STM32F769IDiscovery  1.00
uDANTE Audio Networking with STM32F7 DISCO board
arm_mat_trans_f32.c
Go to the documentation of this file.
1 /* ----------------------------------------------------------------------
2 * Copyright (C) 2010-2014 ARM Limited. All rights reserved.
3 *
4 * $Date: 19. March 2015
5 * $Revision: V.1.4.5
6 *
7 * Project: CMSIS DSP Library
8 * Title: arm_mat_trans_f32.c
9 *
10 * Description: Floating-point matrix transpose.
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 
49 #include "arm_math.h"
50 
70  const arm_matrix_instance_f32 * pSrc,
72 {
73  float32_t *pIn = pSrc->pData; /* input data matrix pointer */
74  float32_t *pOut = pDst->pData; /* output data matrix pointer */
75  float32_t *px; /* Temporary output data matrix pointer */
76  uint16_t nRows = pSrc->numRows; /* number of rows */
77  uint16_t nColumns = pSrc->numCols; /* number of columns */
78 
79 #ifndef ARM_MATH_CM0_FAMILY
80 
81  /* Run the below code for Cortex-M4 and Cortex-M3 */
82 
83  uint16_t blkCnt, i = 0u, row = nRows; /* loop counters */
84  arm_status status; /* status of matrix transpose */
85 
86 
87 #ifdef ARM_MATH_MATRIX_CHECK
88 
89 
90  /* Check for matrix mismatch condition */
91  if((pSrc->numRows != pDst->numCols) || (pSrc->numCols != pDst->numRows))
92  {
93  /* Set status as ARM_MATH_SIZE_MISMATCH */
94  status = ARM_MATH_SIZE_MISMATCH;
95  }
96  else
97 #endif /* #ifdef ARM_MATH_MATRIX_CHECK */
98 
99  {
100  /* Matrix transpose by exchanging the rows with columns */
101  /* row loop */
102  do
103  {
104  /* Loop Unrolling */
105  blkCnt = nColumns >> 2;
106 
107  /* The pointer px is set to starting address of the column being processed */
108  px = pOut + i;
109 
110  /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
111  ** a second loop below computes the remaining 1 to 3 samples. */
112  while(blkCnt > 0u) /* column loop */
113  {
114  /* Read and store the input element in the destination */
115  *px = *pIn++;
116 
117  /* Update the pointer px to point to the next row of the transposed matrix */
118  px += nRows;
119 
120  /* Read and store the input element in the destination */
121  *px = *pIn++;
122 
123  /* Update the pointer px to point to the next row of the transposed matrix */
124  px += nRows;
125 
126  /* Read and store the input element in the destination */
127  *px = *pIn++;
128 
129  /* Update the pointer px to point to the next row of the transposed matrix */
130  px += nRows;
131 
132  /* Read and store the input element in the destination */
133  *px = *pIn++;
134 
135  /* Update the pointer px to point to the next row of the transposed matrix */
136  px += nRows;
137 
138  /* Decrement the column loop counter */
139  blkCnt--;
140  }
141 
142  /* Perform matrix transpose for last 3 samples here. */
143  blkCnt = nColumns % 0x4u;
144 
145  while(blkCnt > 0u)
146  {
147  /* Read and store the input element in the destination */
148  *px = *pIn++;
149 
150  /* Update the pointer px to point to the next row of the transposed matrix */
151  px += nRows;
152 
153  /* Decrement the column loop counter */
154  blkCnt--;
155  }
156 
157 #else
158 
159  /* Run the below code for Cortex-M0 */
160 
161  uint16_t col, i = 0u, row = nRows; /* loop counters */
162  arm_status status; /* status of matrix transpose */
163 
164 
165 #ifdef ARM_MATH_MATRIX_CHECK
166 
167  /* Check for matrix mismatch condition */
168  if((pSrc->numRows != pDst->numCols) || (pSrc->numCols != pDst->numRows))
169  {
170  /* Set status as ARM_MATH_SIZE_MISMATCH */
171  status = ARM_MATH_SIZE_MISMATCH;
172  }
173  else
174 #endif /* #ifdef ARM_MATH_MATRIX_CHECK */
175 
176  {
177  /* Matrix transpose by exchanging the rows with columns */
178  /* row loop */
179  do
180  {
181  /* The pointer px is set to starting address of the column being processed */
182  px = pOut + i;
183 
184  /* Initialize column loop counter */
185  col = nColumns;
186 
187  while(col > 0u)
188  {
189  /* Read and store the input element in the destination */
190  *px = *pIn++;
191 
192  /* Update the pointer px to point to the next row of the transposed matrix */
193  px += nRows;
194 
195  /* Decrement the column loop counter */
196  col--;
197  }
198 
199 #endif /* #ifndef ARM_MATH_CM0_FAMILY */
200 
201  i++;
202 
203  /* Decrement the row loop counter */
204  row--;
205 
206  } while(row > 0u); /* row loop end */
207 
208  /* Set status as ARM_MATH_SUCCESS */
209  status = ARM_MATH_SUCCESS;
210  }
211 
212  /* Return to application */
213  return (status);
214 }
215 
arm_status arm_mat_trans_f32(const arm_matrix_instance_f32 *pSrc, arm_matrix_instance_f32 *pDst)
Floating-point matrix transpose.
float float32_t
32-bit floating-point type definition.
Definition: arm_math.h:407
Instance structure for the floating-point matrix structure.
Definition: arm_math.h:1369
arm_status
Error status returned by some functions in the library.
Definition: arm_math.h:373