| 1 |
148 |
jeremybenn |
/* -------------------------------------------------------------- */
|
| 2 |
|
|
/* (C)Copyright 2001,2008, */
|
| 3 |
|
|
/* International Business Machines Corporation, */
|
| 4 |
|
|
/* Sony Computer Entertainment, Incorporated, */
|
| 5 |
|
|
/* Toshiba Corporation, */
|
| 6 |
|
|
/* */
|
| 7 |
|
|
/* All Rights Reserved. */
|
| 8 |
|
|
/* */
|
| 9 |
|
|
/* Redistribution and use in source and binary forms, with or */
|
| 10 |
|
|
/* without modification, are permitted provided that the */
|
| 11 |
|
|
/* following conditions are met: */
|
| 12 |
|
|
/* */
|
| 13 |
|
|
/* - Redistributions of source code must retain the above copyright*/
|
| 14 |
|
|
/* notice, this list of conditions and the following disclaimer. */
|
| 15 |
|
|
/* */
|
| 16 |
|
|
/* - Redistributions in binary form must reproduce the above */
|
| 17 |
|
|
/* copyright notice, this list of conditions and the following */
|
| 18 |
|
|
/* disclaimer in the documentation and/or other materials */
|
| 19 |
|
|
/* provided with the distribution. */
|
| 20 |
|
|
/* */
|
| 21 |
|
|
/* - Neither the name of IBM Corporation nor the names of its */
|
| 22 |
|
|
/* contributors may be used to endorse or promote products */
|
| 23 |
|
|
/* derived from this software without specific prior written */
|
| 24 |
|
|
/* permission. */
|
| 25 |
|
|
/* */
|
| 26 |
|
|
/* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND */
|
| 27 |
|
|
/* CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, */
|
| 28 |
|
|
/* INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF */
|
| 29 |
|
|
/* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE */
|
| 30 |
|
|
/* DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR */
|
| 31 |
|
|
/* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, */
|
| 32 |
|
|
/* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT */
|
| 33 |
|
|
/* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; */
|
| 34 |
|
|
/* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) */
|
| 35 |
|
|
/* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN */
|
| 36 |
|
|
/* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR */
|
| 37 |
|
|
/* OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, */
|
| 38 |
|
|
/* EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */
|
| 39 |
|
|
/* -------------------------------------------------------------- */
|
| 40 |
|
|
/* PROLOG END TAG zYx */
|
| 41 |
|
|
#ifdef __SPU__
|
| 42 |
|
|
#ifndef _FLOORF4_H_
|
| 43 |
|
|
#define _FLOORF4_H_ 1
|
| 44 |
|
|
|
| 45 |
|
|
#include <spu_intrinsics.h>
|
| 46 |
|
|
|
| 47 |
|
|
|
| 48 |
|
|
/*
|
| 49 |
|
|
* FUNCTION
|
| 50 |
|
|
* vector float _floorf4(vector float value)
|
| 51 |
|
|
*
|
| 52 |
|
|
* DESCRIPTION
|
| 53 |
|
|
* The _floorf4 routine rounds a vector of input values "value" downwards
|
| 54 |
|
|
* to their nearest integer returning the result as a vector of floats.
|
| 55 |
|
|
*
|
| 56 |
|
|
* The full range form (default) provides floor computation on
|
| 57 |
|
|
* all IEEE floating point values. The floor of NANs remain NANs.
|
| 58 |
|
|
* The floor of denorms results in zero.
|
| 59 |
|
|
*
|
| 60 |
|
|
*/
|
| 61 |
|
|
static __inline vector float _floorf4(vector float value)
|
| 62 |
|
|
{
|
| 63 |
|
|
|
| 64 |
|
|
/* FULL FLOATING-POINT RANGE
|
| 65 |
|
|
*/
|
| 66 |
|
|
vec_int4 exp, shift;
|
| 67 |
|
|
vec_uint4 mask, frac_mask, addend, insert, pos;
|
| 68 |
|
|
vec_float4 out;
|
| 69 |
|
|
|
| 70 |
|
|
/* This function generates the following component
|
| 71 |
|
|
* based upon the inputs.
|
| 72 |
|
|
*
|
| 73 |
|
|
* mask = bits of the input that need to be replaced.
|
| 74 |
|
|
* insert = value of the bits that need to be replaced
|
| 75 |
|
|
* addend = value to be added to perform function.
|
| 76 |
|
|
*
|
| 77 |
|
|
* These are applied as follows:.
|
| 78 |
|
|
*
|
| 79 |
|
|
* out = ((in & mask) | insert) + addend
|
| 80 |
|
|
*/
|
| 81 |
|
|
pos = spu_cmpgt((vec_int4)value, -1);
|
| 82 |
|
|
exp = spu_and(spu_rlmask((vec_int4)value, -23), 0xFF);
|
| 83 |
|
|
|
| 84 |
|
|
shift = spu_sub(127, exp);
|
| 85 |
|
|
|
| 86 |
|
|
frac_mask = spu_and(spu_rlmask(spu_splats((unsigned int)0x7FFFFF), shift),
|
| 87 |
|
|
spu_cmpgt((vec_int4)shift, -31));
|
| 88 |
|
|
|
| 89 |
|
|
mask = spu_orc(frac_mask, spu_cmpgt(exp, 126));
|
| 90 |
|
|
|
| 91 |
|
|
addend = spu_andc(spu_andc(spu_add(mask, 1), pos), spu_cmpeq(spu_and((vec_uint4)value, mask), 0));
|
| 92 |
|
|
|
| 93 |
|
|
insert = spu_andc(spu_andc(spu_splats((unsigned int)0xBF800000), pos),
|
| 94 |
|
|
spu_cmpgt((vec_uint4)spu_add(exp, -1), 126));
|
| 95 |
|
|
|
| 96 |
|
|
out = (vec_float4)spu_add(spu_sel((vec_uint4)value, insert, mask), addend);
|
| 97 |
|
|
|
| 98 |
|
|
/* Preserve orignal sign bit (for -0 case)
|
| 99 |
|
|
*/
|
| 100 |
|
|
out = spu_sel(out, value, spu_splats((unsigned int)0x80000000));
|
| 101 |
|
|
|
| 102 |
|
|
return (out);
|
| 103 |
|
|
}
|
| 104 |
|
|
#endif /* _FLOORF4_H_ */
|
| 105 |
|
|
#endif /* __SPU__ */
|
