How to output float from sc_fixed type class, using sc_trace?

[sc_fox]

How to output "float (double)" from sc_fixed type class using sc_trace?

 

I have an existing model (all in floating point design, using "sc_signal<double>"), and am trying to convert it to its fixed point one (i.e. using sc_signal<sc_fixed< ,,,, > >").

 

sc_signal<double> ch_a;

sc_signal<sc_fixed<16,8, SC_RND, SC_SAT> > ch_a_fix;

 

sc_trace(tf,  ch_a, "ch_a" );  // floating point channel tracing, it traces signal in "double".

sc_trace(tf,  ch_a_fix, "ch_a_fix );  // fixed point channel tracing: it traces signal in "integer numbers".

 

The question is, how can I trace the sc_fixed type class (sc_signal<sc_fixed<...:> >) in floating point numbers?

For waveform viewer-wise, I'm using gtkwave for now.

 

I have found one posting that has the same question that I have now. But it doesn't have any replies there.

 

 

http://www.accellera.org/Discussion_Forums/helpforum/archive/msg/msg?list_name=help_forum&monthdir=200802&msg=msg00035.html

 

 

Many thanks in advance.

[Philipp A Hartmann]

I'm afraid there is no builtin way to do this. You can't just trace the value returned from ch_a_fix.to_double() as this is returned as a temporary double (and you can only trace persistent objects).

You may need to use a helper process doing the translation. Something like (untested):

  double ch_a_fix_tr;
  void update_ch_a_fix() {
    ch_a_fix_tr = ch_a_fix.to_double();
  }
  // in constructor ...
  SC_METHOD(update_ch_a_fix);
    sensitive << ch_a_fix;

and then trace ch_a_fix_tr instead.

Of course, you can wrap all of this in a helper class and implement your own sc_trace_fixed_as_double function based on it.

hth,

Philipp

[sc_fox]

Thank you, Philipp, for your answer. I appreciate it.

 

Just one more thing,

 

Because "ch_a_fix" is "channel" defined as:

sc_signal<sc_fixed<...> > ch_a_fix;

 

"sc_signal" class doesn't have "to_double()" as its member function. So I had compile error with the code in your message.

So I have modified your code slightly as follows:

 

 

double ch_a_fix_tr;
void update_ch_a_fix() {
  ch_a_fix_tr = ch_a_fix.read().to_double();  // just added sc_signal<T>.read(), to get the value.
}
// in constructor ...
SC_METHOD(update_ch_a_fix);
sensitive << ch_a_fix;

 

 

I'm not sure if that is the right way to do, but it looks working. 

 

Philipp, do you have any comments on this?

Thanks, again.

[Philipp A Hartmann]

"sc_signal" class doesn't have "to_double()" as its member function. So I had compile error with the code in your message.

... 

  ch_a_fix_tr = ch_a_fix.read().to_double();  // just added sc_signal<T>.read(), to get the value.

Your solution is correct. That's why my example said "untested". Sorry.

/Philipp