From: Shalom.Bresticker@motorola.com
Date: Wed Aug 14 2002 - 12:53:39 PDT
Precedence: bulk
I agree.
The LRM says something like that if any of the RHS operands changes, then the
RHS is re-evaluated and assigned to the LHS.
But that is an implementation description.
A functional description would be that if the value of the RHS changes, then its
new value is assigned to the LHS.
That allows partial execution, etc.
Of course, this explanation could get complicated if the RHS is a function whose
value is not dependent only on the values of its arguments.
Shalom
On Mon, 12 Aug 2002, Steven Sharp wrote:
> A test with Verilog-XL indicates that if index changes but mem[index]
> doesn't, then it does still evaluate the expression (at least if there
> is a function call with a side effect in the expression). Were you
> testing with a different simulator, such as VCS?
>
> I think this is one of those things that is implementation-dependent.
This archive was generated by hypermail 2.1.4
: Thu Oct 10 2002 - 09:24:27 PDT
and
sponsored by Boyd Technology, Inc.