On Tue, Jun 11, 2002 at 11:10:55PM -0500, Craig A. Berry wrote:
> At 6:48 AM +0300 6/12/02, Jarkko Hietaniemi wrote:
> >The current state of the test is as follows:
>
> <snip>
>
> >How does this look with your patch?
>
> With slight mods so we can see which option we're getting:
>
> $ type dbltest.c
> #include <stdio.h>
> #include "config.h"
>
> int main() {
> #ifdef HAS_LONG_DOUBLE
> # if defined(PERL_PRIfldbl) && (LONG_DOUBLESIZE > DOUBLESIZE)
> long double val = 7.0;
> printf("really long: %5.3" PERL_PRIfldbl "\n",val);
> # else
> double val = 7.0;
> printf("not so long: %5.3f\n",val);
> # endif
> #endif
> }
> $ cc dbltest
> $ link dbltest
> $ run dbltest
> really long: 7.000
Excellent.
I wonder what'll happen with 128 and 256 bit processors,
will we have long long doubles...or maybe the number of
longs doubles, long long long long doubles...
--
$jhi++; # http://www.iki.fi/jhi/
# There is this special biologist word we use for 'stable'.
# It is 'dead'. -- Jack Cohen
