RE: [sv-ec] Mantis 2949 - clocking inout

Jonathan,

If it's obvious, then no need for a note. I withdraw my concern.

-- Brad

-----Original Message-----
From: owner-sv-ec@eda.org [mailto:owner-sv-ec@eda.org] On Behalf Of Jonathan Bromley
Sent: Monday, October 11, 2010 1:17 PM
To: brad_pierce@acm.org
Cc: sv-ec@eda.org
Subject: [sv-ec] Mantis 2949 - clocking inout

  hi Brad

At the Champions' meeting you made the following comment on the
current 2949 proposal:

    I infer from this change that writing to an inout clockvar
    has no effect on a subsequent read from it. If that's
    correct, please consider adding a note about it.

At SV-EC today we agreed to ask you to expand on what you are
asking for here, as it wasn't clear to us how such a note
would help.

As the proposal author I'm opposed to adding such a note,
on the grounds that it might oversimplify the situation
and could easily be misleading, whereas the proposal's
description of an inout (as two distinct but homonymous
clockvars attached to the same signal) allows your implied
question to be answered unambiguously: a write to the output
clockvar will update its signal at the appropriate future
time, and that update will in due course be reflected to
the value seen when reading the input clockvar. There is
no way (and there never was a way) for user code to read
back the scheduled, pending or driving value that has
been written to an output or inout clockvar.

On the other hand, anything that helps clarity and
understandability is good, so if you feel strongly that
some clarification is needed then I'd appreciate it if
you could nudge me gently in the right direction.

Thanks

Jonathan Bromley

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