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_fir_decimate_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_fir_decimate_q31.c
emilmont 1:fdd22bb7aa52 9 *
emilmont 2:da51fb522205 10 * Description: Q31 FIR Decimator.
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 /**
emilmont 1:fdd22bb7aa52 39 * @ingroup groupFilters
emilmont 1:fdd22bb7aa52 40 */
emilmont 1:fdd22bb7aa52 41
emilmont 1:fdd22bb7aa52 42 /**
emilmont 1:fdd22bb7aa52 43 * @addtogroup FIR_decimate
emilmont 1:fdd22bb7aa52 44 * @{
emilmont 1:fdd22bb7aa52 45 */
emilmont 1:fdd22bb7aa52 46
emilmont 1:fdd22bb7aa52 47 /**
emilmont 1:fdd22bb7aa52 48 * @brief Processing function for the Q31 FIR decimator.
emilmont 1:fdd22bb7aa52 49 * @param[in] *S points to an instance of the Q31 FIR decimator structure.
emilmont 1:fdd22bb7aa52 50 * @param[in] *pSrc points to the block of input data.
emilmont 1:fdd22bb7aa52 51 * @param[out] *pDst points to the block of output data
emilmont 1:fdd22bb7aa52 52 * @param[in] blockSize number of input samples to process per call.
emilmont 1:fdd22bb7aa52 53 * @return none
emilmont 1:fdd22bb7aa52 54 *
emilmont 1:fdd22bb7aa52 55 * <b>Scaling and Overflow Behavior:</b>
emilmont 1:fdd22bb7aa52 56 * \par
emilmont 1:fdd22bb7aa52 57 * The function is implemented using an internal 64-bit accumulator.
emilmont 1:fdd22bb7aa52 58 * The accumulator has a 2.62 format and maintains full precision of the intermediate multiplication results but provides only a single guard bit.
emilmont 1:fdd22bb7aa52 59 * Thus, if the accumulator result overflows it wraps around rather than clip.
emilmont 1:fdd22bb7aa52 60 * In order to avoid overflows completely the input signal must be scaled down by log2(numTaps) bits (where log2 is read as log to the base 2).
emilmont 1:fdd22bb7aa52 61 * After all multiply-accumulates are performed, the 2.62 accumulator is truncated to 1.32 format and then saturated to 1.31 format.
emilmont 1:fdd22bb7aa52 62 *
emilmont 1:fdd22bb7aa52 63 * \par
emilmont 1:fdd22bb7aa52 64 * Refer to the function <code>arm_fir_decimate_fast_q31()</code> for a faster but less precise implementation of this function for Cortex-M3 and Cortex-M4.
emilmont 1:fdd22bb7aa52 65 */
emilmont 1:fdd22bb7aa52 66
emilmont 1:fdd22bb7aa52 67 void arm_fir_decimate_q31(
emilmont 1:fdd22bb7aa52 68 const arm_fir_decimate_instance_q31 * S,
emilmont 1:fdd22bb7aa52 69 q31_t * pSrc,
emilmont 1:fdd22bb7aa52 70 q31_t * pDst,
emilmont 1:fdd22bb7aa52 71 uint32_t blockSize)
emilmont 1:fdd22bb7aa52 72 {
emilmont 1:fdd22bb7aa52 73 q31_t *pState = S->pState; /* State pointer */
emilmont 1:fdd22bb7aa52 74 q31_t *pCoeffs = S->pCoeffs; /* Coefficient pointer */
emilmont 1:fdd22bb7aa52 75 q31_t *pStateCurnt; /* Points to the current sample of the state */
emilmont 1:fdd22bb7aa52 76 q31_t x0, c0; /* Temporary variables to hold state and coefficient values */
emilmont 1:fdd22bb7aa52 77 q31_t *px; /* Temporary pointers for state buffer */
emilmont 1:fdd22bb7aa52 78 q31_t *pb; /* Temporary pointers for coefficient buffer */
emilmont 1:fdd22bb7aa52 79 q63_t sum0; /* Accumulator */
emilmont 1:fdd22bb7aa52 80 uint32_t numTaps = S->numTaps; /* Number of taps */
emilmont 1:fdd22bb7aa52 81 uint32_t i, tapCnt, blkCnt, outBlockSize = blockSize / S->M; /* Loop counters */
emilmont 1:fdd22bb7aa52 82
emilmont 1:fdd22bb7aa52 83
emilmont 1:fdd22bb7aa52 84 #ifndef ARM_MATH_CM0
emilmont 1:fdd22bb7aa52 85
emilmont 1:fdd22bb7aa52 86 /* Run the below code for Cortex-M4 and Cortex-M3 */
emilmont 1:fdd22bb7aa52 87
emilmont 1:fdd22bb7aa52 88 /* S->pState buffer contains previous frame (numTaps - 1) samples */
emilmont 1:fdd22bb7aa52 89 /* pStateCurnt points to the location where the new input data should be written */
emilmont 1:fdd22bb7aa52 90 pStateCurnt = S->pState + (numTaps - 1u);
emilmont 1:fdd22bb7aa52 91
emilmont 1:fdd22bb7aa52 92 /* Total number of output samples to be computed */
emilmont 1:fdd22bb7aa52 93 blkCnt = outBlockSize;
emilmont 1:fdd22bb7aa52 94
emilmont 1:fdd22bb7aa52 95 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 96 {
emilmont 1:fdd22bb7aa52 97 /* Copy decimation factor number of new input samples into the state buffer */
emilmont 1:fdd22bb7aa52 98 i = S->M;
emilmont 1:fdd22bb7aa52 99
emilmont 1:fdd22bb7aa52 100 do
emilmont 1:fdd22bb7aa52 101 {
emilmont 1:fdd22bb7aa52 102 *pStateCurnt++ = *pSrc++;
emilmont 1:fdd22bb7aa52 103
emilmont 1:fdd22bb7aa52 104 } while(--i);
emilmont 1:fdd22bb7aa52 105
emilmont 1:fdd22bb7aa52 106 /* Set accumulator to zero */
emilmont 1:fdd22bb7aa52 107 sum0 = 0;
emilmont 1:fdd22bb7aa52 108
emilmont 1:fdd22bb7aa52 109 /* Initialize state pointer */
emilmont 1:fdd22bb7aa52 110 px = pState;
emilmont 1:fdd22bb7aa52 111
emilmont 1:fdd22bb7aa52 112 /* Initialize coeff pointer */
emilmont 1:fdd22bb7aa52 113 pb = pCoeffs;
emilmont 1:fdd22bb7aa52 114
emilmont 1:fdd22bb7aa52 115 /* Loop unrolling. Process 4 taps at a time. */
emilmont 1:fdd22bb7aa52 116 tapCnt = numTaps >> 2;
emilmont 1:fdd22bb7aa52 117
emilmont 1:fdd22bb7aa52 118 /* Loop over the number of taps. Unroll by a factor of 4.
emilmont 1:fdd22bb7aa52 119 ** Repeat until we've computed numTaps-4 coefficients. */
emilmont 1:fdd22bb7aa52 120 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 121 {
emilmont 1:fdd22bb7aa52 122 /* Read the b[numTaps-1] coefficient */
emilmont 1:fdd22bb7aa52 123 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 124
emilmont 1:fdd22bb7aa52 125 /* Read x[n-numTaps-1] sample */
emilmont 1:fdd22bb7aa52 126 x0 = *(px++);
emilmont 1:fdd22bb7aa52 127
emilmont 1:fdd22bb7aa52 128 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 129 sum0 += (q63_t) x0 *c0;
emilmont 1:fdd22bb7aa52 130
emilmont 1:fdd22bb7aa52 131 /* Read the b[numTaps-2] coefficient */
emilmont 1:fdd22bb7aa52 132 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 133
emilmont 1:fdd22bb7aa52 134 /* Read x[n-numTaps-2] sample */
emilmont 1:fdd22bb7aa52 135 x0 = *(px++);
emilmont 1:fdd22bb7aa52 136
emilmont 1:fdd22bb7aa52 137 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 138 sum0 += (q63_t) x0 *c0;
emilmont 1:fdd22bb7aa52 139
emilmont 1:fdd22bb7aa52 140 /* Read the b[numTaps-3] coefficient */
emilmont 1:fdd22bb7aa52 141 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 142
emilmont 1:fdd22bb7aa52 143 /* Read x[n-numTaps-3] sample */
emilmont 1:fdd22bb7aa52 144 x0 = *(px++);
emilmont 1:fdd22bb7aa52 145
emilmont 1:fdd22bb7aa52 146 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 147 sum0 += (q63_t) x0 *c0;
emilmont 1:fdd22bb7aa52 148
emilmont 1:fdd22bb7aa52 149 /* Read the b[numTaps-4] coefficient */
emilmont 1:fdd22bb7aa52 150 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 151
emilmont 1:fdd22bb7aa52 152 /* Read x[n-numTaps-4] sample */
emilmont 1:fdd22bb7aa52 153 x0 = *(px++);
emilmont 1:fdd22bb7aa52 154
emilmont 1:fdd22bb7aa52 155 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 156 sum0 += (q63_t) x0 *c0;
emilmont 1:fdd22bb7aa52 157
emilmont 1:fdd22bb7aa52 158 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 159 tapCnt--;
emilmont 1:fdd22bb7aa52 160 }
emilmont 1:fdd22bb7aa52 161
emilmont 1:fdd22bb7aa52 162 /* If the filter length is not a multiple of 4, compute the remaining filter taps */
emilmont 1:fdd22bb7aa52 163 tapCnt = numTaps % 0x4u;
emilmont 1:fdd22bb7aa52 164
emilmont 1:fdd22bb7aa52 165 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 166 {
emilmont 1:fdd22bb7aa52 167 /* Read coefficients */
emilmont 1:fdd22bb7aa52 168 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 169
emilmont 1:fdd22bb7aa52 170 /* Fetch 1 state variable */
emilmont 1:fdd22bb7aa52 171 x0 = *(px++);
emilmont 1:fdd22bb7aa52 172
emilmont 1:fdd22bb7aa52 173 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 174 sum0 += (q63_t) x0 *c0;
emilmont 1:fdd22bb7aa52 175
emilmont 1:fdd22bb7aa52 176 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 177 tapCnt--;
emilmont 1:fdd22bb7aa52 178 }
emilmont 1:fdd22bb7aa52 179
emilmont 1:fdd22bb7aa52 180 /* Advance the state pointer by the decimation factor
emilmont 1:fdd22bb7aa52 181 * to process the next group of decimation factor number samples */
emilmont 1:fdd22bb7aa52 182 pState = pState + S->M;
emilmont 1:fdd22bb7aa52 183
emilmont 1:fdd22bb7aa52 184 /* The result is in the accumulator, store in the destination buffer. */
emilmont 1:fdd22bb7aa52 185 *pDst++ = (q31_t) (sum0 >> 31);
emilmont 1:fdd22bb7aa52 186
emilmont 1:fdd22bb7aa52 187 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 188 blkCnt--;
emilmont 1:fdd22bb7aa52 189 }
emilmont 1:fdd22bb7aa52 190
emilmont 1:fdd22bb7aa52 191 /* Processing is complete.
emilmont 1:fdd22bb7aa52 192 ** Now copy the last numTaps - 1 samples to the satrt of the state buffer.
emilmont 1:fdd22bb7aa52 193 ** This prepares the state buffer for the next function call. */
emilmont 1:fdd22bb7aa52 194
emilmont 1:fdd22bb7aa52 195 /* Points to the start of the state buffer */
emilmont 1:fdd22bb7aa52 196 pStateCurnt = S->pState;
emilmont 1:fdd22bb7aa52 197
emilmont 1:fdd22bb7aa52 198 i = (numTaps - 1u) >> 2u;
emilmont 1:fdd22bb7aa52 199
emilmont 1:fdd22bb7aa52 200 /* copy data */
emilmont 1:fdd22bb7aa52 201 while(i > 0u)
emilmont 1:fdd22bb7aa52 202 {
emilmont 1:fdd22bb7aa52 203 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 204 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 205 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 206 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 207
emilmont 1:fdd22bb7aa52 208 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 209 i--;
emilmont 1:fdd22bb7aa52 210 }
emilmont 1:fdd22bb7aa52 211
emilmont 1:fdd22bb7aa52 212 i = (numTaps - 1u) % 0x04u;
emilmont 1:fdd22bb7aa52 213
emilmont 1:fdd22bb7aa52 214 /* copy data */
emilmont 1:fdd22bb7aa52 215 while(i > 0u)
emilmont 1:fdd22bb7aa52 216 {
emilmont 1:fdd22bb7aa52 217 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 218
emilmont 1:fdd22bb7aa52 219 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 220 i--;
emilmont 1:fdd22bb7aa52 221 }
emilmont 1:fdd22bb7aa52 222
emilmont 1:fdd22bb7aa52 223 #else
emilmont 1:fdd22bb7aa52 224
emilmont 1:fdd22bb7aa52 225 /* Run the below code for Cortex-M0 */
emilmont 1:fdd22bb7aa52 226
emilmont 1:fdd22bb7aa52 227 /* S->pState buffer contains previous frame (numTaps - 1) samples */
emilmont 1:fdd22bb7aa52 228 /* pStateCurnt points to the location where the new input data should be written */
emilmont 1:fdd22bb7aa52 229 pStateCurnt = S->pState + (numTaps - 1u);
emilmont 1:fdd22bb7aa52 230
emilmont 1:fdd22bb7aa52 231 /* Total number of output samples to be computed */
emilmont 1:fdd22bb7aa52 232 blkCnt = outBlockSize;
emilmont 1:fdd22bb7aa52 233
emilmont 1:fdd22bb7aa52 234 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 235 {
emilmont 1:fdd22bb7aa52 236 /* Copy decimation factor number of new input samples into the state buffer */
emilmont 1:fdd22bb7aa52 237 i = S->M;
emilmont 1:fdd22bb7aa52 238
emilmont 1:fdd22bb7aa52 239 do
emilmont 1:fdd22bb7aa52 240 {
emilmont 1:fdd22bb7aa52 241 *pStateCurnt++ = *pSrc++;
emilmont 1:fdd22bb7aa52 242
emilmont 1:fdd22bb7aa52 243 } while(--i);
emilmont 1:fdd22bb7aa52 244
emilmont 1:fdd22bb7aa52 245 /* Set accumulator to zero */
emilmont 1:fdd22bb7aa52 246 sum0 = 0;
emilmont 1:fdd22bb7aa52 247
emilmont 1:fdd22bb7aa52 248 /* Initialize state pointer */
emilmont 1:fdd22bb7aa52 249 px = pState;
emilmont 1:fdd22bb7aa52 250
emilmont 1:fdd22bb7aa52 251 /* Initialize coeff pointer */
emilmont 1:fdd22bb7aa52 252 pb = pCoeffs;
emilmont 1:fdd22bb7aa52 253
emilmont 1:fdd22bb7aa52 254 tapCnt = numTaps;
emilmont 1:fdd22bb7aa52 255
emilmont 1:fdd22bb7aa52 256 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 257 {
emilmont 1:fdd22bb7aa52 258 /* Read coefficients */
emilmont 1:fdd22bb7aa52 259 c0 = *pb++;
emilmont 1:fdd22bb7aa52 260
emilmont 1:fdd22bb7aa52 261 /* Fetch 1 state variable */
emilmont 1:fdd22bb7aa52 262 x0 = *px++;
emilmont 1:fdd22bb7aa52 263
emilmont 1:fdd22bb7aa52 264 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 265 sum0 += (q63_t) x0 *c0;
emilmont 1:fdd22bb7aa52 266
emilmont 1:fdd22bb7aa52 267 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 268 tapCnt--;
emilmont 1:fdd22bb7aa52 269 }
emilmont 1:fdd22bb7aa52 270
emilmont 1:fdd22bb7aa52 271 /* Advance the state pointer by the decimation factor
emilmont 1:fdd22bb7aa52 272 * to process the next group of decimation factor number samples */
emilmont 1:fdd22bb7aa52 273 pState = pState + S->M;
emilmont 1:fdd22bb7aa52 274
emilmont 1:fdd22bb7aa52 275 /* The result is in the accumulator, store in the destination buffer. */
emilmont 1:fdd22bb7aa52 276 *pDst++ = (q31_t) (sum0 >> 31);
emilmont 1:fdd22bb7aa52 277
emilmont 1:fdd22bb7aa52 278 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 279 blkCnt--;
emilmont 1:fdd22bb7aa52 280 }
emilmont 1:fdd22bb7aa52 281
emilmont 1:fdd22bb7aa52 282 /* Processing is complete.
emilmont 1:fdd22bb7aa52 283 ** Now copy the last numTaps - 1 samples to the start of the state buffer.
emilmont 1:fdd22bb7aa52 284 ** This prepares the state buffer for the next function call. */
emilmont 1:fdd22bb7aa52 285
emilmont 1:fdd22bb7aa52 286 /* Points to the start of the state buffer */
emilmont 1:fdd22bb7aa52 287 pStateCurnt = S->pState;
emilmont 1:fdd22bb7aa52 288
emilmont 1:fdd22bb7aa52 289 i = numTaps - 1u;
emilmont 1:fdd22bb7aa52 290
emilmont 1:fdd22bb7aa52 291 /* copy data */
emilmont 1:fdd22bb7aa52 292 while(i > 0u)
emilmont 1:fdd22bb7aa52 293 {
emilmont 1:fdd22bb7aa52 294 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 295
emilmont 1:fdd22bb7aa52 296 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 297 i--;
emilmont 1:fdd22bb7aa52 298 }
emilmont 1:fdd22bb7aa52 299
emilmont 1:fdd22bb7aa52 300 #endif /* #ifndef ARM_MATH_CM0 */
emilmont 1:fdd22bb7aa52 301
emilmont 1:fdd22bb7aa52 302 }
emilmont 1:fdd22bb7aa52 303
emilmont 1:fdd22bb7aa52 304 /**
emilmont 1:fdd22bb7aa52 305 * @} end of FIR_decimate group
emilmont 1:fdd22bb7aa52 306 */