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_lattice_f32.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_lattice_f32.c
emilmont 1:fdd22bb7aa52 9 *
emilmont 2:da51fb522205 10 * Description: Processing function for the floating-point FIR Lattice 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 /**
emilmont 1:fdd22bb7aa52 39 * @ingroup groupFilters
emilmont 1:fdd22bb7aa52 40 */
emilmont 1:fdd22bb7aa52 41
emilmont 1:fdd22bb7aa52 42 /**
emilmont 1:fdd22bb7aa52 43 * @defgroup FIR_Lattice Finite Impulse Response (FIR) Lattice Filters
emilmont 1:fdd22bb7aa52 44 *
emilmont 1:fdd22bb7aa52 45 * This set of functions implements Finite Impulse Response (FIR) lattice filters
emilmont 1:fdd22bb7aa52 46 * for Q15, Q31 and floating-point data types. Lattice filters are used in a
emilmont 1:fdd22bb7aa52 47 * variety of adaptive filter applications. The filter structure is feedforward and
emilmont 1:fdd22bb7aa52 48 * the net impulse response is finite length.
emilmont 1:fdd22bb7aa52 49 * The functions operate on blocks
emilmont 1:fdd22bb7aa52 50 * of input and output data and each call to the function processes
emilmont 1:fdd22bb7aa52 51 * <code>blockSize</code> samples through the filter. <code>pSrc</code> and
emilmont 1:fdd22bb7aa52 52 * <code>pDst</code> point to input and output arrays containing <code>blockSize</code> values.
emilmont 1:fdd22bb7aa52 53 *
emilmont 1:fdd22bb7aa52 54 * \par Algorithm:
emilmont 1:fdd22bb7aa52 55 * \image html FIRLattice.gif "Finite Impulse Response Lattice filter"
emilmont 1:fdd22bb7aa52 56 * The following difference equation is implemented:
emilmont 1:fdd22bb7aa52 57 * <pre>
emilmont 1:fdd22bb7aa52 58 * f0[n] = g0[n] = x[n]
emilmont 1:fdd22bb7aa52 59 * fm[n] = fm-1[n] + km * gm-1[n-1] for m = 1, 2, ...M
emilmont 1:fdd22bb7aa52 60 * gm[n] = km * fm-1[n] + gm-1[n-1] for m = 1, 2, ...M
emilmont 1:fdd22bb7aa52 61 * y[n] = fM[n]
emilmont 1:fdd22bb7aa52 62 * </pre>
emilmont 1:fdd22bb7aa52 63 * \par
emilmont 1:fdd22bb7aa52 64 * <code>pCoeffs</code> points to tha array of reflection coefficients of size <code>numStages</code>.
emilmont 1:fdd22bb7aa52 65 * Reflection Coefficients are stored in the following order.
emilmont 1:fdd22bb7aa52 66 * \par
emilmont 1:fdd22bb7aa52 67 * <pre>
emilmont 1:fdd22bb7aa52 68 * {k1, k2, ..., kM}
emilmont 1:fdd22bb7aa52 69 * </pre>
emilmont 1:fdd22bb7aa52 70 * where M is number of stages
emilmont 1:fdd22bb7aa52 71 * \par
emilmont 1:fdd22bb7aa52 72 * <code>pState</code> points to a state array of size <code>numStages</code>.
emilmont 1:fdd22bb7aa52 73 * The state variables (g values) hold previous inputs and are stored in the following order.
emilmont 1:fdd22bb7aa52 74 * <pre>
emilmont 1:fdd22bb7aa52 75 * {g0[n], g1[n], g2[n] ...gM-1[n]}
emilmont 1:fdd22bb7aa52 76 * </pre>
emilmont 1:fdd22bb7aa52 77 * The state variables are updated after each block of data is processed; the coefficients are untouched.
emilmont 1:fdd22bb7aa52 78 * \par Instance Structure
emilmont 1:fdd22bb7aa52 79 * The coefficients and state variables for a filter are stored together in an instance data structure.
emilmont 1:fdd22bb7aa52 80 * A separate instance structure must be defined for each filter.
emilmont 1:fdd22bb7aa52 81 * Coefficient arrays may be shared among several instances while state variable arrays cannot be shared.
emilmont 1:fdd22bb7aa52 82 * There are separate instance structure declarations for each of the 3 supported data types.
emilmont 1:fdd22bb7aa52 83 *
emilmont 1:fdd22bb7aa52 84 * \par Initialization Functions
emilmont 1:fdd22bb7aa52 85 * There is also an associated initialization function for each data type.
emilmont 1:fdd22bb7aa52 86 * The initialization function performs the following operations:
emilmont 1:fdd22bb7aa52 87 * - Sets the values of the internal structure fields.
emilmont 1:fdd22bb7aa52 88 * - Zeros out the values in the state buffer.
emilmont 1:fdd22bb7aa52 89 *
emilmont 1:fdd22bb7aa52 90 * \par
emilmont 1:fdd22bb7aa52 91 * Use of the initialization function is optional.
emilmont 1:fdd22bb7aa52 92 * However, if the initialization function is used, then the instance structure cannot be placed into a const data section.
emilmont 1:fdd22bb7aa52 93 * To place an instance structure into a const data section, the instance structure must be manually initialized.
emilmont 1:fdd22bb7aa52 94 * Set the values in the state buffer to zeros and then manually initialize the instance structure as follows:
emilmont 1:fdd22bb7aa52 95 * <pre>
emilmont 1:fdd22bb7aa52 96 *arm_fir_lattice_instance_f32 S = {numStages, pState, pCoeffs};
emilmont 1:fdd22bb7aa52 97 *arm_fir_lattice_instance_q31 S = {numStages, pState, pCoeffs};
emilmont 1:fdd22bb7aa52 98 *arm_fir_lattice_instance_q15 S = {numStages, pState, pCoeffs};
emilmont 1:fdd22bb7aa52 99 * </pre>
emilmont 1:fdd22bb7aa52 100 * \par
emilmont 1:fdd22bb7aa52 101 * where <code>numStages</code> is the number of stages in the filter; <code>pState</code> is the address of the state buffer;
emilmont 1:fdd22bb7aa52 102 * <code>pCoeffs</code> is the address of the coefficient buffer.
emilmont 1:fdd22bb7aa52 103 * \par Fixed-Point Behavior
emilmont 1:fdd22bb7aa52 104 * Care must be taken when using the fixed-point versions of the FIR Lattice filter functions.
emilmont 1:fdd22bb7aa52 105 * In particular, the overflow and saturation behavior of the accumulator used in each function must be considered.
emilmont 1:fdd22bb7aa52 106 * Refer to the function specific documentation below for usage guidelines.
emilmont 1:fdd22bb7aa52 107 */
emilmont 1:fdd22bb7aa52 108
emilmont 1:fdd22bb7aa52 109 /**
emilmont 1:fdd22bb7aa52 110 * @addtogroup FIR_Lattice
emilmont 1:fdd22bb7aa52 111 * @{
emilmont 1:fdd22bb7aa52 112 */
emilmont 1:fdd22bb7aa52 113
emilmont 1:fdd22bb7aa52 114
emilmont 1:fdd22bb7aa52 115 /**
emilmont 1:fdd22bb7aa52 116 * @brief Processing function for the floating-point FIR lattice filter.
emilmont 1:fdd22bb7aa52 117 * @param[in] *S points to an instance of the floating-point FIR lattice structure.
emilmont 1:fdd22bb7aa52 118 * @param[in] *pSrc points to the block of input data.
emilmont 1:fdd22bb7aa52 119 * @param[out] *pDst points to the block of output data
emilmont 1:fdd22bb7aa52 120 * @param[in] blockSize number of samples to process.
emilmont 1:fdd22bb7aa52 121 * @return none.
emilmont 1:fdd22bb7aa52 122 */
emilmont 1:fdd22bb7aa52 123
emilmont 1:fdd22bb7aa52 124 void arm_fir_lattice_f32(
emilmont 1:fdd22bb7aa52 125 const arm_fir_lattice_instance_f32 * S,
emilmont 1:fdd22bb7aa52 126 float32_t * pSrc,
emilmont 1:fdd22bb7aa52 127 float32_t * pDst,
emilmont 1:fdd22bb7aa52 128 uint32_t blockSize)
emilmont 1:fdd22bb7aa52 129 {
emilmont 1:fdd22bb7aa52 130 float32_t *pState; /* State pointer */
emilmont 1:fdd22bb7aa52 131 float32_t *pCoeffs = S->pCoeffs; /* Coefficient pointer */
emilmont 1:fdd22bb7aa52 132 float32_t *px; /* temporary state pointer */
emilmont 1:fdd22bb7aa52 133 float32_t *pk; /* temporary coefficient pointer */
emilmont 1:fdd22bb7aa52 134
emilmont 1:fdd22bb7aa52 135
emilmont 1:fdd22bb7aa52 136 #ifndef ARM_MATH_CM0
emilmont 1:fdd22bb7aa52 137
emilmont 1:fdd22bb7aa52 138 /* Run the below code for Cortex-M4 and Cortex-M3 */
emilmont 1:fdd22bb7aa52 139
emilmont 1:fdd22bb7aa52 140 float32_t fcurr1, fnext1, gcurr1, gnext1; /* temporary variables for first sample in loop unrolling */
emilmont 1:fdd22bb7aa52 141 float32_t fcurr2, fnext2, gnext2; /* temporary variables for second sample in loop unrolling */
emilmont 1:fdd22bb7aa52 142 float32_t fcurr3, fnext3, gnext3; /* temporary variables for third sample in loop unrolling */
emilmont 1:fdd22bb7aa52 143 float32_t fcurr4, fnext4, gnext4; /* temporary variables for fourth sample in loop unrolling */
emilmont 1:fdd22bb7aa52 144 uint32_t numStages = S->numStages; /* Number of stages in the filter */
emilmont 1:fdd22bb7aa52 145 uint32_t blkCnt, stageCnt; /* temporary variables for counts */
emilmont 1:fdd22bb7aa52 146
emilmont 1:fdd22bb7aa52 147 gcurr1 = 0.0f;
emilmont 1:fdd22bb7aa52 148 pState = &S->pState[0];
emilmont 1:fdd22bb7aa52 149
emilmont 1:fdd22bb7aa52 150 blkCnt = blockSize >> 2;
emilmont 1:fdd22bb7aa52 151
emilmont 1:fdd22bb7aa52 152 /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
emilmont 1:fdd22bb7aa52 153 a second loop below computes the remaining 1 to 3 samples. */
emilmont 1:fdd22bb7aa52 154 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 155 {
emilmont 1:fdd22bb7aa52 156
emilmont 1:fdd22bb7aa52 157 /* Read two samples from input buffer */
emilmont 1:fdd22bb7aa52 158 /* f0(n) = x(n) */
emilmont 1:fdd22bb7aa52 159 fcurr1 = *pSrc++;
emilmont 1:fdd22bb7aa52 160 fcurr2 = *pSrc++;
emilmont 1:fdd22bb7aa52 161
emilmont 1:fdd22bb7aa52 162 /* Initialize coeff pointer */
emilmont 1:fdd22bb7aa52 163 pk = (pCoeffs);
emilmont 1:fdd22bb7aa52 164
emilmont 1:fdd22bb7aa52 165 /* Initialize state pointer */
emilmont 1:fdd22bb7aa52 166 px = pState;
emilmont 1:fdd22bb7aa52 167
emilmont 1:fdd22bb7aa52 168 /* Read g0(n-1) from state */
emilmont 1:fdd22bb7aa52 169 gcurr1 = *px;
emilmont 1:fdd22bb7aa52 170
emilmont 1:fdd22bb7aa52 171 /* Process first sample for first tap */
emilmont 1:fdd22bb7aa52 172 /* f1(n) = f0(n) + K1 * g0(n-1) */
emilmont 1:fdd22bb7aa52 173 fnext1 = fcurr1 + ((*pk) * gcurr1);
emilmont 1:fdd22bb7aa52 174 /* g1(n) = f0(n) * K1 + g0(n-1) */
emilmont 1:fdd22bb7aa52 175 gnext1 = (fcurr1 * (*pk)) + gcurr1;
emilmont 1:fdd22bb7aa52 176
emilmont 1:fdd22bb7aa52 177 /* Process second sample for first tap */
emilmont 1:fdd22bb7aa52 178 /* for sample 2 processing */
emilmont 1:fdd22bb7aa52 179 fnext2 = fcurr2 + ((*pk) * fcurr1);
emilmont 1:fdd22bb7aa52 180 gnext2 = (fcurr2 * (*pk)) + fcurr1;
emilmont 1:fdd22bb7aa52 181
emilmont 1:fdd22bb7aa52 182 /* Read next two samples from input buffer */
emilmont 1:fdd22bb7aa52 183 /* f0(n+2) = x(n+2) */
emilmont 1:fdd22bb7aa52 184 fcurr3 = *pSrc++;
emilmont 1:fdd22bb7aa52 185 fcurr4 = *pSrc++;
emilmont 1:fdd22bb7aa52 186
emilmont 1:fdd22bb7aa52 187 /* Copy only last input samples into the state buffer
emilmont 1:fdd22bb7aa52 188 which will be used for next four samples processing */
emilmont 1:fdd22bb7aa52 189 *px++ = fcurr4;
emilmont 1:fdd22bb7aa52 190
emilmont 1:fdd22bb7aa52 191 /* Process third sample for first tap */
emilmont 1:fdd22bb7aa52 192 fnext3 = fcurr3 + ((*pk) * fcurr2);
emilmont 1:fdd22bb7aa52 193 gnext3 = (fcurr3 * (*pk)) + fcurr2;
emilmont 1:fdd22bb7aa52 194
emilmont 1:fdd22bb7aa52 195 /* Process fourth sample for first tap */
emilmont 1:fdd22bb7aa52 196 fnext4 = fcurr4 + ((*pk) * fcurr3);
emilmont 1:fdd22bb7aa52 197 gnext4 = (fcurr4 * (*pk++)) + fcurr3;
emilmont 1:fdd22bb7aa52 198
emilmont 1:fdd22bb7aa52 199 /* Update of f values for next coefficient set processing */
emilmont 1:fdd22bb7aa52 200 fcurr1 = fnext1;
emilmont 1:fdd22bb7aa52 201 fcurr2 = fnext2;
emilmont 1:fdd22bb7aa52 202 fcurr3 = fnext3;
emilmont 1:fdd22bb7aa52 203 fcurr4 = fnext4;
emilmont 1:fdd22bb7aa52 204
emilmont 1:fdd22bb7aa52 205 /* Loop unrolling. Process 4 taps at a time . */
emilmont 1:fdd22bb7aa52 206 stageCnt = (numStages - 1u) >> 2u;
emilmont 1:fdd22bb7aa52 207
emilmont 1:fdd22bb7aa52 208 /* Loop over the number of taps. Unroll by a factor of 4.
emilmont 1:fdd22bb7aa52 209 ** Repeat until we've computed numStages-3 coefficients. */
emilmont 1:fdd22bb7aa52 210
emilmont 1:fdd22bb7aa52 211 /* Process 2nd, 3rd, 4th and 5th taps ... here */
emilmont 1:fdd22bb7aa52 212 while(stageCnt > 0u)
emilmont 1:fdd22bb7aa52 213 {
emilmont 1:fdd22bb7aa52 214 /* Read g1(n-1), g3(n-1) .... from state */
emilmont 1:fdd22bb7aa52 215 gcurr1 = *px;
emilmont 1:fdd22bb7aa52 216
emilmont 1:fdd22bb7aa52 217 /* save g1(n) in state buffer */
emilmont 1:fdd22bb7aa52 218 *px++ = gnext4;
emilmont 1:fdd22bb7aa52 219
emilmont 1:fdd22bb7aa52 220 /* Process first sample for 2nd, 6th .. tap */
emilmont 1:fdd22bb7aa52 221 /* Sample processing for K2, K6.... */
emilmont 1:fdd22bb7aa52 222 /* f2(n) = f1(n) + K2 * g1(n-1) */
emilmont 1:fdd22bb7aa52 223 fnext1 = fcurr1 + ((*pk) * gcurr1);
emilmont 1:fdd22bb7aa52 224 /* Process second sample for 2nd, 6th .. tap */
emilmont 1:fdd22bb7aa52 225 /* for sample 2 processing */
emilmont 1:fdd22bb7aa52 226 fnext2 = fcurr2 + ((*pk) * gnext1);
emilmont 1:fdd22bb7aa52 227 /* Process third sample for 2nd, 6th .. tap */
emilmont 1:fdd22bb7aa52 228 fnext3 = fcurr3 + ((*pk) * gnext2);
emilmont 1:fdd22bb7aa52 229 /* Process fourth sample for 2nd, 6th .. tap */
emilmont 1:fdd22bb7aa52 230 fnext4 = fcurr4 + ((*pk) * gnext3);
emilmont 1:fdd22bb7aa52 231
emilmont 1:fdd22bb7aa52 232 /* g2(n) = f1(n) * K2 + g1(n-1) */
emilmont 1:fdd22bb7aa52 233 /* Calculation of state values for next stage */
emilmont 1:fdd22bb7aa52 234 gnext4 = (fcurr4 * (*pk)) + gnext3;
emilmont 1:fdd22bb7aa52 235 gnext3 = (fcurr3 * (*pk)) + gnext2;
emilmont 1:fdd22bb7aa52 236 gnext2 = (fcurr2 * (*pk)) + gnext1;
emilmont 1:fdd22bb7aa52 237 gnext1 = (fcurr1 * (*pk++)) + gcurr1;
emilmont 1:fdd22bb7aa52 238
emilmont 1:fdd22bb7aa52 239
emilmont 1:fdd22bb7aa52 240 /* Read g2(n-1), g4(n-1) .... from state */
emilmont 1:fdd22bb7aa52 241 gcurr1 = *px;
emilmont 1:fdd22bb7aa52 242
emilmont 1:fdd22bb7aa52 243 /* save g2(n) in state buffer */
emilmont 1:fdd22bb7aa52 244 *px++ = gnext4;
emilmont 1:fdd22bb7aa52 245
emilmont 1:fdd22bb7aa52 246 /* Sample processing for K3, K7.... */
emilmont 1:fdd22bb7aa52 247 /* Process first sample for 3rd, 7th .. tap */
emilmont 1:fdd22bb7aa52 248 /* f3(n) = f2(n) + K3 * g2(n-1) */
emilmont 1:fdd22bb7aa52 249 fcurr1 = fnext1 + ((*pk) * gcurr1);
emilmont 1:fdd22bb7aa52 250 /* Process second sample for 3rd, 7th .. tap */
emilmont 1:fdd22bb7aa52 251 fcurr2 = fnext2 + ((*pk) * gnext1);
emilmont 1:fdd22bb7aa52 252 /* Process third sample for 3rd, 7th .. tap */
emilmont 1:fdd22bb7aa52 253 fcurr3 = fnext3 + ((*pk) * gnext2);
emilmont 1:fdd22bb7aa52 254 /* Process fourth sample for 3rd, 7th .. tap */
emilmont 1:fdd22bb7aa52 255 fcurr4 = fnext4 + ((*pk) * gnext3);
emilmont 1:fdd22bb7aa52 256
emilmont 1:fdd22bb7aa52 257 /* Calculation of state values for next stage */
emilmont 1:fdd22bb7aa52 258 /* g3(n) = f2(n) * K3 + g2(n-1) */
emilmont 1:fdd22bb7aa52 259 gnext4 = (fnext4 * (*pk)) + gnext3;
emilmont 1:fdd22bb7aa52 260 gnext3 = (fnext3 * (*pk)) + gnext2;
emilmont 1:fdd22bb7aa52 261 gnext2 = (fnext2 * (*pk)) + gnext1;
emilmont 1:fdd22bb7aa52 262 gnext1 = (fnext1 * (*pk++)) + gcurr1;
emilmont 1:fdd22bb7aa52 263
emilmont 1:fdd22bb7aa52 264
emilmont 1:fdd22bb7aa52 265 /* Read g1(n-1), g3(n-1) .... from state */
emilmont 1:fdd22bb7aa52 266 gcurr1 = *px;
emilmont 1:fdd22bb7aa52 267
emilmont 1:fdd22bb7aa52 268 /* save g3(n) in state buffer */
emilmont 1:fdd22bb7aa52 269 *px++ = gnext4;
emilmont 1:fdd22bb7aa52 270
emilmont 1:fdd22bb7aa52 271 /* Sample processing for K4, K8.... */
emilmont 1:fdd22bb7aa52 272 /* Process first sample for 4th, 8th .. tap */
emilmont 1:fdd22bb7aa52 273 /* f4(n) = f3(n) + K4 * g3(n-1) */
emilmont 1:fdd22bb7aa52 274 fnext1 = fcurr1 + ((*pk) * gcurr1);
emilmont 1:fdd22bb7aa52 275 /* Process second sample for 4th, 8th .. tap */
emilmont 1:fdd22bb7aa52 276 /* for sample 2 processing */
emilmont 1:fdd22bb7aa52 277 fnext2 = fcurr2 + ((*pk) * gnext1);
emilmont 1:fdd22bb7aa52 278 /* Process third sample for 4th, 8th .. tap */
emilmont 1:fdd22bb7aa52 279 fnext3 = fcurr3 + ((*pk) * gnext2);
emilmont 1:fdd22bb7aa52 280 /* Process fourth sample for 4th, 8th .. tap */
emilmont 1:fdd22bb7aa52 281 fnext4 = fcurr4 + ((*pk) * gnext3);
emilmont 1:fdd22bb7aa52 282
emilmont 1:fdd22bb7aa52 283 /* g4(n) = f3(n) * K4 + g3(n-1) */
emilmont 1:fdd22bb7aa52 284 /* Calculation of state values for next stage */
emilmont 1:fdd22bb7aa52 285 gnext4 = (fcurr4 * (*pk)) + gnext3;
emilmont 1:fdd22bb7aa52 286 gnext3 = (fcurr3 * (*pk)) + gnext2;
emilmont 1:fdd22bb7aa52 287 gnext2 = (fcurr2 * (*pk)) + gnext1;
emilmont 1:fdd22bb7aa52 288 gnext1 = (fcurr1 * (*pk++)) + gcurr1;
emilmont 1:fdd22bb7aa52 289
emilmont 1:fdd22bb7aa52 290 /* Read g2(n-1), g4(n-1) .... from state */
emilmont 1:fdd22bb7aa52 291 gcurr1 = *px;
emilmont 1:fdd22bb7aa52 292
emilmont 1:fdd22bb7aa52 293 /* save g4(n) in state buffer */
emilmont 1:fdd22bb7aa52 294 *px++ = gnext4;
emilmont 1:fdd22bb7aa52 295
emilmont 1:fdd22bb7aa52 296 /* Sample processing for K5, K9.... */
emilmont 1:fdd22bb7aa52 297 /* Process first sample for 5th, 9th .. tap */
emilmont 1:fdd22bb7aa52 298 /* f5(n) = f4(n) + K5 * g4(n-1) */
emilmont 1:fdd22bb7aa52 299 fcurr1 = fnext1 + ((*pk) * gcurr1);
emilmont 1:fdd22bb7aa52 300 /* Process second sample for 5th, 9th .. tap */
emilmont 1:fdd22bb7aa52 301 fcurr2 = fnext2 + ((*pk) * gnext1);
emilmont 1:fdd22bb7aa52 302 /* Process third sample for 5th, 9th .. tap */
emilmont 1:fdd22bb7aa52 303 fcurr3 = fnext3 + ((*pk) * gnext2);
emilmont 1:fdd22bb7aa52 304 /* Process fourth sample for 5th, 9th .. tap */
emilmont 1:fdd22bb7aa52 305 fcurr4 = fnext4 + ((*pk) * gnext3);
emilmont 1:fdd22bb7aa52 306
emilmont 1:fdd22bb7aa52 307 /* Calculation of state values for next stage */
emilmont 1:fdd22bb7aa52 308 /* g5(n) = f4(n) * K5 + g4(n-1) */
emilmont 1:fdd22bb7aa52 309 gnext4 = (fnext4 * (*pk)) + gnext3;
emilmont 1:fdd22bb7aa52 310 gnext3 = (fnext3 * (*pk)) + gnext2;
emilmont 1:fdd22bb7aa52 311 gnext2 = (fnext2 * (*pk)) + gnext1;
emilmont 1:fdd22bb7aa52 312 gnext1 = (fnext1 * (*pk++)) + gcurr1;
emilmont 1:fdd22bb7aa52 313
emilmont 1:fdd22bb7aa52 314 stageCnt--;
emilmont 1:fdd22bb7aa52 315 }
emilmont 1:fdd22bb7aa52 316
emilmont 1:fdd22bb7aa52 317 /* If the (filter length -1) is not a multiple of 4, compute the remaining filter taps */
emilmont 1:fdd22bb7aa52 318 stageCnt = (numStages - 1u) % 0x4u;
emilmont 1:fdd22bb7aa52 319
emilmont 1:fdd22bb7aa52 320 while(stageCnt > 0u)
emilmont 1:fdd22bb7aa52 321 {
emilmont 1:fdd22bb7aa52 322 gcurr1 = *px;
emilmont 1:fdd22bb7aa52 323
emilmont 1:fdd22bb7aa52 324 /* save g value in state buffer */
emilmont 1:fdd22bb7aa52 325 *px++ = gnext4;
emilmont 1:fdd22bb7aa52 326
emilmont 1:fdd22bb7aa52 327 /* Process four samples for last three taps here */
emilmont 1:fdd22bb7aa52 328 fnext1 = fcurr1 + ((*pk) * gcurr1);
emilmont 1:fdd22bb7aa52 329 fnext2 = fcurr2 + ((*pk) * gnext1);
emilmont 1:fdd22bb7aa52 330 fnext3 = fcurr3 + ((*pk) * gnext2);
emilmont 1:fdd22bb7aa52 331 fnext4 = fcurr4 + ((*pk) * gnext3);
emilmont 1:fdd22bb7aa52 332
emilmont 1:fdd22bb7aa52 333 /* g1(n) = f0(n) * K1 + g0(n-1) */
emilmont 1:fdd22bb7aa52 334 gnext4 = (fcurr4 * (*pk)) + gnext3;
emilmont 1:fdd22bb7aa52 335 gnext3 = (fcurr3 * (*pk)) + gnext2;
emilmont 1:fdd22bb7aa52 336 gnext2 = (fcurr2 * (*pk)) + gnext1;
emilmont 1:fdd22bb7aa52 337 gnext1 = (fcurr1 * (*pk++)) + gcurr1;
emilmont 1:fdd22bb7aa52 338
emilmont 1:fdd22bb7aa52 339 /* Update of f values for next coefficient set processing */
emilmont 1:fdd22bb7aa52 340 fcurr1 = fnext1;
emilmont 1:fdd22bb7aa52 341 fcurr2 = fnext2;
emilmont 1:fdd22bb7aa52 342 fcurr3 = fnext3;
emilmont 1:fdd22bb7aa52 343 fcurr4 = fnext4;
emilmont 1:fdd22bb7aa52 344
emilmont 1:fdd22bb7aa52 345 stageCnt--;
emilmont 1:fdd22bb7aa52 346
emilmont 1:fdd22bb7aa52 347 }
emilmont 1:fdd22bb7aa52 348
emilmont 1:fdd22bb7aa52 349 /* The results in the 4 accumulators, store in the destination buffer. */
emilmont 1:fdd22bb7aa52 350 /* y(n) = fN(n) */
emilmont 1:fdd22bb7aa52 351 *pDst++ = fcurr1;
emilmont 1:fdd22bb7aa52 352 *pDst++ = fcurr2;
emilmont 1:fdd22bb7aa52 353 *pDst++ = fcurr3;
emilmont 1:fdd22bb7aa52 354 *pDst++ = fcurr4;
emilmont 1:fdd22bb7aa52 355
emilmont 1:fdd22bb7aa52 356 blkCnt--;
emilmont 1:fdd22bb7aa52 357 }
emilmont 1:fdd22bb7aa52 358
emilmont 1:fdd22bb7aa52 359 /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
emilmont 1:fdd22bb7aa52 360 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 361 blkCnt = blockSize % 0x4u;
emilmont 1:fdd22bb7aa52 362
emilmont 1:fdd22bb7aa52 363 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 364 {
emilmont 1:fdd22bb7aa52 365 /* f0(n) = x(n) */
emilmont 1:fdd22bb7aa52 366 fcurr1 = *pSrc++;
emilmont 1:fdd22bb7aa52 367
emilmont 1:fdd22bb7aa52 368 /* Initialize coeff pointer */
emilmont 1:fdd22bb7aa52 369 pk = (pCoeffs);
emilmont 1:fdd22bb7aa52 370
emilmont 1:fdd22bb7aa52 371 /* Initialize state pointer */
emilmont 1:fdd22bb7aa52 372 px = pState;
emilmont 1:fdd22bb7aa52 373
emilmont 1:fdd22bb7aa52 374 /* read g2(n) from state buffer */
emilmont 1:fdd22bb7aa52 375 gcurr1 = *px;
emilmont 1:fdd22bb7aa52 376
emilmont 1:fdd22bb7aa52 377 /* for sample 1 processing */
emilmont 1:fdd22bb7aa52 378 /* f1(n) = f0(n) + K1 * g0(n-1) */
emilmont 1:fdd22bb7aa52 379 fnext1 = fcurr1 + ((*pk) * gcurr1);
emilmont 1:fdd22bb7aa52 380 /* g1(n) = f0(n) * K1 + g0(n-1) */
emilmont 1:fdd22bb7aa52 381 gnext1 = (fcurr1 * (*pk++)) + gcurr1;
emilmont 1:fdd22bb7aa52 382
emilmont 1:fdd22bb7aa52 383 /* save g1(n) in state buffer */
emilmont 1:fdd22bb7aa52 384 *px++ = fcurr1;
emilmont 1:fdd22bb7aa52 385
emilmont 1:fdd22bb7aa52 386 /* f1(n) is saved in fcurr1
emilmont 1:fdd22bb7aa52 387 for next stage processing */
emilmont 1:fdd22bb7aa52 388 fcurr1 = fnext1;
emilmont 1:fdd22bb7aa52 389
emilmont 1:fdd22bb7aa52 390 stageCnt = (numStages - 1u);
emilmont 1:fdd22bb7aa52 391
emilmont 1:fdd22bb7aa52 392 /* stage loop */
emilmont 1:fdd22bb7aa52 393 while(stageCnt > 0u)
emilmont 1:fdd22bb7aa52 394 {
emilmont 1:fdd22bb7aa52 395 /* read g2(n) from state buffer */
emilmont 1:fdd22bb7aa52 396 gcurr1 = *px;
emilmont 1:fdd22bb7aa52 397
emilmont 1:fdd22bb7aa52 398 /* save g1(n) in state buffer */
emilmont 1:fdd22bb7aa52 399 *px++ = gnext1;
emilmont 1:fdd22bb7aa52 400
emilmont 1:fdd22bb7aa52 401 /* Sample processing for K2, K3.... */
emilmont 1:fdd22bb7aa52 402 /* f2(n) = f1(n) + K2 * g1(n-1) */
emilmont 1:fdd22bb7aa52 403 fnext1 = fcurr1 + ((*pk) * gcurr1);
emilmont 1:fdd22bb7aa52 404 /* g2(n) = f1(n) * K2 + g1(n-1) */
emilmont 1:fdd22bb7aa52 405 gnext1 = (fcurr1 * (*pk++)) + gcurr1;
emilmont 1:fdd22bb7aa52 406
emilmont 1:fdd22bb7aa52 407 /* f1(n) is saved in fcurr1
emilmont 1:fdd22bb7aa52 408 for next stage processing */
emilmont 1:fdd22bb7aa52 409 fcurr1 = fnext1;
emilmont 1:fdd22bb7aa52 410
emilmont 1:fdd22bb7aa52 411 stageCnt--;
emilmont 1:fdd22bb7aa52 412
emilmont 1:fdd22bb7aa52 413 }
emilmont 1:fdd22bb7aa52 414
emilmont 1:fdd22bb7aa52 415 /* y(n) = fN(n) */
emilmont 1:fdd22bb7aa52 416 *pDst++ = fcurr1;
emilmont 1:fdd22bb7aa52 417
emilmont 1:fdd22bb7aa52 418 blkCnt--;
emilmont 1:fdd22bb7aa52 419
emilmont 1:fdd22bb7aa52 420 }
emilmont 1:fdd22bb7aa52 421
emilmont 1:fdd22bb7aa52 422 #else
emilmont 1:fdd22bb7aa52 423
emilmont 1:fdd22bb7aa52 424 /* Run the below code for Cortex-M0 */
emilmont 1:fdd22bb7aa52 425
emilmont 1:fdd22bb7aa52 426 float32_t fcurr, fnext, gcurr, gnext; /* temporary variables */
emilmont 1:fdd22bb7aa52 427 uint32_t numStages = S->numStages; /* Length of the filter */
emilmont 1:fdd22bb7aa52 428 uint32_t blkCnt, stageCnt; /* temporary variables for counts */
emilmont 1:fdd22bb7aa52 429
emilmont 1:fdd22bb7aa52 430 pState = &S->pState[0];
emilmont 1:fdd22bb7aa52 431
emilmont 1:fdd22bb7aa52 432 blkCnt = blockSize;
emilmont 1:fdd22bb7aa52 433
emilmont 1:fdd22bb7aa52 434 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 435 {
emilmont 1:fdd22bb7aa52 436 /* f0(n) = x(n) */
emilmont 1:fdd22bb7aa52 437 fcurr = *pSrc++;
emilmont 1:fdd22bb7aa52 438
emilmont 1:fdd22bb7aa52 439 /* Initialize coeff pointer */
emilmont 1:fdd22bb7aa52 440 pk = pCoeffs;
emilmont 1:fdd22bb7aa52 441
emilmont 1:fdd22bb7aa52 442 /* Initialize state pointer */
emilmont 1:fdd22bb7aa52 443 px = pState;
emilmont 1:fdd22bb7aa52 444
emilmont 1:fdd22bb7aa52 445 /* read g0(n-1) from state buffer */
emilmont 1:fdd22bb7aa52 446 gcurr = *px;
emilmont 1:fdd22bb7aa52 447
emilmont 1:fdd22bb7aa52 448 /* for sample 1 processing */
emilmont 1:fdd22bb7aa52 449 /* f1(n) = f0(n) + K1 * g0(n-1) */
emilmont 1:fdd22bb7aa52 450 fnext = fcurr + ((*pk) * gcurr);
emilmont 1:fdd22bb7aa52 451 /* g1(n) = f0(n) * K1 + g0(n-1) */
emilmont 1:fdd22bb7aa52 452 gnext = (fcurr * (*pk++)) + gcurr;
emilmont 1:fdd22bb7aa52 453
emilmont 1:fdd22bb7aa52 454 /* save f0(n) in state buffer */
emilmont 1:fdd22bb7aa52 455 *px++ = fcurr;
emilmont 1:fdd22bb7aa52 456
emilmont 1:fdd22bb7aa52 457 /* f1(n) is saved in fcurr
emilmont 1:fdd22bb7aa52 458 for next stage processing */
emilmont 1:fdd22bb7aa52 459 fcurr = fnext;
emilmont 1:fdd22bb7aa52 460
emilmont 1:fdd22bb7aa52 461 stageCnt = (numStages - 1u);
emilmont 1:fdd22bb7aa52 462
emilmont 1:fdd22bb7aa52 463 /* stage loop */
emilmont 1:fdd22bb7aa52 464 while(stageCnt > 0u)
emilmont 1:fdd22bb7aa52 465 {
emilmont 1:fdd22bb7aa52 466 /* read g2(n) from state buffer */
emilmont 1:fdd22bb7aa52 467 gcurr = *px;
emilmont 1:fdd22bb7aa52 468
emilmont 1:fdd22bb7aa52 469 /* save g1(n) in state buffer */
emilmont 1:fdd22bb7aa52 470 *px++ = gnext;
emilmont 1:fdd22bb7aa52 471
emilmont 1:fdd22bb7aa52 472 /* Sample processing for K2, K3.... */
emilmont 1:fdd22bb7aa52 473 /* f2(n) = f1(n) + K2 * g1(n-1) */
emilmont 1:fdd22bb7aa52 474 fnext = fcurr + ((*pk) * gcurr);
emilmont 1:fdd22bb7aa52 475 /* g2(n) = f1(n) * K2 + g1(n-1) */
emilmont 1:fdd22bb7aa52 476 gnext = (fcurr * (*pk++)) + gcurr;
emilmont 1:fdd22bb7aa52 477
emilmont 1:fdd22bb7aa52 478 /* f1(n) is saved in fcurr1
emilmont 1:fdd22bb7aa52 479 for next stage processing */
emilmont 1:fdd22bb7aa52 480 fcurr = fnext;
emilmont 1:fdd22bb7aa52 481
emilmont 1:fdd22bb7aa52 482 stageCnt--;
emilmont 1:fdd22bb7aa52 483
emilmont 1:fdd22bb7aa52 484 }
emilmont 1:fdd22bb7aa52 485
emilmont 1:fdd22bb7aa52 486 /* y(n) = fN(n) */
emilmont 1:fdd22bb7aa52 487 *pDst++ = fcurr;
emilmont 1:fdd22bb7aa52 488
emilmont 1:fdd22bb7aa52 489 blkCnt--;
emilmont 1:fdd22bb7aa52 490
emilmont 1:fdd22bb7aa52 491 }
emilmont 1:fdd22bb7aa52 492
emilmont 1:fdd22bb7aa52 493 #endif /* #ifndef ARM_MATH_CM0 */
emilmont 1:fdd22bb7aa52 494
emilmont 1:fdd22bb7aa52 495 }
emilmont 1:fdd22bb7aa52 496
emilmont 1:fdd22bb7aa52 497 /**
emilmont 1:fdd22bb7aa52 498 * @} end of FIR_Lattice group
emilmont 1:fdd22bb7aa52 499 */