/* rsqrtd2 - for each of two double slots, compute reciprocal square root.
   Copyright (C) 2006, 2007 Sony Computer Entertainment Inc.
   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 the Sony Computer Entertainment Inc 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.
 */

#ifndef ___SIMD_MATH_RSQRTD2_H___
#define ___SIMD_MATH_RSQRTD2_H___

#include <simdmath.h>
#include <spu_intrinsics.h>

#include <simdmath/isinfd2.h>
#include <simdmath/signbitd2.h>
#include <simdmath/is0denormd2.h>

// 
// Handles exceptional values as follows:
// NaN -> NaN
// (+,-)0 -> (+,-)Inf
// +Inf -> +0
// -Inf -> Nan
// -Finite -> Nan
// Denormal inputs are treated as zero.

static inline vector double
_rsqrtd2 (vector double x)
{
  vec_ullong2 expmask, onemask, signmask, evenexp;
  vec_double2 half, one, man, exp, nexp, y1, y2, y3, zero, inf, nan, result;
  vec_float4  halff, onef, manf, y0f, y1f;

  expmask = spu_splats(0x7ff0000000000000ull);
  onemask = spu_splats(0x0010000000000000ull);
  signmask = spu_splats(0x8000000000000000ull);
  onef = spu_splats(1.0f);
  one = spu_extend( onef );
  halff = spu_splats(0.5f);
  half = spu_extend( halff );

  // Factor input ( mantissa x 2^exponent ) into ( mantissa x 2^(-i) ) and ( 2^(exponent+i) )
  // where i = 0 when exponent is even and i = 1 when exponent is odd.
  // 
  // Compute reciprocal-square-root of second factor by finding -(exponent+i)/2:
  // 
  // biased_exp = 1023 + exponent
  // new_biased_exp = 1023 - (exponent+i)/2 
  //                = 1023 - (biased_exp-1023+i)/2
  //                = (3069 - (biased_exp+i)) / 2

  evenexp = spu_and( (vec_ullong2)x, onemask );
  man = spu_sel( x, (vec_double2)spu_add( spu_splats(0x3fe00000u), (vec_uint4)evenexp ), expmask );

  exp = spu_and( x, (vec_double2)expmask );
  nexp = spu_or( exp, (vec_double2)onemask );
  nexp = (vec_double2)spu_rlmask( spu_sub( (vec_uint4)spu_splats(0xbfd0000000000000ull), (vec_uint4)nexp ), -1 );

  // Compute mantissa part in single precision.
  // Convert back to double and multiply with 2^(-(exponent+i)/2), then
  // do two Newton-Raphson steps for full precision.

  manf = spu_roundtf( man );
  y0f = spu_rsqrte( manf );
  y1f = spu_madd( spu_mul( y0f, halff ), spu_nmsub( y0f, spu_mul( y0f, manf ), onef ), y0f );
  y1 = spu_mul( spu_extend( y1f ), nexp );
  y2 = spu_madd( spu_mul( y1, half ), spu_nmsub( y1, spu_mul( y1, x ), one ), y1 );
  y3 = spu_madd( spu_mul( y2, half ), spu_nmsub( y2, spu_mul( y2, x ), one ), y2 );

  // Choose iterated result or special value.

  zero = spu_and( x, (vec_double2)signmask );
  inf = spu_sel( (vec_double2)expmask, x, signmask );
  nan = (vec_double2)spu_splats(0x7ff8000000000000ull);

  result = spu_sel( y3, zero, _isinfd2 ( x ) );
  result = spu_sel( result, nan, _signbitd2 ( x ) );
  result = spu_sel( result, inf, _is0denormd2 ( x ) );

  return result;
}

#endif