tmk_core/tool/mbed/mbed-sdk/libraries/dsp/cmsis_dsp/BasicMathFunctions/arm_shift_q31.c

   1 /* ----------------------------------------------------------------------
   2 * Copyright (C) 2010-2013 ARM Limited. All rights reserved.
   3 *
   4 * $Date:        17. January 2013
   5 * $Revision:    V1.4.1
   6 *
   7 * Project:          CMSIS DSP Library
   8 * Title:                arm_shift_q31.c
   9 *
  10 * Description:  Shifts the elements of a Q31 vector by a specified number of bits.
  11 *
  12 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
  13 *
  14 * Redistribution and use in source and binary forms, with or without
  15 * modification, are permitted provided that the following conditions
  16 * are met:
  17 *   - Redistributions of source code must retain the above copyright
  18 *     notice, this list of conditions and the following disclaimer.
  19 *   - Redistributions in binary form must reproduce the above copyright
  20 *     notice, this list of conditions and the following disclaimer in
  21 *     the documentation and/or other materials provided with the
  22 *     distribution.
  23 *   - Neither the name of ARM LIMITED nor the names of its contributors
  24 *     may be used to endorse or promote products derived from this
  25 *     software without specific prior written permission.
  26 *
  27 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
  28 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
  29 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
  30 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
  31 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
  32 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
  33 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
  34 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
  35 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  36 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
  37 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
  38 * POSSIBILITY OF SUCH DAMAGE.
  39 * -------------------------------------------------------------------- */
  40
  41 #include "arm_math.h"
  42
  43 /**
  44  * @ingroup groupMath
  45  */
  46 /**
  47  * @defgroup shift Vector Shift
  48  *
  49  * Shifts the elements of a fixed-point vector by a specified number of bits.
  50  * There are separate functions for Q7, Q15, and Q31 data types.
  51  * The underlying algorithm used is:
  52  *
  53  * <pre>
  54  *     pDst[n] = pSrc[n] << shift,   0 <= n < blockSize.
  55  * </pre>
  56  *
  57  * If <code>shift</code> is positive then the elements of the vector are shifted to the left.
  58  * If <code>shift</code> is negative then the elements of the vector are shifted to the right.
  59  *
  60  * The functions support in-place computation allowing the source and destination
  61  * pointers to reference the same memory buffer.
  62  */
  63
  64 /**
  65  * @addtogroup shift
  66  * @{
  67  */
  68
  69 /**
  70  * @brief  Shifts the elements of a Q31 vector a specified number of bits.
  71  * @param[in]  *pSrc points to the input vector
  72  * @param[in]  shiftBits number of bits to shift.  A positive value shifts left; a negative value shifts right.
  73  * @param[out]  *pDst points to the output vector
  74  * @param[in]  blockSize number of samples in the vector
  75  * @return none.
  76  *
  77  *
  78  * <b>Scaling and Overflow Behavior:</b>
  79  * \par
  80  * The function uses saturating arithmetic.
  81  * Results outside of the allowable Q31 range [0x80000000 0x7FFFFFFF] will be saturated.
  82  */
  83
  84 void arm_shift_q31(
  85   q31_t * pSrc,
  86   int8_t shiftBits,
  87   q31_t * pDst,
  88   uint32_t blockSize)
  89 {
  90   uint32_t blkCnt;                               /* loop counter */
  91   uint8_t sign = (shiftBits & 0x80);             /* Sign of shiftBits */
  92
  93 #ifndef ARM_MATH_CM0_FAMILY
  94
  95   q31_t in1, in2, in3, in4;                      /* Temporary input variables */
  96   q31_t out1, out2, out3, out4;                  /* Temporary output variables */
  97
  98   /*loop Unrolling */
  99   blkCnt = blockSize >> 2u;
 100
 101
 102   if(sign == 0u)
 103   {
 104     /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
 105      ** a second loop below computes the remaining 1 to 3 samples. */
 106     while(blkCnt > 0u)
 107     {
 108       /* C = A  << shiftBits */
 109       /* Shift the input and then store the results in the destination buffer. */
 110       in1 = *pSrc;
 111       in2 = *(pSrc + 1);
 112       out1 = in1 << shiftBits;
 113       in3 = *(pSrc + 2);
 114       out2 = in2 << shiftBits;
 115       in4 = *(pSrc + 3);
 116       if(in1 != (out1 >> shiftBits))
 117         out1 = 0x7FFFFFFF ^ (in1 >> 31);
 118
 119       if(in2 != (out2 >> shiftBits))
 120         out2 = 0x7FFFFFFF ^ (in2 >> 31);
 121
 122       *pDst = out1;
 123       out3 = in3 << shiftBits;
 124       *(pDst + 1) = out2;
 125       out4 = in4 << shiftBits;
 126
 127       if(in3 != (out3 >> shiftBits))
 128         out3 = 0x7FFFFFFF ^ (in3 >> 31);
 129
 130       if(in4 != (out4 >> shiftBits))
 131         out4 = 0x7FFFFFFF ^ (in4 >> 31);
 132
 133       *(pDst + 2) = out3;
 134       *(pDst + 3) = out4;
 135
 136       /* Update destination pointer to process next sampels */
 137       pSrc += 4u;
 138       pDst += 4u;
 139
 140       /* Decrement the loop counter */
 141       blkCnt--;
 142     }
 143   }
 144   else
 145   {
 146
 147     /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
 148      ** a second loop below computes the remaining 1 to 3 samples. */
 149     while(blkCnt > 0u)
 150     {
 151       /* C = A >>  shiftBits */
 152       /* Shift the input and then store the results in the destination buffer. */
 153       in1 = *pSrc;
 154       in2 = *(pSrc + 1);
 155       in3 = *(pSrc + 2);
 156       in4 = *(pSrc + 3);
 157
 158       *pDst = (in1 >> -shiftBits);
 159       *(pDst + 1) = (in2 >> -shiftBits);
 160       *(pDst + 2) = (in3 >> -shiftBits);
 161       *(pDst + 3) = (in4 >> -shiftBits);
 162
 163
 164       pSrc += 4u;
 165       pDst += 4u;
 166
 167       blkCnt--;
 168     }
 169
 170   }
 171
 172   /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
 173    ** No loop unrolling is used. */
 174   blkCnt = blockSize % 0x4u;
 175
 176 #else
 177
 178   /* Run the below code for Cortex-M0 */
 179
 180
 181   /* Initialize blkCnt with number of samples */
 182   blkCnt = blockSize;
 183
 184 #endif /* #ifndef ARM_MATH_CM0_FAMILY */
 185
 186
 187   while(blkCnt > 0u)
 188   {
 189     /* C = A (>> or <<) shiftBits */
 190     /* Shift the input and then store the result in the destination buffer. */
 191     *pDst++ = (sign == 0u) ? clip_q63_to_q31((q63_t) * pSrc++ << shiftBits) :
 192       (*pSrc++ >> -shiftBits);
 193
 194     /* Decrement the loop counter */
 195     blkCnt--;
 196   }
 197
 198
 199 }
 200
 201 /**
 202  * @} end of shift group
 203  */