cosf.c 2.93 KB
/**************************************************************************
 *									  *
 *		 Copyright (C) 1994, Silicon Graphics, Inc.		  *
 *									  *
 *  These coded instructions, statements, and computer programs  contain  *
 *  unpublished  proprietary  information of Silicon Graphics, Inc., and  *
 *  are protected by Federal copyright law.  They  may  not be disclosed  *
 *  to  third  parties  or copied or duplicated in any form, in whole or  *
 *  in part, without the prior written consent of Silicon Graphics, Inc.  *
 *									  *
 **************************************************************************/

#include "guint.h"

/* ====================================================================
 * ====================================================================
 *
 * Module: fcos.c
 * $Revision: 1.4 $
 * $Date: 2003/07/15 03:12:04 $
 * $Author: blythe $
 * $Source: /root/leakn64/depot/rf/sw/bbplayer/libultra/monegi/gu/cosf.c,v $
 *
 * Revision history:
 *  09-Jun-93 - Original Version
 *
 * Description:	source code for fcos function
 *
 * ====================================================================
 * ====================================================================
 */

#ifdef __GNUC__
float cosf(float x) __attribute__ ((weak, alias ("__cosf")));
#endif
#pragma weak fcos = __cosf
#pragma weak cosf = __cosf
#define	fcos __cosf

/* coefficients for polynomial approximation of cos on +/- pi/2 */

static const du	P[] =
{
{{0x3ff00000,	0x00000000}},
{{0xbfc55554,	0xbc83656d}},
{{0x3f8110ed,	0x3804c2a0}},
{{0xbf29f6ff,	0xeea56814}},
{{0x3ec5dbdf,	0x0e314bfe}},
};

static const du	rpi =
{{0x3fd45f30,	0x6dc9c883}};

static const du	pihi =
{{0x400921fb,	0x50000000}};

static const du	pilo =
{{0x3e6110b4,	0x611a6263}};

static const fu	zero = {0x00000000};


/* ====================================================================
 *
 * FunctionName		fcos
 *
 * Description		computes cosine of arg
 *
 * ====================================================================
 */

float
fcos( float x )
{
float	absx;
double	dx, xsq, poly;
double	dn;
int	n;
double	result;
int	ix, xpt;


	ix = *(int *)&x;
	xpt = (ix >> 22);
	xpt &= 0x1ff;

	/* xpt is exponent(x) + 1 bit of mantissa */


	if ( xpt < 0x136 )
	{
		/* |x| < 2^28 */

		/* use the standard algorithm from Cody and Waite, doing
		   the computations in double precision
		*/

		absx = ABS(x);

		dx = absx;

		dn = dx*rpi.d + 0.5;
		n = ROUND(dn);
		dn = n;

		dn -= 0.5;

		dx = dx - dn*pihi.d;
		dx = dx - dn*pilo.d;	/* dx = x - (n - 0.5)*pi */

		xsq = dx*dx;

		poly = ((P[4].d*xsq + P[3].d)*xsq + P[2].d)*xsq + P[1].d;

		result = dx + (dx*xsq)*poly;

		/* negate result if n is odd */

		if ( (n & 1) == 0 )
			return ( (float)result );

		return ( -(float)result );
	}

	if ( x != x )
	{
		/* x is a NaN; return a quiet NaN */

#ifdef _IP_NAN_SETS_ERRNO

		*__errnoaddr = EDOM;
#endif
		
		return ( __libm_qnan_f );
	}

	/* just give up and return 0.0 */

	return ( zero.f );
}