<[EMAIL PROTECTED]> writes:

> Thanks Stephen.  I went ahead and changed the library:
> USE ieee.std_logic_arith.ALL;
>
> to:
> USE IEEE.Numeric_STD;
>
> in both the original and the test and that seemed to allow everything to be 
> anylized smoothly. 
>
> Now I have  a new problem at the testbench.  For my for loops (toward the 
> bottom) i have:
>
>         FOR j IN 0 TO 255 LOOP
>           Number1 <= conv_std_logic_vector(j, 8);
>           Even1 <= '0';
>           WAIT FOR 10 ns;
>         END LOOP;  -- j
>        FOR j IN 0 TO 255 LOOP
>           Number1 <= conv_std_logic_vector(j, 8);
>           Even1 <= '1';
>           WAIT FOR 10 ns;
>         END LOOP;  -- j
>
> GHDL doesnt seem to like my conv_std_logic_vector and spits out that
> there is "no declaration" for it.  I think this goes back to when
> you said that IEEE.Numeric_STD differs somewhat from
> std_logic_unsigned but provides the same functionality.  Is there an
> equivilant for the syntax 'conv_std_logic_vector'  or do I have to
> use a different library alltogether?

The specification for the IEEE.Numeric_Std library is provided with
ghdl; see Ghdl-0.25/lib/v93/ieee/numeric_std.v93 in the source
distribution. I'm not clear where this ends up in the installation.

Assuming 'number1' is an 'unsigned' type, you want 'to_unsigned'.

> Oh, also, I tried implementing Synopsis.std_logic_unsigned in its
> place but it seems that I get a no declaration error for that as
> well. Im guessing GHDL doesn't come with the synopsis libraries?

Apparently it does; see Tristan's response. 

However, I don't see the source in the ghdl lib directory, and I've never
used the Synopsis library with ghdl.

-- 
-- Stephe


_______________________________________________
Ghdl-discuss mailing list
[email protected]
https://mail.gna.org/listinfo/ghdl-discuss

Reply via email to