Subject: Re: Fw: 1364.1 pragmas
From: John Michael Williams (jwill@AstraGate.net)
Date: Tue Sep 03 2002 - 23:18:36 PDT
Hi Pat.
If the following parses not as a comment but as
an attribute delimiter, why wouldn't the PLI be able to
access it? For example, "///" or "//(*".
Why does the current parser recognize a Pascal
comment "(*" as part of the language?
Because that's the way the parser is programmed!
Maybe it is a good idea to keep the present
attribute format; maybe not. But, because the parser
would have to be modified to be conforming, isn't
a convincing reason to avoid using verilog-VHDL
portable attribute constructs--at least, it isn't
convincing to me.
There must be a good reason . . ..
--
John
jwill@AstraGate.net
John Michael Williams
Pat Bryant wrote:
>
> John:
>
> It seems that if one uses "token-ized" comments, then you
> cannot have access to the attribute values via the PLI.
> The PLI requires that the simulator/parser be able to recognize
> these "token-ized" comments. Therefore, you don't gain anything
> by these "token-ized" comments. We should keep the attribute
> format currently in Verilog, allowing for modifications
> based upon the current discussion concerning the existence of
> keywords such as "synthesis" should be required.
>
> Just my opinions.
>
> Pat
> ---------------------------------------------------------------
> Patrick [Pat] Bryant
> pkbryant2@attbi.com
> file://home.attbi.com/~pkbryant2
> ----- Original Message -----
> From: John Michael Williams <jwill@AstraGate.net>
> To: <vlog-synth@server.eda.org>
> Sent: Tuesday, September 03, 2002 9:48 AM
> Subject: Re: 1364.1 pragmas
>
> > Hi Dave.
> >...
This archive was generated by hypermail 2b28 : Tue Sep 03 2002 - 23:21:34 PDT