Re: [PATCH v3] target/arm: Implement SVE2 FMMLA

[Richard Henderson]

On 4/22/20 9:55 AM, Stephen Long wrote:
> +    intptr_t opr_sz = simd_oprsz(desc) / (sizeof(TYPE) >> 2);               \
> +                                                                            \
> +    for (s = 0; s < opr_sz; ++s) {                                          \
> +        TYPE *n = vn + s * (sizeof(TYPE) >> 2);                             \
> +        TYPE *m = vm + s * (sizeof(TYPE) >> 2);                             \
> +        TYPE *a = va + s * (sizeof(TYPE) >> 2);                             \
> +        TYPE *d = vd + s * (sizeof(TYPE) >> 2);                             \

Shifting the wrong way.  Need to multiply by 4 not divide.

I've fixed this up, and also expanded the macro to two functions; I think it's
clearer that way in this case.

Applied to my SVE2 branch.  Thanks,


r~