lib/main/CMSIS/DSP/Source/BasicMathFunctions/arm_scale_q7.c

   1 /* ----------------------------------------------------------------------
   2  * Project:      CMSIS DSP Library
   3  * Title:        arm_scale_q7.c
   4  * Description:  Multiplies a Q7 vector by a scalar
   5  *
   6  * $Date:        27. January 2017
   7  * $Revision:    V.1.5.1
   8  *
   9  * Target Processor: Cortex-M cores
  10  * -------------------------------------------------------------------- */
  11 /*
  12  * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
  13  *
  14  * SPDX-License-Identifier: Apache-2.0
  15  *
  16  * Licensed under the Apache License, Version 2.0 (the License); you may
  17  * not use this file except in compliance with the License.
  18  * You may obtain a copy of the License at
  19  *
  20  * www.apache.org/licenses/LICENSE-2.0
  21  *
  22  * Unless required by applicable law or agreed to in writing, software
  23  * distributed under the License is distributed on an AS IS BASIS, WITHOUT
  24  * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  25  * See the License for the specific language governing permissions and
  26  * limitations under the License.
  27  */
  28
  29 #include "arm_math.h"
  30
  31 /**
  32  * @ingroup groupMath
  33  */
  34
  35 /**
  36  * @addtogroup scale
  37  * @{
  38  */
  39
  40 /**
  41  * @brief Multiplies a Q7 vector by a scalar.
  42  * @param[in]       *pSrc points to the input vector
  43  * @param[in]       scaleFract fractional portion of the scale value
  44  * @param[in]       shift number of bits to shift the result by
  45  * @param[out]      *pDst points to the output vector
  46  * @param[in]       blockSize number of samples in the vector
  47  * @return none.
  48  *
  49  * <b>Scaling and Overflow Behavior:</b>
  50  * \par
  51  * The input data <code>*pSrc</code> and <code>scaleFract</code> are in 1.7 format.
  52  * These are multiplied to yield a 2.14 intermediate result and this is shifted with saturation to 1.7 format.
  53  */
  54
  55 void arm_scale_q7(
  56   q7_t * pSrc,
  57   q7_t scaleFract,
  58   int8_t shift,
  59   q7_t * pDst,
  60   uint32_t blockSize)
  61 {
  62   int8_t kShift = 7 - shift;                     /* shift to apply after scaling */
  63   uint32_t blkCnt;                               /* loop counter */
  64
  65 #if defined (ARM_MATH_DSP)
  66
  67 /* Run the below code for Cortex-M4 and Cortex-M3 */
  68   q7_t in1, in2, in3, in4, out1, out2, out3, out4;      /* Temporary variables to store input & output */
  69
  70
  71   /*loop Unrolling */
  72   blkCnt = blockSize >> 2U;
  73
  74
  75   /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
  76    ** a second loop below computes the remaining 1 to 3 samples. */
  77   while (blkCnt > 0U)
  78   {
  79     /* Reading 4 inputs from memory */
  80     in1 = *pSrc++;
  81     in2 = *pSrc++;
  82     in3 = *pSrc++;
  83     in4 = *pSrc++;
  84
  85     /* C = A * scale */
  86     /* Scale the inputs and then store the results in the temporary variables. */
  87     out1 = (q7_t) (__SSAT(((in1) * scaleFract) >> kShift, 8));
  88     out2 = (q7_t) (__SSAT(((in2) * scaleFract) >> kShift, 8));
  89     out3 = (q7_t) (__SSAT(((in3) * scaleFract) >> kShift, 8));
  90     out4 = (q7_t) (__SSAT(((in4) * scaleFract) >> kShift, 8));
  91
  92     /* Packing the individual outputs into 32bit and storing in
  93      * destination buffer in single write */
  94     *__SIMD32(pDst)++ = __PACKq7(out1, out2, out3, out4);
  95
  96     /* Decrement the loop counter */
  97     blkCnt--;
  98   }
  99
 100   /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
 101    ** No loop unrolling is used. */
 102   blkCnt = blockSize % 0x4U;
 103
 104   while (blkCnt > 0U)
 105   {
 106     /* C = A * scale */
 107     /* Scale the input and then store the result in the destination buffer. */
 108     *pDst++ = (q7_t) (__SSAT(((*pSrc++) * scaleFract) >> kShift, 8));
 109
 110     /* Decrement the loop counter */
 111     blkCnt--;
 112   }
 113
 114 #else
 115
 116   /* Run the below code for Cortex-M0 */
 117
 118   /* Initialize blkCnt with number of samples */
 119   blkCnt = blockSize;
 120
 121   while (blkCnt > 0U)
 122   {
 123     /* C = A * scale */
 124     /* Scale the input and then store the result in the destination buffer. */
 125     *pDst++ = (q7_t) (__SSAT((((q15_t) * pSrc++ * scaleFract) >> kShift), 8));
 126
 127     /* Decrement the loop counter */
 128     blkCnt--;
 129   }
 130
 131 #endif /* #if defined (ARM_MATH_DSP) */
 132
 133 }
 134
 135 /**
 136  * @} end of scale group
 137  */