On 25 January 2012 14:22, Dave Tweed <pic@dtweed.com> wrote:

> My memory's a little hazy on this point, but doesn't the C standard state
> that
> compile-time operations are to be done with the largest available size of
> each
> type; i.e., longs for integers and doubles for floating-point?
>

Should be used 'int' type, yes.


> It looks like the compiler in this case assumed that a compile-time
> constant
> calculation was to be done with signed char, which it then sign-extended =
to
> 16 bits for the run-time comparison. At the very least, this violates the
> principle of "least surprise".
>

I think in C18 the default is 'char' -- maybe to save program memory and
runtime --, maybe if integer promotion was used then it would go to the C89
standard (have not checked that). I vaguely remember discussing about such
kind of issues here in this list few years back.

Tamas



>
> -- Dave Tweed
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>



--=20
int main() { char *a,*s,*q; printf(s=3D"int main() { char *a,*s,*q;
printf(s=3D%s%s%s, q=3D%s%s%s%s,s,q,q,a=3D%s%s%s%s,q,q,q,a,a,q); }",
q=3D"\"",s,q,q,a=3D"\\",q,q,q,a,a,q); }
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.