Re: [sv-bc] Query related with signedness of variable.

From: Surya Pratik Saha <spsaha@cal.interrasystems.com>
Date: Tue Mar 22 2011 - 19:45:57 PDT

Hi Greg,
If you are saying "but b2[0] is signed, whereas b3[0] isn't", then I am not sure in which logic you are saying 'var1[0]' is signed. Does the signed keyword of struct not apply as a whole of var1 variable including [1:0] range? I think yes. Then var1 is signed but var1[0] is not. Can you please refer any LRM snippet for your comment.

Regards
Surya

-------- Original Message  --------
Subject: Re: [sv-bc] Query related with signedness of variable.
From: Greg Jaxon <Greg.Jaxon@synopsys.com>
To: Dhiraj Kumar Prasad <dhiraj@cal.interrasystems.com>
Cc: "sv-bc@server.eda.org" <sv-bc@eda.org>
Date: Wednesday, March 23, 2011 12:59:45 AM

On 3/22/2011 8:59 AM, Dhiraj Kumar Prasad wrote:

Hello,

I have a query related with signedness of a variable.

In following example

module test();

typedef struct packed signed {
            bit [1:0] bit1;
}[1:0]S1;
S1 var1;
int i;
initial
begin
    var1[0] = 2'b10;
    i = var1[0];
end
endmodule

Value of i should be 32'b11111111111111111111111111111110 or 
32'b00000000000000000000000000000010.

Value of i should be 32'sb11111111111111111111111111111110
The elements of var1 are packed signed structures.

Different standard simulator are behaving differently.

Should the equivalent of struct be like

typedef bit signed [1:0] b1;
typedef b1[1:0] b2;

or
typedef bit signed [1:0][1:0] b3;

Not sure how you mean "equivalent" here, but b2[0] is signed, whereas b3[0] isn't.

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

--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean. Received on Tue Mar 22 19:46:54 2011