Originally posted by: SystemAdmin

Hi,

how can I switch float point precision used by FPU from extended-double to double in C/C++ code (compiled by xlc)? On Linux it is simple, for example:

fpu_control_t fpu;
_FPU_GETCW(fpu);
fpu &= ~_FPU_EXTENDED;
fpu |= _FPU_DOUBLE;
_FPU_SETCW(fpu);

Is there any switch for xlc or syscall?

thanks
#AIX-Forum