dsPIC30F4013内联汇编浮点乘法

嗨，我试着使用内嵌程序集XC16编译器来执行浮点乘法。该设备是30F4013。第一个值来自查找表，它是一个浮动值，在2和1之间变化，因此，0.9，0.3，0.4等。它与音频样本乘以一个已签名的int。是否有一种在内联汇编中这样做的方法，因为C是非常慢的。我需要执行每一个采样周期乘以48千赫。本质上，结果＝AudioSoint（int）* LUTvalue（浮点）。我知道我可以把所有的浮点都转换成整数和比例，但这对于应用来说是不理想的。谢谢。



      以下为原文

    Hi,

I'm trying to perform a floating point multiply using inline assembly, XC16 compiler. The device is the 30F4013.

The first value comes from a look up table and it's a float value which varies between 2.0 and 1.0, so, 0.9, 0.3, 0.4 etc.

It's getting multiplied with an audio sample which is a signed int.

Is there a way of doing this in inline assembly as doing it is C is woefully slow. I need to perform the multiply every sample period, running at 48Khz.

In essence, Results = AudioSample(INT) * LUTvalue(Float).

I understand I can covert all the floats to integers and scale but that isn't ideal for the application.

Thanks.