-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, Mar 21, 2008 at 03:07:52PM -0700, Bob Blick wrote:
> 
> On Fri, 21 Mar 2008 23:59:36 +0000 (GMT), "sergio masci"
> <smplx@allotrope.net> said:
> 
> > > Also I think volatile declarations work the other way around and will
> > > not protect you from having to use a shadow register for the type of use
> > > you are putting it to.
> > 
> > I would disagree. The way I understand "volatile" to work is that the 
> > value of a variable may change so don't optimise it.
> > 
> > When the change can occure is undefined. Therefore I would argue that 
> > using it as a temporary location (as is being used here) is a violation.
> 
> Interesting. I had thought that protection that way is not guaranteed,
> all that volatile did was prevent the compiler from assuming the
> variable has not changed simply because no nearby code changes it, so
> something like this:
> 
> x = 0;
> while (x);
> 
> wouldn't get optimized out.
> 
> But I want to hear what Walter says.

I think Sergio's point is pretty good; I can't think of an example where
an access to a volitile variable would be required without it's ever
changing value screwing up the result. In the context of embedded
programming, or maybe even any programming, I can't think of a case
where such behavior would be useful as well.

- -- 
peter[:-1]@petertodd.org http://petertodd.org
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFH5cn03bMhDbI9xWQRAtsaAKCXPROqUbBLzwhPhtYQN6AcXz4RsQCgk9dY
6mOVFBK0Dk7vc5LW7YFuwKY=
=rIM0
-----END PGP SIGNATURE-----
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist