diff options
Diffstat (limited to 'newlib/libm/machine/spu/headers/hypotf4.h')
-rw-r--r-- | newlib/libm/machine/spu/headers/hypotf4.h | 139 |
1 files changed, 139 insertions, 0 deletions
diff --git a/newlib/libm/machine/spu/headers/hypotf4.h b/newlib/libm/machine/spu/headers/hypotf4.h new file mode 100644 index 000000000..4c4fb40c1 --- /dev/null +++ b/newlib/libm/machine/spu/headers/hypotf4.h @@ -0,0 +1,139 @@ +/* -------------------------------------------------------------- */ +/* (C)Copyright 2006,2007, */ +/* International Business Machines Corporation */ +/* All Rights Reserved. */ +/* */ +/* Redistribution and use in source and binary forms, with or */ +/* without modification, are permitted provided that the */ +/* following conditions are met: */ +/* */ +/* - Redistributions of source code must retain the above copyright*/ +/* notice, this list of conditions and the following disclaimer. */ +/* */ +/* - Redistributions in binary form must reproduce the above */ +/* copyright notice, this list of conditions and the following */ +/* disclaimer in the documentation and/or other materials */ +/* provided with the distribution. */ +/* */ +/* - Neither the name of IBM Corporation nor the names of its */ +/* contributors may be used to endorse or promote products */ +/* derived from this software without specific prior written */ +/* permission. */ +/* Redistributions of source code must retain the above copyright */ +/* notice, this list of conditions and the following disclaimer. */ +/* */ +/* Redistributions in binary form must reproduce the above */ +/* copyright notice, this list of conditions and the following */ +/* disclaimer in the documentation and/or other materials */ +/* provided with the distribution. */ +/* */ +/* Neither the name of IBM Corporation nor the names of its */ +/* contributors may be used to endorse or promote products */ +/* derived from this software without specific prior written */ +/* permission. */ +/* */ +/* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND */ +/* CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, */ +/* INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF */ +/* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE */ +/* DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR */ +/* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, */ +/* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT */ +/* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; */ +/* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) */ +/* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN */ +/* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR */ +/* OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, */ +/* EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ +/* -------------------------------------------------------------- */ +/* PROLOG END TAG zYx */ +#ifdef __SPU__ +#ifndef _HYPOTF4_H_ +#define _HYPOTF4_H_ 1 + +#include <spu_intrinsics.h> + +#include "sqrtf4.h" + +/* + * FUNCTION + * vector float _hypotf4(vector float x, vector float y) + * + * DESCRIPTION + * The function hypotf4 returns a float vector in which each element is + * the square root of the sum of the squares of the corresponding + * elements of x and y. In other words, each element is sqrt(x^2 + y^2). + * + * The purpose of this function is to avoid overflow during + * intermediate calculations, and therefore it is slower than + * simply calcualting sqrt(x^2 + y^2). + * + * This function is performed by factoring out the larger of the 2 + * input exponents and moving this factor outside of the sqrt calculation. + * This will minimize the possibility of over/underflow when the square + * of the values are calculated. Think of it as normalizing the larger + * input to the range [1,2). + * + * + * Special Cases: + * - hypot(x, +/-0) returns |x| + * - hypot(+/- infinity, y) returns +infinity + * - hypot(+/- infinity, NaN) returns +infinity + * + */ + + +static __inline vector float _hypotf4(vector float x, vector float y) +{ + vector unsigned int emask = spu_splats(0x7F800000u); + vector unsigned int mmask = spu_splats(0x007FFFFFu); + vector signed int bias = spu_splats(0x3F800000); + vector float inf = (vec_float4)spu_splats(0x7F800000); + vector float onef = spu_splats(1.0f); + vector float sbit = spu_splats(-0.0f); + vector float max, max_e, max_m; + vector float min, min_e, min_m; + vector unsigned int xgty; + vector float sum; + vector float result; + + /* Only need absolute values for this function */ + x = spu_andc(x, sbit); + y = spu_andc(y, sbit); + xgty = spu_cmpgt(x,y); + + max = spu_sel(y,x,xgty); + min = spu_sel(x,y,xgty); + + /* Extract exponents and mantissas */ + max_e = (vec_float4)spu_and((vec_uint4)max, emask); + max_m = (vec_float4)spu_and((vec_uint4)max, mmask); + min_e = (vec_float4)spu_and((vec_uint4)min, emask); + min_m = (vec_float4)spu_and((vec_uint4)min, mmask); + + /* Adjust the exponent of the smaller of the 2 input values by + * subtracting max_exp from min_exp. + */ + vec_int4 min_e_int = spu_sub((vec_int4)min_e, (vec_int4)max_e); + min_e = (vec_float4)spu_add(min_e_int, bias); + + /* If the new min exponent is too small, just set it to 0. It + * wouldn't contribute to the final result in either case. + */ + min_e = spu_sel(min_e, sbit, spu_cmpgt(sbit, min_e)); + + /* Combine new exponents with original mantissas */ + max = spu_or(onef, max_m); + min = spu_or(min_e, min_m); + + sum = _sqrtf4(spu_madd(max, max, spu_mul(min, min))); + sum = spu_mul(max_e, sum); + + /* Special case: x = +/- infinity */ + result = spu_sel(sum, inf, spu_cmpeq(x, inf)); + + return result; +} + +#endif /* _HYPOTF4_H_ */ +#endif /* __SPU__ */ |