Merge pull request #11270 from haslinghuis/rename_attr
[betaflight.git] / lib / main / CMSIS / DSP / Source / MatrixFunctions / arm_mat_sub_f32.c
blob7c0b54e9aec944feff31019b4a1dfdbfc3ea7e12
1 /* ----------------------------------------------------------------------
2 * Project: CMSIS DSP Library
3 * Title: arm_mat_sub_f32.c
4 * Description: Floating-point matrix subtraction
6 * $Date: 27. January 2017
7 * $Revision: V.1.5.1
9 * Target Processor: Cortex-M cores
10 * -------------------------------------------------------------------- */
12 * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
14 * SPDX-License-Identifier: Apache-2.0
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
20 * www.apache.org/licenses/LICENSE-2.0
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.
29 #include "arm_math.h"
31 /**
32 * @ingroup groupMatrix
35 /**
36 * @defgroup MatrixSub Matrix Subtraction
38 * Subtract two matrices.
39 * \image html MatrixSubtraction.gif "Subraction of two 3 x 3 matrices"
41 * The functions check to make sure that
42 * <code>pSrcA</code>, <code>pSrcB</code>, and <code>pDst</code> have the same
43 * number of rows and columns.
46 /**
47 * @addtogroup MatrixSub
48 * @{
51 /**
52 * @brief Floating-point matrix subtraction
53 * @param[in] *pSrcA points to the first input matrix structure
54 * @param[in] *pSrcB points to the second input matrix structure
55 * @param[out] *pDst points to output matrix structure
56 * @return The function returns either
57 * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.
60 arm_status arm_mat_sub_f32(
61 const arm_matrix_instance_f32 * pSrcA,
62 const arm_matrix_instance_f32 * pSrcB,
63 arm_matrix_instance_f32 * pDst)
65 float32_t *pIn1 = pSrcA->pData; /* input data matrix pointer A */
66 float32_t *pIn2 = pSrcB->pData; /* input data matrix pointer B */
67 float32_t *pOut = pDst->pData; /* output data matrix pointer */
69 #if defined (ARM_MATH_DSP)
71 float32_t inA1, inA2, inB1, inB2, out1, out2; /* temporary variables */
73 #endif // #if defined (ARM_MATH_DSP)
75 uint32_t numSamples; /* total number of elements in the matrix */
76 uint32_t blkCnt; /* loop counters */
77 arm_status status; /* status of matrix subtraction */
79 #ifdef ARM_MATH_MATRIX_CHECK
80 /* Check for matrix mismatch condition */
81 if ((pSrcA->numRows != pSrcB->numRows) ||
82 (pSrcA->numCols != pSrcB->numCols) ||
83 (pSrcA->numRows != pDst->numRows) || (pSrcA->numCols != pDst->numCols))
85 /* Set status as ARM_MATH_SIZE_MISMATCH */
86 status = ARM_MATH_SIZE_MISMATCH;
88 else
89 #endif /* #ifdef ARM_MATH_MATRIX_CHECK */
91 /* Total number of samples in the input matrix */
92 numSamples = (uint32_t) pSrcA->numRows * pSrcA->numCols;
94 #if defined (ARM_MATH_DSP)
96 /* Run the below code for Cortex-M4 and Cortex-M3 */
98 /* Loop Unrolling */
99 blkCnt = numSamples >> 2U;
101 /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
102 ** a second loop below computes the remaining 1 to 3 samples. */
103 while (blkCnt > 0U)
105 /* C(m,n) = A(m,n) - B(m,n) */
106 /* Subtract and then store the results in the destination buffer. */
107 /* Read values from source A */
108 inA1 = pIn1[0];
110 /* Read values from source B */
111 inB1 = pIn2[0];
113 /* Read values from source A */
114 inA2 = pIn1[1];
116 /* out = sourceA - sourceB */
117 out1 = inA1 - inB1;
119 /* Read values from source B */
120 inB2 = pIn2[1];
122 /* Read values from source A */
123 inA1 = pIn1[2];
125 /* out = sourceA - sourceB */
126 out2 = inA2 - inB2;
128 /* Read values from source B */
129 inB1 = pIn2[2];
131 /* Store result in destination */
132 pOut[0] = out1;
133 pOut[1] = out2;
135 /* Read values from source A */
136 inA2 = pIn1[3];
138 /* Read values from source B */
139 inB2 = pIn2[3];
141 /* out = sourceA - sourceB */
142 out1 = inA1 - inB1;
145 /* out = sourceA - sourceB */
146 out2 = inA2 - inB2;
148 /* Store result in destination */
149 pOut[2] = out1;
151 /* Store result in destination */
152 pOut[3] = out2;
155 /* update pointers to process next sampels */
156 pIn1 += 4U;
157 pIn2 += 4U;
158 pOut += 4U;
160 /* Decrement the loop counter */
161 blkCnt--;
164 /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
165 ** No loop unrolling is used. */
166 blkCnt = numSamples % 0x4U;
168 #else
170 /* Run the below code for Cortex-M0 */
172 /* Initialize blkCnt with number of samples */
173 blkCnt = numSamples;
175 #endif /* #if defined (ARM_MATH_DSP) */
177 while (blkCnt > 0U)
179 /* C(m,n) = A(m,n) - B(m,n) */
180 /* Subtract and then store the results in the destination buffer. */
181 *pOut++ = (*pIn1++) - (*pIn2++);
183 /* Decrement the loop counter */
184 blkCnt--;
187 /* Set status as ARM_MATH_SUCCESS */
188 status = ARM_MATH_SUCCESS;
191 /* Return to application */
192 return (status);
196 * @} end of MatrixSub group