Re: Oneshot timers

Hi Paul,
Didn't see any response to your query, but yes I think that passing 0 to a 
timer should mean it is a oneshot timer, hence it should fire once and only 
fire again when period becomes > 0 in the case of using a variable.

I also think that -ve value for a period should result in a simulator warning, 
stating the period will be treated as 0.0.
I think we should avoid ',,' to represent empty arguments. We only allow this 
in a couple of places, namely laplace/zi and instantiations (port lists, 
parameter lists).

Cheers...
Dave

Paul Floyd wrote:
> Hi
> 
> In the the 2.2 LRM, section 6.7.5.3, nothing is explicitly stated for a 
> non-positive period. My interpretation is that this is the same as for a 
> timer with just one argument, namely a oneshot timer. In particular, a 
> non-positive period is necessary for a oneshot that uses a non-default 
> time_tol. E.g.,
> 
> @(timer(0.2)) <statements>
> 
> would be a timer that fires once at time 0.2 with a default time_tol and
> 
> @(timer(0.2, 0.0, 1e-6)) <statements>
> 
> would be a timer that fires once at time 0.2 with a time_tol of 1e-6.
> 
> Is it worth making this explicit in the LRM? And are both zero and 
> negative values for the period equally valid for a oneshot?
> 
> Lastly, would syntax like
> 
> @(timer(0.2,, 1e-6)) <statements>
> 
> be acceptable and mean the same thing?
> 
> I'd say no since the syntax is given as
> 
> timer ( start_time [ , period [ , time_tol ] ] ) ;
> 
> meaning comma and argument or nothing, not just the comma.
> 
> Regards
> Paul Floyd

-- 
=====================================
-- David Miller
-- Design Technology (Austin)
-- Freescale Semiconductor
-- Ph : 512 996-7377 Fax: x7755
=====================================

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