Subject: Re: [sv-ac] cyclicity and ill-founded forms
From: John Havlicek (john.havlicek@motorola.com)
Date: Mon Feb 10 2003 - 08:23:26 PST
Hi Adam:
The example I gave has a cyclic syntactic dependency.
s depends syntactically on t and x, while t depends syntactically
on s and y. If you form a directed graph from the relation
"depends syntactically on", that graph will have a cycle for these
definitions.
I think Surrendra meant to say that all _cyclic_ definitions are
illegal.
Best regards,
John Havlicek
> Good morning all;
>
> John wrote:
>
> I have not yet found any place in the LRM that mentions
> acyclicity requirements for defined forms. Without
> acyclicity, simple ill-founded forms are possible, e.g.
>
> seq s = (t;x);
> seq t = (s;y);
>
> Is it intended that all cyclic syntactic dependencies be
> disallowed?
>
> Surrendra wrote:
>
> All acyclic definitions are illegal. LRM should be updated to clearly state this
> restriction.
>
>
> In addition an example should be provided for those not currently using terms
> such as 'acyclicity'...
>
>
> Adam Krolnik
> Verification Mgr.
> LSI Logic Corp.
> Plano TX. 75074
>
>
This archive was generated by hypermail 2b28 : Mon Feb 10 2003 - 08:24:13 PST