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_conv_partial_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_conv_partial_fast_q31.c
emilmont 1:fdd22bb7aa52 9 *
emilmont 2:da51fb522205 10 * Description: Fast Q31 Partial convolution.
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 PartialConv
emilmont 1:fdd22bb7aa52 44 * @{
emilmont 1:fdd22bb7aa52 45 */
emilmont 1:fdd22bb7aa52 46
emilmont 1:fdd22bb7aa52 47 /**
emilmont 1:fdd22bb7aa52 48 * @brief Partial convolution 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.
emilmont 1:fdd22bb7aa52 54 * @param[in] firstIndex is the first output sample to start with.
emilmont 1:fdd22bb7aa52 55 * @param[in] numPoints is the number of output points to be computed.
emilmont 1:fdd22bb7aa52 56 * @return Returns either ARM_MATH_SUCCESS if the function completed correctly or ARM_MATH_ARGUMENT_ERROR if the requested subset is not in the range [0 srcALen+srcBLen-2].
emilmont 1:fdd22bb7aa52 57 *
emilmont 1:fdd22bb7aa52 58 * \par
emilmont 1:fdd22bb7aa52 59 * See <code>arm_conv_partial_q31()</code> for a slower implementation of this function which uses a 64-bit accumulator to provide higher precision.
emilmont 1:fdd22bb7aa52 60 */
emilmont 1:fdd22bb7aa52 61
emilmont 1:fdd22bb7aa52 62 arm_status arm_conv_partial_fast_q31(
emilmont 1:fdd22bb7aa52 63 q31_t * pSrcA,
emilmont 1:fdd22bb7aa52 64 uint32_t srcALen,
emilmont 1:fdd22bb7aa52 65 q31_t * pSrcB,
emilmont 1:fdd22bb7aa52 66 uint32_t srcBLen,
emilmont 1:fdd22bb7aa52 67 q31_t * pDst,
emilmont 1:fdd22bb7aa52 68 uint32_t firstIndex,
emilmont 1:fdd22bb7aa52 69 uint32_t numPoints)
emilmont 1:fdd22bb7aa52 70 {
emilmont 1:fdd22bb7aa52 71 q31_t *pIn1; /* inputA pointer */
emilmont 1:fdd22bb7aa52 72 q31_t *pIn2; /* inputB pointer */
emilmont 1:fdd22bb7aa52 73 q31_t *pOut = pDst; /* output pointer */
emilmont 1:fdd22bb7aa52 74 q31_t *px; /* Intermediate inputA pointer */
emilmont 1:fdd22bb7aa52 75 q31_t *py; /* Intermediate inputB pointer */
emilmont 1:fdd22bb7aa52 76 q31_t *pSrc1, *pSrc2; /* Intermediate pointers */
emilmont 1:fdd22bb7aa52 77 q31_t sum, acc0, acc1, acc2, acc3; /* Accumulators */
emilmont 1:fdd22bb7aa52 78 q31_t x0, x1, x2, x3, c0;
emilmont 1:fdd22bb7aa52 79 uint32_t j, k, count, check, blkCnt;
emilmont 1:fdd22bb7aa52 80 int32_t blockSize1, blockSize2, blockSize3; /* loop counters */
emilmont 1:fdd22bb7aa52 81 arm_status status; /* status of Partial convolution */
emilmont 1:fdd22bb7aa52 82
emilmont 1:fdd22bb7aa52 83
emilmont 1:fdd22bb7aa52 84 /* Check for range of output samples to be calculated */
emilmont 1:fdd22bb7aa52 85 if((firstIndex + numPoints) > ((srcALen + (srcBLen - 1u))))
emilmont 1:fdd22bb7aa52 86 {
emilmont 1:fdd22bb7aa52 87 /* Set status as ARM_MATH_ARGUMENT_ERROR */
emilmont 1:fdd22bb7aa52 88 status = ARM_MATH_ARGUMENT_ERROR;
emilmont 1:fdd22bb7aa52 89 }
emilmont 1:fdd22bb7aa52 90 else
emilmont 1:fdd22bb7aa52 91 {
emilmont 1:fdd22bb7aa52 92
emilmont 1:fdd22bb7aa52 93 /* The algorithm implementation is based on the lengths of the inputs. */
emilmont 1:fdd22bb7aa52 94 /* srcB is always made to slide across srcA. */
emilmont 1:fdd22bb7aa52 95 /* So srcBLen is always considered as shorter or equal to srcALen */
emilmont 1:fdd22bb7aa52 96 if(srcALen >= srcBLen)
emilmont 1:fdd22bb7aa52 97 {
emilmont 1:fdd22bb7aa52 98 /* Initialization of inputA pointer */
emilmont 1:fdd22bb7aa52 99 pIn1 = pSrcA;
emilmont 1:fdd22bb7aa52 100
emilmont 1:fdd22bb7aa52 101 /* Initialization of inputB pointer */
emilmont 1:fdd22bb7aa52 102 pIn2 = pSrcB;
emilmont 1:fdd22bb7aa52 103 }
emilmont 1:fdd22bb7aa52 104 else
emilmont 1:fdd22bb7aa52 105 {
emilmont 1:fdd22bb7aa52 106 /* Initialization of inputA pointer */
emilmont 1:fdd22bb7aa52 107 pIn1 = pSrcB;
emilmont 1:fdd22bb7aa52 108
emilmont 1:fdd22bb7aa52 109 /* Initialization of inputB pointer */
emilmont 1:fdd22bb7aa52 110 pIn2 = pSrcA;
emilmont 1:fdd22bb7aa52 111
emilmont 1:fdd22bb7aa52 112 /* srcBLen is always considered as shorter or equal to srcALen */
emilmont 1:fdd22bb7aa52 113 j = srcBLen;
emilmont 1:fdd22bb7aa52 114 srcBLen = srcALen;
emilmont 1:fdd22bb7aa52 115 srcALen = j;
emilmont 1:fdd22bb7aa52 116 }
emilmont 1:fdd22bb7aa52 117
emilmont 1:fdd22bb7aa52 118 /* Conditions to check which loopCounter holds
emilmont 1:fdd22bb7aa52 119 * the first and last indices of the output samples to be calculated. */
emilmont 1:fdd22bb7aa52 120 check = firstIndex + numPoints;
emilmont 1:fdd22bb7aa52 121 blockSize3 = ((int32_t) check - (int32_t) srcALen);
emilmont 1:fdd22bb7aa52 122 blockSize3 = (blockSize3 > 0) ? blockSize3 : 0;
emilmont 1:fdd22bb7aa52 123 blockSize1 = (((int32_t) srcBLen - 1) - (int32_t) firstIndex);
emilmont 1:fdd22bb7aa52 124 blockSize1 = (blockSize1 > 0) ? ((check > (srcBLen - 1u)) ? blockSize1 :
emilmont 1:fdd22bb7aa52 125 (int32_t) numPoints) : 0;
emilmont 1:fdd22bb7aa52 126 blockSize2 = (int32_t) check - ((blockSize3 + blockSize1) +
emilmont 1:fdd22bb7aa52 127 (int32_t) firstIndex);
emilmont 1:fdd22bb7aa52 128 blockSize2 = (blockSize2 > 0) ? blockSize2 : 0;
emilmont 1:fdd22bb7aa52 129
emilmont 1:fdd22bb7aa52 130 /* conv(x,y) at n = x[n] * y[0] + x[n-1] * y[1] + x[n-2] * y[2] + ...+ x[n-N+1] * y[N -1] */
emilmont 1:fdd22bb7aa52 131 /* The function is internally
emilmont 1:fdd22bb7aa52 132 * divided into three stages according to the number of multiplications that has to be
emilmont 1:fdd22bb7aa52 133 * taken place between inputA samples and inputB samples. In the first stage of the
emilmont 1:fdd22bb7aa52 134 * algorithm, the multiplications increase by one for every iteration.
emilmont 1:fdd22bb7aa52 135 * In the second stage of the algorithm, srcBLen number of multiplications are done.
emilmont 1:fdd22bb7aa52 136 * In the third stage of the algorithm, the multiplications decrease by one
emilmont 1:fdd22bb7aa52 137 * for every iteration. */
emilmont 1:fdd22bb7aa52 138
emilmont 1:fdd22bb7aa52 139 /* Set the output pointer to point to the firstIndex
emilmont 1:fdd22bb7aa52 140 * of the output sample to be calculated. */
emilmont 1:fdd22bb7aa52 141 pOut = pDst + firstIndex;
emilmont 1:fdd22bb7aa52 142
emilmont 1:fdd22bb7aa52 143 /* --------------------------
emilmont 1:fdd22bb7aa52 144 * Initializations of stage1
emilmont 1:fdd22bb7aa52 145 * -------------------------*/
emilmont 1:fdd22bb7aa52 146
emilmont 1:fdd22bb7aa52 147 /* sum = x[0] * y[0]
emilmont 1:fdd22bb7aa52 148 * sum = x[0] * y[1] + x[1] * y[0]
emilmont 1:fdd22bb7aa52 149 * ....
emilmont 1:fdd22bb7aa52 150 * sum = x[0] * y[srcBlen - 1] + x[1] * y[srcBlen - 2] +...+ x[srcBLen - 1] * y[0]
emilmont 1:fdd22bb7aa52 151 */
emilmont 1:fdd22bb7aa52 152
emilmont 1:fdd22bb7aa52 153 /* In this stage the MAC operations are increased by 1 for every iteration.
emilmont 1:fdd22bb7aa52 154 The count variable holds the number of MAC operations performed.
emilmont 1:fdd22bb7aa52 155 Since the partial convolution starts from firstIndex
emilmont 1:fdd22bb7aa52 156 Number of Macs to be performed is firstIndex + 1 */
emilmont 1:fdd22bb7aa52 157 count = 1u + firstIndex;
emilmont 1:fdd22bb7aa52 158
emilmont 1:fdd22bb7aa52 159 /* Working pointer of inputA */
emilmont 1:fdd22bb7aa52 160 px = pIn1;
emilmont 1:fdd22bb7aa52 161
emilmont 1:fdd22bb7aa52 162 /* Working pointer of inputB */
emilmont 1:fdd22bb7aa52 163 pSrc2 = pIn2 + firstIndex;
emilmont 1:fdd22bb7aa52 164 py = pSrc2;
emilmont 1:fdd22bb7aa52 165
emilmont 1:fdd22bb7aa52 166 /* ------------------------
emilmont 1:fdd22bb7aa52 167 * Stage1 process
emilmont 1:fdd22bb7aa52 168 * ----------------------*/
emilmont 1:fdd22bb7aa52 169
emilmont 1:fdd22bb7aa52 170 /* The first loop starts here */
emilmont 1:fdd22bb7aa52 171 while(blockSize1 > 0)
emilmont 1:fdd22bb7aa52 172 {
emilmont 1:fdd22bb7aa52 173 /* Accumulator is made zero for every iteration */
emilmont 1:fdd22bb7aa52 174 sum = 0;
emilmont 1:fdd22bb7aa52 175
emilmont 1:fdd22bb7aa52 176 /* Apply loop unrolling and compute 4 MACs simultaneously. */
emilmont 1:fdd22bb7aa52 177 k = count >> 2u;
emilmont 1:fdd22bb7aa52 178
emilmont 1:fdd22bb7aa52 179 /* First part of the processing with loop unrolling. Compute 4 MACs at a time.
emilmont 1:fdd22bb7aa52 180 ** a second loop below computes MACs for the remaining 1 to 3 samples. */
emilmont 1:fdd22bb7aa52 181 while(k > 0u)
emilmont 1:fdd22bb7aa52 182 {
emilmont 1:fdd22bb7aa52 183 /* x[0] * y[srcBLen - 1] */
emilmont 1:fdd22bb7aa52 184 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 185 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 186
emilmont 1:fdd22bb7aa52 187 /* x[1] * y[srcBLen - 2] */
emilmont 1:fdd22bb7aa52 188 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 189 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 190
emilmont 1:fdd22bb7aa52 191 /* x[2] * y[srcBLen - 3] */
emilmont 1:fdd22bb7aa52 192 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 193 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 194
emilmont 1:fdd22bb7aa52 195 /* x[3] * y[srcBLen - 4] */
emilmont 1:fdd22bb7aa52 196 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 197 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 198
emilmont 1:fdd22bb7aa52 199 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 200 k--;
emilmont 1:fdd22bb7aa52 201 }
emilmont 1:fdd22bb7aa52 202
emilmont 1:fdd22bb7aa52 203 /* If the count is not a multiple of 4, compute any remaining MACs here.
emilmont 1:fdd22bb7aa52 204 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 205 k = count % 0x4u;
emilmont 1:fdd22bb7aa52 206
emilmont 1:fdd22bb7aa52 207 while(k > 0u)
emilmont 1:fdd22bb7aa52 208 {
emilmont 1:fdd22bb7aa52 209 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 210 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 211 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 212
emilmont 1:fdd22bb7aa52 213 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 214 k--;
emilmont 1:fdd22bb7aa52 215 }
emilmont 1:fdd22bb7aa52 216
emilmont 1:fdd22bb7aa52 217 /* Store the result in the accumulator in the destination buffer. */
emilmont 1:fdd22bb7aa52 218 *pOut++ = sum << 1;
emilmont 1:fdd22bb7aa52 219
emilmont 1:fdd22bb7aa52 220 /* Update the inputA and inputB pointers for next MAC calculation */
emilmont 1:fdd22bb7aa52 221 py = ++pSrc2;
emilmont 1:fdd22bb7aa52 222 px = pIn1;
emilmont 1:fdd22bb7aa52 223
emilmont 1:fdd22bb7aa52 224 /* Increment the MAC count */
emilmont 1:fdd22bb7aa52 225 count++;
emilmont 1:fdd22bb7aa52 226
emilmont 1:fdd22bb7aa52 227 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 228 blockSize1--;
emilmont 1:fdd22bb7aa52 229 }
emilmont 1:fdd22bb7aa52 230
emilmont 1:fdd22bb7aa52 231 /* --------------------------
emilmont 1:fdd22bb7aa52 232 * Initializations of stage2
emilmont 1:fdd22bb7aa52 233 * ------------------------*/
emilmont 1:fdd22bb7aa52 234
emilmont 1:fdd22bb7aa52 235 /* sum = x[0] * y[srcBLen-1] + x[1] * y[srcBLen-2] +...+ x[srcBLen-1] * y[0]
emilmont 1:fdd22bb7aa52 236 * sum = x[1] * y[srcBLen-1] + x[2] * y[srcBLen-2] +...+ x[srcBLen] * y[0]
emilmont 1:fdd22bb7aa52 237 * ....
emilmont 1:fdd22bb7aa52 238 * sum = x[srcALen-srcBLen-2] * y[srcBLen-1] + x[srcALen] * y[srcBLen-2] +...+ x[srcALen-1] * y[0]
emilmont 1:fdd22bb7aa52 239 */
emilmont 1:fdd22bb7aa52 240
emilmont 1:fdd22bb7aa52 241 /* Working pointer of inputA */
emilmont 1:fdd22bb7aa52 242 px = pIn1;
emilmont 1:fdd22bb7aa52 243
emilmont 1:fdd22bb7aa52 244 /* Working pointer of inputB */
emilmont 1:fdd22bb7aa52 245 pSrc2 = pIn2 + (srcBLen - 1u);
emilmont 1:fdd22bb7aa52 246 py = pSrc2;
emilmont 1:fdd22bb7aa52 247
emilmont 1:fdd22bb7aa52 248 /* count is index by which the pointer pIn1 to be incremented */
emilmont 1:fdd22bb7aa52 249 count = 0u;
emilmont 1:fdd22bb7aa52 250
emilmont 1:fdd22bb7aa52 251 /* -------------------
emilmont 1:fdd22bb7aa52 252 * Stage2 process
emilmont 1:fdd22bb7aa52 253 * ------------------*/
emilmont 1:fdd22bb7aa52 254
emilmont 1:fdd22bb7aa52 255 /* Stage2 depends on srcBLen as in this stage srcBLen number of MACS are performed.
emilmont 1:fdd22bb7aa52 256 * So, to loop unroll over blockSize2,
emilmont 1:fdd22bb7aa52 257 * srcBLen should be greater than or equal to 4 */
emilmont 1:fdd22bb7aa52 258 if(srcBLen >= 4u)
emilmont 1:fdd22bb7aa52 259 {
emilmont 1:fdd22bb7aa52 260 /* Loop unroll over blockSize2 */
emilmont 1:fdd22bb7aa52 261 blkCnt = ((uint32_t) blockSize2 >> 2u);
emilmont 1:fdd22bb7aa52 262
emilmont 1:fdd22bb7aa52 263 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 264 {
emilmont 1:fdd22bb7aa52 265 /* Set all accumulators to zero */
emilmont 1:fdd22bb7aa52 266 acc0 = 0;
emilmont 1:fdd22bb7aa52 267 acc1 = 0;
emilmont 1:fdd22bb7aa52 268 acc2 = 0;
emilmont 1:fdd22bb7aa52 269 acc3 = 0;
emilmont 1:fdd22bb7aa52 270
emilmont 1:fdd22bb7aa52 271 /* read x[0], x[1], x[2] samples */
emilmont 1:fdd22bb7aa52 272 x0 = *(px++);
emilmont 1:fdd22bb7aa52 273 x1 = *(px++);
emilmont 1:fdd22bb7aa52 274 x2 = *(px++);
emilmont 1:fdd22bb7aa52 275
emilmont 1:fdd22bb7aa52 276 /* Apply loop unrolling and compute 4 MACs simultaneously. */
emilmont 1:fdd22bb7aa52 277 k = srcBLen >> 2u;
emilmont 1:fdd22bb7aa52 278
emilmont 1:fdd22bb7aa52 279 /* First part of the processing with loop unrolling. Compute 4 MACs at a time.
emilmont 1:fdd22bb7aa52 280 ** a second loop below computes MACs for the remaining 1 to 3 samples. */
emilmont 1:fdd22bb7aa52 281 do
emilmont 1:fdd22bb7aa52 282 {
emilmont 1:fdd22bb7aa52 283 /* Read y[srcBLen - 1] sample */
emilmont 1:fdd22bb7aa52 284 c0 = *(py--);
emilmont 1:fdd22bb7aa52 285
emilmont 1:fdd22bb7aa52 286 /* Read x[3] sample */
emilmont 1:fdd22bb7aa52 287 x3 = *(px++);
emilmont 1:fdd22bb7aa52 288
emilmont 1:fdd22bb7aa52 289 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 290 /* acc0 += x[0] * y[srcBLen - 1] */
emilmont 1:fdd22bb7aa52 291 acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x0 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 292
emilmont 1:fdd22bb7aa52 293 /* acc1 += x[1] * y[srcBLen - 1] */
emilmont 1:fdd22bb7aa52 294 acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x1 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 295
emilmont 1:fdd22bb7aa52 296 /* acc2 += x[2] * y[srcBLen - 1] */
emilmont 1:fdd22bb7aa52 297 acc2 = (q31_t) ((((q63_t) acc2 << 32) + ((q63_t) x2 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 298
emilmont 1:fdd22bb7aa52 299 /* acc3 += x[3] * y[srcBLen - 1] */
emilmont 1:fdd22bb7aa52 300 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x3 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 301
emilmont 1:fdd22bb7aa52 302 /* Read y[srcBLen - 2] sample */
emilmont 1:fdd22bb7aa52 303 c0 = *(py--);
emilmont 1:fdd22bb7aa52 304
emilmont 1:fdd22bb7aa52 305 /* Read x[4] sample */
emilmont 1:fdd22bb7aa52 306 x0 = *(px++);
emilmont 1:fdd22bb7aa52 307
emilmont 1:fdd22bb7aa52 308 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 309 /* acc0 += x[1] * y[srcBLen - 2] */
emilmont 1:fdd22bb7aa52 310 acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x1 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 311 /* acc1 += x[2] * y[srcBLen - 2] */
emilmont 1:fdd22bb7aa52 312 acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x2 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 313 /* acc2 += x[3] * y[srcBLen - 2] */
emilmont 1:fdd22bb7aa52 314 acc2 = (q31_t) ((((q63_t) acc2 << 32) + ((q63_t) x3 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 315 /* acc3 += x[4] * y[srcBLen - 2] */
emilmont 1:fdd22bb7aa52 316 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x0 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 317
emilmont 1:fdd22bb7aa52 318 /* Read y[srcBLen - 3] sample */
emilmont 1:fdd22bb7aa52 319 c0 = *(py--);
emilmont 1:fdd22bb7aa52 320
emilmont 1:fdd22bb7aa52 321 /* Read x[5] sample */
emilmont 1:fdd22bb7aa52 322 x1 = *(px++);
emilmont 1:fdd22bb7aa52 323
emilmont 1:fdd22bb7aa52 324 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 325 /* acc0 += x[2] * y[srcBLen - 3] */
emilmont 1:fdd22bb7aa52 326 acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x2 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 327 /* acc1 += x[3] * y[srcBLen - 2] */
emilmont 1:fdd22bb7aa52 328 acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x3 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 329 /* acc2 += x[4] * y[srcBLen - 2] */
emilmont 1:fdd22bb7aa52 330 acc2 = (q31_t) ((((q63_t) acc2 << 32) + ((q63_t) x0 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 331 /* acc3 += x[5] * y[srcBLen - 2] */
emilmont 1:fdd22bb7aa52 332 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x1 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 333
emilmont 1:fdd22bb7aa52 334 /* Read y[srcBLen - 4] sample */
emilmont 1:fdd22bb7aa52 335 c0 = *(py--);
emilmont 1:fdd22bb7aa52 336
emilmont 1:fdd22bb7aa52 337 /* Read x[6] sample */
emilmont 1:fdd22bb7aa52 338 x2 = *(px++);
emilmont 1:fdd22bb7aa52 339
emilmont 1:fdd22bb7aa52 340 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 341 /* acc0 += x[3] * y[srcBLen - 4] */
emilmont 1:fdd22bb7aa52 342 acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x3 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 343 /* acc1 += x[4] * y[srcBLen - 4] */
emilmont 1:fdd22bb7aa52 344 acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x0 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 345 /* acc2 += x[5] * y[srcBLen - 4] */
emilmont 1:fdd22bb7aa52 346 acc2 = (q31_t) ((((q63_t) acc2 << 32) + ((q63_t) x1 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 347 /* acc3 += x[6] * y[srcBLen - 4] */
emilmont 1:fdd22bb7aa52 348 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x2 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 349
emilmont 1:fdd22bb7aa52 350
emilmont 1:fdd22bb7aa52 351 } while(--k);
emilmont 1:fdd22bb7aa52 352
emilmont 1:fdd22bb7aa52 353 /* If the srcBLen is not a multiple of 4, compute any remaining MACs here.
emilmont 1:fdd22bb7aa52 354 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 355 k = srcBLen % 0x4u;
emilmont 1:fdd22bb7aa52 356
emilmont 1:fdd22bb7aa52 357 while(k > 0u)
emilmont 1:fdd22bb7aa52 358 {
emilmont 1:fdd22bb7aa52 359 /* Read y[srcBLen - 5] sample */
emilmont 1:fdd22bb7aa52 360 c0 = *(py--);
emilmont 1:fdd22bb7aa52 361
emilmont 1:fdd22bb7aa52 362 /* Read x[7] sample */
emilmont 1:fdd22bb7aa52 363 x3 = *(px++);
emilmont 1:fdd22bb7aa52 364
emilmont 1:fdd22bb7aa52 365 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 366 /* acc0 += x[4] * y[srcBLen - 5] */
emilmont 1:fdd22bb7aa52 367 acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x0 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 368 /* acc1 += x[5] * y[srcBLen - 5] */
emilmont 1:fdd22bb7aa52 369 acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x1 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 370 /* acc2 += x[6] * y[srcBLen - 5] */
emilmont 1:fdd22bb7aa52 371 acc2 = (q31_t) ((((q63_t) acc2 << 32) + ((q63_t) x2 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 372 /* acc3 += x[7] * y[srcBLen - 5] */
emilmont 1:fdd22bb7aa52 373 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x3 * c0)) >> 32);
emilmont 1:fdd22bb7aa52 374
emilmont 1:fdd22bb7aa52 375 /* Reuse the present samples for the next MAC */
emilmont 1:fdd22bb7aa52 376 x0 = x1;
emilmont 1:fdd22bb7aa52 377 x1 = x2;
emilmont 1:fdd22bb7aa52 378 x2 = x3;
emilmont 1:fdd22bb7aa52 379
emilmont 1:fdd22bb7aa52 380 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 381 k--;
emilmont 1:fdd22bb7aa52 382 }
emilmont 1:fdd22bb7aa52 383
emilmont 1:fdd22bb7aa52 384 /* Store the result in the accumulator in the destination buffer. */
emilmont 1:fdd22bb7aa52 385 *pOut++ = (q31_t) (acc0 << 1);
emilmont 1:fdd22bb7aa52 386 *pOut++ = (q31_t) (acc1 << 1);
emilmont 1:fdd22bb7aa52 387 *pOut++ = (q31_t) (acc2 << 1);
emilmont 1:fdd22bb7aa52 388 *pOut++ = (q31_t) (acc3 << 1);
emilmont 1:fdd22bb7aa52 389
emilmont 1:fdd22bb7aa52 390 /* Increment the pointer pIn1 index, count by 4 */
emilmont 1:fdd22bb7aa52 391 count += 4u;
emilmont 1:fdd22bb7aa52 392
emilmont 1:fdd22bb7aa52 393 /* Update the inputA and inputB pointers for next MAC calculation */
emilmont 1:fdd22bb7aa52 394 px = pIn1 + count;
emilmont 1:fdd22bb7aa52 395 py = pSrc2;
emilmont 1:fdd22bb7aa52 396
emilmont 1:fdd22bb7aa52 397 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 398 blkCnt--;
emilmont 1:fdd22bb7aa52 399 }
emilmont 1:fdd22bb7aa52 400
emilmont 1:fdd22bb7aa52 401 /* If the blockSize2 is not a multiple of 4, compute any remaining output samples here.
emilmont 1:fdd22bb7aa52 402 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 403 blkCnt = (uint32_t) blockSize2 % 0x4u;
emilmont 1:fdd22bb7aa52 404
emilmont 1:fdd22bb7aa52 405 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 406 {
emilmont 1:fdd22bb7aa52 407 /* Accumulator is made zero for every iteration */
emilmont 1:fdd22bb7aa52 408 sum = 0;
emilmont 1:fdd22bb7aa52 409
emilmont 1:fdd22bb7aa52 410 /* Apply loop unrolling and compute 4 MACs simultaneously. */
emilmont 1:fdd22bb7aa52 411 k = srcBLen >> 2u;
emilmont 1:fdd22bb7aa52 412
emilmont 1:fdd22bb7aa52 413 /* First part of the processing with loop unrolling. Compute 4 MACs at a time.
emilmont 1:fdd22bb7aa52 414 ** a second loop below computes MACs for the remaining 1 to 3 samples. */
emilmont 1:fdd22bb7aa52 415 while(k > 0u)
emilmont 1:fdd22bb7aa52 416 {
emilmont 1:fdd22bb7aa52 417 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 418 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 419 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 420 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 421 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 422 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 423 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 424 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 425 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 426
emilmont 1:fdd22bb7aa52 427 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 428 k--;
emilmont 1:fdd22bb7aa52 429 }
emilmont 1:fdd22bb7aa52 430
emilmont 1:fdd22bb7aa52 431 /* If the srcBLen is not a multiple of 4, compute any remaining MACs here.
emilmont 1:fdd22bb7aa52 432 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 433 k = srcBLen % 0x4u;
emilmont 1:fdd22bb7aa52 434
emilmont 1:fdd22bb7aa52 435 while(k > 0u)
emilmont 1:fdd22bb7aa52 436 {
emilmont 1:fdd22bb7aa52 437 /* Perform the multiply-accumulate */
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 /* Store the result in the accumulator in the destination buffer. */
emilmont 1:fdd22bb7aa52 446 *pOut++ = sum << 1;
emilmont 1:fdd22bb7aa52 447
emilmont 1:fdd22bb7aa52 448 /* Increment the MAC count */
emilmont 1:fdd22bb7aa52 449 count++;
emilmont 1:fdd22bb7aa52 450
emilmont 1:fdd22bb7aa52 451 /* Update the inputA and inputB pointers for next MAC calculation */
emilmont 1:fdd22bb7aa52 452 px = pIn1 + count;
emilmont 1:fdd22bb7aa52 453 py = pSrc2;
emilmont 1:fdd22bb7aa52 454
emilmont 1:fdd22bb7aa52 455 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 456 blkCnt--;
emilmont 1:fdd22bb7aa52 457 }
emilmont 1:fdd22bb7aa52 458 }
emilmont 1:fdd22bb7aa52 459 else
emilmont 1:fdd22bb7aa52 460 {
emilmont 1:fdd22bb7aa52 461 /* If the srcBLen is not a multiple of 4,
emilmont 1:fdd22bb7aa52 462 * the blockSize2 loop cannot be unrolled by 4 */
emilmont 1:fdd22bb7aa52 463 blkCnt = (uint32_t) blockSize2;
emilmont 1:fdd22bb7aa52 464
emilmont 1:fdd22bb7aa52 465 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 466 {
emilmont 1:fdd22bb7aa52 467 /* Accumulator is made zero for every iteration */
emilmont 1:fdd22bb7aa52 468 sum = 0;
emilmont 1:fdd22bb7aa52 469
emilmont 1:fdd22bb7aa52 470 /* srcBLen number of MACS should be performed */
emilmont 1:fdd22bb7aa52 471 k = srcBLen;
emilmont 1:fdd22bb7aa52 472
emilmont 1:fdd22bb7aa52 473 while(k > 0u)
emilmont 1:fdd22bb7aa52 474 {
emilmont 1:fdd22bb7aa52 475 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 476 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 477 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 478
emilmont 1:fdd22bb7aa52 479 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 480 k--;
emilmont 1:fdd22bb7aa52 481 }
emilmont 1:fdd22bb7aa52 482
emilmont 1:fdd22bb7aa52 483 /* Store the result in the accumulator in the destination buffer. */
emilmont 1:fdd22bb7aa52 484 *pOut++ = sum << 1;
emilmont 1:fdd22bb7aa52 485
emilmont 1:fdd22bb7aa52 486 /* Increment the MAC count */
emilmont 1:fdd22bb7aa52 487 count++;
emilmont 1:fdd22bb7aa52 488
emilmont 1:fdd22bb7aa52 489 /* Update the inputA and inputB pointers for next MAC calculation */
emilmont 1:fdd22bb7aa52 490 px = pIn1 + count;
emilmont 1:fdd22bb7aa52 491 py = pSrc2;
emilmont 1:fdd22bb7aa52 492
emilmont 1:fdd22bb7aa52 493 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 494 blkCnt--;
emilmont 1:fdd22bb7aa52 495 }
emilmont 1:fdd22bb7aa52 496 }
emilmont 1:fdd22bb7aa52 497
emilmont 1:fdd22bb7aa52 498
emilmont 1:fdd22bb7aa52 499 /* --------------------------
emilmont 1:fdd22bb7aa52 500 * Initializations of stage3
emilmont 1:fdd22bb7aa52 501 * -------------------------*/
emilmont 1:fdd22bb7aa52 502
emilmont 1:fdd22bb7aa52 503 /* sum += x[srcALen-srcBLen+1] * y[srcBLen-1] + x[srcALen-srcBLen+2] * y[srcBLen-2] +...+ x[srcALen-1] * y[1]
emilmont 1:fdd22bb7aa52 504 * sum += x[srcALen-srcBLen+2] * y[srcBLen-1] + x[srcALen-srcBLen+3] * y[srcBLen-2] +...+ x[srcALen-1] * y[2]
emilmont 1:fdd22bb7aa52 505 * ....
emilmont 1:fdd22bb7aa52 506 * sum += x[srcALen-2] * y[srcBLen-1] + x[srcALen-1] * y[srcBLen-2]
emilmont 1:fdd22bb7aa52 507 * sum += x[srcALen-1] * y[srcBLen-1]
emilmont 1:fdd22bb7aa52 508 */
emilmont 1:fdd22bb7aa52 509
emilmont 1:fdd22bb7aa52 510 /* In this stage the MAC operations are decreased by 1 for every iteration.
emilmont 1:fdd22bb7aa52 511 The count variable holds the number of MAC operations performed */
emilmont 1:fdd22bb7aa52 512 count = srcBLen - 1u;
emilmont 1:fdd22bb7aa52 513
emilmont 1:fdd22bb7aa52 514 /* Working pointer of inputA */
emilmont 1:fdd22bb7aa52 515 pSrc1 = (pIn1 + srcALen) - (srcBLen - 1u);
emilmont 1:fdd22bb7aa52 516 px = pSrc1;
emilmont 1:fdd22bb7aa52 517
emilmont 1:fdd22bb7aa52 518 /* Working pointer of inputB */
emilmont 1:fdd22bb7aa52 519 pSrc2 = pIn2 + (srcBLen - 1u);
emilmont 1:fdd22bb7aa52 520 py = pSrc2;
emilmont 1:fdd22bb7aa52 521
emilmont 1:fdd22bb7aa52 522 /* -------------------
emilmont 1:fdd22bb7aa52 523 * Stage3 process
emilmont 1:fdd22bb7aa52 524 * ------------------*/
emilmont 1:fdd22bb7aa52 525
emilmont 1:fdd22bb7aa52 526 while(blockSize3 > 0)
emilmont 1:fdd22bb7aa52 527 {
emilmont 1:fdd22bb7aa52 528 /* Accumulator is made zero for every iteration */
emilmont 1:fdd22bb7aa52 529 sum = 0;
emilmont 1:fdd22bb7aa52 530
emilmont 1:fdd22bb7aa52 531 /* Apply loop unrolling and compute 4 MACs simultaneously. */
emilmont 1:fdd22bb7aa52 532 k = count >> 2u;
emilmont 1:fdd22bb7aa52 533
emilmont 1:fdd22bb7aa52 534 /* First part of the processing with loop unrolling. Compute 4 MACs at a time.
emilmont 1:fdd22bb7aa52 535 ** a second loop below computes MACs for the remaining 1 to 3 samples. */
emilmont 1:fdd22bb7aa52 536 while(k > 0u)
emilmont 1:fdd22bb7aa52 537 {
emilmont 1:fdd22bb7aa52 538 /* sum += x[srcALen - srcBLen + 1] * y[srcBLen - 1] */
emilmont 1:fdd22bb7aa52 539 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 540 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 541
emilmont 1:fdd22bb7aa52 542 /* sum += x[srcALen - srcBLen + 2] * y[srcBLen - 2] */
emilmont 1:fdd22bb7aa52 543 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 544 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 545
emilmont 1:fdd22bb7aa52 546 /* sum += x[srcALen - srcBLen + 3] * y[srcBLen - 3] */
emilmont 1:fdd22bb7aa52 547 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 548 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 549
emilmont 1:fdd22bb7aa52 550 /* sum += x[srcALen - srcBLen + 4] * y[srcBLen - 4] */
emilmont 1:fdd22bb7aa52 551 sum = (q31_t) ((((q63_t) sum << 32) +
emilmont 1:fdd22bb7aa52 552 ((q63_t) * px++ * (*py--))) >> 32);
emilmont 1:fdd22bb7aa52 553
emilmont 1:fdd22bb7aa52 554 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 555 k--;
emilmont 1:fdd22bb7aa52 556 }
emilmont 1:fdd22bb7aa52 557
emilmont 1:fdd22bb7aa52 558 /* If the count is not a multiple of 4, compute any remaining MACs here.
emilmont 1:fdd22bb7aa52 559 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 560 k = count % 0x4u;
emilmont 1:fdd22bb7aa52 561
emilmont 1:fdd22bb7aa52 562 while(k > 0u)
emilmont 1:fdd22bb7aa52 563 {
emilmont 1:fdd22bb7aa52 564 /* Perform the multiply-accumulates */
emilmont 1:fdd22bb7aa52 565 /* sum += x[srcALen-1] * y[srcBLen-1] */
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 /* Store the result in the accumulator in the destination buffer. */
emilmont 1:fdd22bb7aa52 574 *pOut++ = sum << 1;
emilmont 1:fdd22bb7aa52 575
emilmont 1:fdd22bb7aa52 576 /* Update the inputA and inputB pointers for next MAC calculation */
emilmont 1:fdd22bb7aa52 577 px = ++pSrc1;
emilmont 1:fdd22bb7aa52 578 py = pSrc2;
emilmont 1:fdd22bb7aa52 579
emilmont 1:fdd22bb7aa52 580 /* Decrement the MAC count */
emilmont 1:fdd22bb7aa52 581 count--;
emilmont 1:fdd22bb7aa52 582
emilmont 1:fdd22bb7aa52 583 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 584 blockSize3--;
emilmont 1:fdd22bb7aa52 585
emilmont 1:fdd22bb7aa52 586 }
emilmont 1:fdd22bb7aa52 587
emilmont 1:fdd22bb7aa52 588 /* set status as ARM_MATH_SUCCESS */
emilmont 1:fdd22bb7aa52 589 status = ARM_MATH_SUCCESS;
emilmont 1:fdd22bb7aa52 590 }
emilmont 1:fdd22bb7aa52 591
emilmont 1:fdd22bb7aa52 592 /* Return to application */
emilmont 1:fdd22bb7aa52 593 return (status);
emilmont 1:fdd22bb7aa52 594
emilmont 1:fdd22bb7aa52 595 }
emilmont 1:fdd22bb7aa52 596
emilmont 1:fdd22bb7aa52 597 /**
emilmont 1:fdd22bb7aa52 598 * @} end of PartialConv group
emilmont 1:fdd22bb7aa52 599 */