arm_negate_q15.c 4.4 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171
  1. /* ----------------------------------------------------------------------
  2. * Project: CMSIS DSP Library
  3. * Title: arm_negate_q15.c
  4. * Description: Negates Q15 vectors
  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 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. @return none
  42. @par Conditions for optimum performance
  43. Input and output buffers should be aligned by 32-bit
  44. @par Scaling and Overflow Behavior
  45. The function uses saturating arithmetic.
  46. The Q15 value -1 (0x8000) is saturated to the maximum allowable positive value 0x7FFF.
  47. */
  48. #if defined(ARM_MATH_MVEI)
  49. #include "arm_helium_utils.h"
  50. void arm_negate_q15(
  51. const q15_t * pSrc,
  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
  63. * Negate and then store the results in the destination buffer.
  64. */
  65. vecSrc = vld1q(pSrc);
  66. vst1q(pDst, vqnegq(vecSrc));
  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, vqnegq(vecSrc), p0);
  86. }
  87. }
  88. #else
  89. void arm_negate_q15(
  90. const q15_t * pSrc,
  91. q15_t * pDst,
  92. uint32_t blockSize)
  93. {
  94. uint32_t blkCnt; /* Loop counter */
  95. q15_t in; /* Temporary input variable */
  96. #if defined (ARM_MATH_LOOPUNROLL)
  97. #if defined (ARM_MATH_DSP)
  98. q31_t in1; /* Temporary input variables */
  99. #endif
  100. /* Loop unrolling: Compute 4 outputs at a time */
  101. blkCnt = blockSize >> 2U;
  102. while (blkCnt > 0U)
  103. {
  104. /* C = -A */
  105. #if defined (ARM_MATH_DSP)
  106. /* Negate and store result in destination buffer (2 samples at a time). */
  107. in1 = read_q15x2_ia ((q15_t **) &pSrc);
  108. write_q15x2_ia (&pDst, __QSUB16(0, in1));
  109. in1 = read_q15x2_ia ((q15_t **) &pSrc);
  110. write_q15x2_ia (&pDst, __QSUB16(0, in1));
  111. #else
  112. in = *pSrc++;
  113. *pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
  114. in = *pSrc++;
  115. *pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
  116. in = *pSrc++;
  117. *pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
  118. in = *pSrc++;
  119. *pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
  120. #endif
  121. /* Decrement loop counter */
  122. blkCnt--;
  123. }
  124. /* Loop unrolling: Compute remaining outputs */
  125. blkCnt = blockSize % 0x4U;
  126. #else
  127. /* Initialize blkCnt with number of samples */
  128. blkCnt = blockSize;
  129. #endif /* #if defined (ARM_MATH_LOOPUNROLL) */
  130. while (blkCnt > 0U)
  131. {
  132. /* C = -A */
  133. /* Negate and store result in destination buffer. */
  134. in = *pSrc++;
  135. *pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
  136. /* Decrement loop counter */
  137. blkCnt--;
  138. }
  139. }
  140. #endif /* defined(ARM_MATH_MVEI) */
  141. /**
  142. @} end of BasicNegate group
  143. */