pigweed / third_party / github / STMicroelectronics / cmsis_core / 7dd288b23bf605a3a2fafa81a29d2c96a2fd83ce / . / DSP_Lib / Source / FilteringFunctions / arm_fir_interpolate_q31.c

/*----------------------------------------------------------------------------- | |

* 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_interpolate_q31.c | |

* | |

* Description: Q31 FIR interpolation. | |

* | |

* Target Processor: Cortex-M4/Cortex-M3/Cortex-M0 | |

* | |

* 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_Interpolate | |

* @{ | |

*/ | |

/** | |

* @brief Processing function for the Q31 FIR interpolator. | |

* @param[in] *S points to an instance of the Q31 FIR interpolator 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 | |

* The function is implemented using an internal 64-bit accumulator. | |

* The accumulator has a 2.62 format and maintains full precision of the intermediate multiplication results but provides only a single guard bit. | |

* Thus, if the accumulator result overflows it wraps around rather than clip. | |

* In order to avoid overflows completely the input signal must be scaled down by <code>1/(numTaps/L)</code>. | |

* since <code>numTaps/L</code> additions occur per output sample. | |

* After all multiply-accumulates are performed, the 2.62 accumulator is truncated to 1.32 format and then saturated to 1.31 format. | |

*/ | |

#ifndef ARM_MATH_CM0_FAMILY | |

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

void arm_fir_interpolate_q31( | |

const arm_fir_interpolate_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 *ptr1, *ptr2; /* Temporary pointers for state and coefficient buffers */ | |

q63_t sum0; /* Accumulators */ | |

q31_t x0, c0; /* Temporary variables to hold state and coefficient values */ | |

uint32_t i, blkCnt, j; /* Loop counters */ | |

uint16_t phaseLen = S->phaseLength, tapCnt; /* Length of each polyphase filter component */ | |

uint32_t blkCntN2; | |

q63_t acc0, acc1; | |

q31_t x1; | |

/* S->pState buffer contains previous frame (phaseLen - 1) samples */ | |

/* pStateCurnt points to the location where the new input data should be written */ | |

pStateCurnt = S->pState + ((q31_t) phaseLen - 1); | |

/* Initialise blkCnt */ | |

blkCnt = blockSize / 2; | |

blkCntN2 = blockSize - (2 * blkCnt); | |

/* Samples loop unrolled by 2 */ | |

while(blkCnt > 0u) | |

{ | |

/* Copy new input sample into the state buffer */ | |

*pStateCurnt++ = *pSrc++; | |

*pStateCurnt++ = *pSrc++; | |

/* Address modifier index of coefficient buffer */ | |

j = 1u; | |

/* Loop over the Interpolation factor. */ | |

i = (S->L); | |

while(i > 0u) | |

{ | |

/* Set accumulator to zero */ | |

acc0 = 0; | |

acc1 = 0; | |

/* Initialize state pointer */ | |

ptr1 = pState; | |

/* Initialize coefficient pointer */ | |

ptr2 = pCoeffs + (S->L - j); | |

/* Loop over the polyPhase length. Unroll by a factor of 4. | |

** Repeat until we've computed numTaps-(4*S->L) coefficients. */ | |

tapCnt = phaseLen >> 2u; | |

x0 = *(ptr1++); | |

while(tapCnt > 0u) | |

{ | |

/* Read the input sample */ | |

x1 = *(ptr1++); | |

/* Read the coefficient */ | |

c0 = *(ptr2); | |

/* Perform the multiply-accumulate */ | |

acc0 += (q63_t) x0 *c0; | |

acc1 += (q63_t) x1 *c0; | |

/* Read the coefficient */ | |

c0 = *(ptr2 + S->L); | |

/* Read the input sample */ | |

x0 = *(ptr1++); | |

/* Perform the multiply-accumulate */ | |

acc0 += (q63_t) x1 *c0; | |

acc1 += (q63_t) x0 *c0; | |

/* Read the coefficient */ | |

c0 = *(ptr2 + S->L * 2); | |

/* Read the input sample */ | |

x1 = *(ptr1++); | |

/* Perform the multiply-accumulate */ | |

acc0 += (q63_t) x0 *c0; | |

acc1 += (q63_t) x1 *c0; | |

/* Read the coefficient */ | |

c0 = *(ptr2 + S->L * 3); | |

/* Read the input sample */ | |

x0 = *(ptr1++); | |

/* Perform the multiply-accumulate */ | |

acc0 += (q63_t) x1 *c0; | |

acc1 += (q63_t) x0 *c0; | |

/* Upsampling is done by stuffing L-1 zeros between each sample. | |

* So instead of multiplying zeros with coefficients, | |

* Increment the coefficient pointer by interpolation factor times. */ | |

ptr2 += 4 * S->L; | |

/* Decrement the loop counter */ | |

tapCnt--; | |

} | |

/* If the polyPhase length is not a multiple of 4, compute the remaining filter taps */ | |

tapCnt = phaseLen % 0x4u; | |

while(tapCnt > 0u) | |

{ | |

/* Read the input sample */ | |

x1 = *(ptr1++); | |

/* Read the coefficient */ | |

c0 = *(ptr2); | |

/* Perform the multiply-accumulate */ | |

acc0 += (q63_t) x0 *c0; | |

acc1 += (q63_t) x1 *c0; | |

/* Increment the coefficient pointer by interpolation factor times. */ | |

ptr2 += S->L; | |

/* update states for next sample processing */ | |

x0 = x1; | |

/* Decrement the loop counter */ | |

tapCnt--; | |

} | |

/* The result is in the accumulator, store in the destination buffer. */ | |

*pDst = (q31_t) (acc0 >> 31); | |

*(pDst + S->L) = (q31_t) (acc1 >> 31); | |

pDst++; | |

/* Increment the address modifier index of coefficient buffer */ | |

j++; | |

/* Decrement the loop counter */ | |

i--; | |

} | |

/* Advance the state pointer by 1 | |

* to process the next group of interpolation factor number samples */ | |

pState = pState + 2; | |

pDst += S->L; | |

/* Decrement the loop counter */ | |

blkCnt--; | |

} | |

/* If the blockSize is not a multiple of 2, compute any remaining output samples here. | |

** No loop unrolling is used. */ | |

blkCnt = blkCntN2; | |

/* Loop over the blockSize. */ | |

while(blkCnt > 0u) | |

{ | |

/* Copy new input sample into the state buffer */ | |

*pStateCurnt++ = *pSrc++; | |

/* Address modifier index of coefficient buffer */ | |

j = 1u; | |

/* Loop over the Interpolation factor. */ | |

i = S->L; | |

while(i > 0u) | |

{ | |

/* Set accumulator to zero */ | |

sum0 = 0; | |

/* Initialize state pointer */ | |

ptr1 = pState; | |

/* Initialize coefficient pointer */ | |

ptr2 = pCoeffs + (S->L - j); | |

/* Loop over the polyPhase length. Unroll by a factor of 4. | |

** Repeat until we've computed numTaps-(4*S->L) coefficients. */ | |

tapCnt = phaseLen >> 2; | |

while(tapCnt > 0u) | |

{ | |

/* Read the coefficient */ | |

c0 = *(ptr2); | |

/* Upsampling is done by stuffing L-1 zeros between each sample. | |

* So instead of multiplying zeros with coefficients, | |

* Increment the coefficient pointer by interpolation factor times. */ | |

ptr2 += S->L; | |

/* Read the input sample */ | |

x0 = *(ptr1++); | |

/* Perform the multiply-accumulate */ | |

sum0 += (q63_t) x0 *c0; | |

/* Read the coefficient */ | |

c0 = *(ptr2); | |

/* Increment the coefficient pointer by interpolation factor times. */ | |

ptr2 += S->L; | |

/* Read the input sample */ | |

x0 = *(ptr1++); | |

/* Perform the multiply-accumulate */ | |

sum0 += (q63_t) x0 *c0; | |

/* Read the coefficient */ | |

c0 = *(ptr2); | |

/* Increment the coefficient pointer by interpolation factor times. */ | |

ptr2 += S->L; | |

/* Read the input sample */ | |

x0 = *(ptr1++); | |

/* Perform the multiply-accumulate */ | |

sum0 += (q63_t) x0 *c0; | |

/* Read the coefficient */ | |

c0 = *(ptr2); | |

/* Increment the coefficient pointer by interpolation factor times. */ | |

ptr2 += S->L; | |

/* Read the input sample */ | |

x0 = *(ptr1++); | |

/* Perform the multiply-accumulate */ | |

sum0 += (q63_t) x0 *c0; | |

/* Decrement the loop counter */ | |

tapCnt--; | |

} | |

/* If the polyPhase length is not a multiple of 4, compute the remaining filter taps */ | |

tapCnt = phaseLen & 0x3u; | |

while(tapCnt > 0u) | |

{ | |

/* Read the coefficient */ | |

c0 = *(ptr2); | |

/* Increment the coefficient pointer by interpolation factor times. */ | |

ptr2 += S->L; | |

/* Read the input sample */ | |

x0 = *(ptr1++); | |

/* Perform the multiply-accumulate */ | |

sum0 += (q63_t) x0 *c0; | |

/* Decrement the loop counter */ | |

tapCnt--; | |

} | |

/* The result is in the accumulator, store in the destination buffer. */ | |

*pDst++ = (q31_t) (sum0 >> 31); | |

/* Increment the address modifier index of coefficient buffer */ | |

j++; | |

/* Decrement the loop counter */ | |

i--; | |

} | |

/* Advance the state pointer by 1 | |

* to process the next group of interpolation factor number samples */ | |

pState = pState + 1; | |

/* Decrement the loop counter */ | |

blkCnt--; | |

} | |

/* Processing is complete. | |

** Now copy the last phaseLen - 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; | |

tapCnt = (phaseLen - 1u) >> 2u; | |

/* copy data */ | |

while(tapCnt > 0u) | |

{ | |

*pStateCurnt++ = *pState++; | |

*pStateCurnt++ = *pState++; | |

*pStateCurnt++ = *pState++; | |

*pStateCurnt++ = *pState++; | |

/* Decrement the loop counter */ | |

tapCnt--; | |

} | |

tapCnt = (phaseLen - 1u) % 0x04u; | |

/* copy data */ | |

while(tapCnt > 0u) | |

{ | |

*pStateCurnt++ = *pState++; | |

/* Decrement the loop counter */ | |

tapCnt--; | |

} | |

} | |

#else | |

void arm_fir_interpolate_q31( | |

const arm_fir_interpolate_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 *ptr1, *ptr2; /* Temporary pointers for state and coefficient buffers */ | |

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

q63_t sum; /* Accumulator */ | |

q31_t x0, c0; /* Temporary variables to hold state and coefficient values */ | |

uint32_t i, blkCnt; /* Loop counters */ | |

uint16_t phaseLen = S->phaseLength, tapCnt; /* Length of each polyphase filter component */ | |

/* S->pState buffer contains previous frame (phaseLen - 1) samples */ | |

/* pStateCurnt points to the location where the new input data should be written */ | |

pStateCurnt = S->pState + ((q31_t) phaseLen - 1); | |

/* Total number of intput samples */ | |

blkCnt = blockSize; | |

/* Loop over the blockSize. */ | |

while(blkCnt > 0u) | |

{ | |

/* Copy new input sample into the state buffer */ | |

*pStateCurnt++ = *pSrc++; | |

/* Loop over the Interpolation factor. */ | |

i = S->L; | |

while(i > 0u) | |

{ | |

/* Set accumulator to zero */ | |

sum = 0; | |

/* Initialize state pointer */ | |

ptr1 = pState; | |

/* Initialize coefficient pointer */ | |

ptr2 = pCoeffs + (i - 1u); | |

tapCnt = phaseLen; | |

while(tapCnt > 0u) | |

{ | |

/* Read the coefficient */ | |

c0 = *(ptr2); | |

/* Increment the coefficient pointer by interpolation factor times. */ | |

ptr2 += S->L; | |

/* Read the input sample */ | |

x0 = *ptr1++; | |

/* Perform the multiply-accumulate */ | |

sum += (q63_t) x0 *c0; | |

/* Decrement the loop counter */ | |

tapCnt--; | |

} | |

/* The result is in the accumulator, store in the destination buffer. */ | |

*pDst++ = (q31_t) (sum >> 31); | |

/* Decrement the loop counter */ | |

i--; | |

} | |

/* Advance the state pointer by 1 | |

* to process the next group of interpolation factor number samples */ | |

pState = pState + 1; | |

/* Decrement the loop counter */ | |

blkCnt--; | |

} | |

/* Processing is complete. | |

** Now copy the last phaseLen - 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; | |

tapCnt = phaseLen - 1u; | |

/* copy data */ | |

while(tapCnt > 0u) | |

{ | |

*pStateCurnt++ = *pState++; | |

/* Decrement the loop counter */ | |

tapCnt--; | |

} | |

} | |

#endif /* #ifndef ARM_MATH_CM0_FAMILY */ | |

/** | |

* @} end of FIR_Interpolate group | |

*/ |