Bresticker, Shalom wrote: > Greg, > >>> Do you have a case where it would matter to a synthesis tool? >> Yes. Are they rare? Once you adopt a coding style that >> relies on type parameters or types declared in interfaces, >> this issue comes up immediately. > > At least for type parameters, it was explained that it should not be a > problem. ?? >> Is the module-specialization forest something that a >> testbench compiler can reconstruct from all the trees of a >> fully instantiated design? If that is hopeless, then perhaps >> the LRM is serving testbench needs here. >> I'd find that argument hard to believe, but I'm not a >> testbench expert. > > On the contrary, I was saying that while the effect on synthesizable > code might be small, it could make writing testbenches much more > difficult. In other words, no one wants instance-specificity - it distinguishes too many types whose provenance is otherwise identical. > On the other hand, SV verification has caught on much more than SV > design, and I have not heard that this has been one of the bigger > problems. What semantics does it provide? Notice how type parameters are the easiest way to find out... Greg -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean.Received on Fri Sep 7 15:43:38 2007
This archive was generated by hypermail 2.1.8 : Fri Sep 07 2007 - 15:44:17 PDT