (as well as one in previous patch into a 
standalone math patch (and provide attribution to wherever they came from).
-boris


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel