Re: [sv-bc] RE: email ballot: Due 8am PDT Friday, May 1

From: Steven Sharp <sharp_at_.....>
Date: Thu Apr 30 2009 - 19:16:51 PDT
>[Alsop, Thomas R] While I agree with the proposal and I believe we are heading 
in the right direction, it actually makes visible some holes that are not 
handled.  Specifically what happens when the number of addresses or words in the 
memory file _exceed_ the finish address.  The previous paragraph describes that 
an error is issued when addresses in the file are out of range.  If we go down 
the path of just issuing a warning when file and memory range are not equal, but 
still writing to the memory, won't people just get used to ignoring the warning. 


Yes, and they have had 20 years to get used to ignoring the warning :-)

> I am just wondering if we should make out of range writes an error like the 
previous paragraph. In other words if the file has too many words in it, less 
than the size of the array.  


The reason that an out-of-range address is described as an error is that
it is supposed to terminate the read operation at that point.  It doesn't
stop the simulation.

If the number of words in the file exceed the finish address, then it doesn't
really matter whether you call it a warning or an error.  The reading stops
at the finish address regardless.

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 Thu Apr 30 19:17:32 2009

This archive was generated by hypermail 2.1.8 : Thu Apr 30 2009 - 19:18:14 PDT