Re: [PATCH 24/36] target/arm: Convert Neon VHADD 3-reg-same insns

[Richard Henderson <richard.henderson@linaro.org>]

On 4/30/20 11:09 AM, Peter Maydell wrote:
> +    for (pass = 0; pass < (a->q ? 4 : 2); pass++) {
> +        tmp = neon_load_reg(a->vn, pass);
> +        tmp2 = neon_load_reg(a->vm, pass);
> +        fn(tmp, tmp, tmp2);
> +        tcg_temp_free_i32(tmp2);
> +        neon_store_reg(a->vd, pass, tmp);
> +    }
> +    return true;
> +}
> +
> +#define DO_3SAME_32(INSN, func)                                         \
> +    static bool trans_##INSN##_S_3s(DisasContext *s, arg_3same *a)      \
> +    {                                                                   \
> +        static NeonGenTwoOpFn * const fns[] = {                         \
> +            gen_helper_neon_##func##_s8,                                \
> +            gen_helper_neon_##func##_s16,                               \
> +            gen_helper_neon_##func##_s32,                               \
> +        };                                                              \
> +        if (a->size > 2) {                                              \
> +            return false;                                               \
> +        }                                                               \
> +        return do_3same_32(s, a, fns[a->size]);                         \
> +    }                                                                   \

Right, I just talked about the .fni4 hook vs the DO_3SAME_64 patch.


r~