arm_cmplx_conj_q15.c

00001 /* ----------------------------------------------------------------------
00002  * Project:      CMSIS DSP Library
00003  * Title:        arm_cmplx_conj_q15.c
00004  * Description:  Q15 complex conjugate
00005  *
00006  * $Date:        27. January 2017
00007  * $Revision:    V.1.5.1
00008  *
00009  * Target Processor: Cortex-M cores
00010  * -------------------------------------------------------------------- */
00011 /*
00012  * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
00013  *
00014  * SPDX-License-Identifier: Apache-2.0
00015  *
00016  * Licensed under the Apache License, Version 2.0 (the License); you may
00017  * not use this file except in compliance with the License.
00018  * You may obtain a copy of the License at
00019  *
00020  * www.apache.org/licenses/LICENSE-2.0
00021  *
00022  * Unless required by applicable law or agreed to in writing, software
00023  * distributed under the License is distributed on an AS IS BASIS, WITHOUT
00024  * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
00025  * See the License for the specific language governing permissions and
00026  * limitations under the License.
00027  */
00028 
00029 #include "arm_math.h"
00030 
00031 /**
00032  * @ingroup groupCmplxMath
00033  */
00034 
00035 /**
00036  * @addtogroup cmplx_conj
00037  * @{
00038  */
00039 
00040 /**
00041  * @brief  Q15 complex conjugate.
00042  * @param  *pSrc points to the input vector
00043  * @param  *pDst points to the output vector
00044  * @param  numSamples number of complex samples in each vector
00045  * @return none.
00046  *
00047  * <b>Scaling and Overflow Behavior:</b>
00048  * \par
00049  * The function uses saturating arithmetic.
00050  * The Q15 value -1 (0x8000) will be saturated to the maximum allowable positive value 0x7FFF.
00051  */
00052 
00053 void arm_cmplx_conj_q15(
00054   q15_t * pSrc,
00055   q15_t * pDst,
00056   uint32_t numSamples)
00057 {
00058 
00059 #if defined (ARM_MATH_DSP)
00060 
00061   /* Run the below code for Cortex-M4 and Cortex-M3 */
00062   uint32_t blkCnt;                               /* loop counter */
00063   q31_t in1, in2, in3, in4;
00064   q31_t zero = 0;
00065 
00066   /*loop Unrolling */
00067   blkCnt = numSamples >> 2U;
00068 
00069   /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
00070    ** a second loop below computes the remaining 1 to 3 samples. */
00071   while (blkCnt > 0U)
00072   {
00073     /* C[0]+jC[1] = A[0]+ j (-1) A[1] */
00074     /* Calculate Complex Conjugate and then store the results in the destination buffer. */
00075     in1 = *__SIMD32(pSrc)++;
00076     in2 = *__SIMD32(pSrc)++;
00077     in3 = *__SIMD32(pSrc)++;
00078     in4 = *__SIMD32(pSrc)++;
00079 
00080 #ifndef ARM_MATH_BIG_ENDIAN
00081 
00082     in1 = __QASX(zero, in1);
00083     in2 = __QASX(zero, in2);
00084     in3 = __QASX(zero, in3);
00085     in4 = __QASX(zero, in4);
00086 
00087 #else
00088 
00089     in1 = __QSAX(zero, in1);
00090     in2 = __QSAX(zero, in2);
00091     in3 = __QSAX(zero, in3);
00092     in4 = __QSAX(zero, in4);
00093 
00094 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
00095 
00096     in1 = ((uint32_t) in1 >> 16) | ((uint32_t) in1 << 16);
00097     in2 = ((uint32_t) in2 >> 16) | ((uint32_t) in2 << 16);
00098     in3 = ((uint32_t) in3 >> 16) | ((uint32_t) in3 << 16);
00099     in4 = ((uint32_t) in4 >> 16) | ((uint32_t) in4 << 16);
00100 
00101     *__SIMD32(pDst)++ = in1;
00102     *__SIMD32(pDst)++ = in2;
00103     *__SIMD32(pDst)++ = in3;
00104     *__SIMD32(pDst)++ = in4;
00105 
00106     /* Decrement the loop counter */
00107     blkCnt--;
00108   }
00109 
00110   /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
00111    ** No loop unrolling is used. */
00112   blkCnt = numSamples % 0x4U;
00113 
00114   while (blkCnt > 0U)
00115   {
00116     /* C[0]+jC[1] = A[0]+ j (-1) A[1] */
00117     /* Calculate Complex Conjugate and then store the results in the destination buffer. */
00118     *pDst++ = *pSrc++;
00119     *pDst++ = __SSAT(-*pSrc++, 16);
00120 
00121     /* Decrement the loop counter */
00122     blkCnt--;
00123   }
00124 
00125 #else
00126 
00127   q15_t in;
00128 
00129   /* Run the below code for Cortex-M0 */
00130 
00131   while (numSamples > 0U)
00132   {
00133     /* realOut + j (imagOut) = realIn+ j (-1) imagIn */
00134     /* Calculate Complex Conjugate and then store the results in the destination buffer. */
00135     *pDst++ = *pSrc++;
00136     in = *pSrc++;
00137     *pDst++ = (in == (q15_t) 0x8000) ? 0x7fff : -in;
00138 
00139     /* Decrement the loop counter */
00140     numSamples--;
00141   }
00142 
00143 #endif /* #if defined (ARM_MATH_DSP) */
00144 
00145 }
00146 
00147 /**
00148  * @} end of cmplx_conj group
00149  */
00150