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_iir_lattice_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?
User | Revision | Line number | New 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_iir_lattice_q15.c |
emilmont | 1:fdd22bb7aa52 | 9 | * |
emilmont | 2:da51fb522205 | 10 | * Description: Q15 IIR lattice filter processing function. |
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 IIR_Lattice |
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 Q15 IIR lattice filter. |
emilmont | 1:fdd22bb7aa52 | 49 | * @param[in] *S points to an instance of the Q15 IIR lattice 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 samples to process. |
emilmont | 1:fdd22bb7aa52 | 53 | * @return none. |
emilmont | 1:fdd22bb7aa52 | 54 | * |
emilmont | 1:fdd22bb7aa52 | 55 | * @details |
emilmont | 1:fdd22bb7aa52 | 56 | * <b>Scaling and Overflow Behavior:</b> |
emilmont | 1:fdd22bb7aa52 | 57 | * \par |
emilmont | 1:fdd22bb7aa52 | 58 | * The function is implemented using a 64-bit internal accumulator. |
emilmont | 1:fdd22bb7aa52 | 59 | * Both coefficients and state variables are represented in 1.15 format and multiplications yield a 2.30 result. |
emilmont | 1:fdd22bb7aa52 | 60 | * The 2.30 intermediate results are accumulated in a 64-bit accumulator in 34.30 format. |
emilmont | 1:fdd22bb7aa52 | 61 | * There is no risk of internal overflow with this approach and the full precision of intermediate multiplications is preserved. |
emilmont | 1:fdd22bb7aa52 | 62 | * After all additions have been performed, the accumulator is truncated to 34.15 format by discarding low 15 bits. |
emilmont | 1:fdd22bb7aa52 | 63 | * Lastly, the accumulator is saturated to yield a result in 1.15 format. |
emilmont | 1:fdd22bb7aa52 | 64 | */ |
emilmont | 1:fdd22bb7aa52 | 65 | |
emilmont | 1:fdd22bb7aa52 | 66 | void arm_iir_lattice_q15( |
emilmont | 1:fdd22bb7aa52 | 67 | const arm_iir_lattice_instance_q15 * S, |
emilmont | 1:fdd22bb7aa52 | 68 | q15_t * pSrc, |
emilmont | 1:fdd22bb7aa52 | 69 | q15_t * pDst, |
emilmont | 1:fdd22bb7aa52 | 70 | uint32_t blockSize) |
emilmont | 1:fdd22bb7aa52 | 71 | { |
emilmont | 1:fdd22bb7aa52 | 72 | |
emilmont | 1:fdd22bb7aa52 | 73 | |
emilmont | 1:fdd22bb7aa52 | 74 | #ifndef ARM_MATH_CM0 |
emilmont | 1:fdd22bb7aa52 | 75 | |
emilmont | 1:fdd22bb7aa52 | 76 | /* Run the below code for Cortex-M4 and Cortex-M3 */ |
emilmont | 1:fdd22bb7aa52 | 77 | |
emilmont | 1:fdd22bb7aa52 | 78 | q31_t fcurr, fnext, gcurr = 0, gnext; /* Temporary variables for lattice stages */ |
emilmont | 1:fdd22bb7aa52 | 79 | q15_t gnext1, gnext2; /* Temporary variables for lattice stages */ |
emilmont | 1:fdd22bb7aa52 | 80 | uint32_t stgCnt; /* Temporary variables for counts */ |
emilmont | 1:fdd22bb7aa52 | 81 | q63_t acc; /* Accumlator */ |
emilmont | 1:fdd22bb7aa52 | 82 | uint32_t blkCnt, tapCnt; /* Temporary variables for counts */ |
emilmont | 1:fdd22bb7aa52 | 83 | q15_t *px1, *px2, *pk, *pv; /* temporary pointers for state and coef */ |
emilmont | 1:fdd22bb7aa52 | 84 | uint32_t numStages = S->numStages; /* number of stages */ |
emilmont | 1:fdd22bb7aa52 | 85 | q15_t *pState; /* State pointer */ |
emilmont | 1:fdd22bb7aa52 | 86 | q15_t *pStateCurnt; /* State current pointer */ |
emilmont | 1:fdd22bb7aa52 | 87 | q15_t out; /* Temporary variable for output */ |
emilmont | 1:fdd22bb7aa52 | 88 | #ifdef UNALIGNED_SUPPORT_DISABLE |
emilmont | 1:fdd22bb7aa52 | 89 | q15_t v1, v2; |
emilmont | 1:fdd22bb7aa52 | 90 | #endif |
emilmont | 1:fdd22bb7aa52 | 91 | q31_t v; /* Temporary variable for ladder coefficient */ |
emilmont | 1:fdd22bb7aa52 | 92 | |
emilmont | 1:fdd22bb7aa52 | 93 | |
emilmont | 1:fdd22bb7aa52 | 94 | blkCnt = blockSize; |
emilmont | 1:fdd22bb7aa52 | 95 | |
emilmont | 1:fdd22bb7aa52 | 96 | pState = &S->pState[0]; |
emilmont | 1:fdd22bb7aa52 | 97 | |
emilmont | 1:fdd22bb7aa52 | 98 | /* Sample processing */ |
emilmont | 1:fdd22bb7aa52 | 99 | while(blkCnt > 0u) |
emilmont | 1:fdd22bb7aa52 | 100 | { |
emilmont | 1:fdd22bb7aa52 | 101 | /* Read Sample from input buffer */ |
emilmont | 1:fdd22bb7aa52 | 102 | /* fN(n) = x(n) */ |
emilmont | 1:fdd22bb7aa52 | 103 | fcurr = *pSrc++; |
emilmont | 1:fdd22bb7aa52 | 104 | |
emilmont | 1:fdd22bb7aa52 | 105 | /* Initialize state read pointer */ |
emilmont | 1:fdd22bb7aa52 | 106 | px1 = pState; |
emilmont | 1:fdd22bb7aa52 | 107 | /* Initialize state write pointer */ |
emilmont | 1:fdd22bb7aa52 | 108 | px2 = pState; |
emilmont | 1:fdd22bb7aa52 | 109 | /* Set accumulator to zero */ |
emilmont | 1:fdd22bb7aa52 | 110 | acc = 0; |
emilmont | 1:fdd22bb7aa52 | 111 | /* Initialize Ladder coeff pointer */ |
emilmont | 1:fdd22bb7aa52 | 112 | pv = &S->pvCoeffs[0]; |
emilmont | 1:fdd22bb7aa52 | 113 | /* Initialize Reflection coeff pointer */ |
emilmont | 1:fdd22bb7aa52 | 114 | pk = &S->pkCoeffs[0]; |
emilmont | 1:fdd22bb7aa52 | 115 | |
emilmont | 1:fdd22bb7aa52 | 116 | |
emilmont | 1:fdd22bb7aa52 | 117 | /* Process sample for first tap */ |
emilmont | 1:fdd22bb7aa52 | 118 | gcurr = *px1++; |
emilmont | 1:fdd22bb7aa52 | 119 | /* fN-1(n) = fN(n) - kN * gN-1(n-1) */ |
emilmont | 1:fdd22bb7aa52 | 120 | fnext = fcurr - (((q31_t) gcurr * (*pk)) >> 15); |
emilmont | 1:fdd22bb7aa52 | 121 | fnext = __SSAT(fnext, 16); |
emilmont | 1:fdd22bb7aa52 | 122 | /* gN(n) = kN * fN-1(n) + gN-1(n-1) */ |
emilmont | 1:fdd22bb7aa52 | 123 | gnext = (((q31_t) fnext * (*pk++)) >> 15) + gcurr; |
emilmont | 1:fdd22bb7aa52 | 124 | gnext = __SSAT(gnext, 16); |
emilmont | 1:fdd22bb7aa52 | 125 | /* write gN(n) into state for next sample processing */ |
emilmont | 1:fdd22bb7aa52 | 126 | *px2++ = (q15_t) gnext; |
emilmont | 1:fdd22bb7aa52 | 127 | /* y(n) += gN(n) * vN */ |
emilmont | 1:fdd22bb7aa52 | 128 | acc += (q31_t) ((gnext * (*pv++))); |
emilmont | 1:fdd22bb7aa52 | 129 | |
emilmont | 1:fdd22bb7aa52 | 130 | |
emilmont | 1:fdd22bb7aa52 | 131 | /* Update f values for next coefficient processing */ |
emilmont | 1:fdd22bb7aa52 | 132 | fcurr = fnext; |
emilmont | 1:fdd22bb7aa52 | 133 | |
emilmont | 1:fdd22bb7aa52 | 134 | /* Loop unrolling. Process 4 taps at a time. */ |
emilmont | 1:fdd22bb7aa52 | 135 | tapCnt = (numStages - 1u) >> 2; |
emilmont | 1:fdd22bb7aa52 | 136 | |
emilmont | 1:fdd22bb7aa52 | 137 | while(tapCnt > 0u) |
emilmont | 1:fdd22bb7aa52 | 138 | { |
emilmont | 1:fdd22bb7aa52 | 139 | |
emilmont | 1:fdd22bb7aa52 | 140 | /* Process sample for 2nd, 6th ...taps */ |
emilmont | 1:fdd22bb7aa52 | 141 | /* Read gN-2(n-1) from state buffer */ |
emilmont | 1:fdd22bb7aa52 | 142 | gcurr = *px1++; |
emilmont | 1:fdd22bb7aa52 | 143 | /* Process sample for 2nd, 6th .. taps */ |
emilmont | 1:fdd22bb7aa52 | 144 | /* fN-2(n) = fN-1(n) - kN-1 * gN-2(n-1) */ |
emilmont | 1:fdd22bb7aa52 | 145 | fnext = fcurr - (((q31_t) gcurr * (*pk)) >> 15); |
emilmont | 1:fdd22bb7aa52 | 146 | fnext = __SSAT(fnext, 16); |
emilmont | 1:fdd22bb7aa52 | 147 | /* gN-1(n) = kN-1 * fN-2(n) + gN-2(n-1) */ |
emilmont | 1:fdd22bb7aa52 | 148 | gnext = (((q31_t) fnext * (*pk++)) >> 15) + gcurr; |
emilmont | 1:fdd22bb7aa52 | 149 | gnext1 = (q15_t) __SSAT(gnext, 16); |
emilmont | 1:fdd22bb7aa52 | 150 | /* write gN-1(n) into state */ |
emilmont | 1:fdd22bb7aa52 | 151 | *px2++ = (q15_t) gnext1; |
emilmont | 1:fdd22bb7aa52 | 152 | |
emilmont | 1:fdd22bb7aa52 | 153 | |
emilmont | 1:fdd22bb7aa52 | 154 | /* Process sample for 3nd, 7th ...taps */ |
emilmont | 1:fdd22bb7aa52 | 155 | /* Read gN-3(n-1) from state */ |
emilmont | 1:fdd22bb7aa52 | 156 | gcurr = *px1++; |
emilmont | 1:fdd22bb7aa52 | 157 | /* Process sample for 3rd, 7th .. taps */ |
emilmont | 1:fdd22bb7aa52 | 158 | /* fN-3(n) = fN-2(n) - kN-2 * gN-3(n-1) */ |
emilmont | 1:fdd22bb7aa52 | 159 | fcurr = fnext - (((q31_t) gcurr * (*pk)) >> 15); |
emilmont | 1:fdd22bb7aa52 | 160 | fcurr = __SSAT(fcurr, 16); |
emilmont | 1:fdd22bb7aa52 | 161 | /* gN-2(n) = kN-2 * fN-3(n) + gN-3(n-1) */ |
emilmont | 1:fdd22bb7aa52 | 162 | gnext = (((q31_t) fcurr * (*pk++)) >> 15) + gcurr; |
emilmont | 1:fdd22bb7aa52 | 163 | gnext2 = (q15_t) __SSAT(gnext, 16); |
emilmont | 1:fdd22bb7aa52 | 164 | /* write gN-2(n) into state */ |
emilmont | 1:fdd22bb7aa52 | 165 | *px2++ = (q15_t) gnext2; |
emilmont | 1:fdd22bb7aa52 | 166 | |
emilmont | 1:fdd22bb7aa52 | 167 | /* Read vN-1 and vN-2 at a time */ |
emilmont | 1:fdd22bb7aa52 | 168 | #ifndef UNALIGNED_SUPPORT_DISABLE |
emilmont | 1:fdd22bb7aa52 | 169 | |
emilmont | 1:fdd22bb7aa52 | 170 | v = *__SIMD32(pv)++; |
emilmont | 1:fdd22bb7aa52 | 171 | |
emilmont | 1:fdd22bb7aa52 | 172 | #else |
emilmont | 1:fdd22bb7aa52 | 173 | |
emilmont | 2:da51fb522205 | 174 | v1 = *pv++; |
emilmont | 2:da51fb522205 | 175 | v2 = *pv++; |
emilmont | 1:fdd22bb7aa52 | 176 | |
emilmont | 1:fdd22bb7aa52 | 177 | #ifndef ARM_MATH_BIG_ENDIAN |
emilmont | 1:fdd22bb7aa52 | 178 | |
emilmont | 2:da51fb522205 | 179 | v = __PKHBT(v1, v2, 16); |
emilmont | 1:fdd22bb7aa52 | 180 | |
emilmont | 1:fdd22bb7aa52 | 181 | #else |
emilmont | 1:fdd22bb7aa52 | 182 | |
emilmont | 2:da51fb522205 | 183 | v = __PKHBT(v2, v1, 16); |
emilmont | 1:fdd22bb7aa52 | 184 | |
emilmont | 2:da51fb522205 | 185 | #endif /* #ifndef ARM_MATH_BIG_ENDIAN */ |
emilmont | 1:fdd22bb7aa52 | 186 | |
emilmont | 2:da51fb522205 | 187 | #endif /* #ifndef UNALIGNED_SUPPORT_DISABLE */ |
emilmont | 1:fdd22bb7aa52 | 188 | |
emilmont | 1:fdd22bb7aa52 | 189 | |
emilmont | 1:fdd22bb7aa52 | 190 | /* Pack gN-1(n) and gN-2(n) */ |
emilmont | 1:fdd22bb7aa52 | 191 | |
emilmont | 1:fdd22bb7aa52 | 192 | #ifndef ARM_MATH_BIG_ENDIAN |
emilmont | 1:fdd22bb7aa52 | 193 | |
emilmont | 1:fdd22bb7aa52 | 194 | gnext = __PKHBT(gnext1, gnext2, 16); |
emilmont | 1:fdd22bb7aa52 | 195 | |
emilmont | 1:fdd22bb7aa52 | 196 | #else |
emilmont | 1:fdd22bb7aa52 | 197 | |
emilmont | 1:fdd22bb7aa52 | 198 | gnext = __PKHBT(gnext2, gnext1, 16); |
emilmont | 1:fdd22bb7aa52 | 199 | |
emilmont | 1:fdd22bb7aa52 | 200 | #endif /* #ifndef ARM_MATH_BIG_ENDIAN */ |
emilmont | 1:fdd22bb7aa52 | 201 | |
emilmont | 1:fdd22bb7aa52 | 202 | /* y(n) += gN-1(n) * vN-1 */ |
emilmont | 1:fdd22bb7aa52 | 203 | /* process for gN-5(n) * vN-5, gN-9(n) * vN-9 ... */ |
emilmont | 1:fdd22bb7aa52 | 204 | /* y(n) += gN-2(n) * vN-2 */ |
emilmont | 1:fdd22bb7aa52 | 205 | /* process for gN-6(n) * vN-6, gN-10(n) * vN-10 ... */ |
emilmont | 1:fdd22bb7aa52 | 206 | acc = __SMLALD(gnext, v, acc); |
emilmont | 1:fdd22bb7aa52 | 207 | |
emilmont | 1:fdd22bb7aa52 | 208 | |
emilmont | 1:fdd22bb7aa52 | 209 | /* Process sample for 4th, 8th ...taps */ |
emilmont | 1:fdd22bb7aa52 | 210 | /* Read gN-4(n-1) from state */ |
emilmont | 1:fdd22bb7aa52 | 211 | gcurr = *px1++; |
emilmont | 1:fdd22bb7aa52 | 212 | /* Process sample for 4th, 8th .. taps */ |
emilmont | 1:fdd22bb7aa52 | 213 | /* fN-4(n) = fN-3(n) - kN-3 * gN-4(n-1) */ |
emilmont | 1:fdd22bb7aa52 | 214 | fnext = fcurr - (((q31_t) gcurr * (*pk)) >> 15); |
emilmont | 1:fdd22bb7aa52 | 215 | fnext = __SSAT(fnext, 16); |
emilmont | 1:fdd22bb7aa52 | 216 | /* gN-3(n) = kN-3 * fN-1(n) + gN-1(n-1) */ |
emilmont | 1:fdd22bb7aa52 | 217 | gnext = (((q31_t) fnext * (*pk++)) >> 15) + gcurr; |
emilmont | 1:fdd22bb7aa52 | 218 | gnext1 = (q15_t) __SSAT(gnext, 16); |
emilmont | 1:fdd22bb7aa52 | 219 | /* write gN-3(n) for the next sample process */ |
emilmont | 1:fdd22bb7aa52 | 220 | *px2++ = (q15_t) gnext1; |
emilmont | 1:fdd22bb7aa52 | 221 | |
emilmont | 1:fdd22bb7aa52 | 222 | |
emilmont | 1:fdd22bb7aa52 | 223 | /* Process sample for 5th, 9th ...taps */ |
emilmont | 1:fdd22bb7aa52 | 224 | /* Read gN-5(n-1) from state */ |
emilmont | 1:fdd22bb7aa52 | 225 | gcurr = *px1++; |
emilmont | 1:fdd22bb7aa52 | 226 | /* Process sample for 5th, 9th .. taps */ |
emilmont | 1:fdd22bb7aa52 | 227 | /* fN-5(n) = fN-4(n) - kN-4 * gN-5(n-1) */ |
emilmont | 1:fdd22bb7aa52 | 228 | fcurr = fnext - (((q31_t) gcurr * (*pk)) >> 15); |
emilmont | 1:fdd22bb7aa52 | 229 | fcurr = __SSAT(fcurr, 16); |
emilmont | 1:fdd22bb7aa52 | 230 | /* gN-4(n) = kN-4 * fN-5(n) + gN-5(n-1) */ |
emilmont | 1:fdd22bb7aa52 | 231 | gnext = (((q31_t) fcurr * (*pk++)) >> 15) + gcurr; |
emilmont | 1:fdd22bb7aa52 | 232 | gnext2 = (q15_t) __SSAT(gnext, 16); |
emilmont | 1:fdd22bb7aa52 | 233 | /* write gN-4(n) for the next sample process */ |
emilmont | 1:fdd22bb7aa52 | 234 | *px2++ = (q15_t) gnext2; |
emilmont | 1:fdd22bb7aa52 | 235 | |
emilmont | 1:fdd22bb7aa52 | 236 | /* Read vN-3 and vN-4 at a time */ |
emilmont | 1:fdd22bb7aa52 | 237 | #ifndef UNALIGNED_SUPPORT_DISABLE |
emilmont | 1:fdd22bb7aa52 | 238 | |
emilmont | 1:fdd22bb7aa52 | 239 | v = *__SIMD32(pv)++; |
emilmont | 1:fdd22bb7aa52 | 240 | |
emilmont | 1:fdd22bb7aa52 | 241 | #else |
emilmont | 1:fdd22bb7aa52 | 242 | |
emilmont | 2:da51fb522205 | 243 | v1 = *pv++; |
emilmont | 2:da51fb522205 | 244 | v2 = *pv++; |
emilmont | 1:fdd22bb7aa52 | 245 | |
emilmont | 1:fdd22bb7aa52 | 246 | #ifndef ARM_MATH_BIG_ENDIAN |
emilmont | 1:fdd22bb7aa52 | 247 | |
emilmont | 2:da51fb522205 | 248 | v = __PKHBT(v1, v2, 16); |
emilmont | 1:fdd22bb7aa52 | 249 | |
emilmont | 1:fdd22bb7aa52 | 250 | #else |
emilmont | 1:fdd22bb7aa52 | 251 | |
emilmont | 2:da51fb522205 | 252 | v = __PKHBT(v2, v1, 16); |
emilmont | 1:fdd22bb7aa52 | 253 | |
emilmont | 2:da51fb522205 | 254 | #endif /* #ifndef ARM_MATH_BIG_ENDIAN */ |
emilmont | 1:fdd22bb7aa52 | 255 | |
emilmont | 2:da51fb522205 | 256 | #endif /* #ifndef UNALIGNED_SUPPORT_DISABLE */ |
emilmont | 1:fdd22bb7aa52 | 257 | |
emilmont | 1:fdd22bb7aa52 | 258 | |
emilmont | 1:fdd22bb7aa52 | 259 | /* Pack gN-3(n) and gN-4(n) */ |
emilmont | 1:fdd22bb7aa52 | 260 | #ifndef ARM_MATH_BIG_ENDIAN |
emilmont | 1:fdd22bb7aa52 | 261 | |
emilmont | 1:fdd22bb7aa52 | 262 | gnext = __PKHBT(gnext1, gnext2, 16); |
emilmont | 1:fdd22bb7aa52 | 263 | |
emilmont | 1:fdd22bb7aa52 | 264 | #else |
emilmont | 1:fdd22bb7aa52 | 265 | |
emilmont | 1:fdd22bb7aa52 | 266 | gnext = __PKHBT(gnext2, gnext1, 16); |
emilmont | 1:fdd22bb7aa52 | 267 | |
emilmont | 1:fdd22bb7aa52 | 268 | #endif /* #ifndef ARM_MATH_BIG_ENDIAN */ |
emilmont | 1:fdd22bb7aa52 | 269 | |
emilmont | 1:fdd22bb7aa52 | 270 | /* y(n) += gN-4(n) * vN-4 */ |
emilmont | 1:fdd22bb7aa52 | 271 | /* process for gN-8(n) * vN-8, gN-12(n) * vN-12 ... */ |
emilmont | 1:fdd22bb7aa52 | 272 | /* y(n) += gN-3(n) * vN-3 */ |
emilmont | 1:fdd22bb7aa52 | 273 | /* process for gN-7(n) * vN-7, gN-11(n) * vN-11 ... */ |
emilmont | 1:fdd22bb7aa52 | 274 | acc = __SMLALD(gnext, v, acc); |
emilmont | 1:fdd22bb7aa52 | 275 | |
emilmont | 1:fdd22bb7aa52 | 276 | tapCnt--; |
emilmont | 1:fdd22bb7aa52 | 277 | |
emilmont | 1:fdd22bb7aa52 | 278 | } |
emilmont | 1:fdd22bb7aa52 | 279 | |
emilmont | 1:fdd22bb7aa52 | 280 | fnext = fcurr; |
emilmont | 1:fdd22bb7aa52 | 281 | |
emilmont | 1:fdd22bb7aa52 | 282 | /* If the filter length is not a multiple of 4, compute the remaining filter taps */ |
emilmont | 1:fdd22bb7aa52 | 283 | tapCnt = (numStages - 1u) % 0x4u; |
emilmont | 1:fdd22bb7aa52 | 284 | |
emilmont | 1:fdd22bb7aa52 | 285 | while(tapCnt > 0u) |
emilmont | 1:fdd22bb7aa52 | 286 | { |
emilmont | 1:fdd22bb7aa52 | 287 | gcurr = *px1++; |
emilmont | 1:fdd22bb7aa52 | 288 | /* Process sample for last taps */ |
emilmont | 1:fdd22bb7aa52 | 289 | fnext = fcurr - (((q31_t) gcurr * (*pk)) >> 15); |
emilmont | 1:fdd22bb7aa52 | 290 | fnext = __SSAT(fnext, 16); |
emilmont | 1:fdd22bb7aa52 | 291 | gnext = (((q31_t) fnext * (*pk++)) >> 15) + gcurr; |
emilmont | 1:fdd22bb7aa52 | 292 | gnext = __SSAT(gnext, 16); |
emilmont | 1:fdd22bb7aa52 | 293 | /* Output samples for last taps */ |
emilmont | 1:fdd22bb7aa52 | 294 | acc += (q31_t) (((q31_t) gnext * (*pv++))); |
emilmont | 1:fdd22bb7aa52 | 295 | *px2++ = (q15_t) gnext; |
emilmont | 1:fdd22bb7aa52 | 296 | fcurr = fnext; |
emilmont | 1:fdd22bb7aa52 | 297 | |
emilmont | 1:fdd22bb7aa52 | 298 | tapCnt--; |
emilmont | 1:fdd22bb7aa52 | 299 | } |
emilmont | 1:fdd22bb7aa52 | 300 | |
emilmont | 1:fdd22bb7aa52 | 301 | /* y(n) += g0(n) * v0 */ |
emilmont | 1:fdd22bb7aa52 | 302 | acc += (q31_t) (((q31_t) fnext * (*pv++))); |
emilmont | 1:fdd22bb7aa52 | 303 | |
emilmont | 1:fdd22bb7aa52 | 304 | out = (q15_t) __SSAT(acc >> 15, 16); |
emilmont | 1:fdd22bb7aa52 | 305 | *px2++ = (q15_t) fnext; |
emilmont | 1:fdd22bb7aa52 | 306 | |
emilmont | 1:fdd22bb7aa52 | 307 | /* write out into pDst */ |
emilmont | 1:fdd22bb7aa52 | 308 | *pDst++ = out; |
emilmont | 1:fdd22bb7aa52 | 309 | |
emilmont | 1:fdd22bb7aa52 | 310 | /* Advance the state pointer by 4 to process the next group of 4 samples */ |
emilmont | 1:fdd22bb7aa52 | 311 | pState = pState + 1u; |
emilmont | 1:fdd22bb7aa52 | 312 | blkCnt--; |
emilmont | 1:fdd22bb7aa52 | 313 | |
emilmont | 1:fdd22bb7aa52 | 314 | } |
emilmont | 1:fdd22bb7aa52 | 315 | |
emilmont | 1:fdd22bb7aa52 | 316 | /* Processing is complete. Now copy last S->numStages samples to start of the buffer |
emilmont | 1:fdd22bb7aa52 | 317 | for the preperation of next frame process */ |
emilmont | 1:fdd22bb7aa52 | 318 | /* Points to the start of the state buffer */ |
emilmont | 1:fdd22bb7aa52 | 319 | pStateCurnt = &S->pState[0]; |
emilmont | 1:fdd22bb7aa52 | 320 | pState = &S->pState[blockSize]; |
emilmont | 1:fdd22bb7aa52 | 321 | |
emilmont | 1:fdd22bb7aa52 | 322 | stgCnt = (numStages >> 2u); |
emilmont | 1:fdd22bb7aa52 | 323 | |
emilmont | 1:fdd22bb7aa52 | 324 | /* copy data */ |
emilmont | 1:fdd22bb7aa52 | 325 | while(stgCnt > 0u) |
emilmont | 1:fdd22bb7aa52 | 326 | { |
emilmont | 1:fdd22bb7aa52 | 327 | #ifndef UNALIGNED_SUPPORT_DISABLE |
emilmont | 1:fdd22bb7aa52 | 328 | |
emilmont | 1:fdd22bb7aa52 | 329 | *__SIMD32(pStateCurnt)++ = *__SIMD32(pState)++; |
emilmont | 1:fdd22bb7aa52 | 330 | *__SIMD32(pStateCurnt)++ = *__SIMD32(pState)++; |
emilmont | 1:fdd22bb7aa52 | 331 | |
emilmont | 1:fdd22bb7aa52 | 332 | #else |
emilmont | 1:fdd22bb7aa52 | 333 | |
emilmont | 1:fdd22bb7aa52 | 334 | *pStateCurnt++ = *pState++; |
emilmont | 1:fdd22bb7aa52 | 335 | *pStateCurnt++ = *pState++; |
emilmont | 1:fdd22bb7aa52 | 336 | *pStateCurnt++ = *pState++; |
emilmont | 1:fdd22bb7aa52 | 337 | *pStateCurnt++ = *pState++; |
emilmont | 1:fdd22bb7aa52 | 338 | |
emilmont | 2:da51fb522205 | 339 | #endif /* #ifndef UNALIGNED_SUPPORT_DISABLE */ |
emilmont | 1:fdd22bb7aa52 | 340 | |
emilmont | 1:fdd22bb7aa52 | 341 | /* Decrement the loop counter */ |
emilmont | 1:fdd22bb7aa52 | 342 | stgCnt--; |
emilmont | 1:fdd22bb7aa52 | 343 | |
emilmont | 1:fdd22bb7aa52 | 344 | } |
emilmont | 1:fdd22bb7aa52 | 345 | |
emilmont | 1:fdd22bb7aa52 | 346 | /* Calculation of count for remaining q15_t data */ |
emilmont | 1:fdd22bb7aa52 | 347 | stgCnt = (numStages) % 0x4u; |
emilmont | 1:fdd22bb7aa52 | 348 | |
emilmont | 1:fdd22bb7aa52 | 349 | /* copy data */ |
emilmont | 1:fdd22bb7aa52 | 350 | while(stgCnt > 0u) |
emilmont | 1:fdd22bb7aa52 | 351 | { |
emilmont | 1:fdd22bb7aa52 | 352 | *pStateCurnt++ = *pState++; |
emilmont | 1:fdd22bb7aa52 | 353 | |
emilmont | 1:fdd22bb7aa52 | 354 | /* Decrement the loop counter */ |
emilmont | 1:fdd22bb7aa52 | 355 | stgCnt--; |
emilmont | 1:fdd22bb7aa52 | 356 | } |
emilmont | 1:fdd22bb7aa52 | 357 | |
emilmont | 1:fdd22bb7aa52 | 358 | #else |
emilmont | 1:fdd22bb7aa52 | 359 | |
emilmont | 1:fdd22bb7aa52 | 360 | /* Run the below code for Cortex-M0 */ |
emilmont | 1:fdd22bb7aa52 | 361 | |
emilmont | 1:fdd22bb7aa52 | 362 | q31_t fcurr, fnext = 0, gcurr = 0, gnext; /* Temporary variables for lattice stages */ |
emilmont | 1:fdd22bb7aa52 | 363 | uint32_t stgCnt; /* Temporary variables for counts */ |
emilmont | 1:fdd22bb7aa52 | 364 | q63_t acc; /* Accumlator */ |
emilmont | 1:fdd22bb7aa52 | 365 | uint32_t blkCnt, tapCnt; /* Temporary variables for counts */ |
emilmont | 1:fdd22bb7aa52 | 366 | q15_t *px1, *px2, *pk, *pv; /* temporary pointers for state and coef */ |
emilmont | 1:fdd22bb7aa52 | 367 | uint32_t numStages = S->numStages; /* number of stages */ |
emilmont | 1:fdd22bb7aa52 | 368 | q15_t *pState; /* State pointer */ |
emilmont | 1:fdd22bb7aa52 | 369 | q15_t *pStateCurnt; /* State current pointer */ |
emilmont | 1:fdd22bb7aa52 | 370 | q15_t out; /* Temporary variable for output */ |
emilmont | 1:fdd22bb7aa52 | 371 | |
emilmont | 1:fdd22bb7aa52 | 372 | |
emilmont | 1:fdd22bb7aa52 | 373 | blkCnt = blockSize; |
emilmont | 1:fdd22bb7aa52 | 374 | |
emilmont | 1:fdd22bb7aa52 | 375 | pState = &S->pState[0]; |
emilmont | 1:fdd22bb7aa52 | 376 | |
emilmont | 1:fdd22bb7aa52 | 377 | /* Sample processing */ |
emilmont | 1:fdd22bb7aa52 | 378 | while(blkCnt > 0u) |
emilmont | 1:fdd22bb7aa52 | 379 | { |
emilmont | 1:fdd22bb7aa52 | 380 | /* Read Sample from input buffer */ |
emilmont | 1:fdd22bb7aa52 | 381 | /* fN(n) = x(n) */ |
emilmont | 1:fdd22bb7aa52 | 382 | fcurr = *pSrc++; |
emilmont | 1:fdd22bb7aa52 | 383 | |
emilmont | 1:fdd22bb7aa52 | 384 | /* Initialize state read pointer */ |
emilmont | 1:fdd22bb7aa52 | 385 | px1 = pState; |
emilmont | 1:fdd22bb7aa52 | 386 | /* Initialize state write pointer */ |
emilmont | 1:fdd22bb7aa52 | 387 | px2 = pState; |
emilmont | 1:fdd22bb7aa52 | 388 | /* Set accumulator to zero */ |
emilmont | 1:fdd22bb7aa52 | 389 | acc = 0; |
emilmont | 1:fdd22bb7aa52 | 390 | /* Initialize Ladder coeff pointer */ |
emilmont | 1:fdd22bb7aa52 | 391 | pv = &S->pvCoeffs[0]; |
emilmont | 1:fdd22bb7aa52 | 392 | /* Initialize Reflection coeff pointer */ |
emilmont | 1:fdd22bb7aa52 | 393 | pk = &S->pkCoeffs[0]; |
emilmont | 1:fdd22bb7aa52 | 394 | |
emilmont | 1:fdd22bb7aa52 | 395 | tapCnt = numStages; |
emilmont | 1:fdd22bb7aa52 | 396 | |
emilmont | 1:fdd22bb7aa52 | 397 | while(tapCnt > 0u) |
emilmont | 1:fdd22bb7aa52 | 398 | { |
emilmont | 1:fdd22bb7aa52 | 399 | gcurr = *px1++; |
emilmont | 1:fdd22bb7aa52 | 400 | /* Process sample */ |
emilmont | 1:fdd22bb7aa52 | 401 | /* fN-1(n) = fN(n) - kN * gN-1(n-1) */ |
emilmont | 1:fdd22bb7aa52 | 402 | fnext = fcurr - ((gcurr * (*pk)) >> 15); |
emilmont | 1:fdd22bb7aa52 | 403 | fnext = __SSAT(fnext, 16); |
emilmont | 1:fdd22bb7aa52 | 404 | /* gN(n) = kN * fN-1(n) + gN-1(n-1) */ |
emilmont | 1:fdd22bb7aa52 | 405 | gnext = ((fnext * (*pk++)) >> 15) + gcurr; |
emilmont | 1:fdd22bb7aa52 | 406 | gnext = __SSAT(gnext, 16); |
emilmont | 1:fdd22bb7aa52 | 407 | /* Output samples */ |
emilmont | 1:fdd22bb7aa52 | 408 | /* y(n) += gN(n) * vN */ |
emilmont | 1:fdd22bb7aa52 | 409 | acc += (q31_t) ((gnext * (*pv++))); |
emilmont | 1:fdd22bb7aa52 | 410 | /* write gN(n) into state for next sample processing */ |
emilmont | 1:fdd22bb7aa52 | 411 | *px2++ = (q15_t) gnext; |
emilmont | 1:fdd22bb7aa52 | 412 | /* Update f values for next coefficient processing */ |
emilmont | 1:fdd22bb7aa52 | 413 | fcurr = fnext; |
emilmont | 1:fdd22bb7aa52 | 414 | |
emilmont | 1:fdd22bb7aa52 | 415 | tapCnt--; |
emilmont | 1:fdd22bb7aa52 | 416 | } |
emilmont | 1:fdd22bb7aa52 | 417 | |
emilmont | 1:fdd22bb7aa52 | 418 | /* y(n) += g0(n) * v0 */ |
emilmont | 1:fdd22bb7aa52 | 419 | acc += (q31_t) ((fnext * (*pv++))); |
emilmont | 1:fdd22bb7aa52 | 420 | |
emilmont | 1:fdd22bb7aa52 | 421 | out = (q15_t) __SSAT(acc >> 15, 16); |
emilmont | 1:fdd22bb7aa52 | 422 | *px2++ = (q15_t) fnext; |
emilmont | 1:fdd22bb7aa52 | 423 | |
emilmont | 1:fdd22bb7aa52 | 424 | /* write out into pDst */ |
emilmont | 1:fdd22bb7aa52 | 425 | *pDst++ = out; |
emilmont | 1:fdd22bb7aa52 | 426 | |
emilmont | 1:fdd22bb7aa52 | 427 | /* Advance the state pointer by 1 to process the next group of samples */ |
emilmont | 1:fdd22bb7aa52 | 428 | pState = pState + 1u; |
emilmont | 1:fdd22bb7aa52 | 429 | blkCnt--; |
emilmont | 1:fdd22bb7aa52 | 430 | |
emilmont | 1:fdd22bb7aa52 | 431 | } |
emilmont | 1:fdd22bb7aa52 | 432 | |
emilmont | 1:fdd22bb7aa52 | 433 | /* Processing is complete. Now copy last S->numStages samples to start of the buffer |
emilmont | 1:fdd22bb7aa52 | 434 | for the preperation of next frame process */ |
emilmont | 1:fdd22bb7aa52 | 435 | /* Points to the start of the state buffer */ |
emilmont | 1:fdd22bb7aa52 | 436 | pStateCurnt = &S->pState[0]; |
emilmont | 1:fdd22bb7aa52 | 437 | pState = &S->pState[blockSize]; |
emilmont | 1:fdd22bb7aa52 | 438 | |
emilmont | 1:fdd22bb7aa52 | 439 | stgCnt = numStages; |
emilmont | 1:fdd22bb7aa52 | 440 | |
emilmont | 1:fdd22bb7aa52 | 441 | /* copy data */ |
emilmont | 1:fdd22bb7aa52 | 442 | while(stgCnt > 0u) |
emilmont | 1:fdd22bb7aa52 | 443 | { |
emilmont | 1:fdd22bb7aa52 | 444 | *pStateCurnt++ = *pState++; |
emilmont | 1:fdd22bb7aa52 | 445 | |
emilmont | 1:fdd22bb7aa52 | 446 | /* Decrement the loop counter */ |
emilmont | 1:fdd22bb7aa52 | 447 | stgCnt--; |
emilmont | 1:fdd22bb7aa52 | 448 | } |
emilmont | 1:fdd22bb7aa52 | 449 | |
emilmont | 1:fdd22bb7aa52 | 450 | #endif /* #ifndef ARM_MATH_CM0 */ |
emilmont | 1:fdd22bb7aa52 | 451 | |
emilmont | 1:fdd22bb7aa52 | 452 | } |
emilmont | 1:fdd22bb7aa52 | 453 | |
emilmont | 1:fdd22bb7aa52 | 454 | |
emilmont | 1:fdd22bb7aa52 | 455 | |
emilmont | 1:fdd22bb7aa52 | 456 | |
emilmont | 1:fdd22bb7aa52 | 457 | /** |
emilmont | 1:fdd22bb7aa52 | 458 | * @} end of IIR_Lattice group |
emilmont | 1:fdd22bb7aa52 | 459 | */ |