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_f32.c@4:9cee975aadce, 2014-06-23 (annotated)

Committer:
mbed_official
Date:
Mon Jun 23 09:30:09 2014 +0100
Revision:
4:9cee975aadce
Parent:
3:7a284390b0ce
Child:
5:3762170b6d4d
Synchronized with git revision 6e7c7bcec41226f536474daae3c13d49e4c0e865



Full URL: https://github.com/mbedmicro/mbed/commit/6e7c7bcec41226f536474daae3c13d49e4c0e865/



Fix signed unsigned compare in dsp library

Who changed what in which revision?

UserRevisionLine numberNew contents of line
emilmont 1:fdd22bb7aa52 1 /* ----------------------------------------------------------------------
mbed_official 3:7a284390b0ce 2 * Copyright (C) 2010-2013 ARM Limited. All rights reserved.
emilmont 1:fdd22bb7aa52 3 *
mbed_official 3:7a284390b0ce 4 * $Date: 17. January 2013
mbed_official 3:7a284390b0ce 5 * $Revision: V1.4.1
emilmont 1:fdd22bb7aa52 6 *
emilmont 2:da51fb522205 7 * Project: CMSIS DSP Library
emilmont 2:da51fb522205 8 * Title: arm_fir_decimate_f32.c
emilmont 1:fdd22bb7aa52 9 *
emilmont 2:da51fb522205 10 * Description: FIR decimation for floating-point sequences.
emilmont 1:fdd22bb7aa52 11 *
emilmont 1:fdd22bb7aa52 12 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
emilmont 1:fdd22bb7aa52 13 *
mbed_official 3:7a284390b0ce 14 * Redistribution and use in source and binary forms, with or without
mbed_official 3:7a284390b0ce 15 * modification, are permitted provided that the following conditions
mbed_official 3:7a284390b0ce 16 * are met:
mbed_official 3:7a284390b0ce 17 * - Redistributions of source code must retain the above copyright
mbed_official 3:7a284390b0ce 18 * notice, this list of conditions and the following disclaimer.
mbed_official 3:7a284390b0ce 19 * - Redistributions in binary form must reproduce the above copyright
mbed_official 3:7a284390b0ce 20 * notice, this list of conditions and the following disclaimer in
mbed_official 3:7a284390b0ce 21 * the documentation and/or other materials provided with the
mbed_official 3:7a284390b0ce 22 * distribution.
mbed_official 3:7a284390b0ce 23 * - Neither the name of ARM LIMITED nor the names of its contributors
mbed_official 3:7a284390b0ce 24 * may be used to endorse or promote products derived from this
mbed_official 3:7a284390b0ce 25 * software without specific prior written permission.
mbed_official 3:7a284390b0ce 26 *
mbed_official 3:7a284390b0ce 27 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
mbed_official 3:7a284390b0ce 28 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
mbed_official 3:7a284390b0ce 29 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
mbed_official 3:7a284390b0ce 30 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
mbed_official 3:7a284390b0ce 31 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
mbed_official 3:7a284390b0ce 32 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
mbed_official 3:7a284390b0ce 33 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
mbed_official 3:7a284390b0ce 34 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
mbed_official 3:7a284390b0ce 35 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
mbed_official 3:7a284390b0ce 36 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
mbed_official 3:7a284390b0ce 37 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
mbed_official 3:7a284390b0ce 38 * POSSIBILITY OF SUCH DAMAGE.
emilmont 1:fdd22bb7aa52 39 * -------------------------------------------------------------------- */
emilmont 1:fdd22bb7aa52 40
emilmont 1:fdd22bb7aa52 41 #include "arm_math.h"
emilmont 1:fdd22bb7aa52 42
emilmont 1:fdd22bb7aa52 43 /**
emilmont 1:fdd22bb7aa52 44 * @ingroup groupFilters
emilmont 1:fdd22bb7aa52 45 */
emilmont 1:fdd22bb7aa52 46
emilmont 1:fdd22bb7aa52 47 /**
emilmont 1:fdd22bb7aa52 48 * @defgroup FIR_decimate Finite Impulse Response (FIR) Decimator
emilmont 1:fdd22bb7aa52 49 *
emilmont 1:fdd22bb7aa52 50 * These functions combine an FIR filter together with a decimator.
emilmont 1:fdd22bb7aa52 51 * They are used in multirate systems for reducing the sample rate of a signal without introducing aliasing distortion.
emilmont 1:fdd22bb7aa52 52 * Conceptually, the functions are equivalent to the block diagram below:
emilmont 1:fdd22bb7aa52 53 * \image html FIRDecimator.gif "Components included in the FIR Decimator functions"
emilmont 1:fdd22bb7aa52 54 * When decimating by a factor of <code>M</code>, the signal should be prefiltered by a lowpass filter with a normalized
emilmont 1:fdd22bb7aa52 55 * cutoff frequency of <code>1/M</code> in order to prevent aliasing distortion.
emilmont 1:fdd22bb7aa52 56 * The user of the function is responsible for providing the filter coefficients.
emilmont 1:fdd22bb7aa52 57 *
emilmont 1:fdd22bb7aa52 58 * The FIR decimator functions provided in the CMSIS DSP Library combine the FIR filter and the decimator in an efficient manner.
emilmont 1:fdd22bb7aa52 59 * Instead of calculating all of the FIR filter outputs and discarding <code>M-1</code> out of every <code>M</code>, only the
emilmont 1:fdd22bb7aa52 60 * samples output by the decimator are computed.
emilmont 1:fdd22bb7aa52 61 * The functions operate on blocks of input and output data.
emilmont 1:fdd22bb7aa52 62 * <code>pSrc</code> points to an array of <code>blockSize</code> input values and
emilmont 1:fdd22bb7aa52 63 * <code>pDst</code> points to an array of <code>blockSize/M</code> output values.
emilmont 1:fdd22bb7aa52 64 * In order to have an integer number of output samples <code>blockSize</code>
emilmont 1:fdd22bb7aa52 65 * must always be a multiple of the decimation factor <code>M</code>.
emilmont 1:fdd22bb7aa52 66 *
emilmont 1:fdd22bb7aa52 67 * The library provides separate functions for Q15, Q31 and floating-point data types.
emilmont 1:fdd22bb7aa52 68 *
emilmont 1:fdd22bb7aa52 69 * \par Algorithm:
emilmont 1:fdd22bb7aa52 70 * The FIR portion of the algorithm uses the standard form filter:
emilmont 1:fdd22bb7aa52 71 * <pre>
emilmont 1:fdd22bb7aa52 72 * y[n] = b[0] * x[n] + b[1] * x[n-1] + b[2] * x[n-2] + ...+ b[numTaps-1] * x[n-numTaps+1]
emilmont 1:fdd22bb7aa52 73 * </pre>
emilmont 1:fdd22bb7aa52 74 * where, <code>b[n]</code> are the filter coefficients.
emilmont 1:fdd22bb7aa52 75 * \par
emilmont 1:fdd22bb7aa52 76 * The <code>pCoeffs</code> points to a coefficient array of size <code>numTaps</code>.
emilmont 1:fdd22bb7aa52 77 * Coefficients are stored in time reversed order.
emilmont 1:fdd22bb7aa52 78 * \par
emilmont 1:fdd22bb7aa52 79 * <pre>
emilmont 1:fdd22bb7aa52 80 * {b[numTaps-1], b[numTaps-2], b[N-2], ..., b[1], b[0]}
emilmont 1:fdd22bb7aa52 81 * </pre>
emilmont 1:fdd22bb7aa52 82 * \par
emilmont 1:fdd22bb7aa52 83 * <code>pState</code> points to a state array of size <code>numTaps + blockSize - 1</code>.
emilmont 1:fdd22bb7aa52 84 * Samples in the state buffer are stored in the order:
emilmont 1:fdd22bb7aa52 85 * \par
emilmont 1:fdd22bb7aa52 86 * <pre>
emilmont 1:fdd22bb7aa52 87 * {x[n-numTaps+1], x[n-numTaps], x[n-numTaps-1], x[n-numTaps-2]....x[0], x[1], ..., x[blockSize-1]}
emilmont 1:fdd22bb7aa52 88 * </pre>
emilmont 1:fdd22bb7aa52 89 * The state variables are updated after each block of data is processed, the coefficients are untouched.
emilmont 1:fdd22bb7aa52 90 *
emilmont 1:fdd22bb7aa52 91 * \par Instance Structure
emilmont 1:fdd22bb7aa52 92 * The coefficients and state variables for a filter are stored together in an instance data structure.
emilmont 1:fdd22bb7aa52 93 * A separate instance structure must be defined for each filter.
emilmont 1:fdd22bb7aa52 94 * Coefficient arrays may be shared among several instances while state variable array should be allocated separately.
emilmont 1:fdd22bb7aa52 95 * There are separate instance structure declarations for each of the 3 supported data types.
emilmont 1:fdd22bb7aa52 96 *
emilmont 1:fdd22bb7aa52 97 * \par Initialization Functions
emilmont 1:fdd22bb7aa52 98 * There is also an associated initialization function for each data type.
emilmont 1:fdd22bb7aa52 99 * The initialization function performs the following operations:
emilmont 1:fdd22bb7aa52 100 * - Sets the values of the internal structure fields.
emilmont 1:fdd22bb7aa52 101 * - Zeros out the values in the state buffer.
emilmont 1:fdd22bb7aa52 102 * - Checks to make sure that the size of the input is a multiple of the decimation factor.
mbed_official 3:7a284390b0ce 103 * To do this manually without calling the init function, assign the follow subfields of the instance structure:
mbed_official 3:7a284390b0ce 104 * numTaps, pCoeffs, M (decimation factor), pState. Also set all of the values in pState to zero.
emilmont 1:fdd22bb7aa52 105 *
emilmont 1:fdd22bb7aa52 106 * \par
emilmont 1:fdd22bb7aa52 107 * Use of the initialization function is optional.
emilmont 1:fdd22bb7aa52 108 * However, if the initialization function is used, then the instance structure cannot be placed into a const data section.
emilmont 1:fdd22bb7aa52 109 * To place an instance structure into a const data section, the instance structure must be manually initialized.
emilmont 1:fdd22bb7aa52 110 * The code below statically initializes each of the 3 different data type filter instance structures
emilmont 1:fdd22bb7aa52 111 * <pre>
emilmont 1:fdd22bb7aa52 112 *arm_fir_decimate_instance_f32 S = {M, numTaps, pCoeffs, pState};
emilmont 1:fdd22bb7aa52 113 *arm_fir_decimate_instance_q31 S = {M, numTaps, pCoeffs, pState};
emilmont 1:fdd22bb7aa52 114 *arm_fir_decimate_instance_q15 S = {M, numTaps, pCoeffs, pState};
emilmont 1:fdd22bb7aa52 115 * </pre>
emilmont 1:fdd22bb7aa52 116 * where <code>M</code> is the decimation factor; <code>numTaps</code> is the number of filter coefficients in the filter;
emilmont 1:fdd22bb7aa52 117 * <code>pCoeffs</code> is the address of the coefficient buffer;
emilmont 1:fdd22bb7aa52 118 * <code>pState</code> is the address of the state buffer.
emilmont 1:fdd22bb7aa52 119 * Be sure to set the values in the state buffer to zeros when doing static initialization.
emilmont 1:fdd22bb7aa52 120 *
emilmont 1:fdd22bb7aa52 121 * \par Fixed-Point Behavior
emilmont 1:fdd22bb7aa52 122 * Care must be taken when using the fixed-point versions of the FIR decimate filter functions.
emilmont 1:fdd22bb7aa52 123 * In particular, the overflow and saturation behavior of the accumulator used in each function must be considered.
emilmont 1:fdd22bb7aa52 124 * Refer to the function specific documentation below for usage guidelines.
emilmont 1:fdd22bb7aa52 125 */
emilmont 1:fdd22bb7aa52 126
emilmont 1:fdd22bb7aa52 127 /**
emilmont 1:fdd22bb7aa52 128 * @addtogroup FIR_decimate
emilmont 1:fdd22bb7aa52 129 * @{
emilmont 1:fdd22bb7aa52 130 */
emilmont 1:fdd22bb7aa52 131
emilmont 1:fdd22bb7aa52 132 /**
emilmont 1:fdd22bb7aa52 133 * @brief Processing function for the floating-point FIR decimator.
emilmont 1:fdd22bb7aa52 134 * @param[in] *S points to an instance of the floating-point FIR decimator structure.
emilmont 1:fdd22bb7aa52 135 * @param[in] *pSrc points to the block of input data.
emilmont 1:fdd22bb7aa52 136 * @param[out] *pDst points to the block of output data.
emilmont 1:fdd22bb7aa52 137 * @param[in] blockSize number of input samples to process per call.
emilmont 1:fdd22bb7aa52 138 * @return none.
emilmont 1:fdd22bb7aa52 139 */
emilmont 1:fdd22bb7aa52 140
emilmont 1:fdd22bb7aa52 141 void arm_fir_decimate_f32(
emilmont 1:fdd22bb7aa52 142 const arm_fir_decimate_instance_f32 * S,
emilmont 1:fdd22bb7aa52 143 float32_t * pSrc,
emilmont 1:fdd22bb7aa52 144 float32_t * pDst,
emilmont 1:fdd22bb7aa52 145 uint32_t blockSize)
emilmont 1:fdd22bb7aa52 146 {
emilmont 1:fdd22bb7aa52 147 float32_t *pState = S->pState; /* State pointer */
emilmont 1:fdd22bb7aa52 148 float32_t *pCoeffs = S->pCoeffs; /* Coefficient pointer */
emilmont 1:fdd22bb7aa52 149 float32_t *pStateCurnt; /* Points to the current sample of the state */
emilmont 1:fdd22bb7aa52 150 float32_t *px, *pb; /* Temporary pointers for state and coefficient buffers */
emilmont 1:fdd22bb7aa52 151 float32_t sum0; /* Accumulator */
emilmont 1:fdd22bb7aa52 152 float32_t x0, c0; /* Temporary variables to hold state and coefficient values */
emilmont 1:fdd22bb7aa52 153 uint32_t numTaps = S->numTaps; /* Number of filter coefficients in the filter */
emilmont 1:fdd22bb7aa52 154 uint32_t i, tapCnt, blkCnt, outBlockSize = blockSize / S->M; /* Loop counters */
emilmont 1:fdd22bb7aa52 155
mbed_official 3:7a284390b0ce 156 #ifndef ARM_MATH_CM0_FAMILY
emilmont 1:fdd22bb7aa52 157
emilmont 1:fdd22bb7aa52 158 uint32_t blkCntN4;
emilmont 1:fdd22bb7aa52 159 float32_t *px0, *px1, *px2, *px3;
emilmont 1:fdd22bb7aa52 160 float32_t acc0, acc1, acc2, acc3;
emilmont 1:fdd22bb7aa52 161 float32_t x1, x2, x3;
emilmont 1:fdd22bb7aa52 162
emilmont 1:fdd22bb7aa52 163 /* Run the below code for Cortex-M4 and Cortex-M3 */
emilmont 1:fdd22bb7aa52 164
emilmont 1:fdd22bb7aa52 165 /* S->pState buffer contains previous frame (numTaps - 1) samples */
emilmont 1:fdd22bb7aa52 166 /* pStateCurnt points to the location where the new input data should be written */
emilmont 1:fdd22bb7aa52 167 pStateCurnt = S->pState + (numTaps - 1u);
emilmont 1:fdd22bb7aa52 168
emilmont 1:fdd22bb7aa52 169 /* Total number of output samples to be computed */
emilmont 1:fdd22bb7aa52 170 blkCnt = outBlockSize / 4;
emilmont 1:fdd22bb7aa52 171 blkCntN4 = outBlockSize - (4 * blkCnt);
emilmont 1:fdd22bb7aa52 172
emilmont 1:fdd22bb7aa52 173 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 174 {
emilmont 1:fdd22bb7aa52 175 /* Copy 4 * decimation factor number of new input samples into the state buffer */
emilmont 1:fdd22bb7aa52 176 i = 4 * S->M;
emilmont 1:fdd22bb7aa52 177
emilmont 1:fdd22bb7aa52 178 do
emilmont 1:fdd22bb7aa52 179 {
emilmont 1:fdd22bb7aa52 180 *pStateCurnt++ = *pSrc++;
emilmont 1:fdd22bb7aa52 181
emilmont 1:fdd22bb7aa52 182 } while(--i);
emilmont 1:fdd22bb7aa52 183
emilmont 1:fdd22bb7aa52 184 /* Set accumulators to zero */
emilmont 1:fdd22bb7aa52 185 acc0 = 0.0f;
emilmont 1:fdd22bb7aa52 186 acc1 = 0.0f;
emilmont 1:fdd22bb7aa52 187 acc2 = 0.0f;
emilmont 1:fdd22bb7aa52 188 acc3 = 0.0f;
emilmont 1:fdd22bb7aa52 189
emilmont 1:fdd22bb7aa52 190 /* Initialize state pointer for all the samples */
emilmont 1:fdd22bb7aa52 191 px0 = pState;
emilmont 1:fdd22bb7aa52 192 px1 = pState + S->M;
emilmont 1:fdd22bb7aa52 193 px2 = pState + 2 * S->M;
emilmont 1:fdd22bb7aa52 194 px3 = pState + 3 * S->M;
emilmont 1:fdd22bb7aa52 195
emilmont 1:fdd22bb7aa52 196 /* Initialize coeff pointer */
emilmont 1:fdd22bb7aa52 197 pb = pCoeffs;
emilmont 1:fdd22bb7aa52 198
emilmont 1:fdd22bb7aa52 199 /* Loop unrolling. Process 4 taps at a time. */
emilmont 1:fdd22bb7aa52 200 tapCnt = numTaps >> 2;
emilmont 1:fdd22bb7aa52 201
emilmont 1:fdd22bb7aa52 202 /* Loop over the number of taps. Unroll by a factor of 4.
emilmont 1:fdd22bb7aa52 203 ** Repeat until we've computed numTaps-4 coefficients. */
emilmont 1:fdd22bb7aa52 204
emilmont 1:fdd22bb7aa52 205 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 206 {
emilmont 1:fdd22bb7aa52 207 /* Read the b[numTaps-1] coefficient */
emilmont 1:fdd22bb7aa52 208 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 209
emilmont 1:fdd22bb7aa52 210 /* Read x[n-numTaps-1] sample for acc0 */
emilmont 1:fdd22bb7aa52 211 x0 = *(px0++);
emilmont 1:fdd22bb7aa52 212 /* Read x[n-numTaps-1] sample for acc1 */
emilmont 1:fdd22bb7aa52 213 x1 = *(px1++);
emilmont 1:fdd22bb7aa52 214 /* Read x[n-numTaps-1] sample for acc2 */
emilmont 1:fdd22bb7aa52 215 x2 = *(px2++);
emilmont 1:fdd22bb7aa52 216 /* Read x[n-numTaps-1] sample for acc3 */
emilmont 1:fdd22bb7aa52 217 x3 = *(px3++);
emilmont 1:fdd22bb7aa52 218
emilmont 1:fdd22bb7aa52 219 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 220 acc0 += x0 * c0;
emilmont 1:fdd22bb7aa52 221 acc1 += x1 * c0;
emilmont 1:fdd22bb7aa52 222 acc2 += x2 * c0;
emilmont 1:fdd22bb7aa52 223 acc3 += x3 * c0;
emilmont 1:fdd22bb7aa52 224
emilmont 1:fdd22bb7aa52 225 /* Read the b[numTaps-2] coefficient */
emilmont 1:fdd22bb7aa52 226 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 227
emilmont 1:fdd22bb7aa52 228 /* Read x[n-numTaps-2] sample for acc0, acc1, acc2, acc3 */
emilmont 1:fdd22bb7aa52 229 x0 = *(px0++);
emilmont 1:fdd22bb7aa52 230 x1 = *(px1++);
emilmont 1:fdd22bb7aa52 231 x2 = *(px2++);
emilmont 1:fdd22bb7aa52 232 x3 = *(px3++);
emilmont 1:fdd22bb7aa52 233
emilmont 1:fdd22bb7aa52 234 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 235 acc0 += x0 * c0;
emilmont 1:fdd22bb7aa52 236 acc1 += x1 * c0;
emilmont 1:fdd22bb7aa52 237 acc2 += x2 * c0;
emilmont 1:fdd22bb7aa52 238 acc3 += x3 * c0;
emilmont 1:fdd22bb7aa52 239
emilmont 1:fdd22bb7aa52 240 /* Read the b[numTaps-3] coefficient */
emilmont 1:fdd22bb7aa52 241 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 242
emilmont 1:fdd22bb7aa52 243 /* Read x[n-numTaps-3] sample acc0, acc1, acc2, acc3 */
emilmont 1:fdd22bb7aa52 244 x0 = *(px0++);
emilmont 1:fdd22bb7aa52 245 x1 = *(px1++);
emilmont 1:fdd22bb7aa52 246 x2 = *(px2++);
emilmont 1:fdd22bb7aa52 247 x3 = *(px3++);
emilmont 1:fdd22bb7aa52 248
emilmont 1:fdd22bb7aa52 249 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 250 acc0 += x0 * c0;
emilmont 1:fdd22bb7aa52 251 acc1 += x1 * c0;
emilmont 1:fdd22bb7aa52 252 acc2 += x2 * c0;
emilmont 1:fdd22bb7aa52 253 acc3 += x3 * c0;
emilmont 1:fdd22bb7aa52 254
emilmont 1:fdd22bb7aa52 255 /* Read the b[numTaps-4] coefficient */
emilmont 1:fdd22bb7aa52 256 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 257
emilmont 1:fdd22bb7aa52 258 /* Read x[n-numTaps-4] sample acc0, acc1, acc2, acc3 */
emilmont 1:fdd22bb7aa52 259 x0 = *(px0++);
emilmont 1:fdd22bb7aa52 260 x1 = *(px1++);
emilmont 1:fdd22bb7aa52 261 x2 = *(px2++);
emilmont 1:fdd22bb7aa52 262 x3 = *(px3++);
emilmont 1:fdd22bb7aa52 263
emilmont 1:fdd22bb7aa52 264 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 265 acc0 += x0 * c0;
emilmont 1:fdd22bb7aa52 266 acc1 += x1 * c0;
emilmont 1:fdd22bb7aa52 267 acc2 += x2 * c0;
emilmont 1:fdd22bb7aa52 268 acc3 += x3 * c0;
emilmont 1:fdd22bb7aa52 269
emilmont 1:fdd22bb7aa52 270 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 271 tapCnt--;
emilmont 1:fdd22bb7aa52 272 }
emilmont 1:fdd22bb7aa52 273
emilmont 1:fdd22bb7aa52 274 /* If the filter length is not a multiple of 4, compute the remaining filter taps */
emilmont 1:fdd22bb7aa52 275 tapCnt = numTaps % 0x4u;
emilmont 1:fdd22bb7aa52 276
emilmont 1:fdd22bb7aa52 277 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 278 {
emilmont 1:fdd22bb7aa52 279 /* Read coefficients */
emilmont 1:fdd22bb7aa52 280 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 281
emilmont 1:fdd22bb7aa52 282 /* Fetch state variables for acc0, acc1, acc2, acc3 */
emilmont 1:fdd22bb7aa52 283 x0 = *(px0++);
emilmont 1:fdd22bb7aa52 284 x1 = *(px1++);
emilmont 1:fdd22bb7aa52 285 x2 = *(px2++);
emilmont 1:fdd22bb7aa52 286 x3 = *(px3++);
emilmont 1:fdd22bb7aa52 287
emilmont 1:fdd22bb7aa52 288 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 289 acc0 += x0 * c0;
emilmont 1:fdd22bb7aa52 290 acc1 += x1 * c0;
emilmont 1:fdd22bb7aa52 291 acc2 += x2 * c0;
emilmont 1:fdd22bb7aa52 292 acc3 += x3 * c0;
emilmont 1:fdd22bb7aa52 293
emilmont 1:fdd22bb7aa52 294 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 295 tapCnt--;
emilmont 1:fdd22bb7aa52 296 }
emilmont 1:fdd22bb7aa52 297
emilmont 1:fdd22bb7aa52 298 /* Advance the state pointer by the decimation factor
emilmont 1:fdd22bb7aa52 299 * to process the next group of decimation factor number samples */
emilmont 1:fdd22bb7aa52 300 pState = pState + 4 * S->M;
emilmont 1:fdd22bb7aa52 301
emilmont 1:fdd22bb7aa52 302 /* The result is in the accumulator, store in the destination buffer. */
emilmont 1:fdd22bb7aa52 303 *pDst++ = acc0;
emilmont 1:fdd22bb7aa52 304 *pDst++ = acc1;
emilmont 1:fdd22bb7aa52 305 *pDst++ = acc2;
emilmont 1:fdd22bb7aa52 306 *pDst++ = acc3;
emilmont 1:fdd22bb7aa52 307
emilmont 1:fdd22bb7aa52 308 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 309 blkCnt--;
emilmont 1:fdd22bb7aa52 310 }
emilmont 1:fdd22bb7aa52 311
emilmont 1:fdd22bb7aa52 312 while(blkCntN4 > 0u)
emilmont 1:fdd22bb7aa52 313 {
emilmont 1:fdd22bb7aa52 314 /* Copy decimation factor number of new input samples into the state buffer */
emilmont 1:fdd22bb7aa52 315 i = S->M;
emilmont 1:fdd22bb7aa52 316
emilmont 1:fdd22bb7aa52 317 do
emilmont 1:fdd22bb7aa52 318 {
emilmont 1:fdd22bb7aa52 319 *pStateCurnt++ = *pSrc++;
emilmont 1:fdd22bb7aa52 320
emilmont 1:fdd22bb7aa52 321 } while(--i);
emilmont 1:fdd22bb7aa52 322
emilmont 1:fdd22bb7aa52 323 /* Set accumulator to zero */
emilmont 1:fdd22bb7aa52 324 sum0 = 0.0f;
emilmont 1:fdd22bb7aa52 325
emilmont 1:fdd22bb7aa52 326 /* Initialize state pointer */
emilmont 1:fdd22bb7aa52 327 px = pState;
emilmont 1:fdd22bb7aa52 328
emilmont 1:fdd22bb7aa52 329 /* Initialize coeff pointer */
emilmont 1:fdd22bb7aa52 330 pb = pCoeffs;
emilmont 1:fdd22bb7aa52 331
emilmont 1:fdd22bb7aa52 332 /* Loop unrolling. Process 4 taps at a time. */
emilmont 1:fdd22bb7aa52 333 tapCnt = numTaps >> 2;
emilmont 1:fdd22bb7aa52 334
emilmont 1:fdd22bb7aa52 335 /* Loop over the number of taps. Unroll by a factor of 4.
emilmont 1:fdd22bb7aa52 336 ** Repeat until we've computed numTaps-4 coefficients. */
emilmont 1:fdd22bb7aa52 337 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 338 {
emilmont 1:fdd22bb7aa52 339 /* Read the b[numTaps-1] coefficient */
emilmont 1:fdd22bb7aa52 340 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 341
emilmont 1:fdd22bb7aa52 342 /* Read x[n-numTaps-1] sample */
emilmont 1:fdd22bb7aa52 343 x0 = *(px++);
emilmont 1:fdd22bb7aa52 344
emilmont 1:fdd22bb7aa52 345 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 346 sum0 += x0 * c0;
emilmont 1:fdd22bb7aa52 347
emilmont 1:fdd22bb7aa52 348 /* Read the b[numTaps-2] coefficient */
emilmont 1:fdd22bb7aa52 349 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 350
emilmont 1:fdd22bb7aa52 351 /* Read x[n-numTaps-2] sample */
emilmont 1:fdd22bb7aa52 352 x0 = *(px++);
emilmont 1:fdd22bb7aa52 353
emilmont 1:fdd22bb7aa52 354 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 355 sum0 += x0 * c0;
emilmont 1:fdd22bb7aa52 356
emilmont 1:fdd22bb7aa52 357 /* Read the b[numTaps-3] coefficient */
emilmont 1:fdd22bb7aa52 358 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 359
emilmont 1:fdd22bb7aa52 360 /* Read x[n-numTaps-3] sample */
emilmont 1:fdd22bb7aa52 361 x0 = *(px++);
emilmont 1:fdd22bb7aa52 362
emilmont 1:fdd22bb7aa52 363 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 364 sum0 += x0 * c0;
emilmont 1:fdd22bb7aa52 365
emilmont 1:fdd22bb7aa52 366 /* Read the b[numTaps-4] coefficient */
emilmont 1:fdd22bb7aa52 367 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 368
emilmont 1:fdd22bb7aa52 369 /* Read x[n-numTaps-4] sample */
emilmont 1:fdd22bb7aa52 370 x0 = *(px++);
emilmont 1:fdd22bb7aa52 371
emilmont 1:fdd22bb7aa52 372 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 373 sum0 += x0 * c0;
emilmont 1:fdd22bb7aa52 374
emilmont 1:fdd22bb7aa52 375 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 376 tapCnt--;
emilmont 1:fdd22bb7aa52 377 }
emilmont 1:fdd22bb7aa52 378
emilmont 1:fdd22bb7aa52 379 /* If the filter length is not a multiple of 4, compute the remaining filter taps */
emilmont 1:fdd22bb7aa52 380 tapCnt = numTaps % 0x4u;
emilmont 1:fdd22bb7aa52 381
emilmont 1:fdd22bb7aa52 382 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 383 {
emilmont 1:fdd22bb7aa52 384 /* Read coefficients */
emilmont 1:fdd22bb7aa52 385 c0 = *(pb++);
emilmont 1:fdd22bb7aa52 386
emilmont 1:fdd22bb7aa52 387 /* Fetch 1 state variable */
emilmont 1:fdd22bb7aa52 388 x0 = *(px++);
emilmont 1:fdd22bb7aa52 389
emilmont 1:fdd22bb7aa52 390 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 391 sum0 += x0 * c0;
emilmont 1:fdd22bb7aa52 392
emilmont 1:fdd22bb7aa52 393 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 394 tapCnt--;
emilmont 1:fdd22bb7aa52 395 }
emilmont 1:fdd22bb7aa52 396
emilmont 1:fdd22bb7aa52 397 /* Advance the state pointer by the decimation factor
emilmont 1:fdd22bb7aa52 398 * to process the next group of decimation factor number samples */
emilmont 1:fdd22bb7aa52 399 pState = pState + S->M;
emilmont 1:fdd22bb7aa52 400
emilmont 1:fdd22bb7aa52 401 /* The result is in the accumulator, store in the destination buffer. */
emilmont 1:fdd22bb7aa52 402 *pDst++ = sum0;
emilmont 1:fdd22bb7aa52 403
emilmont 1:fdd22bb7aa52 404 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 405 blkCntN4--;
emilmont 1:fdd22bb7aa52 406 }
emilmont 1:fdd22bb7aa52 407
emilmont 1:fdd22bb7aa52 408 /* Processing is complete.
emilmont 1:fdd22bb7aa52 409 ** Now copy the last numTaps - 1 samples to the satrt of the state buffer.
emilmont 1:fdd22bb7aa52 410 ** This prepares the state buffer for the next function call. */
emilmont 1:fdd22bb7aa52 411
emilmont 1:fdd22bb7aa52 412 /* Points to the start of the state buffer */
emilmont 1:fdd22bb7aa52 413 pStateCurnt = S->pState;
emilmont 1:fdd22bb7aa52 414
emilmont 1:fdd22bb7aa52 415 i = (numTaps - 1u) >> 2;
emilmont 1:fdd22bb7aa52 416
emilmont 1:fdd22bb7aa52 417 /* copy data */
emilmont 1:fdd22bb7aa52 418 while(i > 0u)
emilmont 1:fdd22bb7aa52 419 {
emilmont 1:fdd22bb7aa52 420 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 421 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 422 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 423 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 424
emilmont 1:fdd22bb7aa52 425 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 426 i--;
emilmont 1:fdd22bb7aa52 427 }
emilmont 1:fdd22bb7aa52 428
emilmont 1:fdd22bb7aa52 429 i = (numTaps - 1u) % 0x04u;
emilmont 1:fdd22bb7aa52 430
emilmont 1:fdd22bb7aa52 431 /* copy data */
emilmont 1:fdd22bb7aa52 432 while(i > 0u)
emilmont 1:fdd22bb7aa52 433 {
emilmont 1:fdd22bb7aa52 434 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 435
emilmont 1:fdd22bb7aa52 436 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 437 i--;
emilmont 1:fdd22bb7aa52 438 }
emilmont 1:fdd22bb7aa52 439
emilmont 1:fdd22bb7aa52 440 #else
emilmont 1:fdd22bb7aa52 441
emilmont 1:fdd22bb7aa52 442 /* Run the below code for Cortex-M0 */
emilmont 1:fdd22bb7aa52 443
emilmont 1:fdd22bb7aa52 444 /* S->pState buffer contains previous frame (numTaps - 1) samples */
emilmont 1:fdd22bb7aa52 445 /* pStateCurnt points to the location where the new input data should be written */
emilmont 1:fdd22bb7aa52 446 pStateCurnt = S->pState + (numTaps - 1u);
emilmont 1:fdd22bb7aa52 447
emilmont 1:fdd22bb7aa52 448 /* Total number of output samples to be computed */
emilmont 1:fdd22bb7aa52 449 blkCnt = outBlockSize;
emilmont 1:fdd22bb7aa52 450
emilmont 1:fdd22bb7aa52 451 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 452 {
emilmont 1:fdd22bb7aa52 453 /* Copy decimation factor number of new input samples into the state buffer */
emilmont 1:fdd22bb7aa52 454 i = S->M;
emilmont 1:fdd22bb7aa52 455
emilmont 1:fdd22bb7aa52 456 do
emilmont 1:fdd22bb7aa52 457 {
emilmont 1:fdd22bb7aa52 458 *pStateCurnt++ = *pSrc++;
emilmont 1:fdd22bb7aa52 459
emilmont 1:fdd22bb7aa52 460 } while(--i);
emilmont 1:fdd22bb7aa52 461
emilmont 1:fdd22bb7aa52 462 /* Set accumulator to zero */
emilmont 1:fdd22bb7aa52 463 sum0 = 0.0f;
emilmont 1:fdd22bb7aa52 464
emilmont 1:fdd22bb7aa52 465 /* Initialize state pointer */
emilmont 1:fdd22bb7aa52 466 px = pState;
emilmont 1:fdd22bb7aa52 467
emilmont 1:fdd22bb7aa52 468 /* Initialize coeff pointer */
emilmont 1:fdd22bb7aa52 469 pb = pCoeffs;
emilmont 1:fdd22bb7aa52 470
emilmont 1:fdd22bb7aa52 471 tapCnt = numTaps;
emilmont 1:fdd22bb7aa52 472
emilmont 1:fdd22bb7aa52 473 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 474 {
emilmont 1:fdd22bb7aa52 475 /* Read coefficients */
emilmont 1:fdd22bb7aa52 476 c0 = *pb++;
emilmont 1:fdd22bb7aa52 477
emilmont 1:fdd22bb7aa52 478 /* Fetch 1 state variable */
emilmont 1:fdd22bb7aa52 479 x0 = *px++;
emilmont 1:fdd22bb7aa52 480
emilmont 1:fdd22bb7aa52 481 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 482 sum0 += x0 * c0;
emilmont 1:fdd22bb7aa52 483
emilmont 1:fdd22bb7aa52 484 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 485 tapCnt--;
emilmont 1:fdd22bb7aa52 486 }
emilmont 1:fdd22bb7aa52 487
emilmont 1:fdd22bb7aa52 488 /* Advance the state pointer by the decimation factor
emilmont 1:fdd22bb7aa52 489 * to process the next group of decimation factor number samples */
emilmont 1:fdd22bb7aa52 490 pState = pState + S->M;
emilmont 1:fdd22bb7aa52 491
emilmont 1:fdd22bb7aa52 492 /* The result is in the accumulator, store in the destination buffer. */
emilmont 1:fdd22bb7aa52 493 *pDst++ = sum0;
emilmont 1:fdd22bb7aa52 494
emilmont 1:fdd22bb7aa52 495 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 496 blkCnt--;
emilmont 1:fdd22bb7aa52 497 }
emilmont 1:fdd22bb7aa52 498
emilmont 1:fdd22bb7aa52 499 /* Processing is complete.
emilmont 1:fdd22bb7aa52 500 ** Now copy the last numTaps - 1 samples to the start of the state buffer.
emilmont 1:fdd22bb7aa52 501 ** This prepares the state buffer for the next function call. */
emilmont 1:fdd22bb7aa52 502
emilmont 1:fdd22bb7aa52 503 /* Points to the start of the state buffer */
emilmont 1:fdd22bb7aa52 504 pStateCurnt = S->pState;
emilmont 1:fdd22bb7aa52 505
emilmont 1:fdd22bb7aa52 506 /* Copy numTaps number of values */
emilmont 1:fdd22bb7aa52 507 i = (numTaps - 1u);
emilmont 1:fdd22bb7aa52 508
emilmont 1:fdd22bb7aa52 509 /* copy data */
emilmont 1:fdd22bb7aa52 510 while(i > 0u)
emilmont 1:fdd22bb7aa52 511 {
emilmont 1:fdd22bb7aa52 512 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 513
emilmont 1:fdd22bb7aa52 514 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 515 i--;
emilmont 1:fdd22bb7aa52 516 }
emilmont 1:fdd22bb7aa52 517
mbed_official 3:7a284390b0ce 518 #endif /* #ifndef ARM_MATH_CM0_FAMILY */
emilmont 1:fdd22bb7aa52 519
emilmont 1:fdd22bb7aa52 520 }
emilmont 1:fdd22bb7aa52 521
emilmont 1:fdd22bb7aa52 522 /**
emilmont 1:fdd22bb7aa52 523 * @} end of FIR_decimate group
emilmont 1:fdd22bb7aa52 524 */