RE: [sv-bc] Pre-Proposal to handle X-problems in RTL coding

From: Steven Sharp <sharp_at_.....>
Date: Mon Nov 19 2007 - 14:45:52 PST
I was going to wait until I had a chance to read the pre-proposal
before commenting.  However, perhaps it is best that I voice my
concerns.

I don't think that this idea of conservatively X-ing out all the
possibly modified variables is very practical.  By the time you
make it conservative enough to be fairly sure it is not optimistic,
I think it will be so pessimistic as to be unusable in anything but
simple situations.

I think this could be taking the same path as unique/priority, sounding
good on paper, but running into problems in real life.  Only I think it
is likely to be much worse.

This would be a good time to fall back on the principle of requiring an
implementation to prove out the concept before adding it to the standard.
If any implementors think that this concept is promising, they can work
out the issues with it before we standardize anything.  If none of the
implementors think it is that promising, then it probably isn't.  At any
rate, I don't think we should be trying to throw something this big into
the standard at this late date, without more time to consider problems.

Steven Sharp
sharp@cadence.com


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.
Received on Mon Nov 19 14:46:09 2007

This archive was generated by hypermail 2.1.8 : Mon Nov 19 2007 - 14:46:20 PST