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_biquad_cascade_df1_q15.c@2:da51fb522205, 2013-05-30 (annotated)

Committer:
emilmont
Date:
Thu May 30 17:10:11 2013 +0100
Revision:
2:da51fb522205
Parent:
1:fdd22bb7aa52
Child:
3:7a284390b0ce
Keep "cmsis-dsp" module in synch with its source

Who changed what in which revision?

UserRevisionLine numberNew contents of line
emilmont 1:fdd22bb7aa52 1 /* ----------------------------------------------------------------------
emilmont 1:fdd22bb7aa52 2 * Copyright (C) 2010 ARM Limited. All rights reserved.
emilmont 1:fdd22bb7aa52 3 *
emilmont 1:fdd22bb7aa52 4 * $Date: 15. February 2012
emilmont 2:da51fb522205 5 * $Revision: V1.1.0
emilmont 1:fdd22bb7aa52 6 *
emilmont 2:da51fb522205 7 * Project: CMSIS DSP Library
emilmont 2:da51fb522205 8 * Title: arm_biquad_cascade_df1_q15.c
emilmont 1:fdd22bb7aa52 9 *
emilmont 2:da51fb522205 10 * Description: Processing function for the
emilmont 2:da51fb522205 11 * Q15 Biquad cascade DirectFormI(DF1) filter.
emilmont 1:fdd22bb7aa52 12 *
emilmont 1:fdd22bb7aa52 13 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
emilmont 1:fdd22bb7aa52 14 *
emilmont 1:fdd22bb7aa52 15 * Version 1.1.0 2012/02/15
emilmont 1:fdd22bb7aa52 16 * Updated with more optimizations, bug fixes and minor API changes.
emilmont 1:fdd22bb7aa52 17 *
emilmont 1:fdd22bb7aa52 18 * Version 1.0.10 2011/7/15
emilmont 1:fdd22bb7aa52 19 * Big Endian support added and Merged M0 and M3/M4 Source code.
emilmont 1:fdd22bb7aa52 20 *
emilmont 1:fdd22bb7aa52 21 * Version 1.0.3 2010/11/29
emilmont 1:fdd22bb7aa52 22 * Re-organized the CMSIS folders and updated documentation.
emilmont 1:fdd22bb7aa52 23 *
emilmont 1:fdd22bb7aa52 24 * Version 1.0.2 2010/11/11
emilmont 1:fdd22bb7aa52 25 * Documentation updated.
emilmont 1:fdd22bb7aa52 26 *
emilmont 1:fdd22bb7aa52 27 * Version 1.0.1 2010/10/05
emilmont 1:fdd22bb7aa52 28 * Production release and review comments incorporated.
emilmont 1:fdd22bb7aa52 29 *
emilmont 1:fdd22bb7aa52 30 * Version 1.0.0 2010/09/20
emilmont 1:fdd22bb7aa52 31 * Production release and review comments incorporated.
emilmont 1:fdd22bb7aa52 32 *
emilmont 1:fdd22bb7aa52 33 * Version 0.0.5 2010/04/26
emilmont 2:da51fb522205 34 * incorporated review comments and updated with latest CMSIS layer
emilmont 1:fdd22bb7aa52 35 *
emilmont 1:fdd22bb7aa52 36 * Version 0.0.3 2010/03/10
emilmont 1:fdd22bb7aa52 37 * Initial version
emilmont 1:fdd22bb7aa52 38 * -------------------------------------------------------------------- */
emilmont 1:fdd22bb7aa52 39
emilmont 1:fdd22bb7aa52 40 #include "arm_math.h"
emilmont 1:fdd22bb7aa52 41
emilmont 1:fdd22bb7aa52 42 /**
emilmont 1:fdd22bb7aa52 43 * @ingroup groupFilters
emilmont 1:fdd22bb7aa52 44 */
emilmont 1:fdd22bb7aa52 45
emilmont 1:fdd22bb7aa52 46 /**
emilmont 1:fdd22bb7aa52 47 * @addtogroup BiquadCascadeDF1
emilmont 1:fdd22bb7aa52 48 * @{
emilmont 1:fdd22bb7aa52 49 */
emilmont 1:fdd22bb7aa52 50
emilmont 1:fdd22bb7aa52 51 /**
emilmont 1:fdd22bb7aa52 52 * @brief Processing function for the Q15 Biquad cascade filter.
emilmont 1:fdd22bb7aa52 53 * @param[in] *S points to an instance of the Q15 Biquad cascade structure.
emilmont 1:fdd22bb7aa52 54 * @param[in] *pSrc points to the block of input data.
emilmont 1:fdd22bb7aa52 55 * @param[out] *pDst points to the location where the output result is written.
emilmont 1:fdd22bb7aa52 56 * @param[in] blockSize number of samples to process per call.
emilmont 1:fdd22bb7aa52 57 * @return none.
emilmont 1:fdd22bb7aa52 58 *
emilmont 1:fdd22bb7aa52 59 *
emilmont 1:fdd22bb7aa52 60 * <b>Scaling and Overflow Behavior:</b>
emilmont 1:fdd22bb7aa52 61 * \par
emilmont 1:fdd22bb7aa52 62 * The function is implemented using a 64-bit internal accumulator.
emilmont 1:fdd22bb7aa52 63 * Both coefficients and state variables are represented in 1.15 format and multiplications yield a 2.30 result.
emilmont 1:fdd22bb7aa52 64 * The 2.30 intermediate results are accumulated in a 64-bit accumulator in 34.30 format.
emilmont 1:fdd22bb7aa52 65 * There is no risk of internal overflow with this approach and the full precision of intermediate multiplications is preserved.
emilmont 1:fdd22bb7aa52 66 * The accumulator is then shifted by <code>postShift</code> bits to truncate the result to 1.15 format by discarding the low 16 bits.
emilmont 1:fdd22bb7aa52 67 * Finally, the result is saturated to 1.15 format.
emilmont 1:fdd22bb7aa52 68 *
emilmont 1:fdd22bb7aa52 69 * \par
emilmont 1:fdd22bb7aa52 70 * Refer to the function <code>arm_biquad_cascade_df1_fast_q15()</code> for a faster but less precise implementation of this filter for Cortex-M3 and Cortex-M4.
emilmont 1:fdd22bb7aa52 71 */
emilmont 1:fdd22bb7aa52 72
emilmont 1:fdd22bb7aa52 73 void arm_biquad_cascade_df1_q15(
emilmont 1:fdd22bb7aa52 74 const arm_biquad_casd_df1_inst_q15 * S,
emilmont 1:fdd22bb7aa52 75 q15_t * pSrc,
emilmont 1:fdd22bb7aa52 76 q15_t * pDst,
emilmont 1:fdd22bb7aa52 77 uint32_t blockSize)
emilmont 1:fdd22bb7aa52 78 {
emilmont 1:fdd22bb7aa52 79
emilmont 1:fdd22bb7aa52 80
emilmont 1:fdd22bb7aa52 81 #ifndef ARM_MATH_CM0
emilmont 1:fdd22bb7aa52 82
emilmont 1:fdd22bb7aa52 83 /* Run the below code for Cortex-M4 and Cortex-M3 */
emilmont 1:fdd22bb7aa52 84
emilmont 1:fdd22bb7aa52 85 q15_t *pIn = pSrc; /* Source pointer */
emilmont 1:fdd22bb7aa52 86 q15_t *pOut = pDst; /* Destination pointer */
emilmont 1:fdd22bb7aa52 87 q31_t in; /* Temporary variable to hold input value */
emilmont 1:fdd22bb7aa52 88 q31_t out; /* Temporary variable to hold output value */
emilmont 1:fdd22bb7aa52 89 q31_t b0; /* Temporary variable to hold bo value */
emilmont 1:fdd22bb7aa52 90 q31_t b1, a1; /* Filter coefficients */
emilmont 1:fdd22bb7aa52 91 q31_t state_in, state_out; /* Filter state variables */
emilmont 1:fdd22bb7aa52 92 q31_t acc_l, acc_h;
emilmont 1:fdd22bb7aa52 93 q63_t acc; /* Accumulator */
emilmont 1:fdd22bb7aa52 94 int32_t lShift = (15 - (int32_t) S->postShift); /* Post shift */
emilmont 1:fdd22bb7aa52 95 q15_t *pState = S->pState; /* State pointer */
emilmont 1:fdd22bb7aa52 96 q15_t *pCoeffs = S->pCoeffs; /* Coefficient pointer */
emilmont 1:fdd22bb7aa52 97 uint32_t sample, stage = (uint32_t) S->numStages; /* Stage loop counter */
emilmont 1:fdd22bb7aa52 98 int32_t uShift = (32 - lShift);
emilmont 1:fdd22bb7aa52 99
emilmont 1:fdd22bb7aa52 100 do
emilmont 1:fdd22bb7aa52 101 {
emilmont 1:fdd22bb7aa52 102 /* Read the b0 and 0 coefficients using SIMD */
emilmont 1:fdd22bb7aa52 103 b0 = *__SIMD32(pCoeffs)++;
emilmont 1:fdd22bb7aa52 104
emilmont 1:fdd22bb7aa52 105 /* Read the b1 and b2 coefficients using SIMD */
emilmont 1:fdd22bb7aa52 106 b1 = *__SIMD32(pCoeffs)++;
emilmont 1:fdd22bb7aa52 107
emilmont 1:fdd22bb7aa52 108 /* Read the a1 and a2 coefficients using SIMD */
emilmont 1:fdd22bb7aa52 109 a1 = *__SIMD32(pCoeffs)++;
emilmont 1:fdd22bb7aa52 110
emilmont 1:fdd22bb7aa52 111 /* Read the input state values from the state buffer: x[n-1], x[n-2] */
emilmont 1:fdd22bb7aa52 112 state_in = *__SIMD32(pState)++;
emilmont 1:fdd22bb7aa52 113
emilmont 1:fdd22bb7aa52 114 /* Read the output state values from the state buffer: y[n-1], y[n-2] */
emilmont 1:fdd22bb7aa52 115 state_out = *__SIMD32(pState)--;
emilmont 1:fdd22bb7aa52 116
emilmont 1:fdd22bb7aa52 117 /* Apply loop unrolling and compute 2 output values simultaneously. */
emilmont 1:fdd22bb7aa52 118 /* The variable acc hold output values that are being computed:
emilmont 1:fdd22bb7aa52 119 *
emilmont 1:fdd22bb7aa52 120 * acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2]
emilmont 1:fdd22bb7aa52 121 * acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2]
emilmont 1:fdd22bb7aa52 122 */
emilmont 1:fdd22bb7aa52 123 sample = blockSize >> 1u;
emilmont 1:fdd22bb7aa52 124
emilmont 1:fdd22bb7aa52 125 /* First part of the processing with loop unrolling. Compute 2 outputs at a time.
emilmont 1:fdd22bb7aa52 126 ** a second loop below computes the remaining 1 sample. */
emilmont 1:fdd22bb7aa52 127 while(sample > 0u)
emilmont 1:fdd22bb7aa52 128 {
emilmont 1:fdd22bb7aa52 129
emilmont 1:fdd22bb7aa52 130 /* Read the input */
emilmont 1:fdd22bb7aa52 131 in = *__SIMD32(pIn)++;
emilmont 1:fdd22bb7aa52 132
emilmont 1:fdd22bb7aa52 133 /* out = b0 * x[n] + 0 * 0 */
emilmont 1:fdd22bb7aa52 134 out = __SMUAD(b0, in);
emilmont 1:fdd22bb7aa52 135
emilmont 1:fdd22bb7aa52 136 /* acc += b1 * x[n-1] + b2 * x[n-2] + out */
emilmont 1:fdd22bb7aa52 137 acc = __SMLALD(b1, state_in, out);
emilmont 1:fdd22bb7aa52 138 /* acc += a1 * y[n-1] + a2 * y[n-2] */
emilmont 1:fdd22bb7aa52 139 acc = __SMLALD(a1, state_out, acc);
emilmont 1:fdd22bb7aa52 140
emilmont 1:fdd22bb7aa52 141 /* The result is converted from 3.29 to 1.31 if postShift = 1, and then saturation is applied */
emilmont 1:fdd22bb7aa52 142 /* Calc lower part of acc */
emilmont 1:fdd22bb7aa52 143 acc_l = acc & 0xffffffff;
emilmont 1:fdd22bb7aa52 144
emilmont 1:fdd22bb7aa52 145 /* Calc upper part of acc */
emilmont 1:fdd22bb7aa52 146 acc_h = (acc >> 32) & 0xffffffff;
emilmont 1:fdd22bb7aa52 147
emilmont 1:fdd22bb7aa52 148 /* Apply shift for lower part of acc and upper part of acc */
emilmont 1:fdd22bb7aa52 149 out = (uint32_t) acc_l >> lShift | acc_h << uShift;
emilmont 1:fdd22bb7aa52 150
emilmont 1:fdd22bb7aa52 151 out = __SSAT(out, 16);
emilmont 1:fdd22bb7aa52 152
emilmont 1:fdd22bb7aa52 153 /* Every time after the output is computed state should be updated. */
emilmont 1:fdd22bb7aa52 154 /* The states should be updated as: */
emilmont 1:fdd22bb7aa52 155 /* Xn2 = Xn1 */
emilmont 1:fdd22bb7aa52 156 /* Xn1 = Xn */
emilmont 1:fdd22bb7aa52 157 /* Yn2 = Yn1 */
emilmont 1:fdd22bb7aa52 158 /* Yn1 = acc */
emilmont 1:fdd22bb7aa52 159 /* x[n-N], x[n-N-1] are packed together to make state_in of type q31 */
emilmont 1:fdd22bb7aa52 160 /* y[n-N], y[n-N-1] are packed together to make state_out of type q31 */
emilmont 1:fdd22bb7aa52 161
emilmont 1:fdd22bb7aa52 162 #ifndef ARM_MATH_BIG_ENDIAN
emilmont 1:fdd22bb7aa52 163
emilmont 1:fdd22bb7aa52 164 state_in = __PKHBT(in, state_in, 16);
emilmont 1:fdd22bb7aa52 165 state_out = __PKHBT(out, state_out, 16);
emilmont 1:fdd22bb7aa52 166
emilmont 1:fdd22bb7aa52 167 #else
emilmont 1:fdd22bb7aa52 168
emilmont 1:fdd22bb7aa52 169 state_in = __PKHBT(state_in >> 16, (in >> 16), 16);
emilmont 1:fdd22bb7aa52 170 state_out = __PKHBT(state_out >> 16, (out), 16);
emilmont 1:fdd22bb7aa52 171
emilmont 1:fdd22bb7aa52 172 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
emilmont 1:fdd22bb7aa52 173
emilmont 1:fdd22bb7aa52 174 /* out = b0 * x[n] + 0 * 0 */
emilmont 1:fdd22bb7aa52 175 out = __SMUADX(b0, in);
emilmont 1:fdd22bb7aa52 176 /* acc += b1 * x[n-1] + b2 * x[n-2] + out */
emilmont 1:fdd22bb7aa52 177 acc = __SMLALD(b1, state_in, out);
emilmont 1:fdd22bb7aa52 178 /* acc += a1 * y[n-1] + a2 * y[n-2] */
emilmont 1:fdd22bb7aa52 179 acc = __SMLALD(a1, state_out, acc);
emilmont 1:fdd22bb7aa52 180
emilmont 1:fdd22bb7aa52 181 /* The result is converted from 3.29 to 1.31 if postShift = 1, and then saturation is applied */
emilmont 1:fdd22bb7aa52 182 /* Calc lower part of acc */
emilmont 1:fdd22bb7aa52 183 acc_l = acc & 0xffffffff;
emilmont 1:fdd22bb7aa52 184
emilmont 1:fdd22bb7aa52 185 /* Calc upper part of acc */
emilmont 1:fdd22bb7aa52 186 acc_h = (acc >> 32) & 0xffffffff;
emilmont 1:fdd22bb7aa52 187
emilmont 1:fdd22bb7aa52 188 /* Apply shift for lower part of acc and upper part of acc */
emilmont 1:fdd22bb7aa52 189 out = (uint32_t) acc_l >> lShift | acc_h << uShift;
emilmont 1:fdd22bb7aa52 190
emilmont 1:fdd22bb7aa52 191 out = __SSAT(out, 16);
emilmont 1:fdd22bb7aa52 192
emilmont 1:fdd22bb7aa52 193 /* Store the output in the destination buffer. */
emilmont 1:fdd22bb7aa52 194
emilmont 1:fdd22bb7aa52 195 #ifndef ARM_MATH_BIG_ENDIAN
emilmont 1:fdd22bb7aa52 196
emilmont 1:fdd22bb7aa52 197 *__SIMD32(pOut)++ = __PKHBT(state_out, out, 16);
emilmont 1:fdd22bb7aa52 198
emilmont 1:fdd22bb7aa52 199 #else
emilmont 1:fdd22bb7aa52 200
emilmont 1:fdd22bb7aa52 201 *__SIMD32(pOut)++ = __PKHBT(out, state_out >> 16, 16);
emilmont 1:fdd22bb7aa52 202
emilmont 1:fdd22bb7aa52 203 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
emilmont 1:fdd22bb7aa52 204
emilmont 1:fdd22bb7aa52 205 /* Every time after the output is computed state should be updated. */
emilmont 1:fdd22bb7aa52 206 /* The states should be updated as: */
emilmont 1:fdd22bb7aa52 207 /* Xn2 = Xn1 */
emilmont 1:fdd22bb7aa52 208 /* Xn1 = Xn */
emilmont 1:fdd22bb7aa52 209 /* Yn2 = Yn1 */
emilmont 1:fdd22bb7aa52 210 /* Yn1 = acc */
emilmont 1:fdd22bb7aa52 211 /* x[n-N], x[n-N-1] are packed together to make state_in of type q31 */
emilmont 1:fdd22bb7aa52 212 /* y[n-N], y[n-N-1] are packed together to make state_out of type q31 */
emilmont 1:fdd22bb7aa52 213 #ifndef ARM_MATH_BIG_ENDIAN
emilmont 1:fdd22bb7aa52 214
emilmont 1:fdd22bb7aa52 215 state_in = __PKHBT(in >> 16, state_in, 16);
emilmont 1:fdd22bb7aa52 216 state_out = __PKHBT(out, state_out, 16);
emilmont 1:fdd22bb7aa52 217
emilmont 1:fdd22bb7aa52 218 #else
emilmont 1:fdd22bb7aa52 219
emilmont 1:fdd22bb7aa52 220 state_in = __PKHBT(state_in >> 16, in, 16);
emilmont 1:fdd22bb7aa52 221 state_out = __PKHBT(state_out >> 16, out, 16);
emilmont 1:fdd22bb7aa52 222
emilmont 1:fdd22bb7aa52 223 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
emilmont 1:fdd22bb7aa52 224
emilmont 1:fdd22bb7aa52 225
emilmont 1:fdd22bb7aa52 226 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 227 sample--;
emilmont 1:fdd22bb7aa52 228
emilmont 1:fdd22bb7aa52 229 }
emilmont 1:fdd22bb7aa52 230
emilmont 1:fdd22bb7aa52 231 /* If the blockSize is not a multiple of 2, compute any remaining output samples here.
emilmont 1:fdd22bb7aa52 232 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 233
emilmont 1:fdd22bb7aa52 234 if((blockSize & 0x1u) != 0u)
emilmont 1:fdd22bb7aa52 235 {
emilmont 1:fdd22bb7aa52 236 /* Read the input */
emilmont 1:fdd22bb7aa52 237 in = *pIn++;
emilmont 1:fdd22bb7aa52 238
emilmont 1:fdd22bb7aa52 239 /* out = b0 * x[n] + 0 * 0 */
emilmont 1:fdd22bb7aa52 240
emilmont 1:fdd22bb7aa52 241 #ifndef ARM_MATH_BIG_ENDIAN
emilmont 1:fdd22bb7aa52 242
emilmont 1:fdd22bb7aa52 243 out = __SMUAD(b0, in);
emilmont 1:fdd22bb7aa52 244
emilmont 1:fdd22bb7aa52 245 #else
emilmont 1:fdd22bb7aa52 246
emilmont 1:fdd22bb7aa52 247 out = __SMUADX(b0, in);
emilmont 1:fdd22bb7aa52 248
emilmont 1:fdd22bb7aa52 249 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
emilmont 1:fdd22bb7aa52 250
emilmont 1:fdd22bb7aa52 251 /* acc = b1 * x[n-1] + b2 * x[n-2] + out */
emilmont 1:fdd22bb7aa52 252 acc = __SMLALD(b1, state_in, out);
emilmont 1:fdd22bb7aa52 253 /* acc += a1 * y[n-1] + a2 * y[n-2] */
emilmont 1:fdd22bb7aa52 254 acc = __SMLALD(a1, state_out, acc);
emilmont 1:fdd22bb7aa52 255
emilmont 1:fdd22bb7aa52 256 /* The result is converted from 3.29 to 1.31 if postShift = 1, and then saturation is applied */
emilmont 1:fdd22bb7aa52 257 /* Calc lower part of acc */
emilmont 1:fdd22bb7aa52 258 acc_l = acc & 0xffffffff;
emilmont 1:fdd22bb7aa52 259
emilmont 1:fdd22bb7aa52 260 /* Calc upper part of acc */
emilmont 1:fdd22bb7aa52 261 acc_h = (acc >> 32) & 0xffffffff;
emilmont 1:fdd22bb7aa52 262
emilmont 1:fdd22bb7aa52 263 /* Apply shift for lower part of acc and upper part of acc */
emilmont 1:fdd22bb7aa52 264 out = (uint32_t) acc_l >> lShift | acc_h << uShift;
emilmont 1:fdd22bb7aa52 265
emilmont 1:fdd22bb7aa52 266 out = __SSAT(out, 16);
emilmont 1:fdd22bb7aa52 267
emilmont 1:fdd22bb7aa52 268 /* Store the output in the destination buffer. */
emilmont 1:fdd22bb7aa52 269 *pOut++ = (q15_t) out;
emilmont 1:fdd22bb7aa52 270
emilmont 1:fdd22bb7aa52 271 /* Every time after the output is computed state should be updated. */
emilmont 1:fdd22bb7aa52 272 /* The states should be updated as: */
emilmont 1:fdd22bb7aa52 273 /* Xn2 = Xn1 */
emilmont 1:fdd22bb7aa52 274 /* Xn1 = Xn */
emilmont 1:fdd22bb7aa52 275 /* Yn2 = Yn1 */
emilmont 1:fdd22bb7aa52 276 /* Yn1 = acc */
emilmont 1:fdd22bb7aa52 277 /* x[n-N], x[n-N-1] are packed together to make state_in of type q31 */
emilmont 1:fdd22bb7aa52 278 /* y[n-N], y[n-N-1] are packed together to make state_out of type q31 */
emilmont 1:fdd22bb7aa52 279
emilmont 1:fdd22bb7aa52 280 #ifndef ARM_MATH_BIG_ENDIAN
emilmont 1:fdd22bb7aa52 281
emilmont 1:fdd22bb7aa52 282 state_in = __PKHBT(in, state_in, 16);
emilmont 1:fdd22bb7aa52 283 state_out = __PKHBT(out, state_out, 16);
emilmont 1:fdd22bb7aa52 284
emilmont 1:fdd22bb7aa52 285 #else
emilmont 1:fdd22bb7aa52 286
emilmont 1:fdd22bb7aa52 287 state_in = __PKHBT(state_in >> 16, in, 16);
emilmont 1:fdd22bb7aa52 288 state_out = __PKHBT(state_out >> 16, out, 16);
emilmont 1:fdd22bb7aa52 289
emilmont 1:fdd22bb7aa52 290 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
emilmont 1:fdd22bb7aa52 291
emilmont 1:fdd22bb7aa52 292 }
emilmont 1:fdd22bb7aa52 293
emilmont 1:fdd22bb7aa52 294 /* The first stage goes from the input wire to the output wire. */
emilmont 1:fdd22bb7aa52 295 /* Subsequent numStages occur in-place in the output wire */
emilmont 1:fdd22bb7aa52 296 pIn = pDst;
emilmont 1:fdd22bb7aa52 297
emilmont 1:fdd22bb7aa52 298 /* Reset the output pointer */
emilmont 1:fdd22bb7aa52 299 pOut = pDst;
emilmont 1:fdd22bb7aa52 300
emilmont 1:fdd22bb7aa52 301 /* Store the updated state variables back into the state array */
emilmont 1:fdd22bb7aa52 302 *__SIMD32(pState)++ = state_in;
emilmont 1:fdd22bb7aa52 303 *__SIMD32(pState)++ = state_out;
emilmont 1:fdd22bb7aa52 304
emilmont 1:fdd22bb7aa52 305
emilmont 1:fdd22bb7aa52 306 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 307 stage--;
emilmont 1:fdd22bb7aa52 308
emilmont 1:fdd22bb7aa52 309 } while(stage > 0u);
emilmont 1:fdd22bb7aa52 310
emilmont 1:fdd22bb7aa52 311 #else
emilmont 1:fdd22bb7aa52 312
emilmont 1:fdd22bb7aa52 313 /* Run the below code for Cortex-M0 */
emilmont 1:fdd22bb7aa52 314
emilmont 1:fdd22bb7aa52 315 q15_t *pIn = pSrc; /* Source pointer */
emilmont 1:fdd22bb7aa52 316 q15_t *pOut = pDst; /* Destination pointer */
emilmont 1:fdd22bb7aa52 317 q15_t b0, b1, b2, a1, a2; /* Filter coefficients */
emilmont 1:fdd22bb7aa52 318 q15_t Xn1, Xn2, Yn1, Yn2; /* Filter state variables */
emilmont 1:fdd22bb7aa52 319 q15_t Xn; /* temporary input */
emilmont 1:fdd22bb7aa52 320 q63_t acc; /* Accumulator */
emilmont 1:fdd22bb7aa52 321 int32_t shift = (15 - (int32_t) S->postShift); /* Post shift */
emilmont 1:fdd22bb7aa52 322 q15_t *pState = S->pState; /* State pointer */
emilmont 1:fdd22bb7aa52 323 q15_t *pCoeffs = S->pCoeffs; /* Coefficient pointer */
emilmont 1:fdd22bb7aa52 324 uint32_t sample, stage = (uint32_t) S->numStages; /* Stage loop counter */
emilmont 1:fdd22bb7aa52 325
emilmont 1:fdd22bb7aa52 326 do
emilmont 1:fdd22bb7aa52 327 {
emilmont 1:fdd22bb7aa52 328 /* Reading the coefficients */
emilmont 1:fdd22bb7aa52 329 b0 = *pCoeffs++;
emilmont 1:fdd22bb7aa52 330 b1 = *pCoeffs++;
emilmont 1:fdd22bb7aa52 331 b2 = *pCoeffs++;
emilmont 1:fdd22bb7aa52 332 a1 = *pCoeffs++;
emilmont 1:fdd22bb7aa52 333 a2 = *pCoeffs++;
emilmont 1:fdd22bb7aa52 334
emilmont 1:fdd22bb7aa52 335 /* Reading the state values */
emilmont 1:fdd22bb7aa52 336 Xn1 = pState[0];
emilmont 1:fdd22bb7aa52 337 Xn2 = pState[1];
emilmont 1:fdd22bb7aa52 338 Yn1 = pState[2];
emilmont 1:fdd22bb7aa52 339 Yn2 = pState[3];
emilmont 1:fdd22bb7aa52 340
emilmont 1:fdd22bb7aa52 341 /* The variables acc holds the output value that is computed:
emilmont 1:fdd22bb7aa52 342 * acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2]
emilmont 1:fdd22bb7aa52 343 */
emilmont 1:fdd22bb7aa52 344
emilmont 1:fdd22bb7aa52 345 sample = blockSize;
emilmont 1:fdd22bb7aa52 346
emilmont 1:fdd22bb7aa52 347 while(sample > 0u)
emilmont 1:fdd22bb7aa52 348 {
emilmont 1:fdd22bb7aa52 349 /* Read the input */
emilmont 1:fdd22bb7aa52 350 Xn = *pIn++;
emilmont 1:fdd22bb7aa52 351
emilmont 1:fdd22bb7aa52 352 /* acc = b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] */
emilmont 1:fdd22bb7aa52 353 /* acc = b0 * x[n] */
emilmont 1:fdd22bb7aa52 354 acc = (q31_t) b0 *Xn;
emilmont 1:fdd22bb7aa52 355
emilmont 1:fdd22bb7aa52 356 /* acc += b1 * x[n-1] */
emilmont 1:fdd22bb7aa52 357 acc += (q31_t) b1 *Xn1;
emilmont 1:fdd22bb7aa52 358 /* acc += b[2] * x[n-2] */
emilmont 1:fdd22bb7aa52 359 acc += (q31_t) b2 *Xn2;
emilmont 1:fdd22bb7aa52 360 /* acc += a1 * y[n-1] */
emilmont 1:fdd22bb7aa52 361 acc += (q31_t) a1 *Yn1;
emilmont 1:fdd22bb7aa52 362 /* acc += a2 * y[n-2] */
emilmont 1:fdd22bb7aa52 363 acc += (q31_t) a2 *Yn2;
emilmont 1:fdd22bb7aa52 364
emilmont 1:fdd22bb7aa52 365 /* The result is converted to 1.31 */
emilmont 1:fdd22bb7aa52 366 acc = __SSAT((acc >> shift), 16);
emilmont 1:fdd22bb7aa52 367
emilmont 1:fdd22bb7aa52 368 /* Every time after the output is computed state should be updated. */
emilmont 1:fdd22bb7aa52 369 /* The states should be updated as: */
emilmont 1:fdd22bb7aa52 370 /* Xn2 = Xn1 */
emilmont 1:fdd22bb7aa52 371 /* Xn1 = Xn */
emilmont 1:fdd22bb7aa52 372 /* Yn2 = Yn1 */
emilmont 1:fdd22bb7aa52 373 /* Yn1 = acc */
emilmont 1:fdd22bb7aa52 374 Xn2 = Xn1;
emilmont 1:fdd22bb7aa52 375 Xn1 = Xn;
emilmont 1:fdd22bb7aa52 376 Yn2 = Yn1;
emilmont 1:fdd22bb7aa52 377 Yn1 = (q15_t) acc;
emilmont 1:fdd22bb7aa52 378
emilmont 1:fdd22bb7aa52 379 /* Store the output in the destination buffer. */
emilmont 1:fdd22bb7aa52 380 *pOut++ = (q15_t) acc;
emilmont 1:fdd22bb7aa52 381
emilmont 1:fdd22bb7aa52 382 /* decrement the loop counter */
emilmont 1:fdd22bb7aa52 383 sample--;
emilmont 1:fdd22bb7aa52 384 }
emilmont 1:fdd22bb7aa52 385
emilmont 1:fdd22bb7aa52 386 /* The first stage goes from the input buffer to the output buffer. */
emilmont 1:fdd22bb7aa52 387 /* Subsequent stages occur in-place in the output buffer */
emilmont 1:fdd22bb7aa52 388 pIn = pDst;
emilmont 1:fdd22bb7aa52 389
emilmont 1:fdd22bb7aa52 390 /* Reset to destination pointer */
emilmont 1:fdd22bb7aa52 391 pOut = pDst;
emilmont 1:fdd22bb7aa52 392
emilmont 1:fdd22bb7aa52 393 /* Store the updated state variables back into the pState array */
emilmont 1:fdd22bb7aa52 394 *pState++ = Xn1;
emilmont 1:fdd22bb7aa52 395 *pState++ = Xn2;
emilmont 1:fdd22bb7aa52 396 *pState++ = Yn1;
emilmont 1:fdd22bb7aa52 397 *pState++ = Yn2;
emilmont 1:fdd22bb7aa52 398
emilmont 1:fdd22bb7aa52 399 } while(--stage);
emilmont 1:fdd22bb7aa52 400
emilmont 1:fdd22bb7aa52 401 #endif /* #ifndef ARM_MATH_CM0 */
emilmont 1:fdd22bb7aa52 402
emilmont 1:fdd22bb7aa52 403 }
emilmont 1:fdd22bb7aa52 404
emilmont 1:fdd22bb7aa52 405
emilmont 1:fdd22bb7aa52 406 /**
emilmont 1:fdd22bb7aa52 407 * @} end of BiquadCascadeDF1 group
emilmont 1:fdd22bb7aa52 408 */