mbed official / mbed-dsp

CMSIS DSP library

Dependents:   KL25Z_FFT_Demo Hat_Board_v5_1 KL25Z_FFT_Demo_tony KL25Z_FFT_Demo_tony ... more

Fork of mbed-dsp by mbed official

cmsis_dsp/FilteringFunctions/arm_lms_q15.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_lms_q15.c
emilmont 1:fdd22bb7aa52 9 *
emilmont 2:da51fb522205 10 * Description: Processing function for the Q15 LMS filter.
emilmont 1:fdd22bb7aa52 11 *
emilmont 1:fdd22bb7aa52 12 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
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 * Version 0.0.7 2010/06/10
emilmont 1:fdd22bb7aa52 33 * Misra-C changes done
emilmont 1:fdd22bb7aa52 34 * -------------------------------------------------------------------- */
emilmont 1:fdd22bb7aa52 35
emilmont 1:fdd22bb7aa52 36 #include "arm_math.h"
emilmont 1:fdd22bb7aa52 37 /**
emilmont 1:fdd22bb7aa52 38 * @ingroup groupFilters
emilmont 1:fdd22bb7aa52 39 */
emilmont 1:fdd22bb7aa52 40
emilmont 1:fdd22bb7aa52 41 /**
emilmont 1:fdd22bb7aa52 42 * @addtogroup LMS
emilmont 1:fdd22bb7aa52 43 * @{
emilmont 1:fdd22bb7aa52 44 */
emilmont 1:fdd22bb7aa52 45
emilmont 1:fdd22bb7aa52 46 /**
emilmont 1:fdd22bb7aa52 47 * @brief Processing function for Q15 LMS filter.
emilmont 1:fdd22bb7aa52 48 * @param[in] *S points to an instance of the Q15 LMS filter structure.
emilmont 1:fdd22bb7aa52 49 * @param[in] *pSrc points to the block of input data.
emilmont 1:fdd22bb7aa52 50 * @param[in] *pRef points to the block of reference data.
emilmont 1:fdd22bb7aa52 51 * @param[out] *pOut points to the block of output data.
emilmont 1:fdd22bb7aa52 52 * @param[out] *pErr points to the block of error data.
emilmont 1:fdd22bb7aa52 53 * @param[in] blockSize number of samples to process.
emilmont 1:fdd22bb7aa52 54 * @return none.
emilmont 1:fdd22bb7aa52 55 *
emilmont 1:fdd22bb7aa52 56 * \par Scaling and Overflow Behavior:
emilmont 1:fdd22bb7aa52 57 * The function is implemented using a 64-bit internal accumulator.
emilmont 1:fdd22bb7aa52 58 * Both coefficients and state variables are represented in 1.15 format and multiplications yield a 2.30 result.
emilmont 1:fdd22bb7aa52 59 * The 2.30 intermediate results are accumulated in a 64-bit accumulator in 34.30 format.
emilmont 1:fdd22bb7aa52 60 * There is no risk of internal overflow with this approach and the full precision of intermediate multiplications is preserved.
emilmont 1:fdd22bb7aa52 61 * After all additions have been performed, the accumulator is truncated to 34.15 format by discarding low 15 bits.
emilmont 1:fdd22bb7aa52 62 * Lastly, the accumulator is saturated to yield a result in 1.15 format.
emilmont 1:fdd22bb7aa52 63 *
emilmont 1:fdd22bb7aa52 64 * \par
emilmont 2:da51fb522205 65 * In this filter, filter coefficients are updated for each sample and the updation of filter cofficients are saturted.
emilmont 1:fdd22bb7aa52 66 *
emilmont 1:fdd22bb7aa52 67 */
emilmont 1:fdd22bb7aa52 68
emilmont 1:fdd22bb7aa52 69 void arm_lms_q15(
emilmont 1:fdd22bb7aa52 70 const arm_lms_instance_q15 * S,
emilmont 1:fdd22bb7aa52 71 q15_t * pSrc,
emilmont 1:fdd22bb7aa52 72 q15_t * pRef,
emilmont 1:fdd22bb7aa52 73 q15_t * pOut,
emilmont 1:fdd22bb7aa52 74 q15_t * pErr,
emilmont 1:fdd22bb7aa52 75 uint32_t blockSize)
emilmont 1:fdd22bb7aa52 76 {
emilmont 1:fdd22bb7aa52 77 q15_t *pState = S->pState; /* State pointer */
emilmont 1:fdd22bb7aa52 78 uint32_t numTaps = S->numTaps; /* Number of filter coefficients in the filter */
emilmont 1:fdd22bb7aa52 79 q15_t *pCoeffs = S->pCoeffs; /* Coefficient pointer */
emilmont 1:fdd22bb7aa52 80 q15_t *pStateCurnt; /* Points to the current sample of the state */
emilmont 1:fdd22bb7aa52 81 q15_t mu = S->mu; /* Adaptive factor */
emilmont 1:fdd22bb7aa52 82 q15_t *px; /* Temporary pointer for state */
emilmont 1:fdd22bb7aa52 83 q15_t *pb; /* Temporary pointer for coefficient buffer */
emilmont 1:fdd22bb7aa52 84 uint32_t tapCnt, blkCnt; /* Loop counters */
emilmont 1:fdd22bb7aa52 85 q63_t acc; /* Accumulator */
emilmont 1:fdd22bb7aa52 86 q15_t e = 0; /* error of data sample */
emilmont 1:fdd22bb7aa52 87 q15_t alpha; /* Intermediate constant for taps update */
emilmont 1:fdd22bb7aa52 88 q31_t acc_l, acc_h;
emilmont 1:fdd22bb7aa52 89 int32_t lShift = (15 - (int32_t) S->postShift); /* Post shift */
emilmont 1:fdd22bb7aa52 90 int32_t uShift = (32 - lShift);
emilmont 1:fdd22bb7aa52 91
emilmont 1:fdd22bb7aa52 92
emilmont 1:fdd22bb7aa52 93 #ifndef ARM_MATH_CM0
emilmont 1:fdd22bb7aa52 94
emilmont 1:fdd22bb7aa52 95 /* Run the below code for Cortex-M4 and Cortex-M3 */
emilmont 1:fdd22bb7aa52 96
emilmont 1:fdd22bb7aa52 97 q31_t coef; /* Teporary variable for coefficient */
emilmont 1:fdd22bb7aa52 98
emilmont 1:fdd22bb7aa52 99 /* S->pState points to buffer which contains previous frame (numTaps - 1) samples */
emilmont 1:fdd22bb7aa52 100 /* pStateCurnt points to the location where the new input data should be written */
emilmont 1:fdd22bb7aa52 101 pStateCurnt = &(S->pState[(numTaps - 1u)]);
emilmont 1:fdd22bb7aa52 102
emilmont 1:fdd22bb7aa52 103 /* Initializing blkCnt with blockSize */
emilmont 1:fdd22bb7aa52 104 blkCnt = blockSize;
emilmont 1:fdd22bb7aa52 105
emilmont 1:fdd22bb7aa52 106 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 107 {
emilmont 1:fdd22bb7aa52 108 /* Copy the new input sample into the state buffer */
emilmont 1:fdd22bb7aa52 109 *pStateCurnt++ = *pSrc++;
emilmont 1:fdd22bb7aa52 110
emilmont 1:fdd22bb7aa52 111 /* Initialize state pointer */
emilmont 1:fdd22bb7aa52 112 px = pState;
emilmont 1:fdd22bb7aa52 113
emilmont 1:fdd22bb7aa52 114 /* Initialize coefficient pointer */
emilmont 1:fdd22bb7aa52 115 pb = pCoeffs;
emilmont 1:fdd22bb7aa52 116
emilmont 1:fdd22bb7aa52 117 /* Set the accumulator to zero */
emilmont 1:fdd22bb7aa52 118 acc = 0;
emilmont 1:fdd22bb7aa52 119
emilmont 1:fdd22bb7aa52 120 /* Loop unrolling. Process 4 taps at a time. */
emilmont 1:fdd22bb7aa52 121 tapCnt = numTaps >> 2u;
emilmont 1:fdd22bb7aa52 122
emilmont 1:fdd22bb7aa52 123 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 124 {
emilmont 1:fdd22bb7aa52 125 /* acc += b[N] * x[n-N] + b[N-1] * x[n-N-1] */
emilmont 1:fdd22bb7aa52 126 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 127 #ifndef UNALIGNED_SUPPORT_DISABLE
emilmont 1:fdd22bb7aa52 128
emilmont 1:fdd22bb7aa52 129 acc = __SMLALD(*__SIMD32(px)++, (*__SIMD32(pb)++), acc);
emilmont 1:fdd22bb7aa52 130 acc = __SMLALD(*__SIMD32(px)++, (*__SIMD32(pb)++), acc);
emilmont 1:fdd22bb7aa52 131
emilmont 1:fdd22bb7aa52 132 #else
emilmont 1:fdd22bb7aa52 133
emilmont 1:fdd22bb7aa52 134 acc += (q63_t) (((q31_t) (*px++) * (*pb++)));
emilmont 1:fdd22bb7aa52 135 acc += (q63_t) (((q31_t) (*px++) * (*pb++)));
emilmont 1:fdd22bb7aa52 136 acc += (q63_t) (((q31_t) (*px++) * (*pb++)));
emilmont 1:fdd22bb7aa52 137 acc += (q63_t) (((q31_t) (*px++) * (*pb++)));
emilmont 1:fdd22bb7aa52 138
emilmont 1:fdd22bb7aa52 139
emilmont 2:da51fb522205 140 #endif /* #ifndef UNALIGNED_SUPPORT_DISABLE */
emilmont 1:fdd22bb7aa52 141
emilmont 1:fdd22bb7aa52 142 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 143 tapCnt--;
emilmont 1:fdd22bb7aa52 144 }
emilmont 1:fdd22bb7aa52 145
emilmont 1:fdd22bb7aa52 146 /* If the filter length is not a multiple of 4, compute the remaining filter taps */
emilmont 1:fdd22bb7aa52 147 tapCnt = numTaps % 0x4u;
emilmont 1:fdd22bb7aa52 148
emilmont 1:fdd22bb7aa52 149 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 150 {
emilmont 1:fdd22bb7aa52 151 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 152 acc += (q63_t) (((q31_t) (*px++) * (*pb++)));
emilmont 1:fdd22bb7aa52 153
emilmont 1:fdd22bb7aa52 154 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 155 tapCnt--;
emilmont 1:fdd22bb7aa52 156 }
emilmont 1:fdd22bb7aa52 157
emilmont 1:fdd22bb7aa52 158 /* Calc lower part of acc */
emilmont 1:fdd22bb7aa52 159 acc_l = acc & 0xffffffff;
emilmont 1:fdd22bb7aa52 160
emilmont 1:fdd22bb7aa52 161 /* Calc upper part of acc */
emilmont 1:fdd22bb7aa52 162 acc_h = (acc >> 32) & 0xffffffff;
emilmont 1:fdd22bb7aa52 163
emilmont 1:fdd22bb7aa52 164 /* Apply shift for lower part of acc and upper part of acc */
emilmont 1:fdd22bb7aa52 165 acc = (uint32_t) acc_l >> lShift | acc_h << uShift;
emilmont 1:fdd22bb7aa52 166
emilmont 1:fdd22bb7aa52 167 /* Converting the result to 1.15 format and saturate the output */
emilmont 1:fdd22bb7aa52 168 acc = __SSAT(acc, 16);
emilmont 1:fdd22bb7aa52 169
emilmont 1:fdd22bb7aa52 170 /* Store the result from accumulator into the destination buffer. */
emilmont 1:fdd22bb7aa52 171 *pOut++ = (q15_t) acc;
emilmont 1:fdd22bb7aa52 172
emilmont 1:fdd22bb7aa52 173 /* Compute and store error */
emilmont 1:fdd22bb7aa52 174 e = *pRef++ - (q15_t) acc;
emilmont 1:fdd22bb7aa52 175
emilmont 1:fdd22bb7aa52 176 *pErr++ = (q15_t) e;
emilmont 1:fdd22bb7aa52 177
emilmont 1:fdd22bb7aa52 178 /* Compute alpha i.e. intermediate constant for taps update */
emilmont 1:fdd22bb7aa52 179 alpha = (q15_t) (((q31_t) e * (mu)) >> 15);
emilmont 1:fdd22bb7aa52 180
emilmont 1:fdd22bb7aa52 181 /* Initialize state pointer */
emilmont 1:fdd22bb7aa52 182 /* Advance state pointer by 1 for the next sample */
emilmont 1:fdd22bb7aa52 183 px = pState++;
emilmont 1:fdd22bb7aa52 184
emilmont 1:fdd22bb7aa52 185 /* Initialize coefficient pointer */
emilmont 1:fdd22bb7aa52 186 pb = pCoeffs;
emilmont 1:fdd22bb7aa52 187
emilmont 1:fdd22bb7aa52 188 /* Loop unrolling. Process 4 taps at a time. */
emilmont 1:fdd22bb7aa52 189 tapCnt = numTaps >> 2u;
emilmont 1:fdd22bb7aa52 190
emilmont 1:fdd22bb7aa52 191 /* Update filter coefficients */
emilmont 1:fdd22bb7aa52 192 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 193 {
emilmont 1:fdd22bb7aa52 194 coef = (q31_t) * pb + (((q31_t) alpha * (*px++)) >> 15);
emilmont 1:fdd22bb7aa52 195 *pb++ = (q15_t) __SSAT((coef), 16);
emilmont 1:fdd22bb7aa52 196 coef = (q31_t) * pb + (((q31_t) alpha * (*px++)) >> 15);
emilmont 1:fdd22bb7aa52 197 *pb++ = (q15_t) __SSAT((coef), 16);
emilmont 1:fdd22bb7aa52 198 coef = (q31_t) * pb + (((q31_t) alpha * (*px++)) >> 15);
emilmont 1:fdd22bb7aa52 199 *pb++ = (q15_t) __SSAT((coef), 16);
emilmont 1:fdd22bb7aa52 200 coef = (q31_t) * pb + (((q31_t) alpha * (*px++)) >> 15);
emilmont 1:fdd22bb7aa52 201 *pb++ = (q15_t) __SSAT((coef), 16);
emilmont 1:fdd22bb7aa52 202
emilmont 1:fdd22bb7aa52 203 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 204 tapCnt--;
emilmont 1:fdd22bb7aa52 205 }
emilmont 1:fdd22bb7aa52 206
emilmont 1:fdd22bb7aa52 207 /* If the filter length is not a multiple of 4, compute the remaining filter taps */
emilmont 1:fdd22bb7aa52 208 tapCnt = numTaps % 0x4u;
emilmont 1:fdd22bb7aa52 209
emilmont 1:fdd22bb7aa52 210 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 211 {
emilmont 1:fdd22bb7aa52 212 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 213 coef = (q31_t) * pb + (((q31_t) alpha * (*px++)) >> 15);
emilmont 1:fdd22bb7aa52 214 *pb++ = (q15_t) __SSAT((coef), 16);
emilmont 1:fdd22bb7aa52 215
emilmont 1:fdd22bb7aa52 216 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 217 tapCnt--;
emilmont 1:fdd22bb7aa52 218 }
emilmont 1:fdd22bb7aa52 219
emilmont 1:fdd22bb7aa52 220 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 221 blkCnt--;
emilmont 1:fdd22bb7aa52 222
emilmont 1:fdd22bb7aa52 223 }
emilmont 1:fdd22bb7aa52 224
emilmont 1:fdd22bb7aa52 225 /* Processing is complete. Now copy the last numTaps - 1 samples to the
emilmont 1:fdd22bb7aa52 226 satrt of the state buffer. This prepares the state buffer for the
emilmont 1:fdd22bb7aa52 227 next function call. */
emilmont 1:fdd22bb7aa52 228
emilmont 1:fdd22bb7aa52 229 /* Points to the start of the pState buffer */
emilmont 1:fdd22bb7aa52 230 pStateCurnt = S->pState;
emilmont 1:fdd22bb7aa52 231
emilmont 1:fdd22bb7aa52 232 /* Calculation of count for copying integer writes */
emilmont 1:fdd22bb7aa52 233 tapCnt = (numTaps - 1u) >> 2;
emilmont 1:fdd22bb7aa52 234
emilmont 1:fdd22bb7aa52 235 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 236 {
emilmont 1:fdd22bb7aa52 237
emilmont 1:fdd22bb7aa52 238 #ifndef UNALIGNED_SUPPORT_DISABLE
emilmont 1:fdd22bb7aa52 239
emilmont 1:fdd22bb7aa52 240 *__SIMD32(pStateCurnt)++ = *__SIMD32(pState)++;
emilmont 1:fdd22bb7aa52 241 *__SIMD32(pStateCurnt)++ = *__SIMD32(pState)++;
emilmont 1:fdd22bb7aa52 242 #else
emilmont 1:fdd22bb7aa52 243 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 244 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 245 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 246 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 247 #endif
emilmont 1:fdd22bb7aa52 248
emilmont 1:fdd22bb7aa52 249 tapCnt--;
emilmont 1:fdd22bb7aa52 250
emilmont 1:fdd22bb7aa52 251 }
emilmont 1:fdd22bb7aa52 252
emilmont 1:fdd22bb7aa52 253 /* Calculation of count for remaining q15_t data */
emilmont 1:fdd22bb7aa52 254 tapCnt = (numTaps - 1u) % 0x4u;
emilmont 1:fdd22bb7aa52 255
emilmont 1:fdd22bb7aa52 256 /* copy data */
emilmont 1:fdd22bb7aa52 257 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 258 {
emilmont 1:fdd22bb7aa52 259 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 260
emilmont 1:fdd22bb7aa52 261 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 262 tapCnt--;
emilmont 1:fdd22bb7aa52 263 }
emilmont 1:fdd22bb7aa52 264
emilmont 1:fdd22bb7aa52 265 #else
emilmont 1:fdd22bb7aa52 266
emilmont 1:fdd22bb7aa52 267 /* Run the below code for Cortex-M0 */
emilmont 1:fdd22bb7aa52 268
emilmont 1:fdd22bb7aa52 269 /* S->pState points to buffer which contains previous frame (numTaps - 1) samples */
emilmont 1:fdd22bb7aa52 270 /* pStateCurnt points to the location where the new input data should be written */
emilmont 1:fdd22bb7aa52 271 pStateCurnt = &(S->pState[(numTaps - 1u)]);
emilmont 1:fdd22bb7aa52 272
emilmont 1:fdd22bb7aa52 273 /* Loop over blockSize number of values */
emilmont 1:fdd22bb7aa52 274 blkCnt = blockSize;
emilmont 1:fdd22bb7aa52 275
emilmont 1:fdd22bb7aa52 276 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 277 {
emilmont 1:fdd22bb7aa52 278 /* Copy the new input sample into the state buffer */
emilmont 1:fdd22bb7aa52 279 *pStateCurnt++ = *pSrc++;
emilmont 1:fdd22bb7aa52 280
emilmont 1:fdd22bb7aa52 281 /* Initialize pState pointer */
emilmont 1:fdd22bb7aa52 282 px = pState;
emilmont 1:fdd22bb7aa52 283
emilmont 1:fdd22bb7aa52 284 /* Initialize pCoeffs pointer */
emilmont 1:fdd22bb7aa52 285 pb = pCoeffs;
emilmont 1:fdd22bb7aa52 286
emilmont 1:fdd22bb7aa52 287 /* Set the accumulator to zero */
emilmont 1:fdd22bb7aa52 288 acc = 0;
emilmont 1:fdd22bb7aa52 289
emilmont 1:fdd22bb7aa52 290 /* Loop over numTaps number of values */
emilmont 1:fdd22bb7aa52 291 tapCnt = numTaps;
emilmont 1:fdd22bb7aa52 292
emilmont 1:fdd22bb7aa52 293 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 294 {
emilmont 1:fdd22bb7aa52 295 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 296 acc += (q63_t) ((q31_t) (*px++) * (*pb++));
emilmont 1:fdd22bb7aa52 297
emilmont 1:fdd22bb7aa52 298 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 299 tapCnt--;
emilmont 1:fdd22bb7aa52 300 }
emilmont 1:fdd22bb7aa52 301
emilmont 1:fdd22bb7aa52 302 /* Calc lower part of acc */
emilmont 1:fdd22bb7aa52 303 acc_l = acc & 0xffffffff;
emilmont 1:fdd22bb7aa52 304
emilmont 1:fdd22bb7aa52 305 /* Calc upper part of acc */
emilmont 1:fdd22bb7aa52 306 acc_h = (acc >> 32) & 0xffffffff;
emilmont 1:fdd22bb7aa52 307
emilmont 1:fdd22bb7aa52 308 /* Apply shift for lower part of acc and upper part of acc */
emilmont 1:fdd22bb7aa52 309 acc = (uint32_t) acc_l >> lShift | acc_h << uShift;
emilmont 1:fdd22bb7aa52 310
emilmont 1:fdd22bb7aa52 311 /* Converting the result to 1.15 format and saturate the output */
emilmont 1:fdd22bb7aa52 312 acc = __SSAT(acc, 16);
emilmont 1:fdd22bb7aa52 313
emilmont 1:fdd22bb7aa52 314 /* Store the result from accumulator into the destination buffer. */
emilmont 1:fdd22bb7aa52 315 *pOut++ = (q15_t) acc;
emilmont 1:fdd22bb7aa52 316
emilmont 1:fdd22bb7aa52 317 /* Compute and store error */
emilmont 1:fdd22bb7aa52 318 e = *pRef++ - (q15_t) acc;
emilmont 1:fdd22bb7aa52 319
emilmont 1:fdd22bb7aa52 320 *pErr++ = (q15_t) e;
emilmont 1:fdd22bb7aa52 321
emilmont 1:fdd22bb7aa52 322 /* Compute alpha i.e. intermediate constant for taps update */
emilmont 1:fdd22bb7aa52 323 alpha = (q15_t) (((q31_t) e * (mu)) >> 15);
emilmont 1:fdd22bb7aa52 324
emilmont 1:fdd22bb7aa52 325 /* Initialize pState pointer */
emilmont 1:fdd22bb7aa52 326 /* Advance state pointer by 1 for the next sample */
emilmont 1:fdd22bb7aa52 327 px = pState++;
emilmont 1:fdd22bb7aa52 328
emilmont 1:fdd22bb7aa52 329 /* Initialize pCoeffs pointer */
emilmont 1:fdd22bb7aa52 330 pb = pCoeffs;
emilmont 1:fdd22bb7aa52 331
emilmont 1:fdd22bb7aa52 332 /* Loop over numTaps number of values */
emilmont 1:fdd22bb7aa52 333 tapCnt = numTaps;
emilmont 1:fdd22bb7aa52 334
emilmont 1:fdd22bb7aa52 335 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 336 {
emilmont 1:fdd22bb7aa52 337 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 338 *pb++ += (q15_t) (((q31_t) alpha * (*px++)) >> 15);
emilmont 1:fdd22bb7aa52 339
emilmont 1:fdd22bb7aa52 340 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 341 tapCnt--;
emilmont 1:fdd22bb7aa52 342 }
emilmont 1:fdd22bb7aa52 343
emilmont 1:fdd22bb7aa52 344 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 345 blkCnt--;
emilmont 1:fdd22bb7aa52 346
emilmont 1:fdd22bb7aa52 347 }
emilmont 1:fdd22bb7aa52 348
emilmont 1:fdd22bb7aa52 349 /* Processing is complete. Now copy the last numTaps - 1 samples to the
emilmont 1:fdd22bb7aa52 350 start of the state buffer. This prepares the state buffer for the
emilmont 1:fdd22bb7aa52 351 next function call. */
emilmont 1:fdd22bb7aa52 352
emilmont 1:fdd22bb7aa52 353 /* Points to the start of the pState buffer */
emilmont 1:fdd22bb7aa52 354 pStateCurnt = S->pState;
emilmont 1:fdd22bb7aa52 355
emilmont 1:fdd22bb7aa52 356 /* Copy (numTaps - 1u) samples */
emilmont 1:fdd22bb7aa52 357 tapCnt = (numTaps - 1u);
emilmont 1:fdd22bb7aa52 358
emilmont 1:fdd22bb7aa52 359 /* Copy the data */
emilmont 1:fdd22bb7aa52 360 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 361 {
emilmont 1:fdd22bb7aa52 362 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 363
emilmont 1:fdd22bb7aa52 364 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 365 tapCnt--;
emilmont 1:fdd22bb7aa52 366 }
emilmont 1:fdd22bb7aa52 367
emilmont 1:fdd22bb7aa52 368 #endif /* #ifndef ARM_MATH_CM0 */
emilmont 1:fdd22bb7aa52 369
emilmont 1:fdd22bb7aa52 370 }
emilmont 1:fdd22bb7aa52 371
emilmont 1:fdd22bb7aa52 372 /**
emilmont 1:fdd22bb7aa52 373 * @} end of LMS group
emilmont 1:fdd22bb7aa52 374 */