Phone Conference P1450.1 Working Doc Subgroup
Thurs. June 6, 2002 10:00 am PDT
Attendees:
Peter Wohl
Tony Taylor
Dave Kellerman
Jim Teisher
Documents
p1450.1 D13 draft March 21.
D13 review-resolution document June 5.
Agenda
Continue following the current process to review issues identified in the
resolution doc that was updated by Tony.
Discussion
GM-7 There needs to be a
clarification of Scan signal, the Shift block, and the #. It seems that people
have assumed that # means it is a scan signal, and this is where the characters
get shifted in. In reality, the signal must be defined as a scan signal, but
doesn't necessarily have a #. A wfc-list (or a SignalVariable in 1450.1) could
replace the # to provide the scan data.
Questions that were raised:
Is it an error to have a # on a non-scan signal?
Can we pass data to a non-scan signal in a Shift block with a #?
Can loop data and scan data be put on the same vector?
Can the loop data and scan data be collapsed into one construct?
More discussion next meeting.
GM-8 The need for clarification of SignalVariables within a Shift was agree to.
Tony to write the text.
GM-6 A long email of 6/3/02 (from Greg) pointed out the need for additional
semantic definition for how to specify scan data for multiple cores on the same
scan chain. Email suggested use of ParallelPatList & LockStep as a
possible way to implement this. There was some discussion, but we decided to all
take a look at it and review in again next meeting.
GR-22 (originally GR-22-1)
[] - add [ ] to table 5 page 64 1450.0
@ - no discussion
, - we had previously decided to go for space separated list. The example of
max(3, 5) needs to change to max(3 5). This raised the issue that min and max
were only intended to be examples of user defined functions not actually part of
the language, but the way they are in the document they appear to be part of the
language. Do we provide a full definition for min & max, or remove them?
. - no discussion
Review next meeting.
GR-23 (originally GR-22-2) Gordon is correct. The precedence needs to be
changed.
GR-24 (originally GR-22-3) Jim to study and report back 6/20/02. Talk to Gordon
and provide specific examples of problems.
GR-25 (originally GR-22-4)
GR-27 (originally GR-22-6) Is there an ambiguity in the use of WFC or
identifier? Tony to investigate.
GR-26 (originally GR-22-5) It was clear exactly what the concern was. Tony will
investigate.
---------------------------------------
It was decided that only action items that were not related to a issue in the
resolution document will be listed in the minutes. (i.e. If an AI is in the
resolution document is doesn't also go into the minutes AI list.)
Next Meeting
In two weeks; June 20, 2002, 10:00 AM PDT.
Meeting was adjourned at 11:00 PDT.
AIs