mbed official / mbed-dsp

CMSIS DSP library

Dependents:   performance_timer Surfboard_ gps2rtty Capstone ... more

Legacy Warning

This is an mbed 2 library. To learn more about mbed OS 5, visit the docs.

cmsis_dsp/MatrixFunctions/arm_mat_mult_fast_q15.c@5:3762170b6d4d, 2015-11-20 (annotated)

Committer:
mbed_official
Date:
Fri Nov 20 08:45:18 2015 +0000
Revision:
5:3762170b6d4d
Parent:
3:7a284390b0ce 
Synchronized with git revision 2eb940b9a73af188d3004a2575fdfbb05febe62b



Full URL: https://github.com/mbedmicro/mbed/commit/2eb940b9a73af188d3004a2575fdfbb05febe62b/



Added option to build rpc library. closes #1426

Who changed what in which revision?

UserRevisionLine numberNew contents of line
emilmont 1:fdd22bb7aa52 1 /* ----------------------------------------------------------------------
mbed_official 5:3762170b6d4d 2 * Copyright (C) 2010-2014 ARM Limited. All rights reserved.
emilmont 1:fdd22bb7aa52 3 *
mbed_official 5:3762170b6d4d 4 * $Date: 19. March 2015
mbed_official 5:3762170b6d4d 5 * $Revision: V.1.4.5
emilmont 1:fdd22bb7aa52 6 *
emilmont 2:da51fb522205 7 * Project: CMSIS DSP Library
emilmont 2:da51fb522205 8 * Title: arm_mat_mult_fast_q15.c
emilmont 1:fdd22bb7aa52 9 *
emilmont 2:da51fb522205 10 * Description: Q15 matrix multiplication (fast variant)
emilmont 1:fdd22bb7aa52 11 *
emilmont 1:fdd22bb7aa52 12 * Target Processor: Cortex-M4/Cortex-M3
emilmont 1:fdd22bb7aa52 13 *
mbed_official 3:7a284390b0ce 14 * Redistribution and use in source and binary forms, with or without
mbed_official 3:7a284390b0ce 15 * modification, are permitted provided that the following conditions
mbed_official 3:7a284390b0ce 16 * are met:
mbed_official 3:7a284390b0ce 17 * - Redistributions of source code must retain the above copyright
mbed_official 3:7a284390b0ce 18 * notice, this list of conditions and the following disclaimer.
mbed_official 3:7a284390b0ce 19 * - Redistributions in binary form must reproduce the above copyright
mbed_official 3:7a284390b0ce 20 * notice, this list of conditions and the following disclaimer in
mbed_official 3:7a284390b0ce 21 * the documentation and/or other materials provided with the
mbed_official 3:7a284390b0ce 22 * distribution.
mbed_official 3:7a284390b0ce 23 * - Neither the name of ARM LIMITED nor the names of its contributors
mbed_official 3:7a284390b0ce 24 * may be used to endorse or promote products derived from this
mbed_official 3:7a284390b0ce 25 * software without specific prior written permission.
mbed_official 3:7a284390b0ce 26 *
mbed_official 3:7a284390b0ce 27 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
mbed_official 3:7a284390b0ce 28 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
mbed_official 3:7a284390b0ce 29 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
mbed_official 3:7a284390b0ce 30 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
mbed_official 3:7a284390b0ce 31 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
mbed_official 3:7a284390b0ce 32 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
mbed_official 3:7a284390b0ce 33 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
mbed_official 3:7a284390b0ce 34 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
mbed_official 3:7a284390b0ce 35 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
mbed_official 3:7a284390b0ce 36 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
mbed_official 3:7a284390b0ce 37 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
mbed_official 3:7a284390b0ce 38 * POSSIBILITY OF SUCH DAMAGE.
emilmont 1:fdd22bb7aa52 39 * -------------------------------------------------------------------- */
emilmont 1:fdd22bb7aa52 40
emilmont 1:fdd22bb7aa52 41 #include "arm_math.h"
emilmont 1:fdd22bb7aa52 42
emilmont 1:fdd22bb7aa52 43 /**
emilmont 1:fdd22bb7aa52 44 * @ingroup groupMatrix
emilmont 1:fdd22bb7aa52 45 */
emilmont 1:fdd22bb7aa52 46
emilmont 1:fdd22bb7aa52 47 /**
emilmont 1:fdd22bb7aa52 48 * @addtogroup MatrixMult
emilmont 1:fdd22bb7aa52 49 * @{
emilmont 1:fdd22bb7aa52 50 */
emilmont 1:fdd22bb7aa52 51
emilmont 1:fdd22bb7aa52 52
emilmont 1:fdd22bb7aa52 53 /**
emilmont 1:fdd22bb7aa52 54 * @brief Q15 matrix multiplication (fast variant) for Cortex-M3 and Cortex-M4
emilmont 1:fdd22bb7aa52 55 * @param[in] *pSrcA points to the first input matrix structure
emilmont 1:fdd22bb7aa52 56 * @param[in] *pSrcB points to the second input matrix structure
emilmont 1:fdd22bb7aa52 57 * @param[out] *pDst points to output matrix structure
emilmont 2:da51fb522205 58 * @param[in] *pState points to the array for storing intermediate results
emilmont 2:da51fb522205 59 * @return The function returns either
emilmont 1:fdd22bb7aa52 60 * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.
emilmont 1:fdd22bb7aa52 61 *
emilmont 1:fdd22bb7aa52 62 * @details
emilmont 1:fdd22bb7aa52 63 * <b>Scaling and Overflow Behavior:</b>
emilmont 1:fdd22bb7aa52 64 *
emilmont 1:fdd22bb7aa52 65 * \par
emilmont 1:fdd22bb7aa52 66 * The difference between the function arm_mat_mult_q15() and this fast variant is that
emilmont 1:fdd22bb7aa52 67 * the fast variant use a 32-bit rather than a 64-bit accumulator.
emilmont 1:fdd22bb7aa52 68 * The result of each 1.15 x 1.15 multiplication is truncated to
emilmont 1:fdd22bb7aa52 69 * 2.30 format. These intermediate results are accumulated in a 32-bit register in 2.30
emilmont 1:fdd22bb7aa52 70 * format. Finally, the accumulator is saturated and converted to a 1.15 result.
emilmont 1:fdd22bb7aa52 71 *
emilmont 1:fdd22bb7aa52 72 * \par
emilmont 1:fdd22bb7aa52 73 * The fast version has the same overflow behavior as the standard version but provides
emilmont 1:fdd22bb7aa52 74 * less precision since it discards the low 16 bits of each multiplication result.
emilmont 1:fdd22bb7aa52 75 * In order to avoid overflows completely the input signals must be scaled down.
emilmont 1:fdd22bb7aa52 76 * Scale down one of the input matrices by log2(numColsA) bits to
emilmont 1:fdd22bb7aa52 77 * avoid overflows, as a total of numColsA additions are computed internally for each
emilmont 1:fdd22bb7aa52 78 * output element.
emilmont 1:fdd22bb7aa52 79 *
emilmont 1:fdd22bb7aa52 80 * \par
emilmont 1:fdd22bb7aa52 81 * See <code>arm_mat_mult_q15()</code> for a slower implementation of this function
emilmont 1:fdd22bb7aa52 82 * which uses 64-bit accumulation to provide higher precision.
emilmont 1:fdd22bb7aa52 83 */
emilmont 1:fdd22bb7aa52 84
emilmont 1:fdd22bb7aa52 85 arm_status arm_mat_mult_fast_q15(
emilmont 1:fdd22bb7aa52 86 const arm_matrix_instance_q15 * pSrcA,
emilmont 1:fdd22bb7aa52 87 const arm_matrix_instance_q15 * pSrcB,
emilmont 1:fdd22bb7aa52 88 arm_matrix_instance_q15 * pDst,
emilmont 1:fdd22bb7aa52 89 q15_t * pState)
emilmont 1:fdd22bb7aa52 90 {
emilmont 1:fdd22bb7aa52 91 q31_t sum; /* accumulator */
emilmont 1:fdd22bb7aa52 92 q15_t *pSrcBT = pState; /* input data matrix pointer for transpose */
emilmont 1:fdd22bb7aa52 93 q15_t *pInA = pSrcA->pData; /* input data matrix pointer A of Q15 type */
emilmont 1:fdd22bb7aa52 94 q15_t *pInB = pSrcB->pData; /* input data matrix pointer B of Q15 type */
emilmont 1:fdd22bb7aa52 95 q15_t *px; /* Temporary output data matrix pointer */
emilmont 1:fdd22bb7aa52 96 uint16_t numRowsA = pSrcA->numRows; /* number of rows of input matrix A */
emilmont 1:fdd22bb7aa52 97 uint16_t numColsB = pSrcB->numCols; /* number of columns of input matrix B */
emilmont 1:fdd22bb7aa52 98 uint16_t numColsA = pSrcA->numCols; /* number of columns of input matrix A */
emilmont 1:fdd22bb7aa52 99 uint16_t numRowsB = pSrcB->numRows; /* number of rows of input matrix A */
emilmont 1:fdd22bb7aa52 100 uint16_t col, i = 0u, row = numRowsB, colCnt; /* loop counters */
emilmont 1:fdd22bb7aa52 101 arm_status status; /* status of matrix multiplication */
emilmont 1:fdd22bb7aa52 102
emilmont 1:fdd22bb7aa52 103 #ifndef UNALIGNED_SUPPORT_DISABLE
emilmont 1:fdd22bb7aa52 104
emilmont 1:fdd22bb7aa52 105 q31_t in; /* Temporary variable to hold the input value */
emilmont 1:fdd22bb7aa52 106 q31_t inA1, inA2, inB1, inB2;
emilmont 1:fdd22bb7aa52 107
emilmont 1:fdd22bb7aa52 108 #else
emilmont 1:fdd22bb7aa52 109
emilmont 1:fdd22bb7aa52 110 q15_t in; /* Temporary variable to hold the input value */
emilmont 1:fdd22bb7aa52 111 q15_t inA1, inA2, inB1, inB2;
emilmont 1:fdd22bb7aa52 112
emilmont 2:da51fb522205 113 #endif /* #ifndef UNALIGNED_SUPPORT_DISABLE */
emilmont 1:fdd22bb7aa52 114
emilmont 1:fdd22bb7aa52 115 #ifdef ARM_MATH_MATRIX_CHECK
emilmont 1:fdd22bb7aa52 116 /* Check for matrix mismatch condition */
emilmont 1:fdd22bb7aa52 117 if((pSrcA->numCols != pSrcB->numRows) ||
emilmont 1:fdd22bb7aa52 118 (pSrcA->numRows != pDst->numRows) || (pSrcB->numCols != pDst->numCols))
emilmont 1:fdd22bb7aa52 119 {
emilmont 1:fdd22bb7aa52 120 /* Set status as ARM_MATH_SIZE_MISMATCH */
emilmont 1:fdd22bb7aa52 121 status = ARM_MATH_SIZE_MISMATCH;
emilmont 1:fdd22bb7aa52 122 }
emilmont 1:fdd22bb7aa52 123 else
emilmont 1:fdd22bb7aa52 124 #endif
emilmont 1:fdd22bb7aa52 125 {
emilmont 1:fdd22bb7aa52 126 /* Matrix transpose */
emilmont 1:fdd22bb7aa52 127 do
emilmont 1:fdd22bb7aa52 128 {
emilmont 1:fdd22bb7aa52 129 /* Apply loop unrolling and exchange the columns with row elements */
emilmont 1:fdd22bb7aa52 130 col = numColsB >> 2;
emilmont 1:fdd22bb7aa52 131
emilmont 1:fdd22bb7aa52 132 /* The pointer px is set to starting address of the column being processed */
emilmont 1:fdd22bb7aa52 133 px = pSrcBT + i;
emilmont 1:fdd22bb7aa52 134
emilmont 1:fdd22bb7aa52 135 /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
emilmont 1:fdd22bb7aa52 136 ** a second loop below computes the remaining 1 to 3 samples. */
emilmont 1:fdd22bb7aa52 137 while(col > 0u)
emilmont 1:fdd22bb7aa52 138 {
emilmont 1:fdd22bb7aa52 139 #ifndef UNALIGNED_SUPPORT_DISABLE
emilmont 1:fdd22bb7aa52 140 /* Read two elements from the row */
emilmont 1:fdd22bb7aa52 141 in = *__SIMD32(pInB)++;
emilmont 1:fdd22bb7aa52 142
emilmont 1:fdd22bb7aa52 143 /* Unpack and store one element in the destination */
emilmont 1:fdd22bb7aa52 144 #ifndef ARM_MATH_BIG_ENDIAN
emilmont 1:fdd22bb7aa52 145
emilmont 1:fdd22bb7aa52 146 *px = (q15_t) in;
emilmont 1:fdd22bb7aa52 147
emilmont 1:fdd22bb7aa52 148 #else
emilmont 1:fdd22bb7aa52 149
emilmont 1:fdd22bb7aa52 150 *px = (q15_t) ((in & (q31_t) 0xffff0000) >> 16);
emilmont 1:fdd22bb7aa52 151
emilmont 1:fdd22bb7aa52 152 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
emilmont 1:fdd22bb7aa52 153
emilmont 1:fdd22bb7aa52 154 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 155 px += numRowsB;
emilmont 1:fdd22bb7aa52 156
emilmont 1:fdd22bb7aa52 157 /* Unpack and store the second element in the destination */
emilmont 1:fdd22bb7aa52 158 #ifndef ARM_MATH_BIG_ENDIAN
emilmont 1:fdd22bb7aa52 159
emilmont 1:fdd22bb7aa52 160 *px = (q15_t) ((in & (q31_t) 0xffff0000) >> 16);
emilmont 1:fdd22bb7aa52 161
emilmont 1:fdd22bb7aa52 162 #else
emilmont 1:fdd22bb7aa52 163
emilmont 1:fdd22bb7aa52 164 *px = (q15_t) in;
emilmont 1:fdd22bb7aa52 165
emilmont 1:fdd22bb7aa52 166 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
emilmont 1:fdd22bb7aa52 167
emilmont 1:fdd22bb7aa52 168 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 169 px += numRowsB;
emilmont 1:fdd22bb7aa52 170
emilmont 1:fdd22bb7aa52 171 /* Read two elements from the row */
emilmont 1:fdd22bb7aa52 172 in = *__SIMD32(pInB)++;
emilmont 1:fdd22bb7aa52 173
emilmont 1:fdd22bb7aa52 174 /* Unpack and store one element in the destination */
emilmont 1:fdd22bb7aa52 175 #ifndef ARM_MATH_BIG_ENDIAN
emilmont 1:fdd22bb7aa52 176
emilmont 1:fdd22bb7aa52 177 *px = (q15_t) in;
emilmont 1:fdd22bb7aa52 178
emilmont 1:fdd22bb7aa52 179 #else
emilmont 1:fdd22bb7aa52 180
emilmont 1:fdd22bb7aa52 181 *px = (q15_t) ((in & (q31_t) 0xffff0000) >> 16);
emilmont 1:fdd22bb7aa52 182
emilmont 1:fdd22bb7aa52 183 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
emilmont 1:fdd22bb7aa52 184
emilmont 1:fdd22bb7aa52 185 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 186 px += numRowsB;
emilmont 1:fdd22bb7aa52 187
emilmont 1:fdd22bb7aa52 188 /* Unpack and store the second element in the destination */
emilmont 1:fdd22bb7aa52 189
emilmont 1:fdd22bb7aa52 190 #ifndef ARM_MATH_BIG_ENDIAN
emilmont 1:fdd22bb7aa52 191
emilmont 1:fdd22bb7aa52 192 *px = (q15_t) ((in & (q31_t) 0xffff0000) >> 16);
emilmont 1:fdd22bb7aa52 193
emilmont 1:fdd22bb7aa52 194 #else
emilmont 1:fdd22bb7aa52 195
emilmont 1:fdd22bb7aa52 196 *px = (q15_t) in;
emilmont 1:fdd22bb7aa52 197
emilmont 1:fdd22bb7aa52 198 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
emilmont 1:fdd22bb7aa52 199
emilmont 1:fdd22bb7aa52 200 #else
emilmont 1:fdd22bb7aa52 201
emilmont 1:fdd22bb7aa52 202 /* Read one element from the row */
emilmont 1:fdd22bb7aa52 203 in = *pInB++;
emilmont 1:fdd22bb7aa52 204
emilmont 1:fdd22bb7aa52 205 /* Store one element in the destination */
emilmont 1:fdd22bb7aa52 206 *px = in;
emilmont 1:fdd22bb7aa52 207
emilmont 1:fdd22bb7aa52 208 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 209 px += numRowsB;
emilmont 1:fdd22bb7aa52 210
emilmont 1:fdd22bb7aa52 211 /* Read one element from the row */
emilmont 1:fdd22bb7aa52 212 in = *pInB++;
emilmont 1:fdd22bb7aa52 213
emilmont 1:fdd22bb7aa52 214 /* Store one element in the destination */
emilmont 1:fdd22bb7aa52 215 *px = in;
emilmont 1:fdd22bb7aa52 216
emilmont 1:fdd22bb7aa52 217 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 218 px += numRowsB;
emilmont 1:fdd22bb7aa52 219
emilmont 1:fdd22bb7aa52 220 /* Read one element from the row */
emilmont 1:fdd22bb7aa52 221 in = *pInB++;
emilmont 1:fdd22bb7aa52 222
emilmont 1:fdd22bb7aa52 223 /* Store one element in the destination */
emilmont 1:fdd22bb7aa52 224 *px = in;
emilmont 1:fdd22bb7aa52 225
emilmont 1:fdd22bb7aa52 226 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 227 px += numRowsB;
emilmont 1:fdd22bb7aa52 228
emilmont 1:fdd22bb7aa52 229 /* Read one element from the row */
emilmont 1:fdd22bb7aa52 230 in = *pInB++;
emilmont 1:fdd22bb7aa52 231
emilmont 1:fdd22bb7aa52 232 /* Store one element in the destination */
emilmont 1:fdd22bb7aa52 233 *px = in;
emilmont 1:fdd22bb7aa52 234
emilmont 2:da51fb522205 235 #endif /* #ifndef UNALIGNED_SUPPORT_DISABLE */
emilmont 1:fdd22bb7aa52 236
emilmont 2:da51fb522205 237 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 238 px += numRowsB;
emilmont 1:fdd22bb7aa52 239
emilmont 1:fdd22bb7aa52 240 /* Decrement the column loop counter */
emilmont 1:fdd22bb7aa52 241 col--;
emilmont 1:fdd22bb7aa52 242 }
emilmont 1:fdd22bb7aa52 243
emilmont 1:fdd22bb7aa52 244 /* If the columns of pSrcB is not a multiple of 4, compute any remaining output samples here.
emilmont 1:fdd22bb7aa52 245 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 246 col = numColsB % 0x4u;
emilmont 1:fdd22bb7aa52 247
emilmont 1:fdd22bb7aa52 248 while(col > 0u)
emilmont 1:fdd22bb7aa52 249 {
emilmont 1:fdd22bb7aa52 250 /* Read and store the input element in the destination */
emilmont 1:fdd22bb7aa52 251 *px = *pInB++;
emilmont 1:fdd22bb7aa52 252
emilmont 1:fdd22bb7aa52 253 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 254 px += numRowsB;
emilmont 1:fdd22bb7aa52 255
emilmont 1:fdd22bb7aa52 256 /* Decrement the column loop counter */
emilmont 1:fdd22bb7aa52 257 col--;
emilmont 1:fdd22bb7aa52 258 }
emilmont 1:fdd22bb7aa52 259
emilmont 1:fdd22bb7aa52 260 i++;
emilmont 1:fdd22bb7aa52 261
emilmont 1:fdd22bb7aa52 262 /* Decrement the row loop counter */
emilmont 1:fdd22bb7aa52 263 row--;
emilmont 1:fdd22bb7aa52 264
emilmont 1:fdd22bb7aa52 265 } while(row > 0u);
emilmont 1:fdd22bb7aa52 266
emilmont 1:fdd22bb7aa52 267 /* Reset the variables for the usage in the following multiplication process */
emilmont 1:fdd22bb7aa52 268 row = numRowsA;
emilmont 1:fdd22bb7aa52 269 i = 0u;
emilmont 1:fdd22bb7aa52 270 px = pDst->pData;
emilmont 1:fdd22bb7aa52 271
emilmont 1:fdd22bb7aa52 272 /* The following loop performs the dot-product of each row in pSrcA with each column in pSrcB */
emilmont 1:fdd22bb7aa52 273 /* row loop */
emilmont 1:fdd22bb7aa52 274 do
emilmont 1:fdd22bb7aa52 275 {
emilmont 1:fdd22bb7aa52 276 /* For every row wise process, the column loop counter is to be initiated */
emilmont 1:fdd22bb7aa52 277 col = numColsB;
emilmont 1:fdd22bb7aa52 278
emilmont 1:fdd22bb7aa52 279 /* For every row wise process, the pIn2 pointer is set
emilmont 1:fdd22bb7aa52 280 ** to the starting address of the transposed pSrcB data */
emilmont 1:fdd22bb7aa52 281 pInB = pSrcBT;
emilmont 1:fdd22bb7aa52 282
emilmont 1:fdd22bb7aa52 283 /* column loop */
emilmont 1:fdd22bb7aa52 284 do
emilmont 1:fdd22bb7aa52 285 {
emilmont 1:fdd22bb7aa52 286 /* Set the variable sum, that acts as accumulator, to zero */
emilmont 1:fdd22bb7aa52 287 sum = 0;
emilmont 1:fdd22bb7aa52 288
emilmont 1:fdd22bb7aa52 289 /* Apply loop unrolling and compute 2 MACs simultaneously. */
emilmont 1:fdd22bb7aa52 290 colCnt = numColsA >> 2;
emilmont 1:fdd22bb7aa52 291
emilmont 1:fdd22bb7aa52 292 /* Initiate the pointer pIn1 to point to the starting address of the column being processed */
emilmont 1:fdd22bb7aa52 293 pInA = pSrcA->pData + i;
emilmont 1:fdd22bb7aa52 294
emilmont 1:fdd22bb7aa52 295 /* matrix multiplication */
emilmont 1:fdd22bb7aa52 296 while(colCnt > 0u)
emilmont 1:fdd22bb7aa52 297 {
emilmont 1:fdd22bb7aa52 298 /* c(m,n) = a(1,1)*b(1,1) + a(1,2) * b(2,1) + .... + a(m,p)*b(p,n) */
emilmont 1:fdd22bb7aa52 299 #ifndef UNALIGNED_SUPPORT_DISABLE
emilmont 1:fdd22bb7aa52 300
emilmont 1:fdd22bb7aa52 301 inA1 = *__SIMD32(pInA)++;
emilmont 1:fdd22bb7aa52 302 inB1 = *__SIMD32(pInB)++;
emilmont 1:fdd22bb7aa52 303 inA2 = *__SIMD32(pInA)++;
emilmont 1:fdd22bb7aa52 304 inB2 = *__SIMD32(pInB)++;
emilmont 1:fdd22bb7aa52 305
emilmont 1:fdd22bb7aa52 306 sum = __SMLAD(inA1, inB1, sum);
emilmont 1:fdd22bb7aa52 307 sum = __SMLAD(inA2, inB2, sum);
emilmont 1:fdd22bb7aa52 308
emilmont 1:fdd22bb7aa52 309 #else
emilmont 1:fdd22bb7aa52 310
emilmont 1:fdd22bb7aa52 311 inA1 = *pInA++;
emilmont 1:fdd22bb7aa52 312 inB1 = *pInB++;
emilmont 1:fdd22bb7aa52 313 inA2 = *pInA++;
emilmont 1:fdd22bb7aa52 314 sum += inA1 * inB1;
emilmont 1:fdd22bb7aa52 315 inB2 = *pInB++;
emilmont 1:fdd22bb7aa52 316
emilmont 1:fdd22bb7aa52 317 inA1 = *pInA++;
emilmont 1:fdd22bb7aa52 318 inB1 = *pInB++;
emilmont 1:fdd22bb7aa52 319 sum += inA2 * inB2;
emilmont 1:fdd22bb7aa52 320 inA2 = *pInA++;
emilmont 1:fdd22bb7aa52 321 inB2 = *pInB++;
emilmont 1:fdd22bb7aa52 322
emilmont 1:fdd22bb7aa52 323 sum += inA1 * inB1;
emilmont 1:fdd22bb7aa52 324 sum += inA2 * inB2;
emilmont 1:fdd22bb7aa52 325
emilmont 2:da51fb522205 326 #endif /* #ifndef UNALIGNED_SUPPORT_DISABLE */
emilmont 1:fdd22bb7aa52 327
emilmont 1:fdd22bb7aa52 328 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 329 colCnt--;
emilmont 1:fdd22bb7aa52 330 }
emilmont 1:fdd22bb7aa52 331
emilmont 1:fdd22bb7aa52 332 /* process odd column samples */
emilmont 1:fdd22bb7aa52 333 colCnt = numColsA % 0x4u;
emilmont 1:fdd22bb7aa52 334
emilmont 1:fdd22bb7aa52 335 while(colCnt > 0u)
emilmont 1:fdd22bb7aa52 336 {
emilmont 1:fdd22bb7aa52 337 /* c(m,n) = a(1,1)*b(1,1) + a(1,2) * b(2,1) + .... + a(m,p)*b(p,n) */
emilmont 1:fdd22bb7aa52 338 sum += (q31_t) (*pInA++) * (*pInB++);
emilmont 1:fdd22bb7aa52 339
emilmont 1:fdd22bb7aa52 340 colCnt--;
emilmont 1:fdd22bb7aa52 341 }
emilmont 1:fdd22bb7aa52 342
emilmont 1:fdd22bb7aa52 343 /* Saturate and store the result in the destination buffer */
emilmont 1:fdd22bb7aa52 344 *px = (q15_t) (sum >> 15);
emilmont 1:fdd22bb7aa52 345 px++;
emilmont 1:fdd22bb7aa52 346
emilmont 1:fdd22bb7aa52 347 /* Decrement the column loop counter */
emilmont 1:fdd22bb7aa52 348 col--;
emilmont 1:fdd22bb7aa52 349
emilmont 1:fdd22bb7aa52 350 } while(col > 0u);
emilmont 1:fdd22bb7aa52 351
emilmont 1:fdd22bb7aa52 352 i = i + numColsA;
emilmont 1:fdd22bb7aa52 353
emilmont 1:fdd22bb7aa52 354 /* Decrement the row loop counter */
emilmont 1:fdd22bb7aa52 355 row--;
emilmont 1:fdd22bb7aa52 356
emilmont 1:fdd22bb7aa52 357 } while(row > 0u);
emilmont 1:fdd22bb7aa52 358
emilmont 1:fdd22bb7aa52 359 /* set status as ARM_MATH_SUCCESS */
emilmont 1:fdd22bb7aa52 360 status = ARM_MATH_SUCCESS;
emilmont 1:fdd22bb7aa52 361 }
emilmont 1:fdd22bb7aa52 362
emilmont 1:fdd22bb7aa52 363 /* Return to application */
emilmont 1:fdd22bb7aa52 364 return (status);
emilmont 1:fdd22bb7aa52 365 }
emilmont 1:fdd22bb7aa52 366
emilmont 1:fdd22bb7aa52 367 /**
emilmont 1:fdd22bb7aa52 368 * @} end of MatrixMult group
emilmont 1:fdd22bb7aa52 369 */