unsigned long long

[Richard A. Huebner]

Is the unsigned long long primitive data type supported in ANSI
standard C?

I've tried using it a couple of times in standard C, but to no avail.
I'm using both MS VIsual C++ 6, as well as the gcc compiler that comes
with RedHat linux 9.

If not, what data type will yield the largest unsigned integer for me?

Thanks for your help,

Rich

[Tom St Denis]

"Dan Pop" <Da*****@cern.c h> wrote in message
news:bo******** **@sunnews.cern .ch...

In <z4************ *****@news04.bl oor.is.net.cabl e.rogers.com> "Tom St Denis" <to********@iah u.ca> writes:

"Joona I Palaste" <pa*****@cc.hel sinki.fi> wrote in message
news:bo******* ***@oravannahka .helsinki.fi...

Tom St Denis <to********@iah u.ca> scribbled the following:
> "Joona I Palaste" <pa*****@cc.hel sinki.fi> wrote in message
> news:bo******** **@oravannahka. helsinki.fi...
>> > ANSI is old. I for one welcome our new ISO overlords.
>>
>> I would think that the difference between the ANSI and ISO versions of >> the C standard is entirely bureaucratic and has no effect on the
>> technical content.

> Be that as it may "unsigned long long" is part of ISOC AFAIK.

ISO C99, Tom. Not ISO C90.

Tom assumes we're dealing with the latest not the oldest.

An excellent illustration of the point I've made in another thread.

Tom von Tom has ways of making you talk!

Tom

[Jonathan2s6]

Get your Server or Homepage now

Choose which one you prefer.
Redhat, Debian, FreeBSD, Mandrake, Windows, SUSE
http://www.comserver.net

[Joona I Palaste]

Jonathan2s6 <jo*********@ao l.com> scribbled the following:

Get your Server or Homepage now Choose which one you prefer.
Redhat, Debian, FreeBSD, Mandrake, Windows, SUSE

Get lost.

--
/-- Joona Palaste (pa*****@cc.hel sinki.fi) ------------- Finland --------\
\-- http://www.helsinki.fi/~palaste --------------------- rules! --------/
"A bee could, in effect, gather its junk. Llamas (no poor quadripeds) tune
and vow excitedly zooming."
- JIPsoft

[Joe Wright]

Dan Pop wrote:

In <3F***********@ earthlink.net> Joe Wright <jo********@ear thlink.net> writes:

Try this.

/*
Sizes of various things in bits..
*/
#include <stdio.h>
#include <limits.h>

int main(void) {
unsigned int a, b, c;
unsigned long long x, y, z;

c = -1;
b = 1 << (sizeof(int)*CH AR_BIT-1);
a = ~b;

z = -1;
y = 1LL << (sizeof(long long)*CHAR_BIT-1);
x = ~y;

printf("Size of void = %2lu bits\n", sizeof(void) *
CHAR_BIT);
printf("Size of char = %2lu bits\n", sizeof(char) *
CHAR_BIT);
printf("Size of short = %2lu bits\n", sizeof(short) *
CHAR_BIT);
printf("Size of int = %2lu bits\n", sizeof(int) *
CHAR_BIT);
printf("Size of long = %2lu bits\n", sizeof(long) *
CHAR_BIT);
printf("Size of long long = %2lu bits\n", sizeof(long long) *
CHAR_BIT);
printf("Size of int * = %2lu bits\n", sizeof(int *) *
CHAR_BIT);
printf("Size of char * = %2lu bits\n", sizeof(char *) *
CHAR_BIT);
printf("Size of void * = %2lu bits\n", sizeof(void *) *
CHAR_BIT);
printf("Size of float = %2lu bits\n", sizeof(float) *
CHAR_BIT);
printf("Size of double = %2lu bits\n", sizeof(double) *
CHAR_BIT);
printf("Size of long double = %2lu bits\n", sizeof(long double) *
CHAR_BIT);
printf("Max int = %11d\n", a);
printf("Min int = %11d\n", b);
printf("Max unsigned int = %11u\n", c);
printf("Max long long = %20lld\n", x);
printf("Min long long = %20lld\n", y);
printf("Max unsigned long long = %20llu\n", z);
return 0;
}

Try it yourself, with the compiler in conforming mode, this time!

Come on Dan, lighten up. As you have noted elsethread, we are not likely
to have C99 conformance. Vanilla gcc compiles this fine. With -ansi
-pedantic it complains about 'long long' of course. I've already
forgotten why I felt compelled to post code here.
--
Joe Wright http://www.jw-wright.com
"Everything should be made as simple as possible, but not simpler."
--- Albert Einstein ---

[Peter Nilsson]

Da*****@cern.ch (Dan Pop) wrote in message news:<bo******* ***@sunnews.cer n.ch>...

In <bo**********@s parta.btinterne t.com> Richard Heathfield <do******@addre ss.co.uk.invali d> writes:

<snip>
Apart from Eric's comments, you may find it a good idea to cast the
expression sizeof(char) * CHAR_BIT to unsigned long before passing it to a
variadic function such as printf.

Or, more generally, to the actual type expected by the respective
conversion description.

Unless you're using a conforming C99 implementation, there is no
conversion description that is guarantee to properly handle a size_t
value. Even worse, in the case of sizeof(type) * CHAR_BIT, the type
of the expression can be either size_t or int.

....or unsigned int.

--
Peter

[Dan Pop]

In <63************ **************@ posting.google. com> ai***@acay.com. au (Peter Nilsson) writes:

Da*****@cern.c h (Dan Pop) wrote in message news:<bo******* ***@sunnews.cer n.ch>...

In <bo**********@s parta.btinterne t.com> Richard Heathfield <do******@addre ss.co.uk.invali d> writes:

> <snip>
>Apart from Eric's comments, you may find it a good idea to cast the
>expression sizeof(char) * CHAR_BIT to unsigned long before passing it to a
>variadic function such as printf.

Or, more generally, to the actual type expected by the respective
conversion description.

Unless you're using a conforming C99 implementation, there is no
conversion description that is guarantee to properly handle a size_t
value. Even worse, in the case of sizeof(type) * CHAR_BIT, the type
of the expression can be either size_t or int.

...or unsigned int.

ONLY if size_t is an alias of unsigned int (or a shorter unsigned type
with identical properties).

Dan
--
Dan Pop
DESY Zeuthen, RZ group
Email: Da*****@ifh.de

[Dan Pop]

In <3F***********@ earthlink.net> Joe Wright <jo********@ear thlink.net> writes:

Dan Pop wrote:

In <3F***********@ earthlink.net> Joe Wright <jo********@ear thlink.net> writes:

>Try this.
>
>/*
> Sizes of various things in bits..
>*/
>#include <stdio.h>
>#include <limits.h>
>
>int main(void) {
> unsigned int a, b, c;
> unsigned long long x, y, z;
>
> c = -1;
> b = 1 << (sizeof(int)*CH AR_BIT-1);
> a = ~b;
>
> z = -1;
> y = 1LL << (sizeof(long long)*CHAR_BIT-1);
> x = ~y;
>
> printf("Size of void = %2lu bits\n", sizeof(void) *
>CHAR_BIT);
> printf("Size of char = %2lu bits\n", sizeof(char) *
>CHAR_BIT);
> printf("Size of short = %2lu bits\n", sizeof(short) *
>CHAR_BIT);
> printf("Size of int = %2lu bits\n", sizeof(int) *
>CHAR_BIT);
> printf("Size of long = %2lu bits\n", sizeof(long) *
>CHAR_BIT);
> printf("Size of long long = %2lu bits\n", sizeof(long long) *
>CHAR_BIT);
> printf("Size of int * = %2lu bits\n", sizeof(int *) *
>CHAR_BIT);
> printf("Size of char * = %2lu bits\n", sizeof(char *) *
>CHAR_BIT);
> printf("Size of void * = %2lu bits\n", sizeof(void *) *
>CHAR_BIT);
> printf("Size of float = %2lu bits\n", sizeof(float) *
>CHAR_BIT);
> printf("Size of double = %2lu bits\n", sizeof(double) *
>CHAR_BIT);
> printf("Size of long double = %2lu bits\n", sizeof(long double) *
>CHAR_BIT);
> printf("Max int = %11d\n", a);
> printf("Min int = %11d\n", b);
> printf("Max unsigned int = %11u\n", c);
> printf("Max long long = %20lld\n", x);
> printf("Min long long = %20lld\n", y);
> printf("Max unsigned long long = %20llu\n", z);
> return 0;
>}
Try it yourself, with the compiler in conforming mode, this time!

Come on Dan, lighten up. As you have noted elsethread, we are not likely
to have C99 conformance. Vanilla gcc compiles this fine.

Vanilla gcc doesn't compile C, and this is NOT a gratuitously pedantic
remark.
With -ansi
-pedantic it complains about 'long long' of course. I've already
forgotten why I felt compelled to post code here.

Even with -std=c99 -pedantic your code fails to compile cleanly, because
of sizeof(void) which is nonsense in C: void is an incomplete type.

Dan
--
Dan Pop
DESY Zeuthen, RZ group
Email: Da*****@ifh.de

[Peter Nilsson]

Da*****@cern.ch (Dan Pop) wrote in message news:<bp******* ****@sunnews.ce rn.ch>...

In <63************ **************@ posting.google. com> ai***@acay.com. au (Peter Nilsson) writes:

Da*****@cern.c h (Dan Pop) wrote in message news:<bo******* ***@sunnews.cer n.ch>...

In <bo**********@s parta.btinterne t.com> Richard Heathfield <do******@addre ss.co.uk.invali d> writes:
> <snip>
>Apart from Eric's comments, you may find it a good idea to cast the
>expression sizeof(char) * CHAR_BIT to unsigned long before passing it to a
>variadic function such as printf.

Or, more generally, to the actual type expected by the respective
conversion description.

Unless you're using a conforming C99 implementation, there is no
conversion description that is guarantee to properly handle a size_t
value. Even worse, in the case of sizeof(type) * CHAR_BIT, the type
of the expression can be either size_t or int.

...or unsigned int.

ONLY if size_t is an alias of unsigned int (or a shorter unsigned type
with identical properties).

What properties would need to be identical?

--
Peter

[Joe Wright]

Dan Pop wrote:

In <3F***********@ earthlink.net> Joe Wright <jo********@ear thlink.net> writes:

Dan Pop wrote:

In <3F***********@ earthlink.net> Joe Wright <jo********@ear thlink.net> writes:

>Try this.
> [ code snipped ]
Try it yourself, with the compiler in conforming mode, this time!

Come on Dan, lighten up. As you have noted elsethread, we are not likely
to have C99 conformance. Vanilla gcc compiles this fine.

Vanilla gcc doesn't compile C, and this is NOT a gratuitously pedantic
remark.

With -ansi
-pedantic it complains about 'long long' of course. I've already
forgotten why I felt compelled to post code here.

Even with -std=c99 -pedantic your code fails to compile cleanly, because
of sizeof(void) which is nonsense in C: void is an incomplete type.

You are completely right. I shouldn't ask you to lighten up on me. I
made at least four 'mistakes' in that code and you came down hard on two
of them. Thank you. I need to be more careful in what I write and what I
post.

I promise to be more careful in the future. Please keep trying to catch
me out.

Thanks.
--
Joe Wright http://www.jw-wright.com
"Everything should be made as simple as possible, but not simpler."
--- Albert Einstein ---

[Dan Pop]

In <63************ **************@ posting.google. com> ai***@acay.com. au (Peter Nilsson) writes:

Da*****@cern.c h (Dan Pop) wrote in message news:<bp******* ****@sunnews.ce rn.ch>...

In <63************ **************@ posting.google. com> ai***@acay.com. au (Peter Nilsson) writes:

>Da*****@cern.c h (Dan Pop) wrote in message news:<bo******* ***@sunnews.cer n.ch>...
>> In <bo**********@s parta.btinterne t.com> Richard Heathfield <do******@addre ss.co.uk.invali d> writes:
>> > <snip>
>> >Apart from Eric's comments, you may find it a good idea to cast the
>> >expression sizeof(char) * CHAR_BIT to unsigned long before passing it to a
>> >variadic function such as printf.
>>
>> Or, more generally, to the actual type expected by the respective
>> conversion description.
>>
>> Unless you're using a conforming C99 implementation, there is no
>> conversion description that is guarantee to properly handle a size_t
>> value. Even worse, in the case of sizeof(type) * CHAR_BIT, the type
>> of the expression can be either size_t or int.
>
>...or unsigned int.

ONLY if size_t is an alias of unsigned int (or a shorter unsigned type
with identical properties).

What properties would need to be identical?

The range.

Dan
--
Dan Pop
DESY Zeuthen, RZ group
Email: Da*****@ifh.de