Subj : Re: Software Job Market Myths
To   : comp.programming,comp.software-eng
From : Richard Heathfield
Date : Fri Aug 19 2005 09:02 am

Chris Sonnack wrote:

> Richard Heathfield writes:
> 
>>>> curious as to whether you can find the /other/ robustness
>>>> problem that I was describing here.
>>> 
>>> It might prove impossible to derefernce s1 or s2 without
>>> catching a SEGV.
>> 
>> Curiosity satisfied. Thank you.
> 
> Because they might be NULL?

Yes. There are three states a pointer can be in:

1) null
2) indeterminate
3) points to object or function

Alas, in C you cannot easily and portably test function parameters to 
distinguish between indeterminate (invalid) pointers and legitimate 
pointers to object or function. But you /can/ test for NULL. Whether you do 
or not...

> Isn't this one of those religious issues?

It's more of a design issue, really. I can certainly see that an intelligent 
programmer might choose, on some occasions, /not/ to check for NULL ("we've 
used the best algorithms in the book, we've pared the code to the bone, 
we're 0.00000000137 seconds over budget, we're only ever passing stack 
addresses to this function, and it's called in a tight loop - if we drop 
the NULL check, we not only hit budget but gain a tiny bit of slack, too"), 
and yet might on other occasions wish to retain it.


> I recall going
> several painful rounds with Dan Pop over this one.  He was
> adament that the function should fault--as (IIRC) the usual
> C lib functions do.

Yes, I remember. I think it was a silly argument, to be honest, since the 
"right" answer depends on circumstances.

> I was (and very much am) of the opinion that no low-level
> routine of *mine* will allow a fault, if at all possible
> (some of my stuff helps run manufacturing production lines,
> and a software crash is not viewed kindly).

I would tend to prefer an assertion, on the grounds that a program that lets 
a null pointer get that far is almost certainly screwed. But I can 
certainly see your point.

-- 
Richard Heathfield
"Usenet is a strange place" - dmr 29/7/1999
http://www.cpax.org.uk
mail: rjh at above domain

.