blob: d8e01f96cd7ee8121b89497ed878f668f288b20d [file] [log] [blame]
/* ===-- fixsfdi.c - Implement __fixsfdi -----------------------------------===
*
* The LLVM Compiler Infrastructure
*
* This file is distributed under the University of Illinois Open Source
* License. See LICENSE.TXT for details.
*
* ===----------------------------------------------------------------------===
*
* This file implements __fixsfdi for the compiler_rt library.
*
* ===----------------------------------------------------------------------===
*/
#if !defined(__GNUC__) || __GNUC__ < 4 || (__GNUC__ == 4 && __GNUC_MINOR__ < 7) || defined(ARCH_X86)
// ARM gcc >= 4.7 implements this in libgcc
#include "int_lib.h"
/* Returns: convert a to a signed long long, rounding toward zero. */
/* Assumption: float is a IEEE 32 bit floating point type
* su_int is a 32 bit integral type
* value in float is representable in di_int (no range checking performed)
*/
/* seee eeee emmm mmmm mmmm mmmm mmmm mmmm */
di_int
__fixsfdi(float a)
{
float_bits fb;
fb.f = a;
int e = ((fb.u & 0x7F800000) >> 23) - 127;
if (e < 0)
return 0;
di_int s = (si_int)(fb.u & 0x80000000) >> 31;
di_int r = (fb.u & 0x007FFFFF) | 0x00800000;
if (e > 23)
r <<= (e - 23);
else
r >>= (23 - e);
return (r ^ s) - s;
}
#endif