Re: [PATCH 2/4] target/m68k: pass sign directly into make_quotient()

[Richard Henderson]

On 1/2/23 02:10, Mark Cave-Ayland wrote:
> > > I think you need an "abs(floatx80_to_int32())" in both cases as you do in PATCH 
> > > 4
> >
> > Or in fact
> >
> >      sign = extractFloatx80Sign(res);
> >      quot = floatx80_to_int32(floatx80_abs(res->d), status);
> >      make_quotient(env, sign, quot);
>
> Thanks for the suggestion. Just out of curiosity, how does moving the abs to before the 
> integer conversion make a difference here? Is it because floatx80_to_int32() can fail in 
> some circumstances because of the sign of the result?

It's a simple and operation on floats, instead of cmp+cmov on integers.

I think it's a touch clearer as well, having just saved the fp sign, you discard it before 
moving on to the next thing.


r~