Re: [avr-gcc-list] Re: WinAVR 20050214 (gcc 3.4.3) and optimizer bug.

[Russell Shaw]

Larry Barello wrote:
> So, this is all facinating reading, but what, if anything, is a NULL pointer
> in the AVR context?  Address 0 is certainly valid.  Or is it simply
> convention that 0 is "null".

If a pointer is NULL, the actual behind-the-scenes address it represents may
be anything, and is just something a compiler assigns to a pointer object that
is in a known state of "unsettedness".

I don't know what gcc uses for the avr.

> I didn't understand everything that was said, but it seemed clear to me that
> having the construct
>
> char *foo;
>
> if (foo && (*foo !=0))
>
> randomly break seemed pretty harsh.  Was (is) the solution to disable null
> pointer checks?

As in: if(foo == NULL) or: if(foo == 0) or: if(foo)  ?

Definitely not.

> Do I need to add this to my makefile scripts?  Is this a
> lurking problem prior to the latest WinAVR release?

No.
Where the compiler doesn't know whether a value is going to be used as a
pointer, is a cast required. Ie: execl("/bin/sh", "sh", "-c", "date", (char 
*)0);
this function is variadic, so the type of the parameters is unknown from the
prototype.