I've built gcc-4.4.6 on AIX 6.1 and am having a problem building
R-2.13.1 with it. Consider the sample code:
% cat ld.c
#include <math.h>
int
main (void) {
long double x = 0.0, y;
y = nearbyintl (x);
}
% xlc ld.c -lm
% nm -BCpg a.out | grep nearby
268436640 T .nearbyint
% /opt/TWWfsw/gcc44/bin/gcc ld.c -lm
ld: 0711-317 ERROR: Undefined symbol: .__nearbyintl128
ld: 0711-345 Use the -bloadmap or -bnoquiet option to obtain more information.
collect2: ld returned 8 exit status
The problem is that gcc's include-fixed/math.h has:
#define nearbyintl(__x) __nearbyintl128((long double) (__x))
but this is only used if _ISOC99_SOURCE and __LONGDOUBLE128 are
defined. <complex.h> has:
/*
* There are two forms of long double on AIX. The default
* form of long double is the same as a double - 64 bits. There
* is a 128-bit form available with some compilers. If that compiler
* defines __LONGDOUBLE128, then long doubles are 128-bit instead of
* 64-bit. Since the same library routine cannot be used for 128-bit
* and 64-bit values, the 128-bit routines are renamed and macros are
* used to manage the name spaces. It is not necessarily the case that
* all of the 128-bit versions are available, but the macros are defined
* intentionally since the 64-bit versions can provide incorrect results
* when long double values were expected. If 64-bit versions are required
* in 128-bit mode, then the code needs to invoke the double routines a
* rather than the long double routines.
*/
xlc doesn't define __LONGDOUBLE128 but it seems gcc does because
gcc/config/rs6000/aix61.h has:
/* Default to 128 bit long double. */
#define RS6000_DEFAULT_LONG_DOUBLE_SIZE 128
So, how exactly do you get the sample code above to compile using GCC
on AIX?
--
albert chin (china@xxxxxxxxxxxxxxxxxx)
