arm_negate_q15.c 4.3 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170
  1. /* ----------------------------------------------------------------------
  2. * Project: CMSIS DSP Library
  3. * Title: arm_negate_q15.c
  4. * Description: Negates Q15 vectors
  5. *
  6. * $Date: 23 April 2021
  7. * $Revision: V1.9.0
  8. *
  9. * Target Processor: Cortex-M and Cortex-A cores
  10. * -------------------------------------------------------------------- */
  11. /*
  12. * Copyright (C) 2010-2021 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 "dsp/basic_math_functions.h"
  29. /**
  30. @ingroup groupMath
  31. */
  32. /**
  33. @addtogroup BasicNegate
  34. @{
  35. */
  36. /**
  37. @brief Negates the elements of a Q15 vector.
  38. @param[in] pSrc points to the input vector.
  39. @param[out] pDst points to the output vector.
  40. @param[in] blockSize number of samples in each vector.
  41. @par Conditions for optimum performance
  42. Input and output buffers should be aligned by 32-bit
  43. @par Scaling and Overflow Behavior
  44. The function uses saturating arithmetic.
  45. The Q15 value -1 (0x8000) is saturated to the maximum allowable positive value 0x7FFF.
  46. */
  47. #if defined(ARM_MATH_MVEI) && !defined(ARM_MATH_AUTOVECTORIZE)
  48. #include "arm_helium_utils.h"
  49. void arm_negate_q15(
  50. const q15_t * pSrc,
  51. q15_t * pDst,
  52. uint32_t blockSize)
  53. {
  54. uint32_t blkCnt; /* loop counters */
  55. q15x8_t vecSrc;
  56. /* Compute 8 outputs at a time */
  57. blkCnt = blockSize >> 3;
  58. while (blkCnt > 0U)
  59. {
  60. /*
  61. * C = -A
  62. * Negate and then store the results in the destination buffer.
  63. */
  64. vecSrc = vld1q(pSrc);
  65. vst1q(pDst, vqnegq(vecSrc));
  66. /*
  67. * Decrement the blockSize loop counter
  68. */
  69. blkCnt--;
  70. /*
  71. * advance vector source and destination pointers
  72. */
  73. pSrc += 8;
  74. pDst += 8;
  75. }
  76. /*
  77. * tail
  78. */
  79. blkCnt = blockSize & 7;
  80. if (blkCnt > 0U)
  81. {
  82. mve_pred16_t p0 = vctp16q(blkCnt);
  83. vecSrc = vld1q(pSrc);
  84. vstrhq_p(pDst, vqnegq(vecSrc), p0);
  85. }
  86. }
  87. #else
  88. void arm_negate_q15(
  89. const q15_t * pSrc,
  90. q15_t * pDst,
  91. uint32_t blockSize)
  92. {
  93. uint32_t blkCnt; /* Loop counter */
  94. q15_t in; /* Temporary input variable */
  95. #if defined (ARM_MATH_LOOPUNROLL)
  96. #if defined (ARM_MATH_DSP)
  97. q31_t in1; /* Temporary input variables */
  98. #endif
  99. /* Loop unrolling: Compute 4 outputs at a time */
  100. blkCnt = blockSize >> 2U;
  101. while (blkCnt > 0U)
  102. {
  103. /* C = -A */
  104. #if defined (ARM_MATH_DSP)
  105. /* Negate and store result in destination buffer (2 samples at a time). */
  106. in1 = read_q15x2_ia (&pSrc);
  107. write_q15x2_ia (&pDst, __QSUB16(0, in1));
  108. in1 = read_q15x2_ia (&pSrc);
  109. write_q15x2_ia (&pDst, __QSUB16(0, in1));
  110. #else
  111. in = *pSrc++;
  112. *pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
  113. in = *pSrc++;
  114. *pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
  115. in = *pSrc++;
  116. *pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
  117. in = *pSrc++;
  118. *pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
  119. #endif
  120. /* Decrement loop counter */
  121. blkCnt--;
  122. }
  123. /* Loop unrolling: Compute remaining outputs */
  124. blkCnt = blockSize % 0x4U;
  125. #else
  126. /* Initialize blkCnt with number of samples */
  127. blkCnt = blockSize;
  128. #endif /* #if defined (ARM_MATH_LOOPUNROLL) */
  129. while (blkCnt > 0U)
  130. {
  131. /* C = -A */
  132. /* Negate and store result in destination buffer. */
  133. in = *pSrc++;
  134. *pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
  135. /* Decrement loop counter */
  136. blkCnt--;
  137. }
  138. }
  139. #endif /* defined(ARM_MATH_MVEI) */
  140. /**
  141. @} end of BasicNegate group
  142. */