gEDA-dev: verilog operator precedence question: logical & andreduction |

[Stephan Boettcher <boettcher-at-NOSPAMphysik_dot_uni-kiel_dot_de>]

Hello Stephen,

Here is another contruct that our compiler barfs on:

   input      Tclk;
   output     out;
   reg [35:0] Vsel;
   reg [35:0] bits;

   assign     out = |(bits & Vsel) & Tclk;

and in this case it may have a point.

In Lee: Verilog Quickstart, 2nd Edition, page 80, the operator
precedence table does not list the reduction operators seperately.

That could be read that the bitwise & has higher precedence than
the reduction |.  Iverilog treats the reduction operator with higher
precedence than the bitwise operator.

OTOH, writing it in the reverse order should be unambiguous:

  assign out = Tclk & |(bits & Vsel);

What do the relevant standards have to say about this?

Cheers
Stephan


module test;

   reg [7:0] a;
   reg [7:0] m;
   reg 	     p;

   wire      o = |(a&m) & p;


   always
     begin
	#1 p <= 1;
	#1 $display("%b %b %b %b", a,m,p,o);
	#1 p <= 0;
	#1 $display("%b %b %b %b", a,m,p,o);
	a <= a+a;
     end
   initial
     begin
	p <= 0;
	m <= 8'b 11010110;
	a <= 8'b 00000001;
	#33 $finish();
     end

endmodule // test


Stephen Williams <steve@icarus.com> writes:

> Stephan Boettcher wrote:
>> Hi Steve,
>> 
>> The following code runs fine on iveriog, but some synthesizers
>> complain about it:
>> 
>>    wire      next;
>>    reg [1:0] worda;
>>    wire      read = next | |worda;
>> 
>> Is there an issue with the relevant standards, with some opportunity
>> for a warning by ivl?
>
> There is no ambiguity. || is one token, and | | are two tokens.
> I see no way that a reasonable parser would be able to confuse
> this with logical or (||). The point of the warning might be to
> catch typos, though.
>
>> Cheers
>> Stephan
>> 
>>>    reg [1:0] 		worda;
>>>    wire 		read = next | (|worda);     // 6-22, JC
>>>
>>> // Compile ERROR:  Reduction OR follows bitwise OR operator, consider using brackets to disambiguate.
>>> //   wire 		read = next | |worda;     // 6-22, JC

-- 
Stephan Böttcher                     FAX: +49-431-880-3968
Extraterrestrische Physik            Tel: +49-431-880-2508
I.f.Exp.u.Angew.Physik               mailto:boettcher@physik.uni-kiel.de
Leibnizstr. 11, 24118 Kiel, Germany


_______________________________________________
geda-dev mailing list
geda-dev@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev