Details | Last modification | View Log | RSS feed
| Rev | Author | Line No. | Line |
|---|---|---|---|
| 56 | mjames | 1 | /* ---------------------------------------------------------------------- |
| 2 | * Project: CMSIS DSP Library |
||
| 3 | * Title: arm_mat_scale_f32.c |
||
| 4 | * Description: Multiplies a floating-point matrix 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 groupMatrix |
||
| 33 | */ |
||
| 34 | |||
| 35 | /** |
||
| 36 | * @defgroup MatrixScale Matrix Scale |
||
| 37 | * |
||
| 38 | * Multiplies a matrix by a scalar. This is accomplished by multiplying each element in the |
||
| 39 | * matrix by the scalar. For example: |
||
| 40 | * \image html MatrixScale.gif "Matrix Scaling of a 3 x 3 matrix" |
||
| 41 | * |
||
| 42 | * The function checks to make sure that the input and output matrices are of the same size. |
||
| 43 | * |
||
| 44 | * In the fixed-point Q15 and Q31 functions, <code>scale</code> is represented by |
||
| 45 | * a fractional multiplication <code>scaleFract</code> and an arithmetic shift <code>shift</code>. |
||
| 46 | * The shift allows the gain of the scaling operation to exceed 1.0. |
||
| 47 | * The overall scale factor applied to the fixed-point data is |
||
| 48 | * <pre> |
||
| 49 | * scale = scaleFract * 2^shift. |
||
| 50 | * </pre> |
||
| 51 | */ |
||
| 52 | |||
| 53 | /** |
||
| 54 | * @addtogroup MatrixScale |
||
| 55 | * @{ |
||
| 56 | */ |
||
| 57 | |||
| 58 | /** |
||
| 59 | * @brief Floating-point matrix scaling. |
||
| 60 | * @param[in] *pSrc points to input matrix structure |
||
| 61 | * @param[in] scale scale factor to be applied |
||
| 62 | * @param[out] *pDst points to output matrix structure |
||
| 63 | * @return The function returns either <code>ARM_MATH_SIZE_MISMATCH</code> |
||
| 64 | * or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking. |
||
| 65 | * |
||
| 66 | */ |
||
| 67 | |||
| 68 | arm_status arm_mat_scale_f32( |
||
| 69 | const arm_matrix_instance_f32 * pSrc, |
||
| 70 | float32_t scale, |
||
| 71 | arm_matrix_instance_f32 * pDst) |
||
| 72 | { |
||
| 73 | float32_t *pIn = pSrc->pData; /* input data matrix pointer */ |
||
| 74 | float32_t *pOut = pDst->pData; /* output data matrix pointer */ |
||
| 75 | uint32_t numSamples; /* total number of elements in the matrix */ |
||
| 76 | uint32_t blkCnt; /* loop counters */ |
||
| 77 | arm_status status; /* status of matrix scaling */ |
||
| 78 | |||
| 79 | #if defined (ARM_MATH_DSP) |
||
| 80 | |||
| 81 | float32_t in1, in2, in3, in4; /* temporary variables */ |
||
| 82 | float32_t out1, out2, out3, out4; /* temporary variables */ |
||
| 83 | |||
| 84 | #endif // #if defined (ARM_MATH_DSP) |
||
| 85 | |||
| 86 | #ifdef ARM_MATH_MATRIX_CHECK |
||
| 87 | /* Check for matrix mismatch condition */ |
||
| 88 | if ((pSrc->numRows != pDst->numRows) || (pSrc->numCols != pDst->numCols)) |
||
| 89 | { |
||
| 90 | /* Set status as ARM_MATH_SIZE_MISMATCH */ |
||
| 91 | status = ARM_MATH_SIZE_MISMATCH; |
||
| 92 | } |
||
| 93 | else |
||
| 94 | #endif /* #ifdef ARM_MATH_MATRIX_CHECK */ |
||
| 95 | { |
||
| 96 | /* Total number of samples in the input matrix */ |
||
| 97 | numSamples = (uint32_t) pSrc->numRows * pSrc->numCols; |
||
| 98 | |||
| 99 | #if defined (ARM_MATH_DSP) |
||
| 100 | |||
| 101 | /* Run the below code for Cortex-M4 and Cortex-M3 */ |
||
| 102 | |||
| 103 | /* Loop Unrolling */ |
||
| 104 | blkCnt = numSamples >> 2; |
||
| 105 | |||
| 106 | /* First part of the processing with loop unrolling. Compute 4 outputs at a time. |
||
| 107 | ** a second loop below computes the remaining 1 to 3 samples. */ |
||
| 108 | while (blkCnt > 0U) |
||
| 109 | { |
||
| 110 | /* C(m,n) = A(m,n) * scale */ |
||
| 111 | /* Scaling and results are stored in the destination buffer. */ |
||
| 112 | in1 = pIn[0]; |
||
| 113 | in2 = pIn[1]; |
||
| 114 | in3 = pIn[2]; |
||
| 115 | in4 = pIn[3]; |
||
| 116 | |||
| 117 | out1 = in1 * scale; |
||
| 118 | out2 = in2 * scale; |
||
| 119 | out3 = in3 * scale; |
||
| 120 | out4 = in4 * scale; |
||
| 121 | |||
| 122 | |||
| 123 | pOut[0] = out1; |
||
| 124 | pOut[1] = out2; |
||
| 125 | pOut[2] = out3; |
||
| 126 | pOut[3] = out4; |
||
| 127 | |||
| 128 | /* update pointers to process next sampels */ |
||
| 129 | pIn += 4U; |
||
| 130 | pOut += 4U; |
||
| 131 | |||
| 132 | /* Decrement the numSamples loop counter */ |
||
| 133 | blkCnt--; |
||
| 134 | } |
||
| 135 | |||
| 136 | /* If the numSamples is not a multiple of 4, compute any remaining output samples here. |
||
| 137 | ** No loop unrolling is used. */ |
||
| 138 | blkCnt = numSamples % 0x4U; |
||
| 139 | |||
| 140 | #else |
||
| 141 | |||
| 142 | /* Run the below code for Cortex-M0 */ |
||
| 143 | |||
| 144 | /* Initialize blkCnt with number of samples */ |
||
| 145 | blkCnt = numSamples; |
||
| 146 | |||
| 147 | #endif /* #if defined (ARM_MATH_DSP) */ |
||
| 148 | |||
| 149 | while (blkCnt > 0U) |
||
| 150 | { |
||
| 151 | /* C(m,n) = A(m,n) * scale */ |
||
| 152 | /* The results are stored in the destination buffer. */ |
||
| 153 | *pOut++ = (*pIn++) * scale; |
||
| 154 | |||
| 155 | /* Decrement the loop counter */ |
||
| 156 | blkCnt--; |
||
| 157 | } |
||
| 158 | |||
| 159 | /* Set status as ARM_MATH_SUCCESS */ |
||
| 160 | status = ARM_MATH_SUCCESS; |
||
| 161 | } |
||
| 162 | |||
| 163 | /* Return to application */ |
||
| 164 | return (status); |
||
| 165 | } |
||
| 166 | |||
| 167 | /** |
||
| 168 | * @} end of MatrixScale group |
||
| 169 | */ |