avior:    
 * Care must be taken when using the Q15 and Q31 versions of the LMS filter.    
 * The following issues must be considered:    
 * - Scaling of coefficients    
 * - Overflow and saturation    
 *    
 * \par Scaling of Coefficients:    
 * Filter coefficients are represented as fractional values and    
 * coefficients are restricted to lie in the range <code>[-1 +1)</code>.    
 * The fixed-point functions have an additional scaling parameter <code>postShift</code>.    
 * At the output of the filter's accumulator is a shift register which shifts the result by <code>postShift</code> bits.    
 * This essentially scales the filter coefficients by <code>2^postShift</code> and    
 * allows the filter coefficients to exceed the range <code>[+1 -1)</code>.    
 * The value of <code>postShift</code> is set by the user based on the expected gain through the system being modeled.    
 *    
 * \par Overflow and Saturation:    
 * Overflow and saturation behavior of the fixed-point Q15 and Q31 versions are    
 * described separately as part of the function specific documentation below.    
 */

/**    
 * @addtogroup LMS    
 * @{    
 */

/**           
 * @details           
 * This function operates on floating-point data types.       
 *    
 * @brief Processing function for floating-point LMS filter.    
 * @param[in]  *S points to an instance of the floating-point LMS filter structure.    
 * @param[in]  *pSrc points to the block of input data.    
 * @param[in]  *pRef points to the block of reference data.    
 * @param[out] *pOut points to the block of output data.    
 * @param[out] *pErr points to the block of error data.    
 * @param[in]  blockSize number of samples to process.    
 * @return     none.    
 */

void arm_lms_f32(
  const arm_lms_instance_f32 * S,
  float32_t * pSrc,
  float32_t * pRef,
  float32_t * pOut,
  float32_t * pErr,
  uint32_t blockSize)
{
  float32_t *pState = S->pState;                 /* State pointer */
  float32_t *pCoeffs = S->pCoeffs;               /* Coefficient pointer */
  float32_t *pStateCurnt;                        /* Points to the current sample of the state */
  float32_t *px, *pb;                            /* Temporary pointers for state and coefficient buffers */
  float32_t mu = S->mu;                          /* Adaptive factor */
  uint32_t numTaps = S->numTaps;                 /* Number of filter coefficients in the filter */
  uint32_t tapCnt, blkCnt;                       /* Loop counters */
  float32_t sum, e, d;                           /* accumulator, error, reference data sample */
  float32_t w = 0.0f;                            /* weight factor */

  e = 0.0f;
  d = 0.0f;

  /* S->pState points to state array which contains previous frame (numTaps - 1) samples */
  /* pStateCurnt points to the location where the new input data should be written */
  pStateCurnt = &(S->pState[(numTaps - 1u)]);

  blkCnt = blockSize;


#ifndef ARM_MATH_CM0_FAMILY

  /* Run the below code for Cortex-M4 and Cortex-M3 */

  while(blkCnt > 0u)
  {
    /* Copy the new input sample into the state buffer */
    *pStateCurnt++ = *pSrc++;

    /* Initialize pState pointer */
    px = pState;

    /* Initialize coeff pointer */
    pb = (pCoeffs);

    /* Set the accumulator to zero */
    sum = 0.0f;

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

    while(tapCnt > 0u)
    {
      /* Perform the multiply-accumulate */
      sum += (*px++) * (*pb++);
      sum += (*px++) * (*pb++);
      sum += (*px++) * (*pb++);
      sum += (*px++) * (*pb++);

      /* 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)
    {
      /* Perform the multiply-accumulate */
      sum += (*px++) * (*pb++);

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

    /* The result in the accumulator, store in the destination buffer. */
    *pOut++ = sum;

    /* Compute and store error */
    d = (float32_t) (*pRef++);
    e = d - sum;
    *pErr++ = e;

    /* Calculation of Weighting factor for the updating filter coefficients */
    w = e * mu;

    /* Initialize pState pointer */
    px = pState;

    /* Initialize coeff pointer */
    pb = (pCoeffs);

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

    /* Update filter coefficients */
    while(tapCnt > 0u)
    {
      /* Perform the multiply-accumulate */
      *pb = *pb + (w * (*px++));
      pb++;

      *pb = *pb + (w * (*px++));
      pb++;

      *pb = *pb + (w * (*px++));
      pb++;

      *pb = *pb + (w * (*px++));
      pb++;

      /* 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)
    {
      /* Perform the multiply-accumulate */
      *pb = *pb + (w * (*px++));
      pb++;

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

    /* Advance state pointer by 1 for the next sample */
    pState = pState + 1;

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


  /* 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 pState buffer */
  pStateCurnt = S->pState;

  /* Loop unrolling for (numTaps - 1u) samples copy */
  tapCnt = (numTaps - 1u) >> 2u;

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

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

  /* Calculate remaining number of copies */
  tapCnt = (numTaps - 1u) % 0x4u;

  /* Copy the remaining q31_t data */
  while(tapCnt > 0u)
  {
    *pStateCurnt++ = *pState++;

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

#else

  /* Run the below code for Cortex-M0 */

  while(blkCnt > 0u)
  {
    /* Copy the new input sample into the state buffer */
    *pStateCurnt++ = *pSrc++;

    /* Initialize pState pointer */
    px = pState;

    /* Initialize pCoeffs pointer */
    pb = pCoeffs;

    /* Set the accumulator to zero */
    sum = 0.0f;

    /* Loop over numTaps number of values */
    tapCnt = numTaps;

    while(tapCnt > 0u)
    {
      /* Perform the multiply-accumulate */
      sum += (*px++) * (*pb++);

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

    /* The result is stored in the destination buffer. */
    *pOut++ = sum;

    /* Compute and store error */
    d = (float32_t) (*pRef++);
    e = d - sum;
    *pErr++ = e;

    /* Weighting factor for the LMS version */
    w = e * mu;

    /* Initialize pState pointer */
    px = pState;

    /* Initialize pCoeffs pointer */
    pb = pCoeffs;

    /* Loop over numTaps number of values */
    tapCnt = numTaps;

    while(tapCnt > 0u)
    {
      /* Perform the multiply-accumulate */
      *pb = *pb + (w * (*px++));
      pb++;

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

    /* Advance state pointer by 1 for the next sample */
    pState = pState + 1;

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


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

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

  /*  Copy (numTaps - 1u) samples  */
  tapCnt = (numTaps - 1u);

  /* Copy the data */
  while(tapCnt > 0u)
  {
    *pStateCurnt++ = *pState++;

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

#endif /*   #ifndef ARM_MATH_CM0_FAMILY */

}

/**    
   * @} end of LMS group    
   */