Bit slicing for fixed point types? in systemC

[Dev]

hello,

 

i have jus started working with the systemC implementations for high level synthesis

 

I need to implement the bit slicing paradigm in systemC which i tried with the variable.range() method and it works fine.

 

But when i used with the fixed point types as shown in the code below, it didnt not work.

 

define SC_INCLUDE_FX
#include "systemc.h"

SC_MODULE(slice){


sc_in<sc_fixed<42,12,SC_TRN,SC_SAT> > input
sc_out<sc_fixed<14,14,SC_TRN,SC_SAT> > output

void do_slice(){    //
output=input.read().range(42,28); //ERROR IN THIS LINE
}

SC_CTOR(slice){
    SC_METHOD(do_slice)
            sensitive<<input_re;
             dont_initialize();
}
};

then, i tried of casting the fixed to a float / double value and tried to use it with the range method which also didnt work.

 

So the range() method only works with sc_int and sc_uint?

 

Is there any other way to do bit slicing paradigms for fractional values in systemC?

 

please let me know.

 

Thanks in advance


 

[apfitch]

You can convert the range object to and from sc_int, e.g.

SC_MODULE(Slice){


    sc_in<sc_fixed<42,12> > input;
    sc_out<sc_fixed<14,14> > output;

    void do_slice(){    //
        sc_int<14> temp = sc_int<14>(input.read().range(41,28));
        output.write( sc_fixed<14,14>(temp) ); 
    }

    SC_CTOR(Slice){
        SC_METHOD(do_slice)
        sensitive<<input;
        dont_initialize();
    }
};

The allowed conversions are shown in IEEE 1666-2011.

 

regards

Alan