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/FilteringFunctions/arm_correlate_fast_q31.c@2:da51fb522205, 2013-05-30 (annotated)

Committer:
emilmont
Date:
Thu May 30 17:10:11 2013 +0100
Revision:
2:da51fb522205
Parent:
1:fdd22bb7aa52
Child:
3:7a284390b0ce
Keep "cmsis-dsp" module in synch with its source

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 2:da51fb522205 5 * $Revision: V1.1.0
emilmont 1:fdd22bb7aa52 6 *
emilmont 2:da51fb522205 7 * Project: CMSIS DSP Library
emilmont 2:da51fb522205 8 * Title: arm_correlate_fast_q31.c
emilmont 1:fdd22bb7aa52 9 *
emilmont 2:da51fb522205 10 * Description: Fast Q31 Correlation.
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 1.0.11 2011/10/18
emilmont 1:fdd22bb7aa52 18 * Bug Fix in conv, correlation, partial convolution.
emilmont 1:fdd22bb7aa52 19 *
emilmont 1:fdd22bb7aa52 20 * Version 1.0.10 2011/7/15
emilmont 1:fdd22bb7aa52 21 * Big Endian support added and Merged M0 and M3/M4 Source code.
emilmont 1:fdd22bb7aa52 22 *
emilmont 1:fdd22bb7aa52 23 * Version 1.0.3 2010/11/29
emilmont 1:fdd22bb7aa52 24 * Re-organized the CMSIS folders and updated documentation.
emilmont 1:fdd22bb7aa52 25 *
emilmont 1:fdd22bb7aa52 26 * Version 1.0.2 2010/11/11
emilmont 1:fdd22bb7aa52 27 * Documentation updated.
emilmont 1:fdd22bb7aa52 28 *
emilmont 1:fdd22bb7aa52 29 * Version 1.0.1 2010/10/05
emilmont 1:fdd22bb7aa52 30 * Production release and review comments incorporated.
emilmont 1:fdd22bb7aa52 31 *
emilmont 1:fdd22bb7aa52 32 * Version 1.0.0 2010/09/20
emilmont 1:fdd22bb7aa52 33 * Production release and review comments incorporated.
emilmont 1:fdd22bb7aa52 34 * -------------------------------------------------------------------- */
emilmont 1:fdd22bb7aa52 35
emilmont 1:fdd22bb7aa52 36 #include "arm_math.h"
emilmont 1:fdd22bb7aa52 37
emilmont 1:fdd22bb7aa52 38 /**
emilmont 1:fdd22bb7aa52 39 * @ingroup groupFilters
emilmont 1:fdd22bb7aa52 40 */
emilmont 1:fdd22bb7aa52 41
emilmont 1:fdd22bb7aa52 42 /**
emilmont 1:fdd22bb7aa52 43 * @addtogroup Corr
emilmont 1:fdd22bb7aa52 44 * @{
emilmont 1:fdd22bb7aa52 45 */
emilmont 1:fdd22bb7aa52 46
emilmont 1:fdd22bb7aa52 47 /**
emilmont 1:fdd22bb7aa52 48 * @brief Correlation of Q31 sequences (fast version) for Cortex-M3 and Cortex-M4.
emilmont 1:fdd22bb7aa52 49 * @param[in] *pSrcA points to the first input sequence.
emilmont 1:fdd22bb7aa52 50 * @param[in] srcALen length of the first input sequence.
emilmont 1:fdd22bb7aa52 51 * @param[in] *pSrcB points to the second input sequence.
emilmont 1:fdd22bb7aa52 52 * @param[in] srcBLen length of the second input sequence.
emilmont 1:fdd22bb7aa52 53 * @param[out] *pDst points to the location where the output result is written. Length 2 * max(srcALen, srcBLen) - 1.
emilmont 1:fdd22bb7aa52 54 * @return none.
emilmont 1:fdd22bb7aa52 55 *
emilmont 1:fdd22bb7aa52 56 * @details
emilmont 1:fdd22bb7aa52 57 * <b>Scaling and Overflow Behavior:</b>
emilmont 1:fdd22bb7aa52 58 *
emilmont 1:fdd22bb7aa52 59 * \par
emilmont 1:fdd22bb7aa52 60 * This function is optimized for speed at the expense of fixed-point precision and overflow protection.
emilmont 1:fdd22bb7aa52 61 * The result of each 1.31 x 1.31 multiplication is truncated to 2.30 format.
emilmont 1:fdd22bb7aa52 62 * These intermediate results are accumulated in a 32-bit register in 2.30 format.
emilmont 1:fdd22bb7aa52 63 * Finally, the accumulator is saturated and converted to a 1.31 result.
emilmont 1:fdd22bb7aa52 64 *
emilmont 1:fdd22bb7aa52 65 * \par
emilmont 1:fdd22bb7aa52 66 * The fast version has the same overflow behavior as the standard version but provides less precision since it discards the low 32 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 * The input signals should be scaled down to avoid intermediate overflows.
emilmont 1:fdd22bb7aa52 69 * Scale down one of the inputs by 1/min(srcALen, srcBLen)to avoid overflows since a
emilmont 1:fdd22bb7aa52 70 * maximum of min(srcALen, srcBLen) number of additions is carried internally.
emilmont 1:fdd22bb7aa52 71 *
emilmont 1:fdd22bb7aa52 72 * \par
emilmont 1:fdd22bb7aa52 73 * See <code>arm_correlate_q31()</code> for a slower implementation of this function which uses 64-bit accumulation to provide higher precision.
emilmont 1:fdd22bb7aa52 74 */
emilmont 1:fdd22bb7aa52 75
emilmont 1:fdd22bb7aa52 76 void arm_correlate_fast_q31(
emilmont 1:fdd22bb7aa52 77 q31_t * pSrcA,
emilmont 1:fdd22bb7aa52 78 uint32_t srcALen,
emilmont 1:fdd22bb7aa52 79 q31_t * pSrcB,
emilmont 1:fdd22bb7aa52 80 uint32_t srcBLen,
emilmont 1:fdd22bb7aa52 81 q31_t * pDst)
emilmont 1:fdd22bb7aa52 82 {
emilmont 1:fdd22bb7aa52 83 q31_t *pIn1; /* inputA pointer */
emilmont 1:fdd22bb7aa52 84 q31_t *pIn2; /* inputB pointer */
emilmont 1:fdd22bb7aa52 85 q31_t *pOut = pDst; /* output pointer */
emilmont 1:fdd22bb7aa52 86 q31_t *px; /* Intermediate inputA pointer */
emilmont 1:fdd22bb7aa52 87 q31_t *py; /* Intermediate inputB pointer */
emilmont 1:fdd22bb7aa52 88 q31_t *pSrc1; /* Intermediate pointers */
emilmont 1:fdd22bb7aa52 89 q31_t sum, acc0, acc1, acc2, acc3; /* Accumulators */
emilmont 1:fdd22bb7aa52 90 q31_t x0, x1, x2, x3, c0; /* temporary variables for holding input and coefficient values */
emilmont 1:fdd22bb7aa52 91 uint32_t j, k = 0u, count, blkCnt, outBlockSize, blockSize1, blockSize2, blockSize3; /* loop counter */
emilmont 1:fdd22bb7aa52 92 int32_t inc = 1; /* Destination address modifier */
emilmont 1:fdd22bb7aa52 93
emilmont 1:fdd22bb7aa52 94
emilmont 1:fdd22bb7aa52 95 /* The algorithm implementation is based on the lengths of the inputs. */
emilmont 1:fdd22bb7aa52 96 /* srcB is always made to slide across srcA. */
emilmont 1:fdd22bb7aa52 97 /* So srcBLen is always considered as shorter or equal to srcALen */
emilmont 1:fdd22bb7aa52 98 if(srcALen >= srcBLen)
emilmont 1:fdd22bb7aa52 99 {
emilmont 1:fdd22bb7aa52 100 /* Initialization of inputA pointer */
emilmont 1:fdd22bb7aa52 101 pIn1 = (pSrcA);
emilmont 1:fdd22bb7aa52 102
emilmont 1:fdd22bb7aa52 103 /* Initialization of inputB pointer */
emilmont 1:fdd22bb7aa52 104 pIn2 = (pSrcB);
emilmont 1:fdd22bb7aa52 105
emilmont 1:fdd22bb7aa52 106 /* Number of output samples is calculated */
emilmont 1:fdd22bb7aa52 107 outBlockSize = (2u * srcALen) - 1u;
emilmont 1:fdd22bb7aa52 108
emilmont 1:fdd22bb7aa52 109 /* When srcALen > srcBLen, zero padding is done to srcB
emilmont 1:fdd22bb7aa52 110 * to make their lengths equal.
emilmont 1:fdd22bb7aa52 111 * Instead, (outBlockSize - (srcALen + srcBLen - 1))
emilmont 1:fdd22bb7aa52 112 * number of output samples are made zero */
emilmont 1:fdd22bb7aa52 113 j = outBlockSize - (srcALen + (srcBLen - 1u));
emilmont 1:fdd22bb7aa52 114
emilmont 1:fdd22bb7aa52 115 /* Updating the pointer position to non zero value */
emilmont 1:fdd22bb7aa52 116 pOut += j;
emilmont 1:fdd22bb7aa52 117
emilmont 1:fdd22bb7aa52 118 }
emilmont 1:fdd22bb7aa52 119 else
emilmont 1:fdd22bb7aa52 120 {
emilmont 1:fdd22bb7aa52 121 /* Initialization of inputA pointer */
emilmont 1:fdd22bb7aa52 122 pIn1 = (pSrcB);
emilmont 1:fdd22bb7aa52 123
emilmont 1:fdd22bb7aa52 124 /* Initialization of inputB pointer */
emilmont 1:fdd22bb7aa52 125 pIn2 = (pSrcA);
emilmont 1:fdd22bb7aa52 126
emilmont 1:fdd22bb7aa52 127 /* srcBLen is always considered as shorter or equal to srcALen */
emilmont 1:fdd22bb7aa52 128 j = srcBLen;
emilmont 1:fdd22bb7aa52 129 srcBLen = srcALen;
emilmont 1:fdd22bb7aa52 130 srcALen = j;
emilmont 1:fdd22bb7aa52 131
emilmont 1:fdd22bb7aa52 132 /* CORR(x, y) = Reverse order(CORR(y, x)) */
emilmont 1:fdd22bb7aa52 133 /* Hence set the destination pointer to point to the last output sample */
emilmont 1:fdd22bb7aa52 134 pOut = pDst + ((srcALen + srcBLen) - 2u);
emilmont 1:fdd22bb7aa52 135
emilmont 1:fdd22bb7aa52 136 /* Destination address modifier is set to -1 */
emilmont 1:fdd22bb7aa52 137 inc = -1;
emilmont 1:fdd22bb7aa52 138
emilmont 1:fdd22bb7aa52 139 }
emilmont 1:fdd22bb7aa52 140
emilmont 1:fdd22bb7aa52 141 /* The function is internally
emilmont 1:fdd22bb7aa52 142 * divided into three parts according to the number of multiplications that has to be
emilmont 1:fdd22bb7aa52 143 * taken place between inputA samples and inputB samples. In the first part of the
emilmont 1:fdd22bb7aa52 144 * algorithm, the multiplications increase by one for every iteration.
emilmont 1:fdd22bb7aa52 145 * In the second part of the algorithm, srcBLen number of multiplications are done.
emilmont 1:fdd22bb7aa52 146 * In the third part of the algorithm, the multiplications decrease by one
emilmont 1:fdd22bb7aa52 147 * for every iteration.*/
emilmont 1:fdd22bb7aa52 148 /* The algorithm is implemented in three stages.
emilmont 1:fdd22bb7aa52 149 * The loop counters of each stage is initiated here. */
emilmont 1:fdd22bb7aa52 150 blockSize1 = srcBLen - 1u;
emilmont 1:fdd22bb7aa52 151 blockSize2 = srcALen - (srcBLen - 1u);
emilmont 1:fdd22bb7aa52 152 blockSize3 = blockSize1;
emilmont 1:fdd22bb7aa52 153
emilmont 1:fdd22bb7aa52 154 /* --------------------------
emilmont 1:fdd22bb7aa52 155 * Initializations of stage1
emilmont 1:fdd22bb7aa52 156 * -------------------------*/
emilmont 1:fdd22bb7aa52 157
emilmont 1:fdd22bb7aa52 158 /* sum = x[0] * y[srcBlen - 1]
emilmont 1:fdd22bb7aa52 159 * sum = x[0] * y[srcBlen - 2] + x[1] * y[srcBlen - 1]
emilmont 1:fdd22bb7aa52 160 * ....
emilmont 1:fdd22bb7aa52 161 * sum = x[0] * y[0] + x[1] * y[1] +...+ x[srcBLen - 1] * y[srcBLen - 1]
emilmont 1:fdd22bb7aa52 162 */
emilmont 1:fdd22bb7aa52 163
emilmont 1:fdd22bb7aa52 164 /* In this stage the MAC operations are increased by 1 for every iteration.
emilmont 1:fdd22bb7aa52 165 The count variable holds the number of MAC operations performed */
emilmont 1:fdd22bb7aa52 166 count = 1u;
emilmont 1:fdd22bb7aa52 167
emilmont 1:fdd22bb7aa52 168 /* Working pointer of inputA */
emilmont 1:fdd22bb7aa52 169 px = pIn1;
emilmont 1:fdd22bb7aa52 170
emilmont 1:fdd22bb7aa52 171 /* Working pointer of inputB */
emilmont 1:fdd22bb7aa52 172 pSrc1 = pIn2 + (srcBLen - 1u);
emilmont 1:fdd22bb7aa52 173 py = pSrc1;
emilmont 1:fdd22bb7aa52 174
emilmont 1:fdd22bb7aa52 175 /* ------------------------
emilmont 1:fdd22bb7aa52 176 * Stage1 process
emilmont 1:fdd22bb7aa52 177 * ----------------------*/
emilmont 1:fdd22bb7aa52 178
emilmont 1:fdd22bb7aa52 179 /* The first stage starts here */
emilmont 1:fdd22bb7aa52 180 while(blockSize1 > 0u)
emilmont 1:fdd22bb7aa52 181 {
emilmont 1:fdd22bb7aa52 182 /* Accumulator is made zero for every iteration */
emilmont 1:fdd22bb7aa52 183 sum = 0;
emilmont 1:fdd22bb7aa52 184
emilmont 1:fdd22bb7aa52 185 /* Apply loop unrolling and compute 4 MACs simultaneously. */
emilmont 1:fdd22bb7aa52 186 k = count >> 2;
emilmont 1:fdd22bb7aa52 187
emilmont 1:fdd22bb7aa52 188 /* First part of the processing with loop unrolling. Compute 4 MACs at a time.
emilmont 1:fdd22bb7aa52 189 ** a second loop below computes MACs for the remaining 1 to 3 samples. */
emilmont 1:fdd22bb7aa52 190 while(k > 0u)
emilmont 1:fdd22bb7aa52 191 {
emilmont 1:fdd22bb7aa52 192 /* x[0] * y[srcBLen - 4] */
emilmont 1:fdd22bb7aa52 193 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 194 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 195 /* x[1] * y[srcBLen - 3] */
emilmont 1:fdd22bb7aa52 196 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 197 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 198 /* x[2] * y[srcBLen - 2] */
emilmont 1:fdd22bb7aa52 199 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 200 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 201 /* x[3] * y[srcBLen - 1] */
emilmont 1:fdd22bb7aa52 202 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 203 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 204
emilmont 1:fdd22bb7aa52 205 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 206 k--;
emilmont 1:fdd22bb7aa52 207 }
emilmont 1:fdd22bb7aa52 208
emilmont 1:fdd22bb7aa52 209 /* If the count is not a multiple of 4, compute any remaining MACs here.
emilmont 1:fdd22bb7aa52 210 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 211 k = count % 0x4u;
emilmont 1:fdd22bb7aa52 212
emilmont 1:fdd22bb7aa52 213 while(k > 0u)
emilmont 1:fdd22bb7aa52 214 {
emilmont 1:fdd22bb7aa52 215 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 216 /* x[0] * y[srcBLen - 1] */
emilmont 1:fdd22bb7aa52 217 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 218 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 219
emilmont 1:fdd22bb7aa52 220 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 221 k--;
emilmont 1:fdd22bb7aa52 222 }
emilmont 1:fdd22bb7aa52 223
emilmont 1:fdd22bb7aa52 224 /* Store the result in the accumulator in the destination buffer. */
emilmont 1:fdd22bb7aa52 225 *pOut = sum << 1;
emilmont 1:fdd22bb7aa52 226 /* Destination pointer is updated according to the address modifier, inc */
emilmont 1:fdd22bb7aa52 227 pOut += inc;
emilmont 1:fdd22bb7aa52 228
emilmont 1:fdd22bb7aa52 229 /* Update the inputA and inputB pointers for next MAC calculation */
emilmont 1:fdd22bb7aa52 230 py = pSrc1 - count;
emilmont 1:fdd22bb7aa52 231 px = pIn1;
emilmont 1:fdd22bb7aa52 232
emilmont 1:fdd22bb7aa52 233 /* Increment the MAC count */
emilmont 1:fdd22bb7aa52 234 count++;
emilmont 1:fdd22bb7aa52 235
emilmont 1:fdd22bb7aa52 236 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 237 blockSize1--;
emilmont 1:fdd22bb7aa52 238 }
emilmont 1:fdd22bb7aa52 239
emilmont 1:fdd22bb7aa52 240 /* --------------------------
emilmont 1:fdd22bb7aa52 241 * Initializations of stage2
emilmont 1:fdd22bb7aa52 242 * ------------------------*/
emilmont 1:fdd22bb7aa52 243
emilmont 1:fdd22bb7aa52 244 /* sum = x[0] * y[0] + x[1] * y[1] +...+ x[srcBLen-1] * y[srcBLen-1]
emilmont 1:fdd22bb7aa52 245 * sum = x[1] * y[0] + x[2] * y[1] +...+ x[srcBLen] * y[srcBLen-1]
emilmont 1:fdd22bb7aa52 246 * ....
emilmont 1:fdd22bb7aa52 247 * sum = x[srcALen-srcBLen-2] * y[0] + x[srcALen-srcBLen-1] * y[1] +...+ x[srcALen-1] * y[srcBLen-1]
emilmont 1:fdd22bb7aa52 248 */
emilmont 1:fdd22bb7aa52 249
emilmont 1:fdd22bb7aa52 250 /* Working pointer of inputA */
emilmont 1:fdd22bb7aa52 251 px = pIn1;
emilmont 1:fdd22bb7aa52 252
emilmont 1:fdd22bb7aa52 253 /* Working pointer of inputB */
emilmont 1:fdd22bb7aa52 254 py = pIn2;
emilmont 1:fdd22bb7aa52 255
emilmont 1:fdd22bb7aa52 256 /* count is index by which the pointer pIn1 to be incremented */
emilmont 1:fdd22bb7aa52 257 count = 0u;
emilmont 1:fdd22bb7aa52 258
emilmont 1:fdd22bb7aa52 259 /* -------------------
emilmont 1:fdd22bb7aa52 260 * Stage2 process
emilmont 1:fdd22bb7aa52 261 * ------------------*/
emilmont 1:fdd22bb7aa52 262
emilmont 1:fdd22bb7aa52 263 /* Stage2 depends on srcBLen as in this stage srcBLen number of MACS are performed.
emilmont 1:fdd22bb7aa52 264 * So, to loop unroll over blockSize2,
emilmont 1:fdd22bb7aa52 265 * srcBLen should be greater than or equal to 4 */
emilmont 1:fdd22bb7aa52 266 if(srcBLen >= 4u)
emilmont 1:fdd22bb7aa52 267 {
emilmont 1:fdd22bb7aa52 268 /* Loop unroll over blockSize2, by 4 */
emilmont 1:fdd22bb7aa52 269 blkCnt = blockSize2 >> 2u;
emilmont 1:fdd22bb7aa52 270
emilmont 1:fdd22bb7aa52 271 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 272 {
emilmont 1:fdd22bb7aa52 273 /* Set all accumulators to zero */
emilmont 1:fdd22bb7aa52 274 acc0 = 0;
emilmont 1:fdd22bb7aa52 275 acc1 = 0;
emilmont 1:fdd22bb7aa52 276 acc2 = 0;
emilmont 1:fdd22bb7aa52 277 acc3 = 0;
emilmont 1:fdd22bb7aa52 278
emilmont 1:fdd22bb7aa52 279 /* read x[0], x[1], x[2] samples */
emilmont 1:fdd22bb7aa52 280 x0 = *(px++);
emilmont 1:fdd22bb7aa52 281 x1 = *(px++);
emilmont 1:fdd22bb7aa52 282 x2 = *(px++);
emilmont 1:fdd22bb7aa52 283
emilmont 1:fdd22bb7aa52 284 /* Apply loop unrolling and compute 4 MACs simultaneously. */
emilmont 1:fdd22bb7aa52 285 k = srcBLen >> 2u;
emilmont 1:fdd22bb7aa52 286
emilmont 1:fdd22bb7aa52 287 /* First part of the processing with loop unrolling. Compute 4 MACs at a time.
emilmont 1:fdd22bb7aa52 288 ** a second loop below computes MACs for the remaining 1 to 3 samples. */
emilmont 1:fdd22bb7aa52 289 do
emilmont 1:fdd22bb7aa52 290 {
emilmont 1:fdd22bb7aa52 291 /* Read y[0] sample */
emilmont 1:fdd22bb7aa52 292 c0 = *(py++);
emilmont 1:fdd22bb7aa52 293
emilmont 1:fdd22bb7aa52 294 /* Read x[3] sample */
emilmont 1:fdd22bb7aa52 295 x3 = *(px++);
emilmont 1:fdd22bb7aa52 296
emilmont 1:fdd22bb7aa52 297 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 298 /* acc0 += x[0] * y[0] */
emilmont 1:fdd22bb7aa52 299 acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x0 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 300 /* acc1 += x[1] * y[0] */
emilmont 1:fdd22bb7aa52 301 acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x1 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 302 /* acc2 += x[2] * y[0] */
emilmont 1:fdd22bb7aa52 303 acc2 = (q31_t) ((((q63_t) acc2 << 32) + ((q63_t) x2 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 304 /* acc3 += x[3] * y[0] */
emilmont 1:fdd22bb7aa52 305 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x3 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 306
emilmont 1:fdd22bb7aa52 307 /* Read y[1] sample */
emilmont 1:fdd22bb7aa52 308 c0 = *(py++);
emilmont 1:fdd22bb7aa52 309
emilmont 1:fdd22bb7aa52 310 /* Read x[4] sample */
emilmont 1:fdd22bb7aa52 311 x0 = *(px++);
emilmont 1:fdd22bb7aa52 312
emilmont 1:fdd22bb7aa52 313 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 314 /* acc0 += x[1] * y[1] */
emilmont 1:fdd22bb7aa52 315 acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x1 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 316 /* acc1 += x[2] * y[1] */
emilmont 1:fdd22bb7aa52 317 acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x2 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 318 /* acc2 += x[3] * y[1] */
emilmont 1:fdd22bb7aa52 319 acc2 = (q31_t) ((((q63_t) acc2 << 32) + ((q63_t) x3 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 320 /* acc3 += x[4] * y[1] */
emilmont 1:fdd22bb7aa52 321 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x0 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 322
emilmont 1:fdd22bb7aa52 323 /* Read y[2] sample */
emilmont 1:fdd22bb7aa52 324 c0 = *(py++);
emilmont 1:fdd22bb7aa52 325
emilmont 1:fdd22bb7aa52 326 /* Read x[5] sample */
emilmont 1:fdd22bb7aa52 327 x1 = *(px++);
emilmont 1:fdd22bb7aa52 328
emilmont 1:fdd22bb7aa52 329 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 330 /* acc0 += x[2] * y[2] */
emilmont 1:fdd22bb7aa52 331 acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x2 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 332 /* acc1 += x[3] * y[2] */
emilmont 1:fdd22bb7aa52 333 acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x3 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 334 /* acc2 += x[4] * y[2] */
emilmont 1:fdd22bb7aa52 335 acc2 = (q31_t) ((((q63_t) acc2 << 32) + ((q63_t) x0 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 336 /* acc3 += x[5] * y[2] */
emilmont 1:fdd22bb7aa52 337 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x1 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 338
emilmont 1:fdd22bb7aa52 339 /* Read y[3] sample */
emilmont 1:fdd22bb7aa52 340 c0 = *(py++);
emilmont 1:fdd22bb7aa52 341
emilmont 1:fdd22bb7aa52 342 /* Read x[6] sample */
emilmont 1:fdd22bb7aa52 343 x2 = *(px++);
emilmont 1:fdd22bb7aa52 344
emilmont 1:fdd22bb7aa52 345 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 346 /* acc0 += x[3] * y[3] */
emilmont 1:fdd22bb7aa52 347 acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x3 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 348 /* acc1 += x[4] * y[3] */
emilmont 1:fdd22bb7aa52 349 acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x0 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 350 /* acc2 += x[5] * y[3] */
emilmont 1:fdd22bb7aa52 351 acc2 = (q31_t) ((((q63_t) acc2 << 32) + ((q63_t) x1 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 352 /* acc3 += x[6] * y[3] */
emilmont 1:fdd22bb7aa52 353 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x2 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 354
emilmont 1:fdd22bb7aa52 355
emilmont 1:fdd22bb7aa52 356 } while(--k);
emilmont 1:fdd22bb7aa52 357
emilmont 1:fdd22bb7aa52 358 /* If the srcBLen is not a multiple of 4, compute any remaining MACs here.
emilmont 1:fdd22bb7aa52 359 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 360 k = srcBLen % 0x4u;
emilmont 1:fdd22bb7aa52 361
emilmont 1:fdd22bb7aa52 362 while(k > 0u)
emilmont 1:fdd22bb7aa52 363 {
emilmont 1:fdd22bb7aa52 364 /* Read y[4] sample */
emilmont 1:fdd22bb7aa52 365 c0 = *(py++);
emilmont 1:fdd22bb7aa52 366
emilmont 1:fdd22bb7aa52 367 /* Read x[7] sample */
emilmont 1:fdd22bb7aa52 368 x3 = *(px++);
emilmont 1:fdd22bb7aa52 369
emilmont 1:fdd22bb7aa52 370 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 371 /* acc0 += x[4] * y[4] */
emilmont 1:fdd22bb7aa52 372 acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x0 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 373 /* acc1 += x[5] * y[4] */
emilmont 1:fdd22bb7aa52 374 acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x1 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 375 /* acc2 += x[6] * y[4] */
emilmont 1:fdd22bb7aa52 376 acc2 = (q31_t) ((((q63_t) acc2 << 32) + ((q63_t) x2 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 377 /* acc3 += x[7] * y[4] */
emilmont 1:fdd22bb7aa52 378 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x3 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 379
emilmont 1:fdd22bb7aa52 380 /* Reuse the present samples for the next MAC */
emilmont 1:fdd22bb7aa52 381 x0 = x1;
emilmont 1:fdd22bb7aa52 382 x1 = x2;
emilmont 1:fdd22bb7aa52 383 x2 = x3;
emilmont 1:fdd22bb7aa52 384
emilmont 1:fdd22bb7aa52 385 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 386 k--;
emilmont 1:fdd22bb7aa52 387 }
emilmont 1:fdd22bb7aa52 388
emilmont 1:fdd22bb7aa52 389 /* Store the result in the accumulator in the destination buffer. */
emilmont 1:fdd22bb7aa52 390 *pOut = (q31_t) (acc0 << 1);
emilmont 1:fdd22bb7aa52 391 /* Destination pointer is updated according to the address modifier, inc */
emilmont 1:fdd22bb7aa52 392 pOut += inc;
emilmont 1:fdd22bb7aa52 393
emilmont 1:fdd22bb7aa52 394 *pOut = (q31_t) (acc1 << 1);
emilmont 1:fdd22bb7aa52 395 pOut += inc;
emilmont 1:fdd22bb7aa52 396
emilmont 1:fdd22bb7aa52 397 *pOut = (q31_t) (acc2 << 1);
emilmont 1:fdd22bb7aa52 398 pOut += inc;
emilmont 1:fdd22bb7aa52 399
emilmont 1:fdd22bb7aa52 400 *pOut = (q31_t) (acc3 << 1);
emilmont 1:fdd22bb7aa52 401 pOut += inc;
emilmont 1:fdd22bb7aa52 402
emilmont 1:fdd22bb7aa52 403 /* Increment the pointer pIn1 index, count by 4 */
emilmont 1:fdd22bb7aa52 404 count += 4u;
emilmont 1:fdd22bb7aa52 405
emilmont 1:fdd22bb7aa52 406 /* Update the inputA and inputB pointers for next MAC calculation */
emilmont 1:fdd22bb7aa52 407 px = pIn1 + count;
emilmont 1:fdd22bb7aa52 408 py = pIn2;
emilmont 1:fdd22bb7aa52 409
emilmont 1:fdd22bb7aa52 410
emilmont 1:fdd22bb7aa52 411 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 412 blkCnt--;
emilmont 1:fdd22bb7aa52 413 }
emilmont 1:fdd22bb7aa52 414
emilmont 1:fdd22bb7aa52 415 /* If the blockSize2 is not a multiple of 4, compute any remaining output samples here.
emilmont 1:fdd22bb7aa52 416 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 417 blkCnt = blockSize2 % 0x4u;
emilmont 1:fdd22bb7aa52 418
emilmont 1:fdd22bb7aa52 419 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 420 {
emilmont 1:fdd22bb7aa52 421 /* Accumulator is made zero for every iteration */
emilmont 1:fdd22bb7aa52 422 sum = 0;
emilmont 1:fdd22bb7aa52 423
emilmont 1:fdd22bb7aa52 424 /* Apply loop unrolling and compute 4 MACs simultaneously. */
emilmont 1:fdd22bb7aa52 425 k = srcBLen >> 2u;
emilmont 1:fdd22bb7aa52 426
emilmont 1:fdd22bb7aa52 427 /* First part of the processing with loop unrolling. Compute 4 MACs at a time.
emilmont 1:fdd22bb7aa52 428 ** a second loop below computes MACs for the remaining 1 to 3 samples. */
emilmont 1:fdd22bb7aa52 429 while(k > 0u)
emilmont 1:fdd22bb7aa52 430 {
emilmont 1:fdd22bb7aa52 431 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 432 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 433 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 434 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 435 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 436 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 437 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 438 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 439 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 440
emilmont 1:fdd22bb7aa52 441 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 442 k--;
emilmont 1:fdd22bb7aa52 443 }
emilmont 1:fdd22bb7aa52 444
emilmont 1:fdd22bb7aa52 445 /* If the srcBLen is not a multiple of 4, compute any remaining MACs here.
emilmont 1:fdd22bb7aa52 446 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 447 k = srcBLen % 0x4u;
emilmont 1:fdd22bb7aa52 448
emilmont 1:fdd22bb7aa52 449 while(k > 0u)
emilmont 1:fdd22bb7aa52 450 {
emilmont 1:fdd22bb7aa52 451 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 452 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 453 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 454
emilmont 1:fdd22bb7aa52 455 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 456 k--;
emilmont 1:fdd22bb7aa52 457 }
emilmont 1:fdd22bb7aa52 458
emilmont 1:fdd22bb7aa52 459 /* Store the result in the accumulator in the destination buffer. */
emilmont 1:fdd22bb7aa52 460 *pOut = sum << 1;
emilmont 1:fdd22bb7aa52 461 /* Destination pointer is updated according to the address modifier, inc */
emilmont 1:fdd22bb7aa52 462 pOut += inc;
emilmont 1:fdd22bb7aa52 463
emilmont 1:fdd22bb7aa52 464 /* Increment the MAC count */
emilmont 1:fdd22bb7aa52 465 count++;
emilmont 1:fdd22bb7aa52 466
emilmont 1:fdd22bb7aa52 467 /* Update the inputA and inputB pointers for next MAC calculation */
emilmont 1:fdd22bb7aa52 468 px = pIn1 + count;
emilmont 1:fdd22bb7aa52 469 py = pIn2;
emilmont 1:fdd22bb7aa52 470
emilmont 1:fdd22bb7aa52 471
emilmont 1:fdd22bb7aa52 472 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 473 blkCnt--;
emilmont 1:fdd22bb7aa52 474 }
emilmont 1:fdd22bb7aa52 475 }
emilmont 1:fdd22bb7aa52 476 else
emilmont 1:fdd22bb7aa52 477 {
emilmont 1:fdd22bb7aa52 478 /* If the srcBLen is not a multiple of 4,
emilmont 1:fdd22bb7aa52 479 * the blockSize2 loop cannot be unrolled by 4 */
emilmont 1:fdd22bb7aa52 480 blkCnt = blockSize2;
emilmont 1:fdd22bb7aa52 481
emilmont 1:fdd22bb7aa52 482 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 483 {
emilmont 1:fdd22bb7aa52 484 /* Accumulator is made zero for every iteration */
emilmont 1:fdd22bb7aa52 485 sum = 0;
emilmont 1:fdd22bb7aa52 486
emilmont 1:fdd22bb7aa52 487 /* Loop over srcBLen */
emilmont 1:fdd22bb7aa52 488 k = srcBLen;
emilmont 1:fdd22bb7aa52 489
emilmont 1:fdd22bb7aa52 490 while(k > 0u)
emilmont 1:fdd22bb7aa52 491 {
emilmont 1:fdd22bb7aa52 492 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 493 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 494 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 495
emilmont 1:fdd22bb7aa52 496 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 497 k--;
emilmont 1:fdd22bb7aa52 498 }
emilmont 1:fdd22bb7aa52 499
emilmont 1:fdd22bb7aa52 500 /* Store the result in the accumulator in the destination buffer. */
emilmont 1:fdd22bb7aa52 501 *pOut = sum << 1;
emilmont 1:fdd22bb7aa52 502 /* Destination pointer is updated according to the address modifier, inc */
emilmont 1:fdd22bb7aa52 503 pOut += inc;
emilmont 1:fdd22bb7aa52 504
emilmont 1:fdd22bb7aa52 505 /* Increment the MAC count */
emilmont 1:fdd22bb7aa52 506 count++;
emilmont 1:fdd22bb7aa52 507
emilmont 1:fdd22bb7aa52 508 /* Update the inputA and inputB pointers for next MAC calculation */
emilmont 1:fdd22bb7aa52 509 px = pIn1 + count;
emilmont 1:fdd22bb7aa52 510 py = pIn2;
emilmont 1:fdd22bb7aa52 511
emilmont 1:fdd22bb7aa52 512 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 513 blkCnt--;
emilmont 1:fdd22bb7aa52 514 }
emilmont 1:fdd22bb7aa52 515 }
emilmont 1:fdd22bb7aa52 516
emilmont 1:fdd22bb7aa52 517 /* --------------------------
emilmont 1:fdd22bb7aa52 518 * Initializations of stage3
emilmont 1:fdd22bb7aa52 519 * -------------------------*/
emilmont 1:fdd22bb7aa52 520
emilmont 1:fdd22bb7aa52 521 /* sum += x[srcALen-srcBLen+1] * y[0] + x[srcALen-srcBLen+2] * y[1] +...+ x[srcALen-1] * y[srcBLen-1]
emilmont 1:fdd22bb7aa52 522 * sum += x[srcALen-srcBLen+2] * y[0] + x[srcALen-srcBLen+3] * y[1] +...+ x[srcALen-1] * y[srcBLen-1]
emilmont 1:fdd22bb7aa52 523 * ....
emilmont 1:fdd22bb7aa52 524 * sum += x[srcALen-2] * y[0] + x[srcALen-1] * y[1]
emilmont 1:fdd22bb7aa52 525 * sum += x[srcALen-1] * y[0]
emilmont 1:fdd22bb7aa52 526 */
emilmont 1:fdd22bb7aa52 527
emilmont 1:fdd22bb7aa52 528 /* In this stage the MAC operations are decreased by 1 for every iteration.
emilmont 1:fdd22bb7aa52 529 The count variable holds the number of MAC operations performed */
emilmont 1:fdd22bb7aa52 530 count = srcBLen - 1u;
emilmont 1:fdd22bb7aa52 531
emilmont 1:fdd22bb7aa52 532 /* Working pointer of inputA */
emilmont 1:fdd22bb7aa52 533 pSrc1 = ((pIn1 + srcALen) - srcBLen) + 1u;
emilmont 1:fdd22bb7aa52 534 px = pSrc1;
emilmont 1:fdd22bb7aa52 535
emilmont 1:fdd22bb7aa52 536 /* Working pointer of inputB */
emilmont 1:fdd22bb7aa52 537 py = pIn2;
emilmont 1:fdd22bb7aa52 538
emilmont 1:fdd22bb7aa52 539 /* -------------------
emilmont 1:fdd22bb7aa52 540 * Stage3 process
emilmont 1:fdd22bb7aa52 541 * ------------------*/
emilmont 1:fdd22bb7aa52 542
emilmont 1:fdd22bb7aa52 543 while(blockSize3 > 0u)
emilmont 1:fdd22bb7aa52 544 {
emilmont 1:fdd22bb7aa52 545 /* Accumulator is made zero for every iteration */
emilmont 1:fdd22bb7aa52 546 sum = 0;
emilmont 1:fdd22bb7aa52 547
emilmont 1:fdd22bb7aa52 548 /* Apply loop unrolling and compute 4 MACs simultaneously. */
emilmont 1:fdd22bb7aa52 549 k = count >> 2u;
emilmont 1:fdd22bb7aa52 550
emilmont 1:fdd22bb7aa52 551 /* First part of the processing with loop unrolling. Compute 4 MACs at a time.
emilmont 1:fdd22bb7aa52 552 ** a second loop below computes MACs for the remaining 1 to 3 samples. */
emilmont 1:fdd22bb7aa52 553 while(k > 0u)
emilmont 1:fdd22bb7aa52 554 {
emilmont 1:fdd22bb7aa52 555 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 556 /* sum += x[srcALen - srcBLen + 4] * y[3] */
emilmont 1:fdd22bb7aa52 557 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 558 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 559 /* sum += x[srcALen - srcBLen + 3] * y[2] */
emilmont 1:fdd22bb7aa52 560 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 561 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 562 /* sum += x[srcALen - srcBLen + 2] * y[1] */
emilmont 1:fdd22bb7aa52 563 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 564 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 565 /* sum += x[srcALen - srcBLen + 1] * y[0] */
emilmont 1:fdd22bb7aa52 566 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 567 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 568
emilmont 1:fdd22bb7aa52 569 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 570 k--;
emilmont 1:fdd22bb7aa52 571 }
emilmont 1:fdd22bb7aa52 572
emilmont 1:fdd22bb7aa52 573 /* If the count is not a multiple of 4, compute any remaining MACs here.
emilmont 1:fdd22bb7aa52 574 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 575 k = count % 0x4u;
emilmont 1:fdd22bb7aa52 576
emilmont 1:fdd22bb7aa52 577 while(k > 0u)
emilmont 1:fdd22bb7aa52 578 {
emilmont 1:fdd22bb7aa52 579 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 580 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 581 ((q63_t) * px++ * (*py++))) >> 32);
emilmont 1:fdd22bb7aa52 582
emilmont 1:fdd22bb7aa52 583 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 584 k--;
emilmont 1:fdd22bb7aa52 585 }
emilmont 1:fdd22bb7aa52 586
emilmont 1:fdd22bb7aa52 587 /* Store the result in the accumulator in the destination buffer. */
emilmont 1:fdd22bb7aa52 588 *pOut = sum << 1;
emilmont 1:fdd22bb7aa52 589 /* Destination pointer is updated according to the address modifier, inc */
emilmont 1:fdd22bb7aa52 590 pOut += inc;
emilmont 1:fdd22bb7aa52 591
emilmont 1:fdd22bb7aa52 592 /* Update the inputA and inputB pointers for next MAC calculation */
emilmont 1:fdd22bb7aa52 593 px = ++pSrc1;
emilmont 1:fdd22bb7aa52 594 py = pIn2;
emilmont 1:fdd22bb7aa52 595
emilmont 1:fdd22bb7aa52 596 /* Decrement the MAC count */
emilmont 1:fdd22bb7aa52 597 count--;
emilmont 1:fdd22bb7aa52 598
emilmont 1:fdd22bb7aa52 599 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 600 blockSize3--;
emilmont 1:fdd22bb7aa52 601 }
emilmont 1:fdd22bb7aa52 602
emilmont 1:fdd22bb7aa52 603 }
emilmont 1:fdd22bb7aa52 604
emilmont 1:fdd22bb7aa52 605 /**
emilmont 1:fdd22bb7aa52 606 * @} end of Corr group
emilmont 1:fdd22bb7aa52 607 */