On Mon, Jan 29, 2001 at 01:16:13AM +0100, Matthias Buelow wrote:
> Alexander Mai <[EMAIL PROTECTED]> writes:
> >If you don't add the explicit initialization, so I would assume that
> >on runtime the program has to dig out what kind of variable it is
> >and then assign the 0 value. Isn't there a trade-off between
>
> I was under the impression that for standards compliance it is sufficient
> that the memory bytes are initialized to zero, and because of that you
> cannot rely on, for example, static pointer variables being initialized
> to 0 (NULL) because as we all know, 0 in pointer context may be something
> different than all bits cleared. (I'd say the same applies to floating
> point variables so you probably can only rely on integer and character
> types to be properly initialized to 0.) I'm not a language lawyer, though,
> and unfortunately don't have the C89/C99 Standard text at hand so if
> somebody here knows The Truth, I'd like to be enlightened on this topic.
Citing from my C9X draft copy:
If an object that has automatic storage duration is not initialized explicitly, its
value is
indeterminate. If an object that has static storage duration is not initialized
explicitly,
then:
- if it has pointer type, it is initialized to a null pointer;
- if it has arithmetic type, it is initialized to (positive or unsigned) zero;
- if it is an aggregate, every member is initialized (recursively) according to
these rules;
- if it is a union, the first named member is initialized (recursively) according
to these
rules.
> I don't expect the C runtime to do any initializing for bss variables,
> since the OS already provides cleaned (i.e. all bytes 0) pages on
> allocation (well, maybe PC compiler runtimes on pathetic legacy systems
> like MS-DOS do this themselves but that doesn't matter here anyways).
>
> mkb
Arguments based on operating system standards/ architecture/ etc.
are not helpful here - since one shouldn't rely on.
The lower the requirements we have the better it is.
--
Alexander Mai
[EMAIL PROTECTED]
