ISAC: Minutes from meeting on 28 February 2008

These are available on the website.
Note that our next meeting is NEXT WEEK!

Chuck Swart


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


MINUTES of ISAC meeting held via telecom on 28 February 2008

Present: Peter Ashenden, Chuck Swart, Lance Thompson, Ajay Verikat

Absent: Jim Lewis, Larry Soule

Next meeting: Thursday, March 6, 8 pm Pacific Standard time
              (Friday, March 7, 4 am GMT)

Open Bugzilla Issues:

TOPIC: Bugzilla 62: Matching operators not needed for BOOLEAN

Wording changes accepted.

TOPIC: Bugzilla 184: Condition operator ?? is not predefined for STD_ULOGIC

Wording changes accepted.

TOPIC: Bugzilla 185: Definitions for unconstrained and fully
constrained record types overlap

A small additional wording change was made, otherwise accepted.

TOPIC: Bugzilla 190: logical operator return array index ranges not consistent

It is unfortunate that different standard packages use different rules
for determining array index ranges for unconstrained array return
values, but past experience with changing direction and bounds of
concatenation operators shows that many problems arise for designers
when these are changed, so backward compatibility takes precedence
over uniformity.  No change in this area.

TOPIC: Bugzilla 191: Condition operator should be implicitly applied
to guard expressions

Some side issues were raised concerning references to shared variables
in guard expressions. It was decided that no additional problems
arise and the proposed LRM change is accepted.

TOPIC: Bugzilla 192: Subtype problems for out-mode signal parameters

This is a difficult issue. There is a conflict between the extension
allowing OUT mode signal parameters to be read and the relaxation of
scalar subtype compatibility rules between the formal and the
actual. There are (at least) three possible resolutions, none
especially acceptable.

1. Treat out mode signal parameters the same as inout parameters, ie
require VHDL-2002 strict matching of subtypes.

2. Apply the strict matching rule only to OUT mode parameters which
are read. This may be difficult to compute, requiring analysis of
static call chains. However, proposed rules on automatic calculation
of signals in sensitivity lists may already require a similar
calculation.

3. Prohibit reading of OUT mode signal parameters. It appears that the
major requirement was to allow reading of OUT mode variables and
ports.

No resolution has yet been reached.  

A related discussion involved questioning why reading OUT mode signal
parameters produced the effective value instead of the driving value.