| /* -------------------------------------------------------------- */ |
| /* (C)Copyright 2007,2008, */ |
| /* 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. */ |
| /* */ |
| /* 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 _LOG1PD2_H_ |
| #define _LOG1PD2_H_ 1 |
| |
| #include <spu_intrinsics.h> |
| #include "simdmath.h" |
| |
| #include "logd2.h" |
| #include "divd2.h" |
| |
| |
| |
| #define LOG1PD2_P0 0.0000000000000000000000000e+00 |
| #define LOG1PD2_P1 1.0000000000000000000000000e+00 |
| #define LOG1PD2_P2 2.3771612265431403265836252e+00 |
| #define LOG1PD2_P3 2.0034423569559494104908026e+00 |
| #define LOG1PD2_P4 7.1309327316770110272159400e-01 |
| #define LOG1PD2_P5 9.8219761968547217301228613e-02 |
| #define LOG1PD2_P6 3.4385125174546914139650511e-03 |
| |
| #define LOG1PD2_Q0 1.0000000000000000000000000e+00 |
| #define LOG1PD2_Q1 2.8771612265431403265836252e+00 |
| #define LOG1PD2_Q2 3.1086896368941925317130881e+00 |
| #define LOG1PD2_Q3 1.5583843494335058998956356e+00 |
| #define LOG1PD2_Q4 3.6047236436186669283898709e-01 |
| #define LOG1PD2_Q5 3.2620075387969869884496887e-02 |
| #define LOG1PD2_Q6 6.8047193336239690346356479e-04 |
| |
| |
| /* |
| * FUNCTION |
| * vector double _log1pd2(vector double x) |
| * |
| * DESCRIPTION |
| * The function _log1pd2 computes the natural logarithm of x + 1 |
| * for each of the double word elements of x. |
| * |
| */ |
| |
| static __inline vector double _log1pd2(vector double x) |
| { |
| vector double oned = spu_splats(1.0); |
| vector double rangehi = spu_splats(0.35); |
| vector double rangelo = spu_splats(0.0); |
| vector unsigned long long use_log; |
| vector double pr, qr; |
| vector double eresult; |
| vector double rresult; |
| vector double result; |
| |
| /* Compiler Bug. Replace xbug with x when spu_cmp*() doesn't |
| * modify it's arguments! */ |
| volatile vector double xbug = x; |
| use_log = spu_or(spu_cmpgt(xbug, rangehi), spu_cmpgt(rangelo, xbug)); |
| |
| /* |
| * Calculate directly using log(x+1) |
| */ |
| eresult = _logd2(spu_add(x, oned)); |
| |
| /* |
| * For x in [0.0,0.35], use a rational approximation. |
| */ |
| pr = spu_madd(x, spu_splats(LOG1PD2_P6), spu_splats(LOG1PD2_P5)); |
| qr = spu_madd(x, spu_splats(LOG1PD2_Q6), spu_splats(LOG1PD2_Q5)); |
| pr = spu_madd(pr, x, spu_splats(LOG1PD2_P4)); |
| qr = spu_madd(qr, x, spu_splats(LOG1PD2_Q4)); |
| pr = spu_madd(pr, x, spu_splats(LOG1PD2_P3)); |
| qr = spu_madd(qr, x, spu_splats(LOG1PD2_Q3)); |
| pr = spu_madd(pr, x, spu_splats(LOG1PD2_P2)); |
| qr = spu_madd(qr, x, spu_splats(LOG1PD2_Q2)); |
| pr = spu_madd(pr, x, spu_splats(LOG1PD2_P1)); |
| qr = spu_madd(qr, x, spu_splats(LOG1PD2_Q1)); |
| pr = spu_madd(pr, x, spu_splats(LOG1PD2_P0)); |
| qr = spu_madd(qr, x, spu_splats(LOG1PD2_Q0)); |
| rresult = _divd2(pr, qr); |
| |
| /* |
| * Select either direct calculation or rational approximation. |
| */ |
| result = spu_sel(rresult, eresult, use_log); |
| |
| return result; |
| } |
| |
| #endif /* _LOG1PD2_H_ */ |
| #endif /* __SPU__ */ |