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@1:fdd22bb7aa52, 2012-11-28 (annotated)

Committer:
emilmont
Date:
Wed Nov 28 12:30:09 2012 +0000
Revision:
1:fdd22bb7aa52
Child:
2:da51fb522205
DSP library code

Who changed what in which revision?

UserRevisionLine numberNew contents of line
emilmont 1:fdd22bb7aa52 1 /* ----------------------------------------------------------------------
emilmont 1:fdd22bb7aa52 2 * Copyright (C) 2010 ARM Limited. All rights reserved.
emilmont 1:fdd22bb7aa52 3 *
emilmont 1:fdd22bb7aa52 4 * $Date: 15. February 2012
emilmont 1:fdd22bb7aa52 5 * $Revision: V1.1.0
emilmont 1:fdd22bb7aa52 6 *
emilmont 1:fdd22bb7aa52 7 * Project: CMSIS DSP Library
emilmont 1:fdd22bb7aa52 8 * Title: arm_mat_mult_fast_q15.c
emilmont 1:fdd22bb7aa52 9 *
emilmont 1:fdd22bb7aa52 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 *
emilmont 1:fdd22bb7aa52 14 * Version 1.1.0 2012/02/15
emilmont 1:fdd22bb7aa52 15 * Updated with more optimizations, bug fixes and minor API changes.
emilmont 1:fdd22bb7aa52 16 *
emilmont 1:fdd22bb7aa52 17 * Version 1.0.10 2011/7/15
emilmont 1:fdd22bb7aa52 18 * Big Endian support added and Merged M0 and M3/M4 Source code.
emilmont 1:fdd22bb7aa52 19 *
emilmont 1:fdd22bb7aa52 20 * Version 1.0.3 2010/11/29
emilmont 1:fdd22bb7aa52 21 * Re-organized the CMSIS folders and updated documentation.
emilmont 1:fdd22bb7aa52 22 *
emilmont 1:fdd22bb7aa52 23 * Version 1.0.2 2010/11/11
emilmont 1:fdd22bb7aa52 24 * Documentation updated.
emilmont 1:fdd22bb7aa52 25 *
emilmont 1:fdd22bb7aa52 26 * Version 1.0.1 2010/10/05
emilmont 1:fdd22bb7aa52 27 * Production release and review comments incorporated.
emilmont 1:fdd22bb7aa52 28 *
emilmont 1:fdd22bb7aa52 29 * Version 1.0.0 2010/09/20
emilmont 1:fdd22bb7aa52 30 * Production release and review comments incorporated.
emilmont 1:fdd22bb7aa52 31 * -------------------------------------------------------------------- */
emilmont 1:fdd22bb7aa52 32
emilmont 1:fdd22bb7aa52 33 #include "arm_math.h"
emilmont 1:fdd22bb7aa52 34
emilmont 1:fdd22bb7aa52 35 /**
emilmont 1:fdd22bb7aa52 36 * @ingroup groupMatrix
emilmont 1:fdd22bb7aa52 37 */
emilmont 1:fdd22bb7aa52 38
emilmont 1:fdd22bb7aa52 39 /**
emilmont 1:fdd22bb7aa52 40 * @addtogroup MatrixMult
emilmont 1:fdd22bb7aa52 41 * @{
emilmont 1:fdd22bb7aa52 42 */
emilmont 1:fdd22bb7aa52 43
emilmont 1:fdd22bb7aa52 44
emilmont 1:fdd22bb7aa52 45 /**
emilmont 1:fdd22bb7aa52 46 * @brief Q15 matrix multiplication (fast variant) for Cortex-M3 and Cortex-M4
emilmont 1:fdd22bb7aa52 47 * @param[in] *pSrcA points to the first input matrix structure
emilmont 1:fdd22bb7aa52 48 * @param[in] *pSrcB points to the second input matrix structure
emilmont 1:fdd22bb7aa52 49 * @param[out] *pDst points to output matrix structure
emilmont 1:fdd22bb7aa52 50 * @param[in] *pState points to the array for storing intermediate results
emilmont 1:fdd22bb7aa52 51 * @return The function returns either
emilmont 1:fdd22bb7aa52 52 * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.
emilmont 1:fdd22bb7aa52 53 *
emilmont 1:fdd22bb7aa52 54 * @details
emilmont 1:fdd22bb7aa52 55 * <b>Scaling and Overflow Behavior:</b>
emilmont 1:fdd22bb7aa52 56 *
emilmont 1:fdd22bb7aa52 57 * \par
emilmont 1:fdd22bb7aa52 58 * The difference between the function arm_mat_mult_q15() and this fast variant is that
emilmont 1:fdd22bb7aa52 59 * the fast variant use a 32-bit rather than a 64-bit accumulator.
emilmont 1:fdd22bb7aa52 60 * The result of each 1.15 x 1.15 multiplication is truncated to
emilmont 1:fdd22bb7aa52 61 * 2.30 format. These intermediate results are accumulated in a 32-bit register in 2.30
emilmont 1:fdd22bb7aa52 62 * format. Finally, the accumulator is saturated and converted to a 1.15 result.
emilmont 1:fdd22bb7aa52 63 *
emilmont 1:fdd22bb7aa52 64 * \par
emilmont 1:fdd22bb7aa52 65 * The fast version has the same overflow behavior as the standard version but provides
emilmont 1:fdd22bb7aa52 66 * less precision since it discards the low 16 bits of each multiplication result.
emilmont 1:fdd22bb7aa52 67 * In order to avoid overflows completely the input signals must be scaled down.
emilmont 1:fdd22bb7aa52 68 * Scale down one of the input matrices by log2(numColsA) bits to
emilmont 1:fdd22bb7aa52 69 * avoid overflows, as a total of numColsA additions are computed internally for each
emilmont 1:fdd22bb7aa52 70 * output element.
emilmont 1:fdd22bb7aa52 71 *
emilmont 1:fdd22bb7aa52 72 * \par
emilmont 1:fdd22bb7aa52 73 * See <code>arm_mat_mult_q15()</code> for a slower implementation of this function
emilmont 1:fdd22bb7aa52 74 * which uses 64-bit accumulation to provide higher precision.
emilmont 1:fdd22bb7aa52 75 */
emilmont 1:fdd22bb7aa52 76
emilmont 1:fdd22bb7aa52 77 arm_status arm_mat_mult_fast_q15(
emilmont 1:fdd22bb7aa52 78 const arm_matrix_instance_q15 * pSrcA,
emilmont 1:fdd22bb7aa52 79 const arm_matrix_instance_q15 * pSrcB,
emilmont 1:fdd22bb7aa52 80 arm_matrix_instance_q15 * pDst,
emilmont 1:fdd22bb7aa52 81 q15_t * pState)
emilmont 1:fdd22bb7aa52 82 {
emilmont 1:fdd22bb7aa52 83 q31_t sum; /* accumulator */
emilmont 1:fdd22bb7aa52 84 q15_t *pSrcBT = pState; /* input data matrix pointer for transpose */
emilmont 1:fdd22bb7aa52 85 q15_t *pInA = pSrcA->pData; /* input data matrix pointer A of Q15 type */
emilmont 1:fdd22bb7aa52 86 q15_t *pInB = pSrcB->pData; /* input data matrix pointer B of Q15 type */
emilmont 1:fdd22bb7aa52 87 q15_t *px; /* Temporary output data matrix pointer */
emilmont 1:fdd22bb7aa52 88 uint16_t numRowsA = pSrcA->numRows; /* number of rows of input matrix A */
emilmont 1:fdd22bb7aa52 89 uint16_t numColsB = pSrcB->numCols; /* number of columns of input matrix B */
emilmont 1:fdd22bb7aa52 90 uint16_t numColsA = pSrcA->numCols; /* number of columns of input matrix A */
emilmont 1:fdd22bb7aa52 91 uint16_t numRowsB = pSrcB->numRows; /* number of rows of input matrix A */
emilmont 1:fdd22bb7aa52 92 uint16_t col, i = 0u, row = numRowsB, colCnt; /* loop counters */
emilmont 1:fdd22bb7aa52 93 arm_status status; /* status of matrix multiplication */
emilmont 1:fdd22bb7aa52 94
emilmont 1:fdd22bb7aa52 95 #ifndef UNALIGNED_SUPPORT_DISABLE
emilmont 1:fdd22bb7aa52 96
emilmont 1:fdd22bb7aa52 97 q31_t in; /* Temporary variable to hold the input value */
emilmont 1:fdd22bb7aa52 98 q31_t inA1, inA2, inB1, inB2;
emilmont 1:fdd22bb7aa52 99
emilmont 1:fdd22bb7aa52 100 #else
emilmont 1:fdd22bb7aa52 101
emilmont 1:fdd22bb7aa52 102 q15_t in; /* Temporary variable to hold the input value */
emilmont 1:fdd22bb7aa52 103 q15_t inA1, inA2, inB1, inB2;
emilmont 1:fdd22bb7aa52 104
emilmont 1:fdd22bb7aa52 105 #endif /* #ifndef UNALIGNED_SUPPORT_DISABLE */
emilmont 1:fdd22bb7aa52 106
emilmont 1:fdd22bb7aa52 107 #ifdef ARM_MATH_MATRIX_CHECK
emilmont 1:fdd22bb7aa52 108 /* Check for matrix mismatch condition */
emilmont 1:fdd22bb7aa52 109 if((pSrcA->numCols != pSrcB->numRows) ||
emilmont 1:fdd22bb7aa52 110 (pSrcA->numRows != pDst->numRows) || (pSrcB->numCols != pDst->numCols))
emilmont 1:fdd22bb7aa52 111 {
emilmont 1:fdd22bb7aa52 112 /* Set status as ARM_MATH_SIZE_MISMATCH */
emilmont 1:fdd22bb7aa52 113 status = ARM_MATH_SIZE_MISMATCH;
emilmont 1:fdd22bb7aa52 114 }
emilmont 1:fdd22bb7aa52 115 else
emilmont 1:fdd22bb7aa52 116 #endif
emilmont 1:fdd22bb7aa52 117 {
emilmont 1:fdd22bb7aa52 118 /* Matrix transpose */
emilmont 1:fdd22bb7aa52 119 do
emilmont 1:fdd22bb7aa52 120 {
emilmont 1:fdd22bb7aa52 121 /* Apply loop unrolling and exchange the columns with row elements */
emilmont 1:fdd22bb7aa52 122 col = numColsB >> 2;
emilmont 1:fdd22bb7aa52 123
emilmont 1:fdd22bb7aa52 124 /* The pointer px is set to starting address of the column being processed */
emilmont 1:fdd22bb7aa52 125 px = pSrcBT + i;
emilmont 1:fdd22bb7aa52 126
emilmont 1:fdd22bb7aa52 127 /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
emilmont 1:fdd22bb7aa52 128 ** a second loop below computes the remaining 1 to 3 samples. */
emilmont 1:fdd22bb7aa52 129 while(col > 0u)
emilmont 1:fdd22bb7aa52 130 {
emilmont 1:fdd22bb7aa52 131 #ifndef UNALIGNED_SUPPORT_DISABLE
emilmont 1:fdd22bb7aa52 132 /* Read two elements from the row */
emilmont 1:fdd22bb7aa52 133 in = *__SIMD32(pInB)++;
emilmont 1:fdd22bb7aa52 134
emilmont 1:fdd22bb7aa52 135 /* Unpack and store one element in the destination */
emilmont 1:fdd22bb7aa52 136 #ifndef ARM_MATH_BIG_ENDIAN
emilmont 1:fdd22bb7aa52 137
emilmont 1:fdd22bb7aa52 138 *px = (q15_t) in;
emilmont 1:fdd22bb7aa52 139
emilmont 1:fdd22bb7aa52 140 #else
emilmont 1:fdd22bb7aa52 141
emilmont 1:fdd22bb7aa52 142 *px = (q15_t) ((in & (q31_t) 0xffff0000) >> 16);
emilmont 1:fdd22bb7aa52 143
emilmont 1:fdd22bb7aa52 144 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
emilmont 1:fdd22bb7aa52 145
emilmont 1:fdd22bb7aa52 146 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 147 px += numRowsB;
emilmont 1:fdd22bb7aa52 148
emilmont 1:fdd22bb7aa52 149 /* Unpack and store the second element in the destination */
emilmont 1:fdd22bb7aa52 150 #ifndef ARM_MATH_BIG_ENDIAN
emilmont 1:fdd22bb7aa52 151
emilmont 1:fdd22bb7aa52 152 *px = (q15_t) ((in & (q31_t) 0xffff0000) >> 16);
emilmont 1:fdd22bb7aa52 153
emilmont 1:fdd22bb7aa52 154 #else
emilmont 1:fdd22bb7aa52 155
emilmont 1:fdd22bb7aa52 156 *px = (q15_t) in;
emilmont 1:fdd22bb7aa52 157
emilmont 1:fdd22bb7aa52 158 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
emilmont 1:fdd22bb7aa52 159
emilmont 1:fdd22bb7aa52 160 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 161 px += numRowsB;
emilmont 1:fdd22bb7aa52 162
emilmont 1:fdd22bb7aa52 163 /* Read two elements from the row */
emilmont 1:fdd22bb7aa52 164 in = *__SIMD32(pInB)++;
emilmont 1:fdd22bb7aa52 165
emilmont 1:fdd22bb7aa52 166 /* Unpack and store one element in the destination */
emilmont 1:fdd22bb7aa52 167 #ifndef ARM_MATH_BIG_ENDIAN
emilmont 1:fdd22bb7aa52 168
emilmont 1:fdd22bb7aa52 169 *px = (q15_t) in;
emilmont 1:fdd22bb7aa52 170
emilmont 1:fdd22bb7aa52 171 #else
emilmont 1:fdd22bb7aa52 172
emilmont 1:fdd22bb7aa52 173 *px = (q15_t) ((in & (q31_t) 0xffff0000) >> 16);
emilmont 1:fdd22bb7aa52 174
emilmont 1:fdd22bb7aa52 175 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
emilmont 1:fdd22bb7aa52 176
emilmont 1:fdd22bb7aa52 177 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 178 px += numRowsB;
emilmont 1:fdd22bb7aa52 179
emilmont 1:fdd22bb7aa52 180 /* Unpack and store the second element in the destination */
emilmont 1:fdd22bb7aa52 181
emilmont 1:fdd22bb7aa52 182 #ifndef ARM_MATH_BIG_ENDIAN
emilmont 1:fdd22bb7aa52 183
emilmont 1:fdd22bb7aa52 184 *px = (q15_t) ((in & (q31_t) 0xffff0000) >> 16);
emilmont 1:fdd22bb7aa52 185
emilmont 1:fdd22bb7aa52 186 #else
emilmont 1:fdd22bb7aa52 187
emilmont 1:fdd22bb7aa52 188 *px = (q15_t) in;
emilmont 1:fdd22bb7aa52 189
emilmont 1:fdd22bb7aa52 190 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
emilmont 1:fdd22bb7aa52 191
emilmont 1:fdd22bb7aa52 192 #else
emilmont 1:fdd22bb7aa52 193
emilmont 1:fdd22bb7aa52 194 /* Read one element from the row */
emilmont 1:fdd22bb7aa52 195 in = *pInB++;
emilmont 1:fdd22bb7aa52 196
emilmont 1:fdd22bb7aa52 197 /* Store one element in the destination */
emilmont 1:fdd22bb7aa52 198 *px = in;
emilmont 1:fdd22bb7aa52 199
emilmont 1:fdd22bb7aa52 200 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 201 px += numRowsB;
emilmont 1:fdd22bb7aa52 202
emilmont 1:fdd22bb7aa52 203 /* Read one element from the row */
emilmont 1:fdd22bb7aa52 204 in = *pInB++;
emilmont 1:fdd22bb7aa52 205
emilmont 1:fdd22bb7aa52 206 /* Store one element in the destination */
emilmont 1:fdd22bb7aa52 207 *px = in;
emilmont 1:fdd22bb7aa52 208
emilmont 1:fdd22bb7aa52 209 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 210 px += numRowsB;
emilmont 1:fdd22bb7aa52 211
emilmont 1:fdd22bb7aa52 212 /* Read one element from the row */
emilmont 1:fdd22bb7aa52 213 in = *pInB++;
emilmont 1:fdd22bb7aa52 214
emilmont 1:fdd22bb7aa52 215 /* Store one element in the destination */
emilmont 1:fdd22bb7aa52 216 *px = in;
emilmont 1:fdd22bb7aa52 217
emilmont 1:fdd22bb7aa52 218 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 219 px += numRowsB;
emilmont 1:fdd22bb7aa52 220
emilmont 1:fdd22bb7aa52 221 /* Read one element from the row */
emilmont 1:fdd22bb7aa52 222 in = *pInB++;
emilmont 1:fdd22bb7aa52 223
emilmont 1:fdd22bb7aa52 224 /* Store one element in the destination */
emilmont 1:fdd22bb7aa52 225 *px = in;
emilmont 1:fdd22bb7aa52 226
emilmont 1:fdd22bb7aa52 227 #endif /* #ifndef UNALIGNED_SUPPORT_DISABLE */
emilmont 1:fdd22bb7aa52 228
emilmont 1:fdd22bb7aa52 229 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 230 px += numRowsB;
emilmont 1:fdd22bb7aa52 231
emilmont 1:fdd22bb7aa52 232 /* Decrement the column loop counter */
emilmont 1:fdd22bb7aa52 233 col--;
emilmont 1:fdd22bb7aa52 234 }
emilmont 1:fdd22bb7aa52 235
emilmont 1:fdd22bb7aa52 236 /* If the columns of pSrcB is not a multiple of 4, compute any remaining output samples here.
emilmont 1:fdd22bb7aa52 237 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 238 col = numColsB % 0x4u;
emilmont 1:fdd22bb7aa52 239
emilmont 1:fdd22bb7aa52 240 while(col > 0u)
emilmont 1:fdd22bb7aa52 241 {
emilmont 1:fdd22bb7aa52 242 /* Read and store the input element in the destination */
emilmont 1:fdd22bb7aa52 243 *px = *pInB++;
emilmont 1:fdd22bb7aa52 244
emilmont 1:fdd22bb7aa52 245 /* Update the pointer px to point to the next row of the transposed matrix */
emilmont 1:fdd22bb7aa52 246 px += numRowsB;
emilmont 1:fdd22bb7aa52 247
emilmont 1:fdd22bb7aa52 248 /* Decrement the column loop counter */
emilmont 1:fdd22bb7aa52 249 col--;
emilmont 1:fdd22bb7aa52 250 }
emilmont 1:fdd22bb7aa52 251
emilmont 1:fdd22bb7aa52 252 i++;
emilmont 1:fdd22bb7aa52 253
emilmont 1:fdd22bb7aa52 254 /* Decrement the row loop counter */
emilmont 1:fdd22bb7aa52 255 row--;
emilmont 1:fdd22bb7aa52 256
emilmont 1:fdd22bb7aa52 257 } while(row > 0u);
emilmont 1:fdd22bb7aa52 258
emilmont 1:fdd22bb7aa52 259 /* Reset the variables for the usage in the following multiplication process */
emilmont 1:fdd22bb7aa52 260 row = numRowsA;
emilmont 1:fdd22bb7aa52 261 i = 0u;
emilmont 1:fdd22bb7aa52 262 px = pDst->pData;
emilmont 1:fdd22bb7aa52 263
emilmont 1:fdd22bb7aa52 264 /* The following loop performs the dot-product of each row in pSrcA with each column in pSrcB */
emilmont 1:fdd22bb7aa52 265 /* row loop */
emilmont 1:fdd22bb7aa52 266 do
emilmont 1:fdd22bb7aa52 267 {
emilmont 1:fdd22bb7aa52 268 /* For every row wise process, the column loop counter is to be initiated */
emilmont 1:fdd22bb7aa52 269 col = numColsB;
emilmont 1:fdd22bb7aa52 270
emilmont 1:fdd22bb7aa52 271 /* For every row wise process, the pIn2 pointer is set
emilmont 1:fdd22bb7aa52 272 ** to the starting address of the transposed pSrcB data */
emilmont 1:fdd22bb7aa52 273 pInB = pSrcBT;
emilmont 1:fdd22bb7aa52 274
emilmont 1:fdd22bb7aa52 275 /* column loop */
emilmont 1:fdd22bb7aa52 276 do
emilmont 1:fdd22bb7aa52 277 {
emilmont 1:fdd22bb7aa52 278 /* Set the variable sum, that acts as accumulator, to zero */
emilmont 1:fdd22bb7aa52 279 sum = 0;
emilmont 1:fdd22bb7aa52 280
emilmont 1:fdd22bb7aa52 281 /* Apply loop unrolling and compute 2 MACs simultaneously. */
emilmont 1:fdd22bb7aa52 282 colCnt = numColsA >> 2;
emilmont 1:fdd22bb7aa52 283
emilmont 1:fdd22bb7aa52 284 /* Initiate the pointer pIn1 to point to the starting address of the column being processed */
emilmont 1:fdd22bb7aa52 285 pInA = pSrcA->pData + i;
emilmont 1:fdd22bb7aa52 286
emilmont 1:fdd22bb7aa52 287 /* matrix multiplication */
emilmont 1:fdd22bb7aa52 288 while(colCnt > 0u)
emilmont 1:fdd22bb7aa52 289 {
emilmont 1:fdd22bb7aa52 290 /* c(m,n) = a(1,1)*b(1,1) + a(1,2) * b(2,1) + .... + a(m,p)*b(p,n) */
emilmont 1:fdd22bb7aa52 291 #ifndef UNALIGNED_SUPPORT_DISABLE
emilmont 1:fdd22bb7aa52 292
emilmont 1:fdd22bb7aa52 293 inA1 = *__SIMD32(pInA)++;
emilmont 1:fdd22bb7aa52 294 inB1 = *__SIMD32(pInB)++;
emilmont 1:fdd22bb7aa52 295 inA2 = *__SIMD32(pInA)++;
emilmont 1:fdd22bb7aa52 296 inB2 = *__SIMD32(pInB)++;
emilmont 1:fdd22bb7aa52 297
emilmont 1:fdd22bb7aa52 298 sum = __SMLAD(inA1, inB1, sum);
emilmont 1:fdd22bb7aa52 299 sum = __SMLAD(inA2, inB2, sum);
emilmont 1:fdd22bb7aa52 300
emilmont 1:fdd22bb7aa52 301 #else
emilmont 1:fdd22bb7aa52 302
emilmont 1:fdd22bb7aa52 303 inA1 = *pInA++;
emilmont 1:fdd22bb7aa52 304 inB1 = *pInB++;
emilmont 1:fdd22bb7aa52 305 inA2 = *pInA++;
emilmont 1:fdd22bb7aa52 306 sum += inA1 * inB1;
emilmont 1:fdd22bb7aa52 307 inB2 = *pInB++;
emilmont 1:fdd22bb7aa52 308
emilmont 1:fdd22bb7aa52 309 inA1 = *pInA++;
emilmont 1:fdd22bb7aa52 310 inB1 = *pInB++;
emilmont 1:fdd22bb7aa52 311 sum += inA2 * inB2;
emilmont 1:fdd22bb7aa52 312 inA2 = *pInA++;
emilmont 1:fdd22bb7aa52 313 inB2 = *pInB++;
emilmont 1:fdd22bb7aa52 314
emilmont 1:fdd22bb7aa52 315 sum += inA1 * inB1;
emilmont 1:fdd22bb7aa52 316 sum += inA2 * inB2;
emilmont 1:fdd22bb7aa52 317
emilmont 1:fdd22bb7aa52 318 #endif /* #ifndef UNALIGNED_SUPPORT_DISABLE */
emilmont 1:fdd22bb7aa52 319
emilmont 1:fdd22bb7aa52 320 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 321 colCnt--;
emilmont 1:fdd22bb7aa52 322 }
emilmont 1:fdd22bb7aa52 323
emilmont 1:fdd22bb7aa52 324 /* process odd column samples */
emilmont 1:fdd22bb7aa52 325 colCnt = numColsA % 0x4u;
emilmont 1:fdd22bb7aa52 326
emilmont 1:fdd22bb7aa52 327 while(colCnt > 0u)
emilmont 1:fdd22bb7aa52 328 {
emilmont 1:fdd22bb7aa52 329 /* c(m,n) = a(1,1)*b(1,1) + a(1,2) * b(2,1) + .... + a(m,p)*b(p,n) */
emilmont 1:fdd22bb7aa52 330 sum += (q31_t) (*pInA++) * (*pInB++);
emilmont 1:fdd22bb7aa52 331
emilmont 1:fdd22bb7aa52 332 colCnt--;
emilmont 1:fdd22bb7aa52 333 }
emilmont 1:fdd22bb7aa52 334
emilmont 1:fdd22bb7aa52 335 /* Saturate and store the result in the destination buffer */
emilmont 1:fdd22bb7aa52 336 *px = (q15_t) (sum >> 15);
emilmont 1:fdd22bb7aa52 337 px++;
emilmont 1:fdd22bb7aa52 338
emilmont 1:fdd22bb7aa52 339 /* Decrement the column loop counter */
emilmont 1:fdd22bb7aa52 340 col--;
emilmont 1:fdd22bb7aa52 341
emilmont 1:fdd22bb7aa52 342 } while(col > 0u);
emilmont 1:fdd22bb7aa52 343
emilmont 1:fdd22bb7aa52 344 i = i + numColsA;
emilmont 1:fdd22bb7aa52 345
emilmont 1:fdd22bb7aa52 346 /* Decrement the row loop counter */
emilmont 1:fdd22bb7aa52 347 row--;
emilmont 1:fdd22bb7aa52 348
emilmont 1:fdd22bb7aa52 349 } while(row > 0u);
emilmont 1:fdd22bb7aa52 350
emilmont 1:fdd22bb7aa52 351 /* set status as ARM_MATH_SUCCESS */
emilmont 1:fdd22bb7aa52 352 status = ARM_MATH_SUCCESS;
emilmont 1:fdd22bb7aa52 353 }
emilmont 1:fdd22bb7aa52 354
emilmont 1:fdd22bb7aa52 355 /* Return to application */
emilmont 1:fdd22bb7aa52 356 return (status);
emilmont 1:fdd22bb7aa52 357 }
emilmont 1:fdd22bb7aa52 358
emilmont 1:fdd22bb7aa52 359 /**
emilmont 1:fdd22bb7aa52 360 * @} end of MatrixMult group
emilmont 1:fdd22bb7aa52 361 */