Subject: Re: [sv-ac] alternative proposal to "syn1": remove binary delay
From: John Havlicek (john.havlicek@motorola.com)
Date: Thu Feb 13 2003 - 06:26:39 PST
Surrendra and All:
Indeed, if "[n]s" means "1*[n]#s", where I am writing "#" for
overlapping concatenation (fusion), then
[m][n]s
is equivalent to
1*[m]#1*[n]#s
is equivalent to
[m+n-1]#s
If we want "[n]s" to represent an initial delay "followed by" s
and we also want "[m][n]s" to be equivalent to "[m+n]s", which would
be nice, I think we are pretty much forced to define "[n]s" to
be "1*[n];s".
If I needed to solve this problem now without worrying about
what other people think, I would try to make something like the
following work:
* Use ";" (or some other symbol) for non-overlapping concatenation.
There is no repeat count on this operator.
* Use "#" (or some other symbol) for overlapping concatenation.
There is no repeat count on this operator.
* Use "[n]s" (or some similar collection of symbols) to mean
"1*[n];s". Here n >= 0 is allowed and "[0]s" means "s".
The intuition for "[n]s" is "skip forward n time units from
where you are and start s at that point".
I would also have both an overlapping and a non-overalapping version
of the sequence implication operator.
With these basic constructs, users could write what they want
and it probably wouln't even look that bad.
I can imagine being told that I do not need ";" because I can
use "#[1]" instead and have only one concatenation operator.
I don't like this because I want "#" to be a fusion opertor
that requires a letter to fuse on, while ";" is a non-overlapping
concatenation that can happily join empty words.
Best regards,
John Havlicek
> Hi Cindy,
> I didn't mean that the unary semantics changes, But, the number counting
> becomes challenging. For example,
> sequence s1 = ([n] a;b); // start counting from 1(1 being the start point)
> to n to detect "a"
> sequence s2 = (c ; ([n] a;b)); same as (c ; [n] a;b);
> sequence s3 = (c ;[m] ([n] a ; b)); same as (c ; [m+n-1] a ; b)
> etc.
>
> Surrendra
>
>
>
> **********************************************
> Surrendra A. Dudani
> Synopsys, Inc.
> 377 Simarano Drive
> Suite 300
> Marlboro, MA 01752
>
> Tel: 508-263-8072
> Fax: 508-263-8123
> email: dudani@synopsys.com
> **********************************************
This archive was generated by hypermail 2b28 : Thu Feb 13 2003 - 06:27:21 PST