by Is the declaration of logb in /usr/include/math.h as an int a known
bug, or an obscure feature? It does not match the manual page, and
more importantly, produces incorrect results.
#include <stdio.h>
#include <math.h>
main() {
double f=9.3;
double a;
a = (double)logb(f);
printf("logb(%f)=%f\n",f,a);
exit(0);
logb(9.300000)=1074266112.000000Quote:}
Replacing the #include <math.h> with double logb produces
correct results.
Mike Urban