1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
|
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_mat_trans_f16.c
* Description: Floating-point matrix transpose
*
* $Date: 23 April 2021
* $Revision: V1.9.0
*
* Target Processor: Cortex-M and Cortex-A cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2021 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "dsp/matrix_functions_f16.h"
#if defined(ARM_FLOAT16_SUPPORTED)
/**
@ingroup groupMatrix
*/
/**
@addtogroup MatrixTrans
@{
*/
/**
@brief Floating-point matrix transpose.
@param[in] pSrc points to input matrix
@param[out] pDst points to output matrix
@return execution status
- \ref ARM_MATH_SUCCESS : Operation successful
- \ref ARM_MATH_SIZE_MISMATCH : Matrix size check failed
*/
#if defined(ARM_MATH_MVE_FLOAT16) && !defined(ARM_MATH_AUTOVECTORIZE)
#include "arm_helium_utils.h"
arm_status arm_mat_trans_f16(
const arm_matrix_instance_f16 * pSrc,
arm_matrix_instance_f16 * pDst)
{
arm_status status; /* status of matrix transpose */
#ifdef ARM_MATH_MATRIX_CHECK
/* Check for matrix mismatch condition */
if ((pSrc->numRows != pDst->numCols) ||
(pSrc->numCols != pDst->numRows) )
{
/* Set status as ARM_MATH_SIZE_MISMATCH */
status = ARM_MATH_SIZE_MISMATCH;
}
else
#endif /* #ifdef ARM_MATH_MATRIX_CHECK */
{
if (pDst->numRows == pDst->numCols)
{
if (pDst->numCols == 1)
{
pDst->pData[0] = pSrc->pData[0];
return(ARM_MATH_SUCCESS);
}
if (pDst->numCols == 2)
return arm_mat_trans_16bit_2x2((uint16_t *)pSrc->pData, (uint16_t *)pDst->pData);
if (pDst->numCols == 3)
return arm_mat_trans_16bit_3x3_mve((uint16_t *)pSrc->pData, (uint16_t *)pDst->pData);
if (pDst->numCols == 4)
return arm_mat_trans_16bit_4x4_mve((uint16_t *)pSrc->pData, (uint16_t *)pDst->pData);
}
arm_mat_trans_16bit_generic(pSrc->numRows, pSrc->numCols, (uint16_t *)pSrc->pData, (uint16_t *)pDst->pData);
/* Set status as ARM_MATH_SUCCESS */
status = ARM_MATH_SUCCESS;
}
/* Return to application */
return (status);
}
#else
arm_status arm_mat_trans_f16(
const arm_matrix_instance_f16 * pSrc,
arm_matrix_instance_f16 * pDst)
{
float16_t *pIn = pSrc->pData; /* input data matrix pointer */
float16_t *pOut = pDst->pData; /* output data matrix pointer */
float16_t *px; /* Temporary output data matrix pointer */
uint16_t nRows = pSrc->numRows; /* number of rows */
uint16_t nCols = pSrc->numCols; /* number of columns */
uint32_t col, row = nRows, i = 0U; /* Loop counters */
arm_status status; /* status of matrix transpose */
#ifdef ARM_MATH_MATRIX_CHECK
/* Check for matrix mismatch condition */
if ((pSrc->numRows != pDst->numCols) ||
(pSrc->numCols != pDst->numRows) )
{
/* Set status as ARM_MATH_SIZE_MISMATCH */
status = ARM_MATH_SIZE_MISMATCH;
}
else
#endif /* #ifdef ARM_MATH_MATRIX_CHECK */
{
/* Matrix transpose by exchanging the rows with columns */
/* row loop */
do
{
/* Pointer px is set to starting address of column being processed */
px = pOut + i;
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
col = nCols >> 2U;
while (col > 0U) /* column loop */
{
/* Read and store input element in destination */
*px = *pIn++;
/* Update pointer px to point to next row of transposed matrix */
px += nRows;
*px = *pIn++;
px += nRows;
*px = *pIn++;
px += nRows;
*px = *pIn++;
px += nRows;
/* Decrement column loop counter */
col--;
}
/* Loop unrolling: Compute remaining outputs */
col = nCols % 0x4U;
#else
/* Initialize col with number of samples */
col = nCols;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (col > 0U)
{
/* Read and store input element in destination */
*px = *pIn++;
/* Update pointer px to point to next row of transposed matrix */
px += nRows;
/* Decrement column loop counter */
col--;
}
i++;
/* Decrement row loop counter */
row--;
} while (row > 0U); /* row loop end */
/* Set status as ARM_MATH_SUCCESS */
status = ARM_MATH_SUCCESS;
}
/* Return to application */
return (status);
}
#endif /* defined(ARM_MATH_MVEF) && !defined(ARM_MATH_AUTOVECTORIZE) */
/**
* @} end of MatrixTrans group
*/
#endif /* #if defined(ARM_FLOAT16_SUPPORTED) */
|
