[sv-ec] new proposal for Mantis 1897

See "coverage_computation2.pdf" attached to 
http://eda.org/svdb/view.php?id=1897

As we agreed last SV-EC meeting, this introduces a covergroup 
type_option to select between alternative methods of calculating type 
coverage: as an average of instances, or as the union of instances 
described in the original proposal.  I added a bit of text to explain 
crosses in the "union" case, and I fixed the typo where "coverage" was 
used instead of "get_coverage".

I'm curious to get a reaction to the changed description of 
option.per_instance.  If covergroup type coverage is to be calculated by 
default as an average of instances, then option.per_instance is less 
useful than I previously understood it to be.  I propose to redefine it 
as an enabler for get_inst_coverage() and answer the question what does 
get_inst_coverage() return when option.per_instance==0.

-- Dave Scott
Mentor Graphics


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.