https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68211
--- Comment #8 from Marc Glisse <glisse at gcc dot gnu.org> ---
(In reply to Steven Bosscher from comment #7)
> __m128d y = { x, 0 };
> return _mm_cvtsd_f64(_mm_sqrt_round_sd(y, y,
> _MM_FROUND_TO_POS_INF|_MM_FROUND_NO_EXC));
I don't necessarily advocate for optimizing out an existing explicit mov. Maybe
I should, but there could be cases where mov makes the code faster, and I
haven't experimented enough. I am only asking for a way to explicitly skip it
if I believe I know what I am doing. Of course, if we start optimizing out the
mov, it makes my request useless.
