[sv-ac] Re: Can checker assign value to variable outside its boundary?

However, *nowhere in the LRM does it state that a checker cannot assign a
value to a variable in the checker's boundary. * *Mean to say*
*However, nowhere in the LRM does it state that a checker cannot assign a
value to a variable in the checker's scope.
*
On Sun, Aug 23, 2009 at 11:36 PM, ben cohen <hdlcohen@gmail.com> wrote:

> 17.7.1 states "It shall be illegal to reference a checker variable using
> its hierarchical name in assignments"
> (Note: that means that a module cannot access a checker's variable)  That
> is not my question.
> 17.2 states "Variables used in a checker that are neither formal arguments
> to the checker nor internal variables of the
> checker are resolved according to the scoping rules from the scope in which
> the checker is declared."
> However, *nowhere in the LRM does it state that a checker cannot assign a
> value to a variable in the checker's boundary. *
> *Thus, by omission, it is legal to assign a value to a variable outside
> the checker's boundary.  *
> Maybe it's a misconception on my part, *but all ports of a checker are
> inputs.  The obvious intent in making ports inputs is prevent the assignment
> to objects outside the checker boundary.  *In a way, allowing such
> assignments defeats the requirement that ports must be inputs.
> *Question*: *Can a checker assign a value to a variable in it's scope
> declaration?  *
> Below is an example of my understanding.  However, LRM is unclear on this.
> *module *test;
>   logic *a*, b, c. clk;
>    task t;
>       logic t1;
>       t1<= 1'b1;
>       a <= t1;
>     endtask
>
>   some_other_module *m_other*;  *// has local variable z*
>
>   checker c(logic clk);
>     logic x, y;
>      always (@ (posedge clk)  begin
>        * a <= 1'b0;  // ILLEGAL ??   Can write to outside
> checker boundary? *
> *       t.t1 <= 1'b1;  ILLEGAL  ?? *
>       x <= 1'b1; // OK
>       y <= b;  // OK, can read b
>      *m_other.z <= 1'b1;  // ILLEGAL* ??
>     end
> endchecker : c
> Ben Cohen
>
>

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.