RE: multiple analog blocks

In Verilog 1364-2005, a variable declaration assignment has to be a
constant expression (which allows constant functions, but not
non-constant functions), but is executed as though in an initial
construct.

 

SystemVerilog is a little different. 1800-2005 says,

 

"In Verilog, an initialization value specified as part of the
declaration is executed as if the assignment were

made from an initial block, after simulation has started. In
SystemVerilog, setting the initial value of a

static variable as part of the variable declaration (including static
class members) shall occur before any

initial or always blocks are started.

 

Initial values in SystemVerilog are not constrained to simple constants;
they can include run-time expressions,

including dynamic memory allocation. For example, a static class handle
or a mailbox can be created

and initialized by calling its new method (see 14.3.1), or static
variables can be initialized to random values

by calling the $urandom system task. This requires a special pre-initial
pass at run time."

 

Shalom

 

 

________________________________

From: edaorg@v-ms.com [mailto:edaorg@v-ms.com] 
Sent: Monday, December 25, 2006 12:19 AM
To: verilog-ams@verilog.org
Cc: Bresticker, Shalom
Subject: Re: multiple analog blocks

 

Bresticker, Shalom wrote: 

Are you guys really disagreeing with each other?
 
Shalom
  

I was just looking for some clarification.

The restriction that the functions should be "analog functions" probably
applies to calls in initialization code in an  "initial step" block
within an analog block since that code is involved in solving the
circuit matrix. If the code in question is setting up variables  and
other initial conditions prior to attempting solving the matrix (and
won't be called during the solving process) then there is no requirement
(that I can see) for the function calls being constant. I.e. if the code
is in an initial block rather than an analog block or just in the
variable declaration calls don't have to be constant.

Kev.



 
  

	Dave Miller wrote:
	    

		Kevin Cameron wrote:
		      

				With this we could then simply use
constant functions and then we
				have a ready to go method to perform
complex pre-simulation
				initialisation.
				          

			What's telling the compiler it's a constant
function, and does it
			have to be?
			        

		Any analog user defined function who's arguments are
constant
		(parameters, numbers) is a constant function, as they
can be solved
		pre-simulation with no dependencies on a matrix etc.
		I would love to see more use of constant analog
functions, if only
		      

as
  

		an alternative to users writing function macro's to do
the same
		      

thing.
  

		As with all macro's in any language, debugging them can
sometimes be
		      

a
  

		challenge.
		      

	I don't see the need for the function to be a constant or analog
if it
	is just being used for initialization and is not inside an
analog
	    

block.