Experimental Physics and
Industrial Control System

"Ernest L. Williams Jr." <[email protected]> · Tue, 31 Oct 2006 22:56:33 -0500

On Tue, 2006-10-31 at 15:41 -0600, Eric Norum wrote:
> On Oct 31, 2006, at 3:37 PM, Kay-Uwe Kasemir wrote:
> 
> > Hi:
> >
> > Not re-defining NULL is good.
> >
> > The specific fix for alh or other code that uses var-arg, however,
> > is to ignore the old man pages that might talk about NULL,
> > and always cast 0 to the specific type:
> >
> >    some_func("widgetWidth", (int) 0,
> >              "callback_arg", (void *) 0,
> >              "num_entries", (long) 0,  ...)
> >
> > For all these, NULL might have worked on 32 bit,
> > but it won't now.
> 
> Yep, but as you pointed out this represents a significant amount of  
> work (and therefore is unlikely to ever happen).
Hmmm, this sounds like a labor intensive task for an unsuspecting summer
student. Just kidding :)  
Most summer students want to do java anyway :)

I was mistaken about StripTool.  The patches that I supplied to Ken
Evans were indeed incorporated in the latest version of StripTool.
Tracking the issues with NULL was a pain but was done.  I do admit that
it probably was not as overwhelming as alh. StripTool is working under
linux_x86_64

We are not interested in limiting ourselves to 32-bit machines.
Currently, we have AMD64 and now recently a few Intel DUOs (i.e.
Woodcrest) machines.  Of course, we can run 32-bit code on these
machines and we could do that.

-- Ernest 

> 

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System