Fractional arithemtic on DSP56F803

[weberti]

Hi everybody,

I had to implement a controller (for DSP56F803) which was designed with floating point numbers, but it took too much time to execute. So I started to look and I've found the Fractional arithmetic as a possible solution, but something really goes wrong with the basics.

Can somebody help me if I am on the right track?

I am using Metrowerks CodeWarrior 7.0 to compile and I've found the intrinsics_56800E.h, which is referenced at many sources, but it seems it doesn't work with my assembler. It complains about "unknown assembler instruction mnemonic" for example in case of ".optimize_iasm". 

Is this header can be used with my processor at all!? I guess that is the first thing that I've need to find out. Does anybody have an idea?

If I am on the wrong path, then is there anybody who can give some hint which direction to start?

Thanks in advance,

// Tibor

[weberti]

So it seems. It won't work. The .h is specificly for 56800E (as the name states) and the things I am looking for are not supported on 56F803.

[trytohelp]

Hi,

Intrinsic functions ares supported be DSC tool.

Please have a look to the Targeting_56800.pdf manual.

there is a chapter for that:

Inline Assembly Language and Intrinsic Functions

List of Intrinsic Functions: Definitions and Examples

Hope this will help you.

Regards

Pascal Irrle