arm_fir_decimate_fast_q31.c 11.2 KB
Newer Older
Sebastian Renner committed
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351
/* ----------------------------------------------------------------------    
* Copyright (C) 2010-2014 ARM Limited. All rights reserved.    
*    
* $Date:        19. March 2015
* $Revision: 	V.1.4.5
*    
* Project: 	    CMSIS DSP Library    
* Title:	    arm_fir_decimate_fast_q31.c    
*    
* Description:	Fast Q31 FIR Decimator.    
*    
* Target Processor: Cortex-M4/Cortex-M3
*  
* Redistribution and use in source and binary forms, with or without 
* modification, are permitted provided that the following conditions
* are met:
*   - Redistributions of source code must retain the above copyright
*     notice, this list of conditions and the following disclaimer.
*   - Redistributions in binary form must reproduce the above copyright
*     notice, this list of conditions and the following disclaimer in
*     the documentation and/or other materials provided with the 
*     distribution.
*   - Neither the name of ARM LIMITED nor the names of its contributors
*     may be used to endorse or promote products derived from this
*     software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
* FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
* COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
* BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
* CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
* ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
* POSSIBILITY OF SUCH DAMAGE. 
* -------------------------------------------------------------------- */

#include "arm_math.h"

/**    
 * @ingroup groupFilters    
 */

/**    
 * @addtogroup FIR_decimate    
 * @{    
 */

/**    
 * @brief Processing function for the Q31 FIR decimator (fast variant) for Cortex-M3 and Cortex-M4.    
 * @param[in] *S points to an instance of the Q31 FIR decimator structure.    
 * @param[in] *pSrc points to the block of input data.    
 * @param[out] *pDst points to the block of output data    
 * @param[in] blockSize number of input samples to process per call.    
 * @return none    
 *    
 * <b>Scaling and Overflow Behavior:</b>    
 *    
 * \par    
 * This function is optimized for speed at the expense of fixed-point precision and overflow protection.    
 * The result of each 1.31 x 1.31 multiplication is truncated to 2.30 format.    
 * These intermediate results are added to a 2.30 accumulator.    
 * Finally, the accumulator is saturated and converted to a 1.31 result.    
 * The fast version has the same overflow behavior as the standard version and provides less precision since it discards the low 32 bits of each multiplication result.    
 * In order to avoid overflows completely the input signal must be scaled down by log2(numTaps) bits (where log2 is read as log to the base 2).    
 *    
 * \par    
 * Refer to the function <code>arm_fir_decimate_q31()</code> for a slower implementation of this function which uses a 64-bit accumulator to provide higher precision.    
 * Both the slow and the fast versions use the same instance structure.    
 * Use the function <code>arm_fir_decimate_init_q31()</code> to initialize the filter structure.    
 */

void arm_fir_decimate_fast_q31(
  arm_fir_decimate_instance_q31 * S,
  q31_t * pSrc,
  q31_t * pDst,
  uint32_t blockSize)
{
  q31_t *pState = S->pState;                     /* State pointer */
  q31_t *pCoeffs = S->pCoeffs;                   /* Coefficient pointer */
  q31_t *pStateCurnt;                            /* Points to the current sample of the state */
  q31_t x0, c0;                                  /* Temporary variables to hold state and coefficient values */
  q31_t *px;                                     /* Temporary pointers for state buffer */
  q31_t *pb;                                     /* Temporary pointers for coefficient buffer */
  q31_t sum0;                                    /* Accumulator */
  uint32_t numTaps = S->numTaps;                 /* Number of taps */
  uint32_t i, tapCnt, blkCnt, outBlockSize = blockSize / S->M;  /* Loop counters */
  uint32_t blkCntN2;
  q31_t x1;
  q31_t acc0, acc1;
  q31_t *px0, *px1;

  /* S->pState buffer contains previous frame (numTaps - 1) samples */
  /* pStateCurnt points to the location where the new input data should be written */
  pStateCurnt = S->pState + (numTaps - 1u);

  /* Total number of output samples to be computed */

  blkCnt = outBlockSize / 2;
  blkCntN2 = outBlockSize - (2 * blkCnt);

  while(blkCnt > 0u)
  {
    /* Copy decimation factor number of new input samples into the state buffer */
    i = 2 * S->M;

    do
    {
      *pStateCurnt++ = *pSrc++;

    } while(--i);

    /* Set accumulator to zero */
    acc0 = 0;
    acc1 = 0;

    /* Initialize state pointer */
    px0 = pState;
    px1 = pState + S->M;

    /* Initialize coeff pointer */
    pb = pCoeffs;

    /* Loop unrolling.  Process 4 taps at a time. */
    tapCnt = numTaps >> 2;

    /* Loop over the number of taps.  Unroll by a factor of 4.       
     ** Repeat until we've computed numTaps-4 coefficients. */
    while(tapCnt > 0u)
    {
      /* Read the b[numTaps-1] coefficient */
      c0 = *(pb);

      /* Read x[n-numTaps-1] for sample 0 sample 1 */
      x0 = *(px0);
      x1 = *(px1);

      /* Perform the multiply-accumulate */
      acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x0 * c0)) >> 32);
      acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x1 * c0)) >> 32);

      /* Read the b[numTaps-2] coefficient */
      c0 = *(pb + 1u);

      /* Read x[n-numTaps-2]  for sample 0 sample 1  */
      x0 = *(px0 + 1u);
      x1 = *(px1 + 1u);

      /* Perform the multiply-accumulate */
      acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x0 * c0)) >> 32);
      acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x1 * c0)) >> 32);

      /* Read the b[numTaps-3] coefficient */
      c0 = *(pb + 2u);

      /* Read x[n-numTaps-3]  for sample 0 sample 1 */
      x0 = *(px0 + 2u);
      x1 = *(px1 + 2u);
      pb += 4u;

      /* Perform the multiply-accumulate */
      acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x0 * c0)) >> 32);
      acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x1 * c0)) >> 32);

      /* Read the b[numTaps-4] coefficient */
      c0 = *(pb - 1u);

      /* Read x[n-numTaps-4] for sample 0 sample 1 */
      x0 = *(px0 + 3u);
      x1 = *(px1 + 3u);


      /* Perform the multiply-accumulate */
      acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x0 * c0)) >> 32);
      acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x1 * c0)) >> 32);

      /* update state pointers */
      px0 += 4u;
      px1 += 4u;

      /* Decrement the loop counter */
      tapCnt--;
    }

    /* If the filter length is not a multiple of 4, compute the remaining filter taps */
    tapCnt = numTaps % 0x4u;

    while(tapCnt > 0u)
    {
      /* Read coefficients */
      c0 = *(pb++);

      /* Fetch 1 state variable */
      x0 = *(px0++);
      x1 = *(px1++);

      /* Perform the multiply-accumulate */
      acc0 = (q31_t) ((((q63_t) acc0 << 32) + ((q63_t) x0 * c0)) >> 32);
      acc1 = (q31_t) ((((q63_t) acc1 << 32) + ((q63_t) x1 * c0)) >> 32);

      /* Decrement the loop counter */
      tapCnt--;
    }

    /* Advance the state pointer by the decimation factor       
     * to process the next group of decimation factor number samples */
    pState = pState + S->M * 2;

    /* The result is in the accumulator, store in the destination buffer. */
    *pDst++ = (q31_t) (acc0 << 1);
    *pDst++ = (q31_t) (acc1 << 1);

    /* Decrement the loop counter */
    blkCnt--;
  }

  while(blkCntN2 > 0u)
  {
    /* Copy decimation factor number of new input samples into the state buffer */
    i = S->M;

    do
    {
      *pStateCurnt++ = *pSrc++;

    } while(--i);

    /* Set accumulator to zero */
    sum0 = 0;

    /* Initialize state pointer */
    px = pState;

    /* Initialize coeff pointer */
    pb = pCoeffs;

    /* Loop unrolling.  Process 4 taps at a time. */
    tapCnt = numTaps >> 2;

    /* Loop over the number of taps.  Unroll by a factor of 4.       
     ** Repeat until we've computed numTaps-4 coefficients. */
    while(tapCnt > 0u)
    {
      /* Read the b[numTaps-1] coefficient */
      c0 = *(pb++);

      /* Read x[n-numTaps-1] sample */
      x0 = *(px++);

      /* Perform the multiply-accumulate */
      sum0 = (q31_t) ((((q63_t) sum0 << 32) + ((q63_t) x0 * c0)) >> 32);

      /* Read the b[numTaps-2] coefficient */
      c0 = *(pb++);

      /* Read x[n-numTaps-2] sample */
      x0 = *(px++);

      /* Perform the multiply-accumulate */
      sum0 = (q31_t) ((((q63_t) sum0 << 32) + ((q63_t) x0 * c0)) >> 32);

      /* Read the b[numTaps-3] coefficient */
      c0 = *(pb++);

      /* Read x[n-numTaps-3] sample */
      x0 = *(px++);

      /* Perform the multiply-accumulate */
      sum0 = (q31_t) ((((q63_t) sum0 << 32) + ((q63_t) x0 * c0)) >> 32);

      /* Read the b[numTaps-4] coefficient */
      c0 = *(pb++);

      /* Read x[n-numTaps-4] sample */
      x0 = *(px++);

      /* Perform the multiply-accumulate */
      sum0 = (q31_t) ((((q63_t) sum0 << 32) + ((q63_t) x0 * c0)) >> 32);

      /* Decrement the loop counter */
      tapCnt--;
    }

    /* If the filter length is not a multiple of 4, compute the remaining filter taps */
    tapCnt = numTaps % 0x4u;

    while(tapCnt > 0u)
    {
      /* Read coefficients */
      c0 = *(pb++);

      /* Fetch 1 state variable */
      x0 = *(px++);

      /* Perform the multiply-accumulate */
      sum0 = (q31_t) ((((q63_t) sum0 << 32) + ((q63_t) x0 * c0)) >> 32);

      /* Decrement the loop counter */
      tapCnt--;
    }

    /* Advance the state pointer by the decimation factor       
     * to process the next group of decimation factor number samples */
    pState = pState + S->M;

    /* The result is in the accumulator, store in the destination buffer. */
    *pDst++ = (q31_t) (sum0 << 1);

    /* Decrement the loop counter */
    blkCntN2--;
  }

  /* Processing is complete.       
   ** Now copy the last numTaps - 1 samples to the satrt of the state buffer.       
   ** This prepares the state buffer for the next function call. */

  /* Points to the start of the state buffer */
  pStateCurnt = S->pState;

  i = (numTaps - 1u) >> 2u;

  /* copy data */
  while(i > 0u)
  {
    *pStateCurnt++ = *pState++;
    *pStateCurnt++ = *pState++;
    *pStateCurnt++ = *pState++;
    *pStateCurnt++ = *pState++;

    /* Decrement the loop counter */
    i--;
  }

  i = (numTaps - 1u) % 0x04u;

  /* copy data */
  while(i > 0u)
  {
    *pStateCurnt++ = *pState++;

    /* Decrement the loop counter */
    i--;
  }
}

/**    
 * @} end of FIR_decimate group    
 */