Re: [Qemu-devel] [PATCH 2/4] target-tricore: Added MADD.F and MSUB.F instructions

[Richard Henderson]

On 05/29/2016 03:59 PM, address@hidden wrote:
> +    flags = f_get_excp_flags(env);
> +    f_result = float32_muladd(-arg1, arg2, arg3, flags, &env->fp_status);

Bastian already pointed out that flags here is being used wrong,
but I thought I'd reinforce that "-arg1" is not how floating-point
negation works.  You need to use float_muladd_negate_product as
that 4th argument to float32_muladd.


r~