arm_offset_q15.c 4.5 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168
  1. /* ----------------------------------------------------------------------
  2. * Project: CMSIS DSP Library
  3. * Title: arm_offset_q15.c
  4. * Description: Q15 vector offset
  5. *
  6. * $Date: 18. March 2019
  7. * $Revision: V1.6.0
  8. *
  9. * Target Processor: Cortex-M cores
  10. * -------------------------------------------------------------------- */
  11. /*
  12. * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  13. *
  14. * SPDX-License-Identifier: Apache-2.0
  15. *
  16. * Licensed under the Apache License, Version 2.0 (the License); you may
  17. * not use this file except in compliance with the License.
  18. * You may obtain a copy of the License at
  19. *
  20. * www.apache.org/licenses/LICENSE-2.0
  21. *
  22. * Unless required by applicable law or agreed to in writing, software
  23. * distributed under the License is distributed on an AS IS BASIS, WITHOUT
  24. * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  25. * See the License for the specific language governing permissions and
  26. * limitations under the License.
  27. */
  28. #include "arm_math.h"
  29. /**
  30. @ingroup groupMath
  31. */
  32. /**
  33. @addtogroup BasicOffset
  34. @{
  35. */
  36. /**
  37. @brief Adds a constant offset to a Q15 vector.
  38. @param[in] pSrc points to the input vector
  39. @param[in] offset is the offset to be added
  40. @param[out] pDst points to the output vector
  41. @param[in] blockSize number of samples in each vector
  42. @return none
  43. @par Scaling and Overflow Behavior
  44. The function uses saturating arithmetic.
  45. Results outside of the allowable Q15 range [0x8000 0x7FFF] are saturated.
  46. */
  47. #if defined(ARM_MATH_MVEI)
  48. #include "arm_helium_utils.h"
  49. void arm_offset_q15(
  50. const q15_t * pSrc,
  51. q15_t offset,
  52. q15_t * pDst,
  53. uint32_t blockSize)
  54. {
  55. uint32_t blkCnt; /* loop counters */
  56. q15x8_t vecSrc;
  57. /* Compute 8 outputs at a time */
  58. blkCnt = blockSize >> 3;
  59. while (blkCnt > 0U)
  60. {
  61. /*
  62. * C = A + offset
  63. * Add offset and then store the result in the destination buffer.
  64. */
  65. vecSrc = vld1q(pSrc);
  66. vst1q(pDst, vqaddq(vecSrc, offset));
  67. /*
  68. * Decrement the blockSize loop counter
  69. */
  70. blkCnt--;
  71. /*
  72. * advance vector source and destination pointers
  73. */
  74. pSrc += 8;
  75. pDst += 8;
  76. }
  77. /*
  78. * tail
  79. */
  80. blkCnt = blockSize & 7;
  81. if (blkCnt > 0U)
  82. {
  83. mve_pred16_t p0 = vctp16q(blkCnt);
  84. vecSrc = vld1q(pSrc);
  85. vstrhq_p(pDst, vqaddq(vecSrc, offset), p0);
  86. }
  87. }
  88. #else
  89. void arm_offset_q15(
  90. const q15_t * pSrc,
  91. q15_t offset,
  92. q15_t * pDst,
  93. uint32_t blockSize)
  94. {
  95. uint32_t blkCnt; /* Loop counter */
  96. #if defined (ARM_MATH_LOOPUNROLL)
  97. #if defined (ARM_MATH_DSP)
  98. q31_t offset_packed; /* Offset packed to 32 bit */
  99. /* Offset is packed to 32 bit in order to use SIMD32 for addition */
  100. offset_packed = __PKHBT(offset, offset, 16);
  101. #endif
  102. /* Loop unrolling: Compute 4 outputs at a time */
  103. blkCnt = blockSize >> 2U;
  104. while (blkCnt > 0U)
  105. {
  106. /* C = A + offset */
  107. #if defined (ARM_MATH_DSP)
  108. /* Add offset and store result in destination buffer (2 samples at a time). */
  109. write_q15x2_ia (&pDst, __QADD16(read_q15x2_ia ((q15_t **) &pSrc), offset_packed));
  110. write_q15x2_ia (&pDst, __QADD16(read_q15x2_ia ((q15_t **) &pSrc), offset_packed));
  111. #else
  112. *pDst++ = (q15_t) __SSAT(((q31_t) *pSrc++ + offset), 16);
  113. *pDst++ = (q15_t) __SSAT(((q31_t) *pSrc++ + offset), 16);
  114. *pDst++ = (q15_t) __SSAT(((q31_t) *pSrc++ + offset), 16);
  115. *pDst++ = (q15_t) __SSAT(((q31_t) *pSrc++ + offset), 16);
  116. #endif
  117. /* Decrement loop counter */
  118. blkCnt--;
  119. }
  120. /* Loop unrolling: Compute remaining outputs */
  121. blkCnt = blockSize % 0x4U;
  122. #else
  123. /* Initialize blkCnt with number of samples */
  124. blkCnt = blockSize;
  125. #endif /* #if defined (ARM_MATH_LOOPUNROLL) */
  126. while (blkCnt > 0U)
  127. {
  128. /* C = A + offset */
  129. /* Add offset and store result in destination buffer. */
  130. #if defined (ARM_MATH_DSP)
  131. *pDst++ = (q15_t) __QADD16(*pSrc++, offset);
  132. #else
  133. *pDst++ = (q15_t) __SSAT(((q31_t) *pSrc++ + offset), 16);
  134. #endif
  135. /* Decrement loop counter */
  136. blkCnt--;
  137. }
  138. }
  139. #endif /* defined(ARM_MATH_MVEI) */
  140. /**
  141. @} end of BasicOffset group
  142. */