arm_shift_q31.c 6.3 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203
  1. /* ----------------------------------------------------------------------
  2. * Copyright (C) 2010-2014 ARM Limited. All rights reserved.
  3. *
  4. * $Date: 12. March 2014
  5. * $Revision: V1.4.4
  6. *
  7. * Project: CMSIS DSP Library
  8. * Title: arm_shift_q31.c
  9. *
  10. * Description: Shifts the elements of a Q31 vector by a specified number of bits.
  11. *
  12. * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
  13. *
  14. * Redistribution and use in source and binary forms, with or without
  15. * modification, are permitted provided that the following conditions
  16. * are met:
  17. * - Redistributions of source code must retain the above copyright
  18. * notice, this list of conditions and the following disclaimer.
  19. * - Redistributions in binary form must reproduce the above copyright
  20. * notice, this list of conditions and the following disclaimer in
  21. * the documentation and/or other materials provided with the
  22. * distribution.
  23. * - Neither the name of ARM LIMITED nor the names of its contributors
  24. * may be used to endorse or promote products derived from this
  25. * software without specific prior written permission.
  26. *
  27. * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
  28. * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
  29. * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
  30. * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
  31. * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
  32. * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
  33. * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
  34. * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
  35. * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  36. * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
  37. * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
  38. * POSSIBILITY OF SUCH DAMAGE.
  39. * -------------------------------------------------------------------- */
  40. #include "arm_math.h"
  41. /**
  42. * @ingroup groupMath
  43. */
  44. /**
  45. * @defgroup shift Vector Shift
  46. *
  47. * Shifts the elements of a fixed-point vector by a specified number of bits.
  48. * There are separate functions for Q7, Q15, and Q31 data types.
  49. * The underlying algorithm used is:
  50. *
  51. * <pre>
  52. * pDst[n] = pSrc[n] << shift, 0 <= n < blockSize.
  53. * </pre>
  54. *
  55. * If <code>shift</code> is positive then the elements of the vector are shifted to the left.
  56. * If <code>shift</code> is negative then the elements of the vector are shifted to the right.
  57. *
  58. * The functions support in-place computation allowing the source and destination
  59. * pointers to reference the same memory buffer.
  60. */
  61. /**
  62. * @addtogroup shift
  63. * @{
  64. */
  65. /**
  66. * @brief Shifts the elements of a Q31 vector a specified number of bits.
  67. * @param[in] *pSrc points to the input vector
  68. * @param[in] shiftBits number of bits to shift. A positive value shifts left; a negative value shifts right.
  69. * @param[out] *pDst points to the output vector
  70. * @param[in] blockSize number of samples in the vector
  71. * @return none.
  72. *
  73. *
  74. * <b>Scaling and Overflow Behavior:</b>
  75. * \par
  76. * The function uses saturating arithmetic.
  77. * Results outside of the allowable Q31 range [0x80000000 0x7FFFFFFF] will be saturated.
  78. */
  79. void arm_shift_q31(
  80. q31_t * pSrc,
  81. int8_t shiftBits,
  82. q31_t * pDst,
  83. uint32_t blockSize)
  84. {
  85. uint32_t blkCnt; /* loop counter */
  86. uint8_t sign = (shiftBits & 0x80); /* Sign of shiftBits */
  87. #ifndef ARM_MATH_CM0_FAMILY
  88. q31_t in1, in2, in3, in4; /* Temporary input variables */
  89. q31_t out1, out2, out3, out4; /* Temporary output variables */
  90. /*loop Unrolling */
  91. blkCnt = blockSize >> 2u;
  92. if(sign == 0u)
  93. {
  94. /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
  95. ** a second loop below computes the remaining 1 to 3 samples. */
  96. while(blkCnt > 0u)
  97. {
  98. /* C = A << shiftBits */
  99. /* Shift the input and then store the results in the destination buffer. */
  100. in1 = *pSrc;
  101. in2 = *(pSrc + 1);
  102. out1 = in1 << shiftBits;
  103. in3 = *(pSrc + 2);
  104. out2 = in2 << shiftBits;
  105. in4 = *(pSrc + 3);
  106. if(in1 != (out1 >> shiftBits))
  107. out1 = 0x7FFFFFFF ^ (in1 >> 31);
  108. if(in2 != (out2 >> shiftBits))
  109. out2 = 0x7FFFFFFF ^ (in2 >> 31);
  110. *pDst = out1;
  111. out3 = in3 << shiftBits;
  112. *(pDst + 1) = out2;
  113. out4 = in4 << shiftBits;
  114. if(in3 != (out3 >> shiftBits))
  115. out3 = 0x7FFFFFFF ^ (in3 >> 31);
  116. if(in4 != (out4 >> shiftBits))
  117. out4 = 0x7FFFFFFF ^ (in4 >> 31);
  118. *(pDst + 2) = out3;
  119. *(pDst + 3) = out4;
  120. /* Update destination pointer to process next sampels */
  121. pSrc += 4u;
  122. pDst += 4u;
  123. /* Decrement the loop counter */
  124. blkCnt--;
  125. }
  126. }
  127. else
  128. {
  129. /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
  130. ** a second loop below computes the remaining 1 to 3 samples. */
  131. while(blkCnt > 0u)
  132. {
  133. /* C = A >> shiftBits */
  134. /* Shift the input and then store the results in the destination buffer. */
  135. in1 = *pSrc;
  136. in2 = *(pSrc + 1);
  137. in3 = *(pSrc + 2);
  138. in4 = *(pSrc + 3);
  139. *pDst = (in1 >> -shiftBits);
  140. *(pDst + 1) = (in2 >> -shiftBits);
  141. *(pDst + 2) = (in3 >> -shiftBits);
  142. *(pDst + 3) = (in4 >> -shiftBits);
  143. pSrc += 4u;
  144. pDst += 4u;
  145. blkCnt--;
  146. }
  147. }
  148. /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
  149. ** No loop unrolling is used. */
  150. blkCnt = blockSize % 0x4u;
  151. #else
  152. /* Run the below code for Cortex-M0 */
  153. /* Initialize blkCnt with number of samples */
  154. blkCnt = blockSize;
  155. #endif /* #ifndef ARM_MATH_CM0_FAMILY */
  156. while(blkCnt > 0u)
  157. {
  158. /* C = A (>> or <<) shiftBits */
  159. /* Shift the input and then store the result in the destination buffer. */
  160. *pDst++ = (sign == 0u) ? clip_q63_to_q31((q63_t) * pSrc++ << shiftBits) :
  161. (*pSrc++ >> -shiftBits);
  162. /* Decrement the loop counter */
  163. blkCnt--;
  164. }
  165. }
  166. /**
  167. * @} end of shift group
  168. */