* `f32toi32 $aDst0, $aSrc0`

Convert a *single-precision* floating-point value to a signed integer,
rounding as per `$FP_CTL.RND`.