John Garlitos / Mbed OS EVAL-CN0540-ARDZ

Important changes to repositories hosted on mbed.com

Mbed hosted mercurial repositories are deprecated and are due to be permanently deleted in July 2026.

To keep a copy of this software download the repository Zip archive or clone locally using Mercurial.

It is also possible to export all your personal repositories from the account settings page.

Dependencies:   platform_drivers LTC26X6 AD77681

CMSIS-DSP_Lib/Source/BasicMathFunctions/arm_mult_q31.c@1:9dd7c64b4a64, 2021-12-06 (annotated)

Committer:
jngarlitos
Date:
Mon Dec 06 05:22:28 2021 +0000
Revision:
1:9dd7c64b4a64
EVAL-CN0540-ARDZ mbed example program Initial Commit

Who changed what in which revision?

UserRevisionLine numberNew contents of line
jngarlitos 1:9dd7c64b4a64 1 /* ----------------------------------------------------------------------
jngarlitos 1:9dd7c64b4a64 2 * Copyright (C) 2010-2014 ARM Limited. All rights reserved.
jngarlitos 1:9dd7c64b4a64 3 *
jngarlitos 1:9dd7c64b4a64 4 * $Date: 19. March 2015
jngarlitos 1:9dd7c64b4a64 5 * $Revision: V.1.4.5
jngarlitos 1:9dd7c64b4a64 6 *
jngarlitos 1:9dd7c64b4a64 7 * Project: CMSIS DSP Library
jngarlitos 1:9dd7c64b4a64 8 * Title: arm_mult_q31.c
jngarlitos 1:9dd7c64b4a64 9 *
jngarlitos 1:9dd7c64b4a64 10 * Description: Q31 vector multiplication.
jngarlitos 1:9dd7c64b4a64 11 *
jngarlitos 1:9dd7c64b4a64 12 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
jngarlitos 1:9dd7c64b4a64 13 *
jngarlitos 1:9dd7c64b4a64 14 * Redistribution and use in source and binary forms, with or without
jngarlitos 1:9dd7c64b4a64 15 * modification, are permitted provided that the following conditions
jngarlitos 1:9dd7c64b4a64 16 * are met:
jngarlitos 1:9dd7c64b4a64 17 * - Redistributions of source code must retain the above copyright
jngarlitos 1:9dd7c64b4a64 18 * notice, this list of conditions and the following disclaimer.
jngarlitos 1:9dd7c64b4a64 19 * - Redistributions in binary form must reproduce the above copyright
jngarlitos 1:9dd7c64b4a64 20 * notice, this list of conditions and the following disclaimer in
jngarlitos 1:9dd7c64b4a64 21 * the documentation and/or other materials provided with the
jngarlitos 1:9dd7c64b4a64 22 * distribution.
jngarlitos 1:9dd7c64b4a64 23 * - Neither the name of ARM LIMITED nor the names of its contributors
jngarlitos 1:9dd7c64b4a64 24 * may be used to endorse or promote products derived from this
jngarlitos 1:9dd7c64b4a64 25 * software without specific prior written permission.
jngarlitos 1:9dd7c64b4a64 26 *
jngarlitos 1:9dd7c64b4a64 27 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
jngarlitos 1:9dd7c64b4a64 28 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
jngarlitos 1:9dd7c64b4a64 29 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
jngarlitos 1:9dd7c64b4a64 30 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
jngarlitos 1:9dd7c64b4a64 31 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
jngarlitos 1:9dd7c64b4a64 32 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
jngarlitos 1:9dd7c64b4a64 33 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
jngarlitos 1:9dd7c64b4a64 34 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
jngarlitos 1:9dd7c64b4a64 35 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
jngarlitos 1:9dd7c64b4a64 36 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
jngarlitos 1:9dd7c64b4a64 37 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
jngarlitos 1:9dd7c64b4a64 38 * POSSIBILITY OF SUCH DAMAGE.
jngarlitos 1:9dd7c64b4a64 39 * -------------------------------------------------------------------- */
jngarlitos 1:9dd7c64b4a64 40
jngarlitos 1:9dd7c64b4a64 41 #include "arm_math.h"
jngarlitos 1:9dd7c64b4a64 42
jngarlitos 1:9dd7c64b4a64 43 /**
jngarlitos 1:9dd7c64b4a64 44 * @ingroup groupMath
jngarlitos 1:9dd7c64b4a64 45 */
jngarlitos 1:9dd7c64b4a64 46
jngarlitos 1:9dd7c64b4a64 47 /**
jngarlitos 1:9dd7c64b4a64 48 * @addtogroup BasicMult
jngarlitos 1:9dd7c64b4a64 49 * @{
jngarlitos 1:9dd7c64b4a64 50 */
jngarlitos 1:9dd7c64b4a64 51
jngarlitos 1:9dd7c64b4a64 52 /**
jngarlitos 1:9dd7c64b4a64 53 * @brief Q31 vector multiplication.
jngarlitos 1:9dd7c64b4a64 54 * @param[in] *pSrcA points to the first input vector
jngarlitos 1:9dd7c64b4a64 55 * @param[in] *pSrcB points to the second input vector
jngarlitos 1:9dd7c64b4a64 56 * @param[out] *pDst points to the output vector
jngarlitos 1:9dd7c64b4a64 57 * @param[in] blockSize number of samples in each vector
jngarlitos 1:9dd7c64b4a64 58 * @return none.
jngarlitos 1:9dd7c64b4a64 59 *
jngarlitos 1:9dd7c64b4a64 60 * <b>Scaling and Overflow Behavior:</b>
jngarlitos 1:9dd7c64b4a64 61 * \par
jngarlitos 1:9dd7c64b4a64 62 * The function uses saturating arithmetic.
jngarlitos 1:9dd7c64b4a64 63 * Results outside of the allowable Q31 range[0x80000000 0x7FFFFFFF] will be saturated.
jngarlitos 1:9dd7c64b4a64 64 */
jngarlitos 1:9dd7c64b4a64 65
jngarlitos 1:9dd7c64b4a64 66 void arm_mult_q31(
jngarlitos 1:9dd7c64b4a64 67 q31_t * pSrcA,
jngarlitos 1:9dd7c64b4a64 68 q31_t * pSrcB,
jngarlitos 1:9dd7c64b4a64 69 q31_t * pDst,
jngarlitos 1:9dd7c64b4a64 70 uint32_t blockSize)
jngarlitos 1:9dd7c64b4a64 71 {
jngarlitos 1:9dd7c64b4a64 72 uint32_t blkCnt; /* loop counters */
jngarlitos 1:9dd7c64b4a64 73
jngarlitos 1:9dd7c64b4a64 74 #ifndef ARM_MATH_CM0_FAMILY
jngarlitos 1:9dd7c64b4a64 75
jngarlitos 1:9dd7c64b4a64 76 /* Run the below code for Cortex-M4 and Cortex-M3 */
jngarlitos 1:9dd7c64b4a64 77 q31_t inA1, inA2, inA3, inA4; /* temporary input variables */
jngarlitos 1:9dd7c64b4a64 78 q31_t inB1, inB2, inB3, inB4; /* temporary input variables */
jngarlitos 1:9dd7c64b4a64 79 q31_t out1, out2, out3, out4; /* temporary output variables */
jngarlitos 1:9dd7c64b4a64 80
jngarlitos 1:9dd7c64b4a64 81 /* loop Unrolling */
jngarlitos 1:9dd7c64b4a64 82 blkCnt = blockSize >> 2u;
jngarlitos 1:9dd7c64b4a64 83
jngarlitos 1:9dd7c64b4a64 84 /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
jngarlitos 1:9dd7c64b4a64 85 ** a second loop below computes the remaining 1 to 3 samples. */
jngarlitos 1:9dd7c64b4a64 86 while(blkCnt > 0u)
jngarlitos 1:9dd7c64b4a64 87 {
jngarlitos 1:9dd7c64b4a64 88 /* C = A * B */
jngarlitos 1:9dd7c64b4a64 89 /* Multiply the inputs and then store the results in the destination buffer. */
jngarlitos 1:9dd7c64b4a64 90 inA1 = *pSrcA++;
jngarlitos 1:9dd7c64b4a64 91 inA2 = *pSrcA++;
jngarlitos 1:9dd7c64b4a64 92 inA3 = *pSrcA++;
jngarlitos 1:9dd7c64b4a64 93 inA4 = *pSrcA++;
jngarlitos 1:9dd7c64b4a64 94 inB1 = *pSrcB++;
jngarlitos 1:9dd7c64b4a64 95 inB2 = *pSrcB++;
jngarlitos 1:9dd7c64b4a64 96 inB3 = *pSrcB++;
jngarlitos 1:9dd7c64b4a64 97 inB4 = *pSrcB++;
jngarlitos 1:9dd7c64b4a64 98
jngarlitos 1:9dd7c64b4a64 99 out1 = ((q63_t) inA1 * inB1) >> 32;
jngarlitos 1:9dd7c64b4a64 100 out2 = ((q63_t) inA2 * inB2) >> 32;
jngarlitos 1:9dd7c64b4a64 101 out3 = ((q63_t) inA3 * inB3) >> 32;
jngarlitos 1:9dd7c64b4a64 102 out4 = ((q63_t) inA4 * inB4) >> 32;
jngarlitos 1:9dd7c64b4a64 103
jngarlitos 1:9dd7c64b4a64 104 out1 = __SSAT(out1, 31);
jngarlitos 1:9dd7c64b4a64 105 out2 = __SSAT(out2, 31);
jngarlitos 1:9dd7c64b4a64 106 out3 = __SSAT(out3, 31);
jngarlitos 1:9dd7c64b4a64 107 out4 = __SSAT(out4, 31);
jngarlitos 1:9dd7c64b4a64 108
jngarlitos 1:9dd7c64b4a64 109 *pDst++ = out1 << 1u;
jngarlitos 1:9dd7c64b4a64 110 *pDst++ = out2 << 1u;
jngarlitos 1:9dd7c64b4a64 111 *pDst++ = out3 << 1u;
jngarlitos 1:9dd7c64b4a64 112 *pDst++ = out4 << 1u;
jngarlitos 1:9dd7c64b4a64 113
jngarlitos 1:9dd7c64b4a64 114 /* Decrement the blockSize loop counter */
jngarlitos 1:9dd7c64b4a64 115 blkCnt--;
jngarlitos 1:9dd7c64b4a64 116 }
jngarlitos 1:9dd7c64b4a64 117
jngarlitos 1:9dd7c64b4a64 118 /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
jngarlitos 1:9dd7c64b4a64 119 ** No loop unrolling is used. */
jngarlitos 1:9dd7c64b4a64 120 blkCnt = blockSize % 0x4u;
jngarlitos 1:9dd7c64b4a64 121
jngarlitos 1:9dd7c64b4a64 122 while(blkCnt > 0u)
jngarlitos 1:9dd7c64b4a64 123 {
jngarlitos 1:9dd7c64b4a64 124 /* C = A * B */
jngarlitos 1:9dd7c64b4a64 125 /* Multiply the inputs and then store the results in the destination buffer. */
jngarlitos 1:9dd7c64b4a64 126 inA1 = *pSrcA++;
jngarlitos 1:9dd7c64b4a64 127 inB1 = *pSrcB++;
jngarlitos 1:9dd7c64b4a64 128 out1 = ((q63_t) inA1 * inB1) >> 32;
jngarlitos 1:9dd7c64b4a64 129 out1 = __SSAT(out1, 31);
jngarlitos 1:9dd7c64b4a64 130 *pDst++ = out1 << 1u;
jngarlitos 1:9dd7c64b4a64 131
jngarlitos 1:9dd7c64b4a64 132 /* Decrement the blockSize loop counter */
jngarlitos 1:9dd7c64b4a64 133 blkCnt--;
jngarlitos 1:9dd7c64b4a64 134 }
jngarlitos 1:9dd7c64b4a64 135
jngarlitos 1:9dd7c64b4a64 136 #else
jngarlitos 1:9dd7c64b4a64 137
jngarlitos 1:9dd7c64b4a64 138 /* Run the below code for Cortex-M0 */
jngarlitos 1:9dd7c64b4a64 139
jngarlitos 1:9dd7c64b4a64 140 /* Initialize blkCnt with number of samples */
jngarlitos 1:9dd7c64b4a64 141 blkCnt = blockSize;
jngarlitos 1:9dd7c64b4a64 142
jngarlitos 1:9dd7c64b4a64 143
jngarlitos 1:9dd7c64b4a64 144 while(blkCnt > 0u)
jngarlitos 1:9dd7c64b4a64 145 {
jngarlitos 1:9dd7c64b4a64 146 /* C = A * B */
jngarlitos 1:9dd7c64b4a64 147 /* Multiply the inputs and then store the results in the destination buffer. */
jngarlitos 1:9dd7c64b4a64 148 *pDst++ =
jngarlitos 1:9dd7c64b4a64 149 (q31_t) clip_q63_to_q31(((q63_t) (*pSrcA++) * (*pSrcB++)) >> 31);
jngarlitos 1:9dd7c64b4a64 150
jngarlitos 1:9dd7c64b4a64 151 /* Decrement the blockSize loop counter */
jngarlitos 1:9dd7c64b4a64 152 blkCnt--;
jngarlitos 1:9dd7c64b4a64 153 }
jngarlitos 1:9dd7c64b4a64 154
jngarlitos 1:9dd7c64b4a64 155 #endif /* #ifndef ARM_MATH_CM0_FAMILY */
jngarlitos 1:9dd7c64b4a64 156 }
jngarlitos 1:9dd7c64b4a64 157
jngarlitos 1:9dd7c64b4a64 158 /**
jngarlitos 1:9dd7c64b4a64 159 * @} end of BasicMult group
jngarlitos 1:9dd7c64b4a64 160 */