Re: Midterm Solutions -- from Winter 98/99 Term2

Ed Casas (edc@ece.ubc.ca) Sat, 28 Oct 2000 17:50:43 -0700 

Date: Sat, 28 Oct 2000 17:50:43 -0700
From: Ed Casas <edc@ece.ubc.ca>
Subject: Re: Midterm Solutions -- from Winter 98/99 Term2



> In the architecture, signals "count," "next_count," and
> "count_plus_1" are declared as 4-bit unsigned vectors.  YET, in
> the few lines after the "begin," count is directly equated with
> a 9 (integer value).  This code, as it is written on the
> solution set, thus, does not compile.
> 
> Intead of:
> 
> count_plus_1 <= 0 when count = 9 ...
> 
> it should be count_plus_1 <= 0 when count = conv_unsigned(9,4) ...

This solution has a couple of syntax errors, but your analysis is
not quite right.

You need to look at the std_logic_arith package to see whether
the = (equals) operator is defined for an unsigned value on the
left and an integer value on the right.

The summary sheet for the IEEE std_logic packages which you can
get from the course web page at:

        http://casas.ece.ubc.ca/379/1164pkg.pdf

shows that the `=' operator is defined for these argument types.
You can verify this by looking at the source for the
std_logic_arith package which, for a default MaxPlusII
installation, is at:

        c:\maxplus2\vhdl87\ieee\arith.vhd

which has the following entry:

        function "="(L: UNSIGNED; R: INTEGER) return BOOLEAN;

The real problem is that the type of the value being assigned
(the integer constant 0) does match the type of the signal it is
being assigned to.  The corrected solution would be:

   -- next value in count

   count_plus_1 <=
        conv_unsigned(0,4) when count = 9 else
        count + 1 ;

Thanks for pointing this out.
 
-- 
Ed Casas   edc@ece.ubc.ca   http://casas.ece.ubc.ca   +1 604 822-2592