Razvalyaev/UKSI_TEST
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Чет работает

Browse Source
This commit is contained in:
Razvalyaev
2025-12-16 17:57:59 +03:00
commit 260a6416ad
2906 changed files with 982301 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

63
Drivers/CMSIS/DSP/Source/BasicMathFunctions/BasicMathFunctions.c Normal file
View File

@@ -0,0 +1,63 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: BasicMathFunctions.c
* Description: Combination of all basic math function source files.
*
* $Date: 18. March 2019
* $Revision: V1.0.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_abs_f32.c"
#include "arm_abs_q15.c"
#include "arm_abs_q31.c"
#include "arm_abs_q7.c"
#include "arm_add_f32.c"
#include "arm_add_q15.c"
#include "arm_add_q31.c"
#include "arm_add_q7.c"
#include "arm_dot_prod_f32.c"
#include "arm_dot_prod_q15.c"
#include "arm_dot_prod_q31.c"
#include "arm_dot_prod_q7.c"
#include "arm_mult_f32.c"
#include "arm_mult_q15.c"
#include "arm_mult_q31.c"
#include "arm_mult_q7.c"
#include "arm_negate_f32.c"
#include "arm_negate_q15.c"
#include "arm_negate_q31.c"
#include "arm_negate_q7.c"
#include "arm_offset_f32.c"
#include "arm_offset_q15.c"
#include "arm_offset_q31.c"
#include "arm_offset_q7.c"
#include "arm_scale_f32.c"
#include "arm_scale_q15.c"
#include "arm_scale_q31.c"
#include "arm_scale_q7.c"
#include "arm_shift_q15.c"
#include "arm_shift_q31.c"
#include "arm_shift_q7.c"
#include "arm_sub_f32.c"
#include "arm_sub_q15.c"
#include "arm_sub_q31.c"
#include "arm_sub_q7.c"

16
Drivers/CMSIS/DSP/Source/BasicMathFunctions/CMakeLists.txt Normal file
View File

@@ -0,0 +1,16 @@
cmake_minimum_required (VERSION 3.6)
project(CMSISDSPBasicMath)
file(GLOB SRC "./*_*.c")
add_library(CMSISDSPBasicMath STATIC ${SRC})
configdsp(CMSISDSPBasicMath ..)
### Includes
target_include_directories(CMSISDSPBasicMath PUBLIC "${DSP}/../../Include")

146
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_abs_f32.c Normal file
View File

@@ -0,0 +1,146 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_abs_f32.c
* Description: Floating-point vector absolute value
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
#include <math.h>
/**
@ingroup groupMath
*/
/**
@defgroup BasicAbs Vector Absolute Value
Computes the absolute value of a vector on an element-by-element basis.
<pre>
pDst[n] = abs(pSrc[n]), 0 <= n < blockSize.
</pre>
The functions support in-place computation allowing the source and
destination pointers to reference the same memory buffer.
There are separate functions for floating-point, Q7, Q15, and Q31 data types.
*/
/**
@addtogroup BasicAbs
@{
*/
/**
@brief Floating-point vector absolute value.
@param[in] pSrc points to the input vector
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
*/
void arm_abs_f32(
const float32_t * pSrc,
float32_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined(ARM_MATH_NEON)
float32x4_t vec1;
float32x4_t res;
/* Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = |A| */
/* Calculate absolute values and then store the results in the destination buffer. */
vec1 = vld1q_f32(pSrc);
res = vabsq_f32(vec1);
vst1q_f32(pDst, res);
/* Increment pointers */
pSrc += 4;
pDst += 4;
/* Decrement the loop counter */
blkCnt--;
}
/* Tail */
blkCnt = blockSize & 0x3;
#else
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = |A| */
/* Calculate absolute and store result in destination buffer. */
*pDst++ = fabsf(*pSrc++);
*pDst++ = fabsf(*pSrc++);
*pDst++ = fabsf(*pSrc++);
*pDst++ = fabsf(*pSrc++);
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
#endif /* #if defined(ARM_MATH_NEON) */
while (blkCnt > 0U)
{
/* C = |A| */
/* Calculate absolute and store result in destination buffer. */
*pDst++ = fabsf(*pSrc++);
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicAbs group
*/

132
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_abs_q15.c Normal file
View File

@@ -0,0 +1,132 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_abs_q15.c
* Description: Q15 vector absolute value
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicAbs
@{
*/
/**
@brief Q15 vector absolute value.
@param[in] pSrc points to the input vector
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
The Q15 value -1 (0x8000) will be saturated to the maximum allowable positive value 0x7FFF.
*/
void arm_abs_q15(
const q15_t * pSrc,
q15_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
q15_t in; /* Temporary input variable */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = |A| */
/* Calculate absolute of input (if -1 then saturated to 0x7fff) and store result in destination buffer. */
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q15_t)__QSUB16(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == (q15_t) 0x8000) ? 0x7fff : -in);
#endif
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q15_t)__QSUB16(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == (q15_t) 0x8000) ? 0x7fff : -in);
#endif
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q15_t)__QSUB16(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == (q15_t) 0x8000) ? 0x7fff : -in);
#endif
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q15_t)__QSUB16(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == (q15_t) 0x8000) ? 0x7fff : -in);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = |A| */
/* Calculate absolute of input (if -1 then saturated to 0x7fff) and store result in destination buffer. */
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q15_t)__QSUB16(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == (q15_t) 0x8000) ? 0x7fff : -in);
#endif
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicAbs group
*/

132
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_abs_q31.c Normal file
View File

@@ -0,0 +1,132 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_abs_q31.c
* Description: Q31 vector absolute value
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicAbs
@{
*/
/**
@brief Q31 vector absolute value.
@param[in] pSrc points to the input vector
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
The Q31 value -1 (0x80000000) will be saturated to the maximum allowable positive value 0x7FFFFFFF.
*/
void arm_abs_q31(
const q31_t * pSrc,
q31_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
q31_t in; /* Temporary variable */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = |A| */
/* Calculate absolute of input (if -1 then saturated to 0x7fffffff) and store result in destination buffer. */
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q31_t)__QSUB(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == INT32_MIN) ? INT32_MAX : -in);
#endif
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q31_t)__QSUB(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == INT32_MIN) ? INT32_MAX : -in);
#endif
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q31_t)__QSUB(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == INT32_MIN) ? INT32_MAX : -in);
#endif
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q31_t)__QSUB(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == INT32_MIN) ? INT32_MAX : -in);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = |A| */
/* Calculate absolute of input (if -1 then saturated to 0x7fffffff) and store result in destination buffer. */
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q31_t)__QSUB(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == INT32_MIN) ? INT32_MAX : -in);
#endif
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicAbs group
*/

134
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_abs_q7.c Normal file
View File

@@ -0,0 +1,134 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_abs_q7.c
* Description: Q7 vector absolute value
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicAbs
@{
*/
/**
@brief Q7 vector absolute value.
@param[in] pSrc points to the input vector
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Conditions for optimum performance
Input and output buffers should be aligned by 32-bit
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
The Q7 value -1 (0x80) will be saturated to the maximum allowable positive value 0x7F.
*/
void arm_abs_q7(
const q7_t * pSrc,
q7_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
q7_t in; /* Temporary input variable */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = |A| */
/* Calculate absolute of input (if -1 then saturated to 0x7f) and store result in destination buffer. */
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q7_t)__QSUB(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == (q7_t) 0x80) ? (q7_t) 0x7f : -in);
#endif
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q7_t)__QSUB(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == (q7_t) 0x80) ? (q7_t) 0x7f : -in);
#endif
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q7_t)__QSUB(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == (q7_t) 0x80) ? (q7_t) 0x7f : -in);
#endif
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q7_t)__QSUB(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == (q7_t) 0x80) ? (q7_t) 0x7f : -in);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = |A| */
/* Calculate absolute of input (if -1 then saturated to 0x7f) and store result in destination buffer. */
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (in > 0) ? in : (q7_t) __QSUB(0, in);
#else
*pDst++ = (in > 0) ? in : ((in == (q7_t) 0x80) ? (q7_t) 0x7f : -in);
#endif
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicAbs group
*/

145
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_add_f32.c Normal file
View File

@@ -0,0 +1,145 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_add_f32.c
* Description: Floating-point vector addition
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@defgroup BasicAdd Vector Addition
Element-by-element addition of two vectors.
<pre>
pDst[n] = pSrcA[n] + pSrcB[n], 0 <= n < blockSize.
</pre>
There are separate functions for floating-point, Q7, Q15, and Q31 data types.
*/
/**
@addtogroup BasicAdd
@{
*/
/**
@brief Floating-point vector addition.
@param[in] pSrcA points to first input vector
@param[in] pSrcB points to second input vector
@param[out] pDst points to output vector
@param[in] blockSize number of samples in each vector
@return none
*/
void arm_add_f32(
const float32_t * pSrcA,
const float32_t * pSrcB,
float32_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined(ARM_MATH_NEON)
float32x4_t vec1;
float32x4_t vec2;
float32x4_t res;
/* Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A + B */
/* Add and then store the results in the destination buffer. */
vec1 = vld1q_f32(pSrcA);
vec2 = vld1q_f32(pSrcB);
res = vaddq_f32(vec1, vec2);
vst1q_f32(pDst, res);
/* Increment pointers */
pSrcA += 4;
pSrcB += 4;
pDst += 4;
/* Decrement the loop counter */
blkCnt--;
}
/* Tail */
blkCnt = blockSize & 0x3;
#else
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A + B */
/* Add and store result in destination buffer. */
*pDst++ = (*pSrcA++) + (*pSrcB++);
*pDst++ = (*pSrcA++) + (*pSrcB++);
*pDst++ = (*pSrcA++) + (*pSrcB++);
*pDst++ = (*pSrcA++) + (*pSrcB++);
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
#endif /* #if defined(ARM_MATH_NEON) */
while (blkCnt > 0U)
{
/* C = A + B */
/* Add and store result in destination buffer. */
*pDst++ = (*pSrcA++) + (*pSrcB++);
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicAdd group
*/

126
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_add_q15.c Normal file
View File

@@ -0,0 +1,126 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_add_q15.c
* Description: Q15 vector addition
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicAdd
@{
*/
/**
@brief Q15 vector addition.
@param[in] pSrcA points to the first input vector
@param[in] pSrcB points to the second input vector
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q15 range [0x8000 0x7FFF] are saturated.
*/
void arm_add_q15(
const q15_t * pSrcA,
const q15_t * pSrcB,
q15_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined (ARM_MATH_LOOPUNROLL)
#if defined (ARM_MATH_DSP)
q31_t inA1, inA2;
q31_t inB1, inB2;
#endif
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A + B */
#if defined (ARM_MATH_DSP)
/* read 2 times 2 samples at a time from sourceA */
inA1 = read_q15x2_ia ((q15_t **) &pSrcA);
inA2 = read_q15x2_ia ((q15_t **) &pSrcA);
/* read 2 times 2 samples at a time from sourceB */
inB1 = read_q15x2_ia ((q15_t **) &pSrcB);
inB2 = read_q15x2_ia ((q15_t **) &pSrcB);
/* Add and store 2 times 2 samples at a time */
write_q15x2_ia (&pDst, __QADD16(inA1, inB1));
write_q15x2_ia (&pDst, __QADD16(inA2, inB2));
#else
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrcA++ + *pSrcB++), 16);
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrcA++ + *pSrcB++), 16);
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrcA++ + *pSrcB++), 16);
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrcA++ + *pSrcB++), 16);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A + B */
/* Add and store result in destination buffer. */
#if defined (ARM_MATH_DSP)
*pDst++ = (q15_t) __QADD16(*pSrcA++, *pSrcB++);
#else
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrcA++ + *pSrcB++), 16);
#endif
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicAdd group
*/

108
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_add_q31.c Normal file
View File

@@ -0,0 +1,108 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_add_q31.c
* Description: Q31 vector addition
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicAdd
@{
*/
/**
@brief Q31 vector addition.
@param[in] pSrcA points to the first input vector
@param[in] pSrcB points to the second input vector
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q31 range [0x80000000 0x7FFFFFFF] are saturated.
*/
void arm_add_q31(
const q31_t * pSrcA,
const q31_t * pSrcB,
q31_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A + B */
/* Add and store result in destination buffer. */
*pDst++ = __QADD(*pSrcA++, *pSrcB++);
*pDst++ = __QADD(*pSrcA++, *pSrcB++);
*pDst++ = __QADD(*pSrcA++, *pSrcB++);
*pDst++ = __QADD(*pSrcA++, *pSrcB++);
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A + B */
/* Add and store result in destination buffer. */
*pDst++ = __QADD(*pSrcA++, *pSrcB++);
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicAdd group
*/

109
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_add_q7.c Normal file
View File

@@ -0,0 +1,109 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_add_q7.c
* Description: Q7 vector addition
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicAdd
@{
*/
/**
@brief Q7 vector addition.
@param[in] pSrcA points to the first input vector
@param[in] pSrcB points to the second input vector
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q7 range [0x80 0x7F] are saturated.
*/
void arm_add_q7(
const q7_t * pSrcA,
const q7_t * pSrcB,
q7_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A + B */
#if defined (ARM_MATH_DSP)
/* Add and store result in destination buffer (4 samples at a time). */
write_q7x4_ia (&pDst, __QADD8 (read_q7x4_ia ((q7_t **) &pSrcA), read_q7x4_ia ((q7_t **) &pSrcB)));
#else
*pDst++ = (q7_t) __SSAT ((q15_t) *pSrcA++ + *pSrcB++, 8);
*pDst++ = (q7_t) __SSAT ((q15_t) *pSrcA++ + *pSrcB++, 8);
*pDst++ = (q7_t) __SSAT ((q15_t) *pSrcA++ + *pSrcB++, 8);
*pDst++ = (q7_t) __SSAT ((q15_t) *pSrcA++ + *pSrcB++, 8);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A + B */
/* Add and store result in destination buffer. */
*pDst++ = (q7_t) __SSAT((q15_t) *pSrcA++ + *pSrcB++, 8);
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicAdd group
*/

163
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_dot_prod_f32.c Normal file
View File

@@ -0,0 +1,163 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_dot_prod_f32.c
* Description: Floating-point dot product
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@defgroup BasicDotProd Vector Dot Product
Computes the dot product of two vectors.
The vectors are multiplied element-by-element and then summed.
<pre>
sum = pSrcA[0]*pSrcB[0] + pSrcA[1]*pSrcB[1] + ... + pSrcA[blockSize-1]*pSrcB[blockSize-1]
</pre>
There are separate functions for floating-point, Q7, Q15, and Q31 data types.
*/
/**
@addtogroup BasicDotProd
@{
*/
/**
@brief Dot product of floating-point vectors.
@param[in] pSrcA points to the first input vector.
@param[in] pSrcB points to the second input vector.
@param[in] blockSize number of samples in each vector.
@param[out] result output result returned here.
@return none
*/
void arm_dot_prod_f32(
const float32_t * pSrcA,
const float32_t * pSrcB,
uint32_t blockSize,
float32_t * result)
{
uint32_t blkCnt; /* Loop counter */
float32_t sum = 0.0f; /* Temporary return variable */
#if defined(ARM_MATH_NEON)
float32x4_t vec1;
float32x4_t vec2;
float32x4_t res;
float32x4_t accum = vdupq_n_f32(0);
/* Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
vec1 = vld1q_f32(pSrcA);
vec2 = vld1q_f32(pSrcB);
while (blkCnt > 0U)
{
/* C = A[0]*B[0] + A[1]*B[1] + A[2]*B[2] + ... + A[blockSize-1]*B[blockSize-1] */
/* Calculate dot product and then store the result in a temporary buffer. */
accum = vmlaq_f32(accum, vec1, vec2);
/* Increment pointers */
pSrcA += 4;
pSrcB += 4;
vec1 = vld1q_f32(pSrcA);
vec2 = vld1q_f32(pSrcB);
/* Decrement the loop counter */
blkCnt--;
}
#if __aarch64__
sum = vpadds_f32(vpadd_f32(vget_low_f32(accum), vget_high_f32(accum)));
#else
sum = (vpadd_f32(vget_low_f32(accum), vget_high_f32(accum)))[0] + (vpadd_f32(vget_low_f32(accum), vget_high_f32(accum)))[1];
#endif
/* Tail */
blkCnt = blockSize & 0x3;
#else
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
/* First part of the processing with loop unrolling. Compute 4 outputs at a time.
** a second loop below computes the remaining 1 to 3 samples. */
while (blkCnt > 0U)
{
/* C = A[0]* B[0] + A[1]* B[1] + A[2]* B[2] + .....+ A[blockSize-1]* B[blockSize-1] */
/* Calculate dot product and store result in a temporary buffer. */
sum += (*pSrcA++) * (*pSrcB++);
sum += (*pSrcA++) * (*pSrcB++);
sum += (*pSrcA++) * (*pSrcB++);
sum += (*pSrcA++) * (*pSrcB++);
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
#endif /* #if defined(ARM_MATH_NEON) */
while (blkCnt > 0U)
{
/* C = A[0]* B[0] + A[1]* B[1] + A[2]* B[2] + .....+ A[blockSize-1]* B[blockSize-1] */
/* Calculate dot product and store result in a temporary buffer. */
sum += (*pSrcA++) * (*pSrcB++);
/* Decrement loop counter */
blkCnt--;
}
/* Store result in destination buffer */
*result = sum;
}
/**
@} end of BasicDotProd group
*/

120
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_dot_prod_q15.c Normal file
View File

@@ -0,0 +1,120 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_dot_prod_q15.c
* Description: Q15 dot product
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicDotProd
@{
*/
/**
@brief Dot product of Q15 vectors.
@param[in] pSrcA points to the first input vector
@param[in] pSrcB points to the second input vector
@param[in] blockSize number of samples in each vector
@param[out] result output result returned here
@return none
@par Scaling and Overflow Behavior
The intermediate multiplications are in 1.15 x 1.15 = 2.30 format and these
results are added to a 64-bit accumulator in 34.30 format.
Nonsaturating additions are used and given that there are 33 guard bits in the accumulator
there is no risk of overflow.
The return result is in 34.30 format.
*/
void arm_dot_prod_q15(
const q15_t * pSrcA,
const q15_t * pSrcB,
uint32_t blockSize,
q63_t * result)
{
uint32_t blkCnt; /* Loop counter */
q63_t sum = 0; /* Temporary return variable */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A[0]* B[0] + A[1]* B[1] + A[2]* B[2] + .....+ A[blockSize-1]* B[blockSize-1] */
#if defined (ARM_MATH_DSP)
/* Calculate dot product and store result in a temporary buffer. */
sum = __SMLALD(read_q15x2_ia ((q15_t **) &pSrcA), read_q15x2_ia ((q15_t **) &pSrcB), sum);
sum = __SMLALD(read_q15x2_ia ((q15_t **) &pSrcA), read_q15x2_ia ((q15_t **) &pSrcB), sum);
#else
sum += (q63_t)((q31_t) *pSrcA++ * *pSrcB++);
sum += (q63_t)((q31_t) *pSrcA++ * *pSrcB++);
sum += (q63_t)((q31_t) *pSrcA++ * *pSrcB++);
sum += (q63_t)((q31_t) *pSrcA++ * *pSrcB++);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A[0]* B[0] + A[1]* B[1] + A[2]* B[2] + .....+ A[blockSize-1]* B[blockSize-1] */
/* Calculate dot product and store result in a temporary buffer. */
//#if defined (ARM_MATH_DSP)
// sum = __SMLALD(*pSrcA++, *pSrcB++, sum);
//#else
sum += (q63_t)((q31_t) *pSrcA++ * *pSrcB++);
//#endif
/* Decrement loop counter */
blkCnt--;
}
/* Store result in destination buffer in 34.30 format */
*result = sum;
}
/**
@} end of BasicDotProd group
*/

115
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_dot_prod_q31.c Normal file
View File

@@ -0,0 +1,115 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_dot_prod_q31.c
* Description: Q31 dot product
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicDotProd
@{
*/
/**
@brief Dot product of Q31 vectors.
@param[in] pSrcA points to the first input vector.
@param[in] pSrcB points to the second input vector.
@param[in] blockSize number of samples in each vector.
@param[out] result output result returned here.
@return none
@par Scaling and Overflow Behavior
The intermediate multiplications are in 1.31 x 1.31 = 2.62 format and these
are truncated to 2.48 format by discarding the lower 14 bits.
The 2.48 result is then added without saturation to a 64-bit accumulator in 16.48 format.
There are 15 guard bits in the accumulator and there is no risk of overflow as long as
the length of the vectors is less than 2^16 elements.
The return result is in 16.48 format.
*/
void arm_dot_prod_q31(
const q31_t * pSrcA,
const q31_t * pSrcB,
uint32_t blockSize,
q63_t * result)
{
uint32_t blkCnt; /* Loop counter */
q63_t sum = 0; /* Temporary return variable */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A[0]* B[0] + A[1]* B[1] + A[2]* B[2] + .....+ A[blockSize-1]* B[blockSize-1] */
/* Calculate dot product and store result in a temporary buffer. */
sum += ((q63_t) *pSrcA++ * *pSrcB++) >> 14U;
sum += ((q63_t) *pSrcA++ * *pSrcB++) >> 14U;
sum += ((q63_t) *pSrcA++ * *pSrcB++) >> 14U;
sum += ((q63_t) *pSrcA++ * *pSrcB++) >> 14U;
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A[0]* B[0] + A[1]* B[1] + A[2]* B[2] + .....+ A[blockSize-1]* B[blockSize-1] */
/* Calculate dot product and store result in a temporary buffer. */
sum += ((q63_t) *pSrcA++ * *pSrcB++) >> 14U;
/* Decrement loop counter */
blkCnt--;
}
/* Store result in destination buffer in 16.48 format */
*result = sum;
}
/**
@} end of BasicDotProd group
*/

139
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_dot_prod_q7.c Normal file
View File

@@ -0,0 +1,139 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_dot_prod_q7.c
* Description: Q7 dot product
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicDotProd
@{
*/
/**
@brief Dot product of Q7 vectors.
@param[in] pSrcA points to the first input vector
@param[in] pSrcB points to the second input vector
@param[in] blockSize number of samples in each vector
@param[out] result output result returned here
@return none
@par Scaling and Overflow Behavior
The intermediate multiplications are in 1.7 x 1.7 = 2.14 format and these
results are added to an accumulator in 18.14 format.
Nonsaturating additions are used and there is no danger of wrap around as long as
the vectors are less than 2^18 elements long.
The return result is in 18.14 format.
*/
void arm_dot_prod_q7(
const q7_t * pSrcA,
const q7_t * pSrcB,
uint32_t blockSize,
q31_t * result)
{
uint32_t blkCnt; /* Loop counter */
q31_t sum = 0; /* Temporary return variable */
#if defined (ARM_MATH_LOOPUNROLL)
#if defined (ARM_MATH_DSP)
q31_t input1, input2; /* Temporary variables */
q31_t inA1, inA2, inB1, inB2; /* Temporary variables */
#endif
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A[0]* B[0] + A[1]* B[1] + A[2]* B[2] + .....+ A[blockSize-1]* B[blockSize-1] */
#if defined (ARM_MATH_DSP)
/* read 4 samples at a time from sourceA */
input1 = read_q7x4_ia ((q7_t **) &pSrcA);
/* read 4 samples at a time from sourceB */
input2 = read_q7x4_ia ((q7_t **) &pSrcB);
/* extract two q7_t samples to q15_t samples */
inA1 = __SXTB16(__ROR(input1, 8));
/* extract reminaing two samples */
inA2 = __SXTB16(input1);
/* extract two q7_t samples to q15_t samples */
inB1 = __SXTB16(__ROR(input2, 8));
/* extract reminaing two samples */
inB2 = __SXTB16(input2);
/* multiply and accumulate two samples at a time */
sum = __SMLAD(inA1, inB1, sum);
sum = __SMLAD(inA2, inB2, sum);
#else
sum += (q31_t) ((q15_t) *pSrcA++ * *pSrcB++);
sum += (q31_t) ((q15_t) *pSrcA++ * *pSrcB++);
sum += (q31_t) ((q15_t) *pSrcA++ * *pSrcB++);
sum += (q31_t) ((q15_t) *pSrcA++ * *pSrcB++);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A[0]* B[0] + A[1]* B[1] + A[2]* B[2] + .....+ A[blockSize-1]* B[blockSize-1] */
/* Calculate dot product and store result in a temporary buffer. */
//#if defined (ARM_MATH_DSP)
// sum = __SMLAD(*pSrcA++, *pSrcB++, sum);
//#else
sum += (q31_t) ((q15_t) *pSrcA++ * *pSrcB++);
//#endif
/* Decrement loop counter */
blkCnt--;
}
/* Store result in destination buffer in 18.14 format */
*result = sum;
}
/**
@} end of BasicDotProd group
*/

148
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_mult_f32.c Normal file
View File

@@ -0,0 +1,148 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_mult_f32.c
* Description: Floating-point vector multiplication
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@defgroup BasicMult Vector Multiplication
Element-by-element multiplication of two vectors.
<pre>
pDst[n] = pSrcA[n] * pSrcB[n], 0 <= n < blockSize.
</pre>
There are separate functions for floating-point, Q7, Q15, and Q31 data types.
*/
/**
@addtogroup BasicMult
@{
*/
/**
@brief Floating-point vector multiplication.
@param[in] pSrcA points to the first input vector.
@param[in] pSrcB points to the second input vector.
@param[out] pDst points to the output vector.
@param[in] blockSize number of samples in each vector.
@return none
*/
void arm_mult_f32(
const float32_t * pSrcA,
const float32_t * pSrcB,
float32_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined(ARM_MATH_NEON)
float32x4_t vec1;
float32x4_t vec2;
float32x4_t res;
/* Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A * B */
/* Multiply the inputs and then store the results in the destination buffer. */
vec1 = vld1q_f32(pSrcA);
vec2 = vld1q_f32(pSrcB);
res = vmulq_f32(vec1, vec2);
vst1q_f32(pDst, res);
/* Increment pointers */
pSrcA += 4;
pSrcB += 4;
pDst += 4;
/* Decrement the loop counter */
blkCnt--;
}
/* Tail */
blkCnt = blockSize & 0x3;
#else
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A * B */
/* Multiply inputs and store result in destination buffer. */
*pDst++ = (*pSrcA++) * (*pSrcB++);
*pDst++ = (*pSrcA++) * (*pSrcB++);
*pDst++ = (*pSrcA++) * (*pSrcB++);
*pDst++ = (*pSrcA++) * (*pSrcB++);
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
#endif /* #if defined(ARM_MATH_NEON) */
while (blkCnt > 0U)
{
/* C = A * B */
/* Multiply input and store result in destination buffer. */
*pDst++ = (*pSrcA++) * (*pSrcB++);
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicMult group
*/

143
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_mult_q15.c Normal file
View File

@@ -0,0 +1,143 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_mult_q15.c
* Description: Q15 vector multiplication
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicMult
@{
*/
/**
@brief Q15 vector multiplication
@param[in] pSrcA points to first input vector
@param[in] pSrcB points to second input vector
@param[out] pDst points to output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q15 range [0x8000 0x7FFF] are saturated.
*/
void arm_mult_q15(
const q15_t * pSrcA,
const q15_t * pSrcB,
q15_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined (ARM_MATH_LOOPUNROLL)
#if defined (ARM_MATH_DSP)
q31_t inA1, inA2, inB1, inB2; /* Temporary input variables */
q15_t out1, out2, out3, out4; /* Temporary output variables */
q31_t mul1, mul2, mul3, mul4; /* Temporary variables */
#endif
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A * B */
#if defined (ARM_MATH_DSP)
/* read 2 samples at a time from sourceA */
inA1 = read_q15x2_ia ((q15_t **) &pSrcA);
/* read 2 samples at a time from sourceB */
inB1 = read_q15x2_ia ((q15_t **) &pSrcB);
/* read 2 samples at a time from sourceA */
inA2 = read_q15x2_ia ((q15_t **) &pSrcA);
/* read 2 samples at a time from sourceB */
inB2 = read_q15x2_ia ((q15_t **) &pSrcB);
/* multiply mul = sourceA * sourceB */
mul1 = (q31_t) ((q15_t) (inA1 >> 16) * (q15_t) (inB1 >> 16));
mul2 = (q31_t) ((q15_t) (inA1 ) * (q15_t) (inB1 ));
mul3 = (q31_t) ((q15_t) (inA2 >> 16) * (q15_t) (inB2 >> 16));
mul4 = (q31_t) ((q15_t) (inA2 ) * (q15_t) (inB2 ));
/* saturate result to 16 bit */
out1 = (q15_t) __SSAT(mul1 >> 15, 16);
out2 = (q15_t) __SSAT(mul2 >> 15, 16);
out3 = (q15_t) __SSAT(mul3 >> 15, 16);
out4 = (q15_t) __SSAT(mul4 >> 15, 16);
/* store result to destination */
#ifndef ARM_MATH_BIG_ENDIAN
write_q15x2_ia (&pDst, __PKHBT(out2, out1, 16));
write_q15x2_ia (&pDst, __PKHBT(out4, out3, 16));
#else
write_q15x2_ia (&pDst, __PKHBT(out1, out2, 16));
write_q15x2_ia (&pDst, __PKHBT(out3, out4, 16));
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#else
*pDst++ = (q15_t) __SSAT((((q31_t) (*pSrcA++) * (*pSrcB++)) >> 15), 16);
*pDst++ = (q15_t) __SSAT((((q31_t) (*pSrcA++) * (*pSrcB++)) >> 15), 16);
*pDst++ = (q15_t) __SSAT((((q31_t) (*pSrcA++) * (*pSrcB++)) >> 15), 16);
*pDst++ = (q15_t) __SSAT((((q31_t) (*pSrcA++) * (*pSrcB++)) >> 15), 16);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A * B */
/* Multiply inputs and store result in destination buffer. */
*pDst++ = (q15_t) __SSAT((((q31_t) (*pSrcA++) * (*pSrcB++)) >> 15), 16);
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicMult group
*/

119
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_mult_q31.c Normal file
View File

@@ -0,0 +1,119 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_mult_q31.c
* Description: Q31 vector multiplication
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicMult
@{
*/
/**
@brief Q31 vector multiplication.
@param[in] pSrcA points to the first input vector.
@param[in] pSrcB points to the second input vector.
@param[out] pDst points to the output vector.
@param[in] blockSize number of samples in each vector.
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q31 range[0x80000000 0x7FFFFFFF] are saturated.
*/
void arm_mult_q31(
const q31_t * pSrcA,
const q31_t * pSrcB,
q31_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
q31_t out; /* Temporary output variable */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A * B */
/* Multiply inputs and store result in destination buffer. */
out = ((q63_t) *pSrcA++ * *pSrcB++) >> 32;
out = __SSAT(out, 31);
*pDst++ = out << 1U;
out = ((q63_t) *pSrcA++ * *pSrcB++) >> 32;
out = __SSAT(out, 31);
*pDst++ = out << 1U;
out = ((q63_t) *pSrcA++ * *pSrcB++) >> 32;
out = __SSAT(out, 31);
*pDst++ = out << 1U;
out = ((q63_t) *pSrcA++ * *pSrcB++) >> 32;
out = __SSAT(out, 31);
*pDst++ = out << 1U;
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A * B */
/* Multiply inputs and store result in destination buffer. */
out = ((q63_t) *pSrcA++ * *pSrcB++) >> 32;
out = __SSAT(out, 31);
*pDst++ = out << 1U;
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicMult group
*/

119
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_mult_q7.c Normal file
View File

@@ -0,0 +1,119 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_mult_q7.c
* Description: Q7 vector multiplication
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicMult
@{
*/
/**
@brief Q7 vector multiplication
@param[in] pSrcA points to the first input vector
@param[in] pSrcB points to the second input vector
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q7 range [0x80 0x7F] are saturated.
*/
void arm_mult_q7(
const q7_t * pSrcA,
const q7_t * pSrcB,
q7_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined (ARM_MATH_LOOPUNROLL)
#if defined (ARM_MATH_DSP)
q7_t out1, out2, out3, out4; /* Temporary output variables */
#endif
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A * B */
#if defined (ARM_MATH_DSP)
/* Multiply inputs and store results in temporary variables */
out1 = (q7_t) __SSAT((((q15_t) (*pSrcA++) * (*pSrcB++)) >> 7), 8);
out2 = (q7_t) __SSAT((((q15_t) (*pSrcA++) * (*pSrcB++)) >> 7), 8);
out3 = (q7_t) __SSAT((((q15_t) (*pSrcA++) * (*pSrcB++)) >> 7), 8);
out4 = (q7_t) __SSAT((((q15_t) (*pSrcA++) * (*pSrcB++)) >> 7), 8);
/* Pack and store result in destination buffer (in single write) */
write_q7x4_ia (&pDst, __PACKq7(out1, out2, out3, out4));
#else
*pDst++ = (q7_t) __SSAT((((q15_t) (*pSrcA++) * (*pSrcB++)) >> 7), 8);
*pDst++ = (q7_t) __SSAT((((q15_t) (*pSrcA++) * (*pSrcB++)) >> 7), 8);
*pDst++ = (q7_t) __SSAT((((q15_t) (*pSrcA++) * (*pSrcB++)) >> 7), 8);
*pDst++ = (q7_t) __SSAT((((q15_t) (*pSrcA++) * (*pSrcB++)) >> 7), 8);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A * B */
/* Multiply input and store result in destination buffer. */
*pDst++ = (q7_t) __SSAT((((q15_t) (*pSrcA++) * (*pSrcB++)) >> 7), 8);
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicMult group
*/

145
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_negate_f32.c Normal file
View File

@@ -0,0 +1,145 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_negate_f32.c
* Description: Negates floating-point vectors
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@defgroup BasicNegate Vector Negate
Negates the elements of a vector.
<pre>
pDst[n] = -pSrc[n], 0 <= n < blockSize.
</pre>
The functions support in-place computation allowing the source and
destination pointers to reference the same memory buffer.
There are separate functions for floating-point, Q7, Q15, and Q31 data types.
*/
/**
@addtogroup BasicNegate
@{
*/
/**
@brief Negates the elements of a floating-point vector.
@param[in] pSrc points to input vector.
@param[out] pDst points to output vector.
@param[in] blockSize number of samples in each vector.
@return none
*/
void arm_negate_f32(
const float32_t * pSrc,
float32_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined(ARM_MATH_NEON_EXPERIMENTAL)
float32x4_t vec1;
float32x4_t res;
/* Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = -A */
/* Negate and then store the results in the destination buffer. */
vec1 = vld1q_f32(pSrc);
res = vnegq_f32(vec1);
vst1q_f32(pDst, res);
/* Increment pointers */
pSrc += 4;
pDst += 4;
/* Decrement the loop counter */
blkCnt--;
}
/* Tail */
blkCnt = blockSize & 0x3;
#else
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = -A */
/* Negate and store result in destination buffer. */
*pDst++ = -*pSrc++;
*pDst++ = -*pSrc++;
*pDst++ = -*pSrc++;
*pDst++ = -*pSrc++;
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
#endif /* #if defined(ARM_MATH_NEON_EXPERIMENTAL) */
while (blkCnt > 0U)
{
/* C = -A */
/* Negate and store result in destination buffer. */
*pDst++ = -*pSrc++;
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicNegate group
*/

126
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_negate_q15.c Normal file
View File

@@ -0,0 +1,126 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_negate_q15.c
* Description: Negates Q15 vectors
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicNegate
@{
*/
/**
@brief Negates the elements of a Q15 vector.
@param[in] pSrc points to the input vector.
@param[out] pDst points to the output vector.
@param[in] blockSize number of samples in each vector.
@return none
@par Conditions for optimum performance
Input and output buffers should be aligned by 32-bit
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
The Q15 value -1 (0x8000) is saturated to the maximum allowable positive value 0x7FFF.
*/
void arm_negate_q15(
const q15_t * pSrc,
q15_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
q15_t in; /* Temporary input variable */
#if defined (ARM_MATH_LOOPUNROLL)
#if defined (ARM_MATH_DSP)
q31_t in1; /* Temporary input variables */
#endif
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = -A */
#if defined (ARM_MATH_DSP)
/* Negate and store result in destination buffer (2 samples at a time). */
in1 = read_q15x2_ia ((q15_t **) &pSrc);
write_q15x2_ia (&pDst, __QSUB16(0, in1));
in1 = read_q15x2_ia ((q15_t **) &pSrc);
write_q15x2_ia (&pDst, __QSUB16(0, in1));
#else
in = *pSrc++;
*pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
in = *pSrc++;
*pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
in = *pSrc++;
*pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
in = *pSrc++;
*pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = -A */
/* Negate and store result in destination buffer. */
in = *pSrc++;
*pDst++ = (in == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in;
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicNegate group
*/

132
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_negate_q31.c Normal file
View File

@@ -0,0 +1,132 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_negate_q31.c
* Description: Negates Q31 vectors
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicNegate
@{
*/
/**
@brief Negates the elements of a Q31 vector.
@param[in] pSrc points to the input vector.
@param[out] pDst points to the output vector.
@param[in] blockSize number of samples in each vector.
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
The Q31 value -1 (0x80000000) is saturated to the maximum allowable positive value 0x7FFFFFFF.
*/
void arm_negate_q31(
const q31_t * pSrc,
q31_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
q31_t in; /* Temporary input variable */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = -A */
/* Negate and store result in destination buffer. */
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = __QSUB(0, in);
#else
*pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
#endif
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = __QSUB(0, in);
#else
*pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
#endif
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = __QSUB(0, in);
#else
*pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
#endif
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = __QSUB(0, in);
#else
*pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = -A */
/* Negate and store result in destination buffer. */
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = __QSUB(0, in);
#else
*pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
#endif
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicNegate group
*/

126
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_negate_q7.c Normal file
View File

@@ -0,0 +1,126 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_negate_q7.c
* Description: Negates Q7 vectors
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicNegate
@{
*/
/**
@brief Negates the elements of a Q7 vector.
@param[in] pSrc points to the input vector.
@param[out] pDst points to the output vector.
@param[in] blockSize number of samples in each vector.
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
The Q7 value -1 (0x80) is saturated to the maximum allowable positive value 0x7F.
*/
void arm_negate_q7(
const q7_t * pSrc,
q7_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
q7_t in; /* Temporary input variable */
#if defined (ARM_MATH_LOOPUNROLL)
#if defined (ARM_MATH_DSP)
q31_t in1; /* Temporary input variable */
#endif
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = -A */
#if defined (ARM_MATH_DSP)
/* Negate and store result in destination buffer (4 samples at a time). */
in1 = read_q7x4_ia ((q7_t **) &pSrc);
write_q7x4_ia (&pDst, __QSUB8(0, in1));
#else
in = *pSrc++;
*pDst++ = (in == (q7_t) 0x80) ? (q7_t) 0x7f : -in;
in = *pSrc++;
*pDst++ = (in == (q7_t) 0x80) ? (q7_t) 0x7f : -in;
in = *pSrc++;
*pDst++ = (in == (q7_t) 0x80) ? (q7_t) 0x7f : -in;
in = *pSrc++;
*pDst++ = (in == (q7_t) 0x80) ? (q7_t) 0x7f : -in;
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = -A */
/* Negate and store result in destination buffer. */
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = (q7_t) __QSUB(0, in);
#else
*pDst++ = (in == (q7_t) 0x80) ? (q7_t) 0x7f : -in;
#endif
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicNegate group
*/

147
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_offset_f32.c Normal file
View File

@@ -0,0 +1,147 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_offset_f32.c
* Description: Floating-point vector offset
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@defgroup BasicOffset Vector Offset
Adds a constant offset to each element of a vector.
<pre>
pDst[n] = pSrc[n] + offset, 0 <= n < blockSize.
</pre>
The functions support in-place computation allowing the source and
destination pointers to reference the same memory buffer.
There are separate functions for floating-point, Q7, Q15, and Q31 data types.
*/
/**
@addtogroup BasicOffset
@{
*/
/**
@brief Adds a constant offset to a floating-point vector.
@param[in] pSrc points to the input vector
@param[in] offset is the offset to be added
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
*/
void arm_offset_f32(
const float32_t * pSrc,
float32_t offset,
float32_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined(ARM_MATH_NEON_EXPERIMENTAL)
float32x4_t vec1;
float32x4_t res;
/* Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A + offset */
/* Add offset and then store the results in the destination buffer. */
vec1 = vld1q_f32(pSrc);
res = vaddq_f32(vec1,vdupq_n_f32(offset));
vst1q_f32(pDst, res);
/* Increment pointers */
pSrc += 4;
pDst += 4;
/* Decrement the loop counter */
blkCnt--;
}
/* Tail */
blkCnt = blockSize & 0x3;
#else
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A + offset */
/* Add offset and store result in destination buffer. */
*pDst++ = (*pSrc++) + offset;
*pDst++ = (*pSrc++) + offset;
*pDst++ = (*pSrc++) + offset;
*pDst++ = (*pSrc++) + offset;
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
#endif /* #if defined(ARM_MATH_NEON_EXPERIMENTAL) */
while (blkCnt > 0U)
{
/* C = A + offset */
/* Add offset and store result in destination buffer. */
*pDst++ = (*pSrc++) + offset;
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicOffset group
*/

121
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_offset_q15.c Normal file
View File

@@ -0,0 +1,121 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_offset_q15.c
* Description: Q15 vector offset
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicOffset
@{
*/
/**
@brief Adds a constant offset to a Q15 vector.
@param[in] pSrc points to the input vector
@param[in] offset is the offset to be added
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q15 range [0x8000 0x7FFF] are saturated.
*/
void arm_offset_q15(
const q15_t * pSrc,
q15_t offset,
q15_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined (ARM_MATH_LOOPUNROLL)
#if defined (ARM_MATH_DSP)
q31_t offset_packed; /* Offset packed to 32 bit */
/* Offset is packed to 32 bit in order to use SIMD32 for addition */
offset_packed = __PKHBT(offset, offset, 16);
#endif
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A + offset */
#if defined (ARM_MATH_DSP)
/* Add offset and store result in destination buffer (2 samples at a time). */
write_q15x2_ia (&pDst, __QADD16(read_q15x2_ia ((q15_t **) &pSrc), offset_packed));
write_q15x2_ia (&pDst, __QADD16(read_q15x2_ia ((q15_t **) &pSrc), offset_packed));
#else
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrc++ + offset), 16);
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrc++ + offset), 16);
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrc++ + offset), 16);
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrc++ + offset), 16);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A + offset */
/* Add offset and store result in destination buffer. */
#if defined (ARM_MATH_DSP)
*pDst++ = (q15_t) __QADD16(*pSrc++, offset);
#else
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrc++ + offset), 16);
#endif
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicOffset group
*/

128
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_offset_q31.c Normal file
View File

@@ -0,0 +1,128 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_offset_q31.c
* Description: Q31 vector offset
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicOffset
@{
*/
/**
@brief Adds a constant offset to a Q31 vector.
@param[in] pSrc points to the input vector
@param[in] offset is the offset to be added
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q31 range [0x80000000 0x7FFFFFFF] are saturated.
*/
void arm_offset_q31(
const q31_t * pSrc,
q31_t offset,
q31_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A + offset */
/* Add offset and store result in destination buffer. */
#if defined (ARM_MATH_DSP)
*pDst++ = __QADD(*pSrc++, offset);
#else
*pDst++ = (q31_t) clip_q63_to_q31((q63_t) * pSrc++ + offset);
#endif
#if defined (ARM_MATH_DSP)
*pDst++ = __QADD(*pSrc++, offset);
#else
*pDst++ = (q31_t) clip_q63_to_q31((q63_t) * pSrc++ + offset);
#endif
#if defined (ARM_MATH_DSP)
*pDst++ = __QADD(*pSrc++, offset);
#else
*pDst++ = (q31_t) clip_q63_to_q31((q63_t) * pSrc++ + offset);
#endif
#if defined (ARM_MATH_DSP)
*pDst++ = __QADD(*pSrc++, offset);
#else
*pDst++ = (q31_t) clip_q63_to_q31((q63_t) * pSrc++ + offset);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A + offset */
/* Add offset and store result in destination buffer. */
#if defined (ARM_MATH_DSP)
*pDst++ = __QADD(*pSrc++, offset);
#else
*pDst++ = (q31_t) clip_q63_to_q31((q63_t) * pSrc++ + offset);
#endif
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicOffset group
*/

116
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_offset_q7.c Normal file
View File

@@ -0,0 +1,116 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_offset_q7.c
* Description: Q7 vector offset
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicOffset
@{
*/
/**
@brief Adds a constant offset to a Q7 vector.
@param[in] pSrc points to the input vector
@param[in] offset is the offset to be added
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q7 range [0x80 0x7F] are saturated.
*/
void arm_offset_q7(
const q7_t * pSrc,
q7_t offset,
q7_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined (ARM_MATH_LOOPUNROLL)
#if defined (ARM_MATH_DSP)
q31_t offset_packed; /* Offset packed to 32 bit */
/* Offset is packed to 32 bit in order to use SIMD32 for addition */
offset_packed = __PACKq7(offset, offset, offset, offset);
#endif
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A + offset */
#if defined (ARM_MATH_DSP)
/* Add offset and store result in destination buffer (4 samples at a time). */
write_q7x4_ia (&pDst, __QADD8(read_q7x4_ia ((q7_t **) &pSrc), offset_packed));
#else
*pDst++ = (q7_t) __SSAT(*pSrc++ + offset, 8);
*pDst++ = (q7_t) __SSAT(*pSrc++ + offset, 8);
*pDst++ = (q7_t) __SSAT(*pSrc++ + offset, 8);
*pDst++ = (q7_t) __SSAT(*pSrc++ + offset, 8);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A + offset */
/* Add offset and store result in destination buffer. */
*pDst++ = (q7_t) __SSAT((q15_t) *pSrc++ + offset, 8);
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicOffset group
*/

159
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_scale_f32.c Normal file
View File

@@ -0,0 +1,159 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_scale_f32.c
* Description: Multiplies a floating-point vector by a scalar
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@defgroup BasicScale Vector Scale
Multiply a vector by a scalar value. For floating-point data, the algorithm used is:
<pre>
pDst[n] = pSrc[n] * scale, 0 <= n < blockSize.
</pre>
In the fixed-point Q7, Q15, and Q31 functions, <code>scale</code> is represented by
a fractional multiplication <code>scaleFract</code> and an arithmetic shift <code>shift</code>.
The shift allows the gain of the scaling operation to exceed 1.0.
The algorithm used with fixed-point data is:
<pre>
pDst[n] = (pSrc[n] * scaleFract) << shift, 0 <= n < blockSize.
</pre>
The overall scale factor applied to the fixed-point data is
<pre>
scale = scaleFract * 2^shift.
</pre>
The functions support in-place computation allowing the source and destination
pointers to reference the same memory buffer.
*/
/**
@addtogroup BasicScale
@{
*/
/**
@brief Multiplies a floating-point vector by a scalar.
@param[in] pSrc points to the input vector
@param[in] scale scale factor to be applied
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
*/
void arm_scale_f32(
const float32_t *pSrc,
float32_t scale,
float32_t *pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined(ARM_MATH_NEON_EXPERIMENTAL)
float32x4_t vec1;
float32x4_t res;
/* Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A * scale */
/* Scale the input and then store the results in the destination buffer. */
vec1 = vld1q_f32(pSrc);
res = vmulq_f32(vec1, vdupq_n_f32(scale));
vst1q_f32(pDst, res);
/* Increment pointers */
pSrc += 4;
pDst += 4;
/* Decrement the loop counter */
blkCnt--;
}
/* Tail */
blkCnt = blockSize & 0x3;
#else
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A * scale */
/* Scale input and store result in destination buffer. */
*pDst++ = (*pSrc++) * scale;
*pDst++ = (*pSrc++) * scale;
*pDst++ = (*pSrc++) * scale;
*pDst++ = (*pSrc++) * scale;
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
#endif /* #if defined(ARM_MATH_NEON_EXPERIMENTAL) */
while (blkCnt > 0U)
{
/* C = A * scale */
/* Scale input and store result in destination buffer. */
*pDst++ = (*pSrc++) * scale;
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicScale group
*/

144
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_scale_q15.c Normal file
View File

@@ -0,0 +1,144 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_scale_q15.c
* Description: Multiplies a Q15 vector by a scalar
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicScale
@{
*/
/**
@brief Multiplies a Q15 vector by a scalar.
@param[in] pSrc points to the input vector
@param[in] scaleFract fractional portion of the scale value
@param[in] shift number of bits to shift the result by
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The input data <code>*pSrc</code> and <code>scaleFract</code> are in 1.15 format.
These are multiplied to yield a 2.30 intermediate result and this is shifted with saturation to 1.15 format.
*/
void arm_scale_q15(
const q15_t *pSrc,
q15_t scaleFract,
int8_t shift,
q15_t *pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
int8_t kShift = 15 - shift; /* Shift to apply after scaling */
#if defined (ARM_MATH_LOOPUNROLL)
#if defined (ARM_MATH_DSP)
q31_t inA1, inA2;
q31_t out1, out2, out3, out4; /* Temporary output variables */
q15_t in1, in2, in3, in4; /* Temporary input variables */
#endif
#endif
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A * scale */
#if defined (ARM_MATH_DSP)
/* read 2 times 2 samples at a time from source */
inA1 = read_q15x2_ia ((q15_t **) &pSrc);
inA2 = read_q15x2_ia ((q15_t **) &pSrc);
/* Scale inputs and store result in temporary variables
* in single cycle by packing the outputs */
out1 = (q31_t) ((q15_t) (inA1 >> 16) * scaleFract);
out2 = (q31_t) ((q15_t) (inA1 ) * scaleFract);
out3 = (q31_t) ((q15_t) (inA2 >> 16) * scaleFract);
out4 = (q31_t) ((q15_t) (inA2 ) * scaleFract);
/* apply shifting */
out1 = out1 >> kShift;
out2 = out2 >> kShift;
out3 = out3 >> kShift;
out4 = out4 >> kShift;
/* saturate the output */
in1 = (q15_t) (__SSAT(out1, 16));
in2 = (q15_t) (__SSAT(out2, 16));
in3 = (q15_t) (__SSAT(out3, 16));
in4 = (q15_t) (__SSAT(out4, 16));
/* store result to destination */
write_q15x2_ia (&pDst, __PKHBT(in2, in1, 16));
write_q15x2_ia (&pDst, __PKHBT(in4, in3, 16));
#else
*pDst++ = (q15_t) (__SSAT(((q31_t) *pSrc++ * scaleFract) >> kShift, 16));
*pDst++ = (q15_t) (__SSAT(((q31_t) *pSrc++ * scaleFract) >> kShift, 16));
*pDst++ = (q15_t) (__SSAT(((q31_t) *pSrc++ * scaleFract) >> kShift, 16));
*pDst++ = (q15_t) (__SSAT(((q31_t) *pSrc++ * scaleFract) >> kShift, 16));
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A * scale */
/* Scale input and store result in destination buffer. */
*pDst++ = (q15_t) (__SSAT(((q31_t) *pSrc++ * scaleFract) >> kShift, 16));
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicScale group
*/

191
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_scale_q31.c Normal file
View File

@@ -0,0 +1,191 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_scale_q31.c
* Description: Multiplies a Q31 vector by a scalar
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicScale
@{
*/
/**
@brief Multiplies a Q31 vector by a scalar.
@param[in] pSrc points to the input vector
@param[in] scaleFract fractional portion of the scale value
@param[in] shift number of bits to shift the result by
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The input data <code>*pSrc</code> and <code>scaleFract</code> are in 1.31 format.
These are multiplied to yield a 2.62 intermediate result and this is shifted with saturation to 1.31 format.
*/
void arm_scale_q31(
const q31_t *pSrc,
q31_t scaleFract,
int8_t shift,
q31_t *pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
q31_t in, out; /* Temporary variables */
int8_t kShift = shift + 1; /* Shift to apply after scaling */
int8_t sign = (kShift & 0x80);
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
if (sign == 0U)
{
while (blkCnt > 0U)
{
/* C = A * scale */
/* Scale input and store result in destination buffer. */
in = *pSrc++; /* read input from source */
in = ((q63_t) in * scaleFract) >> 32; /* multiply input with scaler value */
out = in << kShift; /* apply shifting */
if (in != (out >> kShift)) /* saturate the result */
out = 0x7FFFFFFF ^ (in >> 31);
*pDst++ = out; /* Store result destination */
in = *pSrc++;
in = ((q63_t) in * scaleFract) >> 32;
out = in << kShift;
if (in != (out >> kShift))
out = 0x7FFFFFFF ^ (in >> 31);
*pDst++ = out;
in = *pSrc++;
in = ((q63_t) in * scaleFract) >> 32;
out = in << kShift;
if (in != (out >> kShift))
out = 0x7FFFFFFF ^ (in >> 31);
*pDst++ = out;
in = *pSrc++;
in = ((q63_t) in * scaleFract) >> 32;
out = in << kShift;
if (in != (out >> kShift))
out = 0x7FFFFFFF ^ (in >> 31);
*pDst++ = out;
/* Decrement loop counter */
blkCnt--;
}
}
else
{
while (blkCnt > 0U)
{
/* C = A * scale */
/* Scale input and store result in destination buffer. */
in = *pSrc++; /* read four inputs from source */
in = ((q63_t) in * scaleFract) >> 32; /* multiply input with scaler value */
out = in >> -kShift; /* apply shifting */
*pDst++ = out; /* Store result destination */
in = *pSrc++;
in = ((q63_t) in * scaleFract) >> 32;
out = in >> -kShift;
*pDst++ = out;
in = *pSrc++;
in = ((q63_t) in * scaleFract) >> 32;
out = in >> -kShift;
*pDst++ = out;
in = *pSrc++;
in = ((q63_t) in * scaleFract) >> 32;
out = in >> -kShift;
*pDst++ = out;
/* Decrement loop counter */
blkCnt--;
}
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
if (sign == 0U)
{
while (blkCnt > 0U)
{
/* C = A * scale */
/* Scale input and store result in destination buffer. */
in = *pSrc++;
in = ((q63_t) in * scaleFract) >> 32;
out = in << kShift;
if (in != (out >> kShift))
out = 0x7FFFFFFF ^ (in >> 31);
*pDst++ = out;
/* Decrement loop counter */
blkCnt--;
}
}
else
{
while (blkCnt > 0U)
{
/* C = A * scale */
/* Scale input and store result in destination buffer. */
in = *pSrc++;
in = ((q63_t) in * scaleFract) >> 32;
out = in >> -kShift;
*pDst++ = out;
/* Decrement loop counter */
blkCnt--;
}
}
}
/**
@} end of BasicScale group
*/

129
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_scale_q7.c Normal file
View File

@@ -0,0 +1,129 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_scale_q7.c
* Description: Multiplies a Q7 vector by a scalar
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicScale
@{
*/
/**
@brief Multiplies a Q7 vector by a scalar.
@param[in] pSrc points to the input vector
@param[in] scaleFract fractional portion of the scale value
@param[in] shift number of bits to shift the result by
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The input data <code>*pSrc</code> and <code>scaleFract</code> are in 1.7 format.
These are multiplied to yield a 2.14 intermediate result and this is shifted with saturation to 1.7 format.
*/
void arm_scale_q7(
const q7_t * pSrc,
q7_t scaleFract,
int8_t shift,
q7_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
int8_t kShift = 7 - shift; /* Shift to apply after scaling */
#if defined (ARM_MATH_LOOPUNROLL)
#if defined (ARM_MATH_DSP)
q7_t in1, in2, in3, in4; /* Temporary input variables */
q7_t out1, out2, out3, out4; /* Temporary output variables */
#endif
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A * scale */
#if defined (ARM_MATH_DSP)
/* Reading 4 inputs from memory */
in1 = *pSrc++;
in2 = *pSrc++;
in3 = *pSrc++;
in4 = *pSrc++;
/* Scale inputs and store result in the temporary variable. */
out1 = (q7_t) (__SSAT(((in1) * scaleFract) >> kShift, 8));
out2 = (q7_t) (__SSAT(((in2) * scaleFract) >> kShift, 8));
out3 = (q7_t) (__SSAT(((in3) * scaleFract) >> kShift, 8));
out4 = (q7_t) (__SSAT(((in4) * scaleFract) >> kShift, 8));
/* Pack and store result in destination buffer (in single write) */
write_q7x4_ia (&pDst, __PACKq7(out1, out2, out3, out4));
#else
*pDst++ = (q7_t) (__SSAT((((q15_t) *pSrc++ * scaleFract) >> kShift), 8));
*pDst++ = (q7_t) (__SSAT((((q15_t) *pSrc++ * scaleFract) >> kShift), 8));
*pDst++ = (q7_t) (__SSAT((((q15_t) *pSrc++ * scaleFract) >> kShift), 8));
*pDst++ = (q7_t) (__SSAT((((q15_t) *pSrc++ * scaleFract) >> kShift), 8));
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A * scale */
/* Scale input and store result in destination buffer. */
*pDst++ = (q7_t) (__SSAT((((q15_t) *pSrc++ * scaleFract) >> kShift), 8));
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicScale group
*/

201
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_shift_q15.c Normal file
View File

@@ -0,0 +1,201 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_shift_q15.c
* Description: Shifts the elements of a Q15 vector by a specified number of bits
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicShift
@{
*/
/**
@brief Shifts the elements of a Q15 vector a specified number of bits
@param[in] pSrc points to the input vector
@param[in] shiftBits number of bits to shift. A positive value shifts left; a negative value shifts right.
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q15 range [0x8000 0x7FFF] are saturated.
*/
void arm_shift_q15(
const q15_t * pSrc,
int8_t shiftBits,
q15_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
uint8_t sign = (shiftBits & 0x80); /* Sign of shiftBits */
#if defined (ARM_MATH_LOOPUNROLL)
#if defined (ARM_MATH_DSP)
q15_t in1, in2; /* Temporary input variables */
#endif
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
/* If the shift value is positive then do right shift else left shift */
if (sign == 0U)
{
while (blkCnt > 0U)
{
/* C = A << shiftBits */
#if defined (ARM_MATH_DSP)
/* read 2 samples from source */
in1 = *pSrc++;
in2 = *pSrc++;
/* Shift the inputs and then store the results in the destination buffer. */
#ifndef ARM_MATH_BIG_ENDIAN
write_q15x2_ia (&pDst, __PKHBT(__SSAT((in1 << shiftBits), 16),
__SSAT((in2 << shiftBits), 16), 16));
#else
write_q15x2_ia (&pDst, __PKHBT(__SSAT((in2 << shiftBits), 16),
__SSAT((in1 << shiftBits), 16), 16));
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* read 2 samples from source */
in1 = *pSrc++;
in2 = *pSrc++;
#ifndef ARM_MATH_BIG_ENDIAN
write_q15x2_ia (&pDst, __PKHBT(__SSAT((in1 << shiftBits), 16),
__SSAT((in2 << shiftBits), 16), 16));
#else
write_q15x2_ia (&pDst, __PKHBT(__SSAT((in2 << shiftBits), 16),
__SSAT((in1 << shiftBits), 16), 16));
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#else
*pDst++ = __SSAT(((q31_t) *pSrc++ << shiftBits), 16);
*pDst++ = __SSAT(((q31_t) *pSrc++ << shiftBits), 16);
*pDst++ = __SSAT(((q31_t) *pSrc++ << shiftBits), 16);
*pDst++ = __SSAT(((q31_t) *pSrc++ << shiftBits), 16);
#endif
/* Decrement loop counter */
blkCnt--;
}
}
else
{
while (blkCnt > 0U)
{
/* C = A >> shiftBits */
#if defined (ARM_MATH_DSP)
/* read 2 samples from source */
in1 = *pSrc++;
in2 = *pSrc++;
/* Shift the inputs and then store the results in the destination buffer. */
#ifndef ARM_MATH_BIG_ENDIAN
write_q15x2_ia (&pDst, __PKHBT((in1 >> -shiftBits),
(in2 >> -shiftBits), 16));
#else
write_q15x2_ia (&pDst, __PKHBT((in2 >> -shiftBits),
(in1 >> -shiftBits), 16));
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* read 2 samples from source */
in1 = *pSrc++;
in2 = *pSrc++;
#ifndef ARM_MATH_BIG_ENDIAN
write_q15x2_ia (&pDst, __PKHBT((in1 >> -shiftBits),
(in2 >> -shiftBits), 16));
#else
write_q15x2_ia (&pDst, __PKHBT((in2 >> -shiftBits),
(in1 >> -shiftBits), 16));
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#else
*pDst++ = (*pSrc++ >> -shiftBits);
*pDst++ = (*pSrc++ >> -shiftBits);
*pDst++ = (*pSrc++ >> -shiftBits);
*pDst++ = (*pSrc++ >> -shiftBits);
#endif
/* Decrement loop counter */
blkCnt--;
}
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
/* If the shift value is positive then do right shift else left shift */
if (sign == 0U)
{
while (blkCnt > 0U)
{
/* C = A << shiftBits */
/* Shift input and store result in destination buffer. */
*pDst++ = __SSAT(((q31_t) *pSrc++ << shiftBits), 16);
/* Decrement loop counter */
blkCnt--;
}
}
else
{
while (blkCnt > 0U)
{
/* C = A >> shiftBits */
/* Shift input and store result in destination buffer. */
*pDst++ = (*pSrc++ >> -shiftBits);
/* Decrement loop counter */
blkCnt--;
}
}
}
/**
@} end of BasicShift group
*/

181
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_shift_q31.c Normal file
View File

@@ -0,0 +1,181 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_shift_q31.c
* Description: Shifts the elements of a Q31 vector by a specified number of bits
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@defgroup BasicShift Vector Shift
Shifts the elements of a fixed-point vector by a specified number of bits.
There are separate functions for Q7, Q15, and Q31 data types.
The underlying algorithm used is:
<pre>
pDst[n] = pSrc[n] << shift, 0 <= n < blockSize.
</pre>
If <code>shift</code> is positive then the elements of the vector are shifted to the left.
If <code>shift</code> is negative then the elements of the vector are shifted to the right.
The functions support in-place computation allowing the source and destination
pointers to reference the same memory buffer.
*/
/**
@addtogroup BasicShift
@{
*/
/**
@brief Shifts the elements of a Q31 vector a specified number of bits.
@param[in] pSrc points to the input vector
@param[in] shiftBits number of bits to shift. A positive value shifts left; a negative value shifts right.
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in the vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q31 range [0x80000000 0x7FFFFFFF] are saturated.
*/
void arm_shift_q31(
const q31_t * pSrc,
int8_t shiftBits,
q31_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
uint8_t sign = (shiftBits & 0x80); /* Sign of shiftBits */
#if defined (ARM_MATH_LOOPUNROLL)
q31_t in, out; /* Temporary variables */
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
/* If the shift value is positive then do right shift else left shift */
if (sign == 0U)
{
while (blkCnt > 0U)
{
/* C = A << shiftBits */
/* Shift input and store result in destination buffer. */
in = *pSrc++;
out = in << shiftBits;
if (in != (out >> shiftBits))
out = 0x7FFFFFFF ^ (in >> 31);
*pDst++ = out;
in = *pSrc++;
out = in << shiftBits;
if (in != (out >> shiftBits))
out = 0x7FFFFFFF ^ (in >> 31);
*pDst++ = out;
in = *pSrc++;
out = in << shiftBits;
if (in != (out >> shiftBits))
out = 0x7FFFFFFF ^ (in >> 31);
*pDst++ = out;
in = *pSrc++;
out = in << shiftBits;
if (in != (out >> shiftBits))
out = 0x7FFFFFFF ^ (in >> 31);
*pDst++ = out;
/* Decrement loop counter */
blkCnt--;
}
}
else
{
while (blkCnt > 0U)
{
/* C = A >> shiftBits */
/* Shift input and store results in destination buffer. */
*pDst++ = (*pSrc++ >> -shiftBits);
*pDst++ = (*pSrc++ >> -shiftBits);
*pDst++ = (*pSrc++ >> -shiftBits);
*pDst++ = (*pSrc++ >> -shiftBits);
/* Decrement loop counter */
blkCnt--;
}
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
/* If the shift value is positive then do right shift else left shift */
if (sign == 0U)
{
while (blkCnt > 0U)
{
/* C = A << shiftBits */
/* Shift input and store result in destination buffer. */
*pDst++ = clip_q63_to_q31((q63_t) *pSrc++ << shiftBits);
/* Decrement loop counter */
blkCnt--;
}
}
else
{
while (blkCnt > 0U)
{
/* C = A >> shiftBits */
/* Shift input and store result in destination buffer. */
*pDst++ = (*pSrc++ >> -shiftBits);
/* Decrement loop counter */
blkCnt--;
}
}
}
/**
@} end of BasicShift group
*/

175
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_shift_q7.c Normal file
View File

@@ -0,0 +1,175 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_shift_q7.c
* Description: Processing function for the Q7 Shifting
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicShift
@{
*/
/**
@brief Shifts the elements of a Q7 vector a specified number of bits
@param[in] pSrc points to the input vector
@param[in] shiftBits number of bits to shift. A positive value shifts left; a negative value shifts right.
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par onditions for optimum performance
Input and output buffers should be aligned by 32-bit
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q7 range [0x80 0x7F] are saturated.
*/
void arm_shift_q7(
const q7_t * pSrc,
int8_t shiftBits,
q7_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
uint8_t sign = (shiftBits & 0x80); /* Sign of shiftBits */
#if defined (ARM_MATH_LOOPUNROLL)
#if defined (ARM_MATH_DSP)
q7_t in1, in2, in3, in4; /* Temporary input variables */
#endif
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
/* If the shift value is positive then do right shift else left shift */
if (sign == 0U)
{
while (blkCnt > 0U)
{
/* C = A << shiftBits */
#if defined (ARM_MATH_DSP)
/* Read 4 inputs */
in1 = *pSrc++;
in2 = *pSrc++;
in3 = *pSrc++;
in4 = *pSrc++;
/* Pack and store result in destination buffer (in single write) */
write_q7x4_ia (&pDst, __PACKq7(__SSAT((in1 << shiftBits), 8),
__SSAT((in2 << shiftBits), 8),
__SSAT((in3 << shiftBits), 8),
__SSAT((in4 << shiftBits), 8) ));
#else
*pDst++ = (q7_t) __SSAT(((q15_t) *pSrc++ << shiftBits), 8);
*pDst++ = (q7_t) __SSAT(((q15_t) *pSrc++ << shiftBits), 8);
*pDst++ = (q7_t) __SSAT(((q15_t) *pSrc++ << shiftBits), 8);
*pDst++ = (q7_t) __SSAT(((q15_t) *pSrc++ << shiftBits), 8);
#endif
/* Decrement loop counter */
blkCnt--;
}
}
else
{
while (blkCnt > 0U)
{
/* C = A >> shiftBits */
#if defined (ARM_MATH_DSP)
/* Read 4 inputs */
in1 = *pSrc++;
in2 = *pSrc++;
in3 = *pSrc++;
in4 = *pSrc++;
/* Pack and store result in destination buffer (in single write) */
write_q7x4_ia (&pDst, __PACKq7((in1 >> -shiftBits),
(in2 >> -shiftBits),
(in3 >> -shiftBits),
(in4 >> -shiftBits) ));
#else
*pDst++ = (*pSrc++ >> -shiftBits);
*pDst++ = (*pSrc++ >> -shiftBits);
*pDst++ = (*pSrc++ >> -shiftBits);
*pDst++ = (*pSrc++ >> -shiftBits);
#endif
/* Decrement loop counter */
blkCnt--;
}
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
/* If the shift value is positive then do right shift else left shift */
if (sign == 0U)
{
while (blkCnt > 0U)
{
/* C = A << shiftBits */
/* Shift input and store result in destination buffer. */
*pDst++ = (q7_t) __SSAT(((q15_t) *pSrc++ << shiftBits), 8);
/* Decrement loop counter */
blkCnt--;
}
}
else
{
while (blkCnt > 0U)
{
/* C = A >> shiftBits */
/* Shift input and store result in destination buffer. */
*pDst++ = (*pSrc++ >> -shiftBits);
/* Decrement loop counter */
blkCnt--;
}
}
}
/**
@} end of BasicShift group
*/

148
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_sub_f32.c Normal file
View File

@@ -0,0 +1,148 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_sub_f32.c
* Description: Floating-point vector subtraction
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@defgroup BasicSub Vector Subtraction
Element-by-element subtraction of two vectors.
<pre>
pDst[n] = pSrcA[n] - pSrcB[n], 0 <= n < blockSize.
</pre>
There are separate functions for floating-point, Q7, Q15, and Q31 data types.
*/
/**
@addtogroup BasicSub
@{
*/
/**
@brief Floating-point vector subtraction.
@param[in] pSrcA points to the first input vector
@param[in] pSrcB points to the second input vector
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
*/
void arm_sub_f32(
const float32_t * pSrcA,
const float32_t * pSrcB,
float32_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined(ARM_MATH_NEON)
float32x4_t vec1;
float32x4_t vec2;
float32x4_t res;
/* Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A - B */
/* Subtract and then store the results in the destination buffer. */
vec1 = vld1q_f32(pSrcA);
vec2 = vld1q_f32(pSrcB);
res = vsubq_f32(vec1, vec2);
vst1q_f32(pDst, res);
/* Increment pointers */
pSrcA += 4;
pSrcB += 4;
pDst += 4;
/* Decrement the loop counter */
blkCnt--;
}
/* Tail */
blkCnt = blockSize & 0x3;
#else
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A - B */
/* Subtract and store result in destination buffer. */
*pDst++ = (*pSrcA++) - (*pSrcB++);
*pDst++ = (*pSrcA++) - (*pSrcB++);
*pDst++ = (*pSrcA++) - (*pSrcB++);
*pDst++ = (*pSrcA++) - (*pSrcB++);
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
#endif /* #if defined(ARM_MATH_NEON) */
while (blkCnt > 0U)
{
/* C = A - B */
/* Subtract and store result in destination buffer. */
*pDst++ = (*pSrcA++) - (*pSrcB++);
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicSub group
*/

126
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_sub_q15.c Normal file
View File

@@ -0,0 +1,126 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_sub_q15.c
* Description: Q15 vector subtraction
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicSub
@{
*/
/**
@brief Q15 vector subtraction.
@param[in] pSrcA points to the first input vector
@param[in] pSrcB points to the second input vector
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q15 range [0x8000 0x7FFF] are saturated.
*/
void arm_sub_q15(
const q15_t * pSrcA,
const q15_t * pSrcB,
q15_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined (ARM_MATH_LOOPUNROLL)
#if defined (ARM_MATH_DSP)
q31_t inA1, inA2;
q31_t inB1, inB2;
#endif
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A - B */
#if defined (ARM_MATH_DSP)
/* read 2 times 2 samples at a time from sourceA */
inA1 = read_q15x2_ia ((q15_t **) &pSrcA);
inA2 = read_q15x2_ia ((q15_t **) &pSrcA);
/* read 2 times 2 samples at a time from sourceB */
inB1 = read_q15x2_ia ((q15_t **) &pSrcB);
inB2 = read_q15x2_ia ((q15_t **) &pSrcB);
/* Subtract and store 2 times 2 samples at a time */
write_q15x2_ia (&pDst, __QSUB16(inA1, inB1));
write_q15x2_ia (&pDst, __QSUB16(inA2, inB2));
#else
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrcA++ - *pSrcB++), 16);
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrcA++ - *pSrcB++), 16);
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrcA++ - *pSrcB++), 16);
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrcA++ - *pSrcB++), 16);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A - B */
/* Subtract and store result in destination buffer. */
#if defined (ARM_MATH_DSP)
*pDst++ = (q15_t) __QSUB16(*pSrcA++, *pSrcB++);
#else
*pDst++ = (q15_t) __SSAT(((q31_t) *pSrcA++ - *pSrcB++), 16);
#endif
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicSub group
*/

108
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_sub_q31.c Normal file
View File

@@ -0,0 +1,108 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_sub_q31.c
* Description: Q31 vector subtraction
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicSub
@{
*/
/**
@brief Q31 vector subtraction.
@param[in] pSrcA points to the first input vector
@param[in] pSrcB points to the second input vector
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q31 range [0x80000000 0x7FFFFFFF] are saturated.
*/
void arm_sub_q31(
const q31_t * pSrcA,
const q31_t * pSrcB,
q31_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A - B */
/* Subtract and store result in destination buffer. */
*pDst++ = __QSUB(*pSrcA++, *pSrcB++);
*pDst++ = __QSUB(*pSrcA++, *pSrcB++);
*pDst++ = __QSUB(*pSrcA++, *pSrcB++);
*pDst++ = __QSUB(*pSrcA++, *pSrcB++);
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A - B */
/* Subtract and store result in destination buffer. */
*pDst++ = __QSUB(*pSrcA++, *pSrcB++);
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicSub group
*/

109
Drivers/CMSIS/DSP/Source/BasicMathFunctions/arm_sub_q7.c Normal file
View File

@@ -0,0 +1,109 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_sub_q7.c
* Description: Q7 vector subtraction
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupMath
*/
/**
@addtogroup BasicSub
@{
*/
/**
@brief Q7 vector subtraction.
@param[in] pSrcA points to the first input vector
@param[in] pSrcB points to the second input vector
@param[out] pDst points to the output vector
@param[in] blockSize number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
Results outside of the allowable Q7 range [0x80 0x7F] will be saturated.
*/
void arm_sub_q7(
const q7_t * pSrcA,
const q7_t * pSrcB,
q7_t * pDst,
uint32_t blockSize)
{
uint32_t blkCnt; /* Loop counter */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = blockSize >> 2U;
while (blkCnt > 0U)
{
/* C = A - B */
#if defined (ARM_MATH_DSP)
/* Subtract and store result in destination buffer (4 samples at a time). */
write_q7x4_ia (&pDst, __QSUB8(read_q7x4_ia ((q7_t **) &pSrcA), read_q7x4_ia ((q7_t **) &pSrcB)));
#else
*pDst++ = (q7_t) __SSAT((q15_t) *pSrcA++ - *pSrcB++, 8);
*pDst++ = (q7_t) __SSAT((q15_t) *pSrcA++ - *pSrcB++, 8);
*pDst++ = (q7_t) __SSAT((q15_t) *pSrcA++ - *pSrcB++, 8);
*pDst++ = (q7_t) __SSAT((q15_t) *pSrcA++ - *pSrcB++, 8);
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = blockSize % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = blockSize;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C = A - B */
/* Subtract and store result in destination buffer. */
*pDst++ = (q7_t) __SSAT((q15_t) *pSrcA++ - *pSrcB++, 8);
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of BasicSub group
*/

223
Drivers/CMSIS/DSP/Source/CMakeLists.txt Normal file
View File

@@ -0,0 +1,223 @@
cmake_minimum_required (VERSION 3.6)
project(CMSISDSP)
# Needed to find the config modules
list(APPEND CMAKE_MODULE_PATH ${CMAKE_CURRENT_SOURCE_DIR}/..)
# Select which parts of the CMSIS-DSP must be compiled.
# There are some dependencies between the parts but they are not tracked
# by this cmake. So, enabling some functions may require to enable some
# other ones.
option(BASICMATH "Basic Math Functions" ON)
option(COMPLEXMATH "Complex Math Functions" ON)
option(CONTROLLER "Controller Functions" ON)
option(FASTMATH "Fast Math Functions" ON)
option(FILTERING "Filtering Functions" ON)
option(MATRIX "Matrix Functions" ON)
option(STATISTICS "Statistics Functions" ON)
option(SUPPORT "Support Functions" ON)
option(TRANSFORM "Transform Functions" ON)
# When OFF it is the default behavior : all tables are included.
option(CONFIGTABLE "Configuration of table allowed" OFF)
# When CONFIGTABLE is ON, select if all interpolation tables must be included
option(ALLFAST "All interpolation tables included" OFF)
# When CONFIGTABLE is ON, select if all FFT tables must be included
option(ALLFFT "All fft tables included" OFF)
# Features which require inclusion of a data table.
# Since some tables may be big, the corresponding feature can be
# disabled.
# Those options are taken into account only when CONFIGTABLE is ON
option(ARM_COS_F32 "cos f32" OFF)
option(ARM_COS_Q31 "cos q31" OFF)
option(ARM_COS_Q15 "cos q15" OFF)
option(ARM_SIN_F32 "sin f32" OFF)
option(ARM_SIN_Q31 "sin q31" OFF)
option(ARM_SIN_Q15 "sin q15" OFF)
option(ARM_SIN_COS_F32 "sin cos f32" OFF)
option(ARM_SIN_COS_Q31 "sin cos q31" OFF)
option(ARM_LMS_NORM_Q31 "lms norm q31" OFF)
option(ARM_LMS_NORM_Q15 "lms norm q15" OFF)
option(CFFT_F32_16 "cfft f32 16" OFF)
option(CFFT_F32_32 "cfft f32 32" OFF)
option(CFFT_F32_64 "cfft f32 64" OFF)
option(CFFT_F32_128 "cfft f32 128" OFF)
option(CFFT_F32_256 "cfft f32 256" OFF)
option(CFFT_F32_512 "cfft f32 512" OFF)
option(CFFT_F32_1024 "cfft f32 1024" OFF)
option(CFFT_F32_2048 "cfft f32 2048" OFF)
option(CFFT_F32_4096 "cfft f32 4096" OFF)
option(CFFT_Q31_16 "cfft q31 16" OFF)
option(CFFT_Q31_32 "cfft q31 32" OFF)
option(CFFT_Q31_64 "cfft q31 64" OFF)
option(CFFT_Q31_128 "cfft q31 128" OFF)
option(CFFT_Q31_256 "cfft q31 256" OFF)
option(CFFT_Q31_512 "cfft q31 512" OFF)
option(CFFT_Q31_1024 "cfft q31 1024" OFF)
option(CFFT_Q31_2048 "cfft q31 2048" OFF)
option(CFFT_Q31_4096 "cfft q31 4096" OFF)
option(CFFT_Q15_16 "cfft q15 16" OFF)
option(CFFT_Q15_32 "cfft q15 32" OFF)
option(CFFT_Q15_64 "cfft q15 64" OFF)
option(CFFT_Q15_128 "cfft q15 128" OFF)
option(CFFT_Q15_256 "cfft q15 256" OFF)
option(CFFT_Q15_512 "cfft q15 512" OFF)
option(CFFT_Q15_1024 "cfft q15 1024" OFF)
option(CFFT_Q15_2048 "cfft q15 2048" OFF)
option(CFFT_Q15_4096 "cfft q15 4096" OFF)
option(RFFT_FAST_F32_32 "rfft fast f32 32" OFF)
option(RFFT_FAST_F32_64 "rfft fast f32 64" OFF)
option(RFFT_FAST_F32_128 "rfft fast f32 128" OFF)
option(RFFT_FAST_F32_256 "rfft fast f32 256" OFF)
option(RFFT_FAST_F32_512 "rfft fast f32 512" OFF)
option(RFFT_FAST_F32_1024 "rfft fast f32 1024" OFF)
option(RFFT_FAST_F32_2048 "rfft fast f32 2048" OFF)
option(RFFT_FAST_F32_4096 "rfft fast f32 4096" OFF)
option(RFFT_F32_128 "rfft f32 128" OFF)
option(RFFT_F32_512 "rfft f32 512" OFF)
option(RFFT_F32_2048 "rfft f32 2048" OFF)
option(RFFT_F32_8192 "rfft f32 8192" OFF)
option(RFFT_Q31_32 "rfft q31 32" OFF)
option(RFFT_Q31_64 "rfft q31 64" OFF)
option(RFFT_Q31_128 "rfft q31 128" OFF)
option(RFFT_Q31_256 "rfft q31 256" OFF)
option(RFFT_Q31_512 "rfft q31 512" OFF)
option(RFFT_Q31_1024 "rfft q31 1024" OFF)
option(RFFT_Q31_2048 "rfft q31 2048" OFF)
option(RFFT_Q31_4096 "rfft q31 4096" OFF)
option(RFFT_Q31_8192 "rfft q31 8192" OFF)
option(RFFT_Q15_32 "rfft q15 32" OFF)
option(RFFT_Q15_64 "rfft q15 64" OFF)
option(RFFT_Q15_128 "rfft q15 128" OFF)
option(RFFT_Q15_256 "rfft q15 256" OFF)
option(RFFT_Q15_512 "rfft q15 512" OFF)
option(RFFT_Q15_1024 "rfft q15 1024" OFF)
option(RFFT_Q15_2048 "rfft q15 2048" OFF)
option(RFFT_Q15_4096 "rfft q15 4096" OFF)
option(RFFT_Q15_8192 "rfft q15 8192" OFF)
option(DCT4_F32_128 "dct4 f32 128" OFF)
option(DCT4_F32_512 "dct4 f32 512" OFF)
option(DCT4_F32_2048 "dct4 f32 2048" OFF)
option(DCT4_F32_8192 "dct4 f32 8192" OFF)
option(DCT4_Q31_128 "dct4 q31 128" OFF)
option(DCT4_Q31_512 "dct4 q31 512" OFF)
option(DCT4_Q31_2048 "dct4 q31 2048" OFF)
option(DCT4_Q31_8192 "dct4 q31 8192" OFF)
option(DCT4_Q15_128 "dct4 q15 128" OFF)
option(DCT4_Q15_512 "dct4 q15 512" OFF)
option(DCT4_Q15_2048 "dct4 q15 2048" OFF)
option(DCT4_Q15_8192 "dct4 q15 8192" OFF)
###########################
#
# CMSIS DSP
#
###########################
# DSP Sources
SET(DSP ".")
add_library(CMSISDSP INTERFACE)
include(config)
if (BASICMATH)
add_subdirectory(BasicMathFunctions)
target_link_libraries(CMSISDSP INTERFACE CMSISDSPBasicMath)
endif()
if (COMPLEXMATH)
add_subdirectory(ComplexMathFunctions)
target_link_libraries(CMSISDSP INTERFACE CMSISDSPComplexMath)
endif()
if (CONTROLLER)
add_subdirectory(ControllerFunctions)
# Fast tables inclusion is allowed
if (CONFIGTABLE)
target_compile_definitions(CMSISDSPController PUBLIC ARM_FAST_ALLOW_TABLES)
endif()
target_link_libraries(CMSISDSP INTERFACE CMSISDSPController)
endif()
if (FASTMATH)
add_subdirectory(FastMathFunctions)
# Fast tables inclusion is allowed
if (CONFIGTABLE)
target_compile_definitions(CMSISDSPFastMath PUBLIC ARM_FAST_ALLOW_TABLES)
endif()
target_link_libraries(CMSISDSP INTERFACE CMSISDSPFastMath)
endif()
if (FILTERING)
add_subdirectory(FilteringFunctions)
# Fast tables inclusion is allowed
if (CONFIGTABLE)
target_compile_definitions(CMSISDSPFiltering PUBLIC ARM_FAST_ALLOW_TABLES)
endif()
target_link_libraries(CMSISDSP INTERFACE CMSISDSPFiltering)
endif()
if (MATRIX)
add_subdirectory(MatrixFunctions)
target_link_libraries(CMSISDSP INTERFACE CMSISDSPMatrix)
endif()
if (STATISTICS)
add_subdirectory(StatisticsFunctions)
target_link_libraries(CMSISDSP INTERFACE CMSISDSPStatistics)
endif()
if (SUPPORT)
add_subdirectory(SupportFunctions)
target_link_libraries(CMSISDSP INTERFACE CMSISDSPSupport)
endif()
if (TRANSFORM)
add_subdirectory(TransformFunctions)
# FFT tables inclusion is allowed
if (CONFIGTABLE)
target_compile_definitions(CMSISDSPTransform PUBLIC ARM_FFT_ALLOW_TABLES)
endif()
target_link_libraries(CMSISDSP INTERFACE CMSISDSPTransform)
endif()
if (FILTERING OR CONTROLLER OR FASTMATH OR TRANSFORM)
add_subdirectory(CommonTables)
if (TRANSFORM)
# FFT tables inclusion is allowed
if (CONFIGTABLE)
target_compile_definitions(CMSISDSPCommon PUBLIC ARM_FFT_ALLOW_TABLES)
endif()
endif()
if (FILTERING OR CONTROLLER OR FASTMATH)
# Select which tables to include
if (CONFIGTABLE)
target_compile_definitions(CMSISDSPCommon PUBLIC ARM_FAST_ALLOW_TABLES)
endif()
endif()
target_link_libraries(CMSISDSP INTERFACE CMSISDSPCommon)
endif()
### Includes
target_include_directories(CMSISDSP INTERFACE "${DSP}/../Include")

31
Drivers/CMSIS/DSP/Source/CommonTables/CMakeLists.txt Normal file
View File

@@ -0,0 +1,31 @@
cmake_minimum_required (VERSION 3.6)
project(CMSISDSPCommon)
add_library(CMSISDSPCommon STATIC arm_common_tables.c)
if (CONFIGTABLE AND ALLFFT)
target_compile_definitions(CMSISDSPCommon PUBLIC ARM_ALL_FFT_TABLES)
endif()
if (CONFIGTABLE AND ALLFAST)
target_compile_definitions(CMSISDSPCommon PUBLIC ARM_ALL_FAST_TABLES)
endif()
include(fft)
fft(CMSISDSPCommon)
include(interpol)
interpol(CMSISDSPCommon)
target_sources(CMSISDSPCommon PRIVATE arm_const_structs.c)
configdsp(CMSISDSPCommon ..)
### Includes
target_include_directories(CMSISDSPCommon PUBLIC "${DSP}/../../Include")

31
Drivers/CMSIS/DSP/Source/CommonTables/CommonTables.c Normal file
View File

@@ -0,0 +1,31 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: CommonTables.c
* Description: Combination of all common table source files.
*
* $Date: 18. March 2019
* $Revision: V1.0.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_common_tables.c"
#include "arm_const_structs.c"

57215
Drivers/CMSIS/DSP/Source/CommonTables/arm_common_tables.c Normal file
View File

File diff suppressed because it is too large Load Diff

486
Drivers/CMSIS/DSP/Source/CommonTables/arm_const_structs.c Normal file
View File

@@ -0,0 +1,486 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_const_structs.c
* Description: Constant structs that are initialized for user convenience.
* For example, some can be given as arguments to the arm_cfft_f32() or arm_rfft_f32() functions.
*
* $Date: 27. January 2017
* $Revision: V.1.5.1
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_const_structs.h"
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_FFT_ALLOW_TABLES)
/* Floating-point structs */
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_16) && defined(ARM_TABLE_BITREVIDX_FLT_16))
const arm_cfft_instance_f32 arm_cfft_sR_f32_len16 = {
16, twiddleCoef_16, armBitRevIndexTable16, ARMBITREVINDEXTABLE_16_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_32) && defined(ARM_TABLE_BITREVIDX_FLT_32))
const arm_cfft_instance_f32 arm_cfft_sR_f32_len32 = {
32, twiddleCoef_32, armBitRevIndexTable32, ARMBITREVINDEXTABLE_32_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_64) && defined(ARM_TABLE_BITREVIDX_FLT_64))
const arm_cfft_instance_f32 arm_cfft_sR_f32_len64 = {
64, twiddleCoef_64, armBitRevIndexTable64, ARMBITREVINDEXTABLE_64_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_128) && defined(ARM_TABLE_BITREVIDX_FLT_128))
const arm_cfft_instance_f32 arm_cfft_sR_f32_len128 = {
128, twiddleCoef_128, armBitRevIndexTable128, ARMBITREVINDEXTABLE_128_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_256) && defined(ARM_TABLE_BITREVIDX_FLT_256))
const arm_cfft_instance_f32 arm_cfft_sR_f32_len256 = {
256, twiddleCoef_256, armBitRevIndexTable256, ARMBITREVINDEXTABLE_256_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_512) && defined(ARM_TABLE_BITREVIDX_FLT_512))
const arm_cfft_instance_f32 arm_cfft_sR_f32_len512 = {
512, twiddleCoef_512, armBitRevIndexTable512, ARMBITREVINDEXTABLE_512_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_1024) && defined(ARM_TABLE_BITREVIDX_FLT_1024))
const arm_cfft_instance_f32 arm_cfft_sR_f32_len1024 = {
1024, twiddleCoef_1024, armBitRevIndexTable1024, ARMBITREVINDEXTABLE_1024_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_2048) && defined(ARM_TABLE_BITREVIDX_FLT_2048))
const arm_cfft_instance_f32 arm_cfft_sR_f32_len2048 = {
2048, twiddleCoef_2048, armBitRevIndexTable2048, ARMBITREVINDEXTABLE_2048_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_4096) && defined(ARM_TABLE_BITREVIDX_FLT_4096))
const arm_cfft_instance_f32 arm_cfft_sR_f32_len4096 = {
4096, twiddleCoef_4096, armBitRevIndexTable4096, ARMBITREVINDEXTABLE_4096_TABLE_LENGTH
};
#endif
/* Fixed-point structs */
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q31_16) && defined(ARM_TABLE_BITREVIDX_FXT_16))
const arm_cfft_instance_q31 arm_cfft_sR_q31_len16 = {
16, twiddleCoef_16_q31, armBitRevIndexTable_fixed_16, ARMBITREVINDEXTABLE_FIXED_16_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q31_32) && defined(ARM_TABLE_BITREVIDX_FXT_32))
const arm_cfft_instance_q31 arm_cfft_sR_q31_len32 = {
32, twiddleCoef_32_q31, armBitRevIndexTable_fixed_32, ARMBITREVINDEXTABLE_FIXED_32_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q31_64) && defined(ARM_TABLE_BITREVIDX_FXT_64))
const arm_cfft_instance_q31 arm_cfft_sR_q31_len64 = {
64, twiddleCoef_64_q31, armBitRevIndexTable_fixed_64, ARMBITREVINDEXTABLE_FIXED_64_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q31_128) && defined(ARM_TABLE_BITREVIDX_FXT_128))
const arm_cfft_instance_q31 arm_cfft_sR_q31_len128 = {
128, twiddleCoef_128_q31, armBitRevIndexTable_fixed_128, ARMBITREVINDEXTABLE_FIXED_128_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q31_256) && defined(ARM_TABLE_BITREVIDX_FXT_256))
const arm_cfft_instance_q31 arm_cfft_sR_q31_len256 = {
256, twiddleCoef_256_q31, armBitRevIndexTable_fixed_256, ARMBITREVINDEXTABLE_FIXED_256_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q31_512) && defined(ARM_TABLE_BITREVIDX_FXT_512))
const arm_cfft_instance_q31 arm_cfft_sR_q31_len512 = {
512, twiddleCoef_512_q31, armBitRevIndexTable_fixed_512, ARMBITREVINDEXTABLE_FIXED_512_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q31_1024) && defined(ARM_TABLE_BITREVIDX_FXT_1024))
const arm_cfft_instance_q31 arm_cfft_sR_q31_len1024 = {
1024, twiddleCoef_1024_q31, armBitRevIndexTable_fixed_1024, ARMBITREVINDEXTABLE_FIXED_1024_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q31_2048) && defined(ARM_TABLE_BITREVIDX_FXT_2048))
const arm_cfft_instance_q31 arm_cfft_sR_q31_len2048 = {
2048, twiddleCoef_2048_q31, armBitRevIndexTable_fixed_2048, ARMBITREVINDEXTABLE_FIXED_2048_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q31_4096) && defined(ARM_TABLE_BITREVIDX_FXT_4096))
const arm_cfft_instance_q31 arm_cfft_sR_q31_len4096 = {
4096, twiddleCoef_4096_q31, armBitRevIndexTable_fixed_4096, ARMBITREVINDEXTABLE_FIXED_4096_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q15_16) && defined(ARM_TABLE_BITREVIDX_FXT_16))
const arm_cfft_instance_q15 arm_cfft_sR_q15_len16 = {
16, twiddleCoef_16_q15, armBitRevIndexTable_fixed_16, ARMBITREVINDEXTABLE_FIXED_16_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q15_32) && defined(ARM_TABLE_BITREVIDX_FXT_32))
const arm_cfft_instance_q15 arm_cfft_sR_q15_len32 = {
32, twiddleCoef_32_q15, armBitRevIndexTable_fixed_32, ARMBITREVINDEXTABLE_FIXED_32_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q15_64) && defined(ARM_TABLE_BITREVIDX_FXT_64))
const arm_cfft_instance_q15 arm_cfft_sR_q15_len64 = {
64, twiddleCoef_64_q15, armBitRevIndexTable_fixed_64, ARMBITREVINDEXTABLE_FIXED_64_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q15_128) && defined(ARM_TABLE_BITREVIDX_FXT_128))
const arm_cfft_instance_q15 arm_cfft_sR_q15_len128 = {
128, twiddleCoef_128_q15, armBitRevIndexTable_fixed_128, ARMBITREVINDEXTABLE_FIXED_128_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q15_256) && defined(ARM_TABLE_BITREVIDX_FXT_256))
const arm_cfft_instance_q15 arm_cfft_sR_q15_len256 = {
256, twiddleCoef_256_q15, armBitRevIndexTable_fixed_256, ARMBITREVINDEXTABLE_FIXED_256_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q15_512) && defined(ARM_TABLE_BITREVIDX_FXT_512))
const arm_cfft_instance_q15 arm_cfft_sR_q15_len512 = {
512, twiddleCoef_512_q15, armBitRevIndexTable_fixed_512, ARMBITREVINDEXTABLE_FIXED_512_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q15_1024) && defined(ARM_TABLE_BITREVIDX_FXT_1024))
const arm_cfft_instance_q15 arm_cfft_sR_q15_len1024 = {
1024, twiddleCoef_1024_q15, armBitRevIndexTable_fixed_1024, ARMBITREVINDEXTABLE_FIXED_1024_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q15_2048) && defined(ARM_TABLE_BITREVIDX_FXT_2048))
const arm_cfft_instance_q15 arm_cfft_sR_q15_len2048 = {
2048, twiddleCoef_2048_q15, armBitRevIndexTable_fixed_2048, ARMBITREVINDEXTABLE_FIXED_2048_TABLE_LENGTH
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_Q15_4096) && defined(ARM_TABLE_BITREVIDX_FXT_4096))
const arm_cfft_instance_q15 arm_cfft_sR_q15_len4096 = {
4096, twiddleCoef_4096_q15, armBitRevIndexTable_fixed_4096, ARMBITREVINDEXTABLE_FIXED_4096_TABLE_LENGTH
};
#endif
/* Structure for real-value inputs */
/* Floating-point structs */
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_32) && defined(ARM_TABLE_BITREVIDX_FLT_32) && defined(ARM_TABLE_TWIDDLECOEF_RFFT_F32_32))
const arm_rfft_fast_instance_f32 arm_rfft_fast_sR_f32_len32 = {
{ 16, twiddleCoef_32, armBitRevIndexTable32, ARMBITREVINDEXTABLE_16_TABLE_LENGTH },
32U,
(float32_t *)twiddleCoef_rfft_32
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_64) && defined(ARM_TABLE_BITREVIDX_FLT_64) && defined(ARM_TABLE_TWIDDLECOEF_RFFT_F32_64))
const arm_rfft_fast_instance_f32 arm_rfft_fast_sR_f32_len64 = {
{ 32, twiddleCoef_32, armBitRevIndexTable32, ARMBITREVINDEXTABLE_32_TABLE_LENGTH },
64U,
(float32_t *)twiddleCoef_rfft_64
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_128) && defined(ARM_TABLE_BITREVIDX_FLT_128) && defined(ARM_TABLE_TWIDDLECOEF_RFFT_F32_128))
const arm_rfft_fast_instance_f32 arm_rfft_fast_sR_f32_len128 = {
{ 64, twiddleCoef_64, armBitRevIndexTable64, ARMBITREVINDEXTABLE_64_TABLE_LENGTH },
128U,
(float32_t *)twiddleCoef_rfft_128
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_256) && defined(ARM_TABLE_BITREVIDX_FLT_256) && defined(ARM_TABLE_TWIDDLECOEF_RFFT_F32_256))
const arm_rfft_fast_instance_f32 arm_rfft_fast_sR_f32_len256 = {
{ 128, twiddleCoef_128, armBitRevIndexTable128, ARMBITREVINDEXTABLE_128_TABLE_LENGTH },
256U,
(float32_t *)twiddleCoef_rfft_256
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_512) && defined(ARM_TABLE_BITREVIDX_FLT_512) && defined(ARM_TABLE_TWIDDLECOEF_RFFT_F32_512))
const arm_rfft_fast_instance_f32 arm_rfft_fast_sR_f32_len512 = {
{ 256, twiddleCoef_256, armBitRevIndexTable256, ARMBITREVINDEXTABLE_256_TABLE_LENGTH },
512U,
(float32_t *)twiddleCoef_rfft_512
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_1024) && defined(ARM_TABLE_BITREVIDX_FLT_1024) && defined(ARM_TABLE_TWIDDLECOEF_RFFT_F32_1024))
const arm_rfft_fast_instance_f32 arm_rfft_fast_sR_f32_len1024 = {
{ 512, twiddleCoef_512, armBitRevIndexTable512, ARMBITREVINDEXTABLE_512_TABLE_LENGTH },
1024U,
(float32_t *)twiddleCoef_rfft_1024
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_2048) && defined(ARM_TABLE_BITREVIDX_FLT_2048) && defined(ARM_TABLE_TWIDDLECOEF_RFFT_F32_2048))
const arm_rfft_fast_instance_f32 arm_rfft_fast_sR_f32_len2048 = {
{ 1024, twiddleCoef_1024, armBitRevIndexTable1024, ARMBITREVINDEXTABLE_1024_TABLE_LENGTH },
2048U,
(float32_t *)twiddleCoef_rfft_2048
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_TWIDDLECOEF_F32_4096) && defined(ARM_TABLE_BITREVIDX_FLT_4096) && defined(ARM_TABLE_TWIDDLECOEF_RFFT_F32_4096))
const arm_rfft_fast_instance_f32 arm_rfft_fast_sR_f32_len4096 = {
{ 2048, twiddleCoef_2048, armBitRevIndexTable2048, ARMBITREVINDEXTABLE_2048_TABLE_LENGTH },
4096U,
(float32_t *)twiddleCoef_rfft_4096
};
#endif
/* Fixed-point structs */
/* q31_t */
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q31) && defined(ARM_TABLE_TWIDDLECOEF_Q31_16) && defined(ARM_TABLE_BITREVIDX_FXT_16))
const arm_rfft_instance_q31 arm_rfft_sR_q31_len32 = {
32U,
0,
1,
256U,
(q31_t*)realCoefAQ31,
(q31_t*)realCoefBQ31,
&arm_cfft_sR_q31_len16
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q31) && defined(ARM_TABLE_TWIDDLECOEF_Q31_32) && defined(ARM_TABLE_BITREVIDX_FXT_32))
const arm_rfft_instance_q31 arm_rfft_sR_q31_len64 = {
64U,
0,
1,
128U,
(q31_t*)realCoefAQ31,
(q31_t*)realCoefBQ31,
&arm_cfft_sR_q31_len32
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q31) && defined(ARM_TABLE_TWIDDLECOEF_Q31_64) && defined(ARM_TABLE_BITREVIDX_FXT_64))
const arm_rfft_instance_q31 arm_rfft_sR_q31_len128 = {
128U,
0,
1,
64U,
(q31_t*)realCoefAQ31,
(q31_t*)realCoefBQ31,
&arm_cfft_sR_q31_len64
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q31) && defined(ARM_TABLE_TWIDDLECOEF_Q31_128) && defined(ARM_TABLE_BITREVIDX_FXT_128))
const arm_rfft_instance_q31 arm_rfft_sR_q31_len256 = {
256U,
0,
1,
32U,
(q31_t*)realCoefAQ31,
(q31_t*)realCoefBQ31,
&arm_cfft_sR_q31_len128
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q31) && defined(ARM_TABLE_TWIDDLECOEF_Q31_256) && defined(ARM_TABLE_BITREVIDX_FXT_256))
const arm_rfft_instance_q31 arm_rfft_sR_q31_len512 = {
512U,
0,
1,
16U,
(q31_t*)realCoefAQ31,
(q31_t*)realCoefBQ31,
&arm_cfft_sR_q31_len256
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q31) && defined(ARM_TABLE_TWIDDLECOEF_Q31_512) && defined(ARM_TABLE_BITREVIDX_FXT_512))
const arm_rfft_instance_q31 arm_rfft_sR_q31_len1024 = {
1024U,
0,
1,
8U,
(q31_t*)realCoefAQ31,
(q31_t*)realCoefBQ31,
&arm_cfft_sR_q31_len512
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q31) && defined(ARM_TABLE_TWIDDLECOEF_Q31_1024) && defined(ARM_TABLE_BITREVIDX_FXT_1024))
const arm_rfft_instance_q31 arm_rfft_sR_q31_len2048 = {
2048U,
0,
1,
4U,
(q31_t*)realCoefAQ31,
(q31_t*)realCoefBQ31,
&arm_cfft_sR_q31_len1024
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q31) && defined(ARM_TABLE_TWIDDLECOEF_Q31_2048) && defined(ARM_TABLE_BITREVIDX_FXT_2048))
const arm_rfft_instance_q31 arm_rfft_sR_q31_len4096 = {
4096U,
0,
1,
2U,
(q31_t*)realCoefAQ31,
(q31_t*)realCoefBQ31,
&arm_cfft_sR_q31_len2048
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q31) && defined(ARM_TABLE_TWIDDLECOEF_Q31_4096) && defined(ARM_TABLE_BITREVIDX_FXT_4096))
const arm_rfft_instance_q31 arm_rfft_sR_q31_len8192 = {
8192U,
0,
1,
1U,
(q31_t*)realCoefAQ31,
(q31_t*)realCoefBQ31,
&arm_cfft_sR_q31_len4096
};
#endif
/* q15_t */
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q15) && defined(ARM_TABLE_TWIDDLECOEF_Q15_16) && defined(ARM_TABLE_BITREVIDX_FXT_16))
const arm_rfft_instance_q15 arm_rfft_sR_q15_len32 = {
32U,
0,
1,
256U,
(q15_t*)realCoefAQ15,
(q15_t*)realCoefBQ15,
&arm_cfft_sR_q15_len16
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q15) && defined(ARM_TABLE_TWIDDLECOEF_Q15_32) && defined(ARM_TABLE_BITREVIDX_FXT_32))
const arm_rfft_instance_q15 arm_rfft_sR_q15_len64 = {
64U,
0,
1,
128U,
(q15_t*)realCoefAQ15,
(q15_t*)realCoefBQ15,
&arm_cfft_sR_q15_len32
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q15) && defined(ARM_TABLE_TWIDDLECOEF_Q15_64) && defined(ARM_TABLE_BITREVIDX_FXT_64))
const arm_rfft_instance_q15 arm_rfft_sR_q15_len128 = {
128U,
0,
1,
64U,
(q15_t*)realCoefAQ15,
(q15_t*)realCoefBQ15,
&arm_cfft_sR_q15_len64
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q15) && defined(ARM_TABLE_TWIDDLECOEF_Q15_128) && defined(ARM_TABLE_BITREVIDX_FXT_128))
const arm_rfft_instance_q15 arm_rfft_sR_q15_len256 = {
256U,
0,
1,
32U,
(q15_t*)realCoefAQ15,
(q15_t*)realCoefBQ15,
&arm_cfft_sR_q15_len128
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q15) && defined(ARM_TABLE_TWIDDLECOEF_Q15_256) && defined(ARM_TABLE_BITREVIDX_FXT_256))
const arm_rfft_instance_q15 arm_rfft_sR_q15_len512 = {
512U,
0,
1,
16U,
(q15_t*)realCoefAQ15,
(q15_t*)realCoefBQ15,
&arm_cfft_sR_q15_len256
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q15) && defined(ARM_TABLE_TWIDDLECOEF_Q15_512) && defined(ARM_TABLE_BITREVIDX_FXT_512))
const arm_rfft_instance_q15 arm_rfft_sR_q15_len1024 = {
1024U,
0,
1,
8U,
(q15_t*)realCoefAQ15,
(q15_t*)realCoefBQ15,
&arm_cfft_sR_q15_len512
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q15) && defined(ARM_TABLE_TWIDDLECOEF_Q15_1024) && defined(ARM_TABLE_BITREVIDX_FXT_1024))
const arm_rfft_instance_q15 arm_rfft_sR_q15_len2048 = {
2048U,
0,
1,
4U,
(q15_t*)realCoefAQ15,
(q15_t*)realCoefBQ15,
&arm_cfft_sR_q15_len1024
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q15) && defined(ARM_TABLE_TWIDDLECOEF_Q15_2048) && defined(ARM_TABLE_BITREVIDX_FXT_2048))
const arm_rfft_instance_q15 arm_rfft_sR_q15_len4096 = {
4096U,
0,
1,
2U,
(q15_t*)realCoefAQ15,
(q15_t*)realCoefBQ15,
&arm_cfft_sR_q15_len2048
};
#endif
#if !defined(ARM_DSP_CONFIG_TABLES) || defined(ARM_ALL_FFT_TABLES) || (defined(ARM_TABLE_REALCOEF_Q15) && defined(ARM_TABLE_TWIDDLECOEF_Q15_4096) && defined(ARM_TABLE_BITREVIDX_FXT_4096))
const arm_rfft_instance_q15 arm_rfft_sR_q15_len8192 = {
8192U,
0,
1,
1U,
(q15_t*)realCoefAQ15,
(q15_t*)realCoefBQ15,
&arm_cfft_sR_q15_len4096
};
#endif
#endif

16
Drivers/CMSIS/DSP/Source/ComplexMathFunctions/CMakeLists.txt Normal file
View File

@@ -0,0 +1,16 @@
cmake_minimum_required (VERSION 3.6)
project(CMSISDSPComplexMath)
file(GLOB SRC "./*_*.c")
add_library(CMSISDSPComplexMath STATIC ${SRC})
configdsp(CMSISDSPComplexMath ..)
### Includes
target_include_directories(CMSISDSPComplexMath PUBLIC "${DSP}/../../Include")

46
Drivers/CMSIS/DSP/Source/ComplexMathFunctions/ComplexMathFunctions.c Normal file
View File

@@ -0,0 +1,46 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: CompexMathFunctions.c
* Description: Combination of all comlex math function source files.
*
* $Date: 18. March 2019
* $Revision: V1.0.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_cmplx_conj_f32.c"
#include "arm_cmplx_conj_q15.c"
#include "arm_cmplx_conj_q31.c"
#include "arm_cmplx_dot_prod_f32.c"
#include "arm_cmplx_dot_prod_q15.c"
#include "arm_cmplx_dot_prod_q31.c"
#include "arm_cmplx_mag_f32.c"
#include "arm_cmplx_mag_q15.c"
#include "arm_cmplx_mag_q31.c"
#include "arm_cmplx_mag_squared_f32.c"
#include "arm_cmplx_mag_squared_q15.c"
#include "arm_cmplx_mag_squared_q31.c"
#include "arm_cmplx_mult_cmplx_f32.c"
#include "arm_cmplx_mult_cmplx_q15.c"
#include "arm_cmplx_mult_cmplx_q31.c"
#include "arm_cmplx_mult_real_f32.c"
#include "arm_cmplx_mult_real_q15.c"
#include "arm_cmplx_mult_real_q31.c"

161
Drivers/CMSIS/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_f32.c Normal file
View File

@@ -0,0 +1,161 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_cmplx_conj_f32.c
* Description: Floating-point complex conjugate
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupCmplxMath
*/
/**
@defgroup cmplx_conj Complex Conjugate
Conjugates the elements of a complex data vector.
The <code>pSrc</code> points to the source data and
<code>pDst</code> points to the destination data where the result should be written.
<code>numSamples</code> specifies the number of complex samples
and the data in each array is stored in an interleaved fashion
(real, imag, real, imag, ...).
Each array has a total of <code>2*numSamples</code> values.
The underlying algorithm is used:
<pre>
for (n = 0; n < numSamples; n++) {
pDst[(2*n) ] = pSrc[(2*n) ]; // real part
pDst[(2*n)+1] = -pSrc[(2*n)+1]; // imag part
}
</pre>
There are separate functions for floating-point, Q15, and Q31 data types.
*/
/**
@addtogroup cmplx_conj
@{
*/
/**
@brief Floating-point complex conjugate.
@param[in] pSrc points to the input vector
@param[out] pDst points to the output vector
@param[in] numSamples number of samples in each vector
@return none
*/
void arm_cmplx_conj_f32(
const float32_t * pSrc,
float32_t * pDst,
uint32_t numSamples)
{
uint32_t blkCnt; /* Loop counter */
#if defined(ARM_MATH_NEON)
float32x4_t zero;
float32x4x2_t vec;
zero = vdupq_n_f32(0.0);
/* Compute 4 outputs at a time */
blkCnt = numSamples >> 2U;
while (blkCnt > 0U)
{
/* C[0]+jC[1] = A[0]+(-1)*jA[1] */
/* Calculate Complex Conjugate and then store the results in the destination buffer. */
vec = vld2q_f32(pSrc);
vec.val[1] = vsubq_f32(zero,vec.val[1]);
vst2q_f32(pDst,vec);
/* Increment pointers */
pSrc += 8;
pDst += 8;
/* Decrement the loop counter */
blkCnt--;
}
/* Tail */
blkCnt = numSamples & 0x3;
#else
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = numSamples >> 2U;
while (blkCnt > 0U)
{
/* C[0] + jC[1] = A[0]+ j(-1)A[1] */
/* Calculate Complex Conjugate and store result in destination buffer. */
*pDst++ = *pSrc++;
*pDst++ = -*pSrc++;
*pDst++ = *pSrc++;
*pDst++ = -*pSrc++;
*pDst++ = *pSrc++;
*pDst++ = -*pSrc++;
*pDst++ = *pSrc++;
*pDst++ = -*pSrc++;
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = numSamples % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = numSamples;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
#endif /* #if defined (ARM_MATH_NEON) */
while (blkCnt > 0U)
{
/* C[0] + jC[1] = A[0]+ j(-1)A[1] */
/* Calculate Complex Conjugate and store result in destination buffer. */
*pDst++ = *pSrc++;
*pDst++ = -*pSrc++;
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of cmplx_conj group
*/

157
Drivers/CMSIS/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_q15.c Normal file
View File

@@ -0,0 +1,157 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_cmplx_conj_q15.c
* Description: Q15 complex conjugate
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupCmplxMath
*/
/**
@addtogroup cmplx_conj
@{
*/
/**
@brief Q15 complex conjugate.
@param[in] pSrc points to the input vector
@param[out] pDst points to the output vector
@param[in] numSamples number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
The Q15 value -1 (0x8000) is saturated to the maximum allowable positive value 0x7FFF.
*/
void arm_cmplx_conj_q15(
const q15_t * pSrc,
q15_t * pDst,
uint32_t numSamples)
{
uint32_t blkCnt; /* Loop counter */
q31_t in1; /* Temporary input variable */
#if defined (ARM_MATH_LOOPUNROLL) && defined (ARM_MATH_DSP)
q31_t in2, in3, in4; /* Temporary input variables */
#endif
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = numSamples >> 2U;
while (blkCnt > 0U)
{
/* C[0] + jC[1] = A[0]+ j(-1)A[1] */
/* Calculate Complex Conjugate and store result in destination buffer. */
#if defined (ARM_MATH_DSP)
in1 = read_q15x2_ia ((q15_t **) &pSrc);
in2 = read_q15x2_ia ((q15_t **) &pSrc);
in3 = read_q15x2_ia ((q15_t **) &pSrc);
in4 = read_q15x2_ia ((q15_t **) &pSrc);
#ifndef ARM_MATH_BIG_ENDIAN
in1 = __QASX(0, in1);
in2 = __QASX(0, in2);
in3 = __QASX(0, in3);
in4 = __QASX(0, in4);
#else
in1 = __QSAX(0, in1);
in2 = __QSAX(0, in2);
in3 = __QSAX(0, in3);
in4 = __QSAX(0, in4);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
in1 = ((uint32_t) in1 >> 16) | ((uint32_t) in1 << 16);
in2 = ((uint32_t) in2 >> 16) | ((uint32_t) in2 << 16);
in3 = ((uint32_t) in3 >> 16) | ((uint32_t) in3 << 16);
in4 = ((uint32_t) in4 >> 16) | ((uint32_t) in4 << 16);
write_q15x2_ia (&pDst, in1);
write_q15x2_ia (&pDst, in2);
write_q15x2_ia (&pDst, in3);
write_q15x2_ia (&pDst, in4);
#else
*pDst++ = *pSrc++;
in1 = *pSrc++;
*pDst++ = (in1 == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in1;
*pDst++ = *pSrc++;
in1 = *pSrc++;
*pDst++ = (in1 == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in1;
*pDst++ = *pSrc++;
in1 = *pSrc++;
*pDst++ = (in1 == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in1;
*pDst++ = *pSrc++;
in1 = *pSrc++;
*pDst++ = (in1 == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in1;
#endif /* #if defined (ARM_MATH_DSP) */
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = numSamples % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = numSamples;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C[0] + jC[1] = A[0]+ j(-1)A[1] */
/* Calculate Complex Conjugate and store result in destination buffer. */
*pDst++ = *pSrc++;
in1 = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = __SSAT(-in1, 16);
#else
*pDst++ = (in1 == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in1;
#endif
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of cmplx_conj group
*/

137
Drivers/CMSIS/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_q31.c Normal file
View File

@@ -0,0 +1,137 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_cmplx_conj_q31.c
* Description: Q31 complex conjugate
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupCmplxMath
*/
/**
@addtogroup cmplx_conj
@{
*/
/**
@brief Q31 complex conjugate.
@param[in] pSrc points to the input vector
@param[out] pDst points to the output vector
@param[in] numSamples number of samples in each vector
@return none
@par Scaling and Overflow Behavior
The function uses saturating arithmetic.
The Q31 value -1 (0x80000000) is saturated to the maximum allowable positive value 0x7FFFFFFF.
*/
void arm_cmplx_conj_q31(
const q31_t * pSrc,
q31_t * pDst,
uint32_t numSamples)
{
uint32_t blkCnt; /* Loop counter */
q31_t in; /* Temporary input variable */
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = numSamples >> 2U;
while (blkCnt > 0U)
{
/* C[0] + jC[1] = A[0]+ j(-1)A[1] */
/* Calculate Complex Conjugate and store result in destination buffer. */
*pDst++ = *pSrc++;
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = __QSUB(0, in);
#else
*pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
#endif
*pDst++ = *pSrc++;
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = __QSUB(0, in);
#else
*pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
#endif
*pDst++ = *pSrc++;
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = __QSUB(0, in);
#else
*pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
#endif
*pDst++ = *pSrc++;
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = __QSUB(0, in);
#else
*pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
#endif
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = numSamples % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = numSamples;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
/* C[0] + jC[1] = A[0]+ j(-1)A[1] */
/* Calculate Complex Conjugate and store result in destination buffer. */
*pDst++ = *pSrc++;
in = *pSrc++;
#if defined (ARM_MATH_DSP)
*pDst++ = __QSUB(0, in);
#else
*pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
#endif
/* Decrement loop counter */
blkCnt--;
}
}
/**
@} end of cmplx_conj group
*/

233
Drivers/CMSIS/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_f32.c Normal file
View File

@@ -0,0 +1,233 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_cmplx_dot_prod_f32.c
* Description: Floating-point complex dot product
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupCmplxMath
*/
/**
@defgroup cmplx_dot_prod Complex Dot Product
Computes the dot product of two complex vectors.
The vectors are multiplied element-by-element and then summed.
The <code>pSrcA</code> points to the first complex input vector and
<code>pSrcB</code> points to the second complex input vector.
<code>numSamples</code> specifies the number of complex samples
and the data in each array is stored in an interleaved fashion
(real, imag, real, imag, ...).
Each array has a total of <code>2*numSamples</code> values.
The underlying algorithm is used:
<pre>
realResult = 0;
imagResult = 0;
for (n = 0; n < numSamples; n++) {
realResult += pSrcA[(2*n)+0] * pSrcB[(2*n)+0] - pSrcA[(2*n)+1] * pSrcB[(2*n)+1];
imagResult += pSrcA[(2*n)+0] * pSrcB[(2*n)+1] + pSrcA[(2*n)+1] * pSrcB[(2*n)+0];
}
</pre>
There are separate functions for floating-point, Q15, and Q31 data types.
*/
/**
@addtogroup cmplx_dot_prod
@{
*/
/**
@brief Floating-point complex dot product.
@param[in] pSrcA points to the first input vector
@param[in] pSrcB points to the second input vector
@param[in] numSamples number of samples in each vector
@param[out] realResult real part of the result returned here
@param[out] imagResult imaginary part of the result returned here
@return none
*/
void arm_cmplx_dot_prod_f32(
const float32_t * pSrcA,
const float32_t * pSrcB,
uint32_t numSamples,
float32_t * realResult,
float32_t * imagResult)
{
uint32_t blkCnt; /* Loop counter */
float32_t real_sum = 0.0f, imag_sum = 0.0f; /* Temporary result variables */
float32_t a0,b0,c0,d0;
#if defined(ARM_MATH_NEON)
float32x4x2_t vec1,vec2,vec3,vec4;
float32x4_t accR,accI;
float32x2_t accum = vdup_n_f32(0);
accR = vdupq_n_f32(0.0);
accI = vdupq_n_f32(0.0);
/* Loop unrolling: Compute 8 outputs at a time */
blkCnt = numSamples >> 3U;
while (blkCnt > 0U)
{
/* C = (A[0]+jA[1])*(B[0]+jB[1]) + ... */
/* Calculate dot product and then store the result in a temporary buffer. */
vec1 = vld2q_f32(pSrcA);
vec2 = vld2q_f32(pSrcB);
/* Increment pointers */
pSrcA += 8;
pSrcB += 8;
/* Re{C} = Re{A}*Re{B} - Im{A}*Im{B} */
accR = vmlaq_f32(accR,vec1.val[0],vec2.val[0]);
accR = vmlsq_f32(accR,vec1.val[1],vec2.val[1]);
/* Im{C} = Re{A}*Im{B} + Im{A}*Re{B} */
accI = vmlaq_f32(accI,vec1.val[1],vec2.val[0]);
accI = vmlaq_f32(accI,vec1.val[0],vec2.val[1]);
vec3 = vld2q_f32(pSrcA);
vec4 = vld2q_f32(pSrcB);
/* Increment pointers */
pSrcA += 8;
pSrcB += 8;
/* Re{C} = Re{A}*Re{B} - Im{A}*Im{B} */
accR = vmlaq_f32(accR,vec3.val[0],vec4.val[0]);
accR = vmlsq_f32(accR,vec3.val[1],vec4.val[1]);
/* Im{C} = Re{A}*Im{B} + Im{A}*Re{B} */
accI = vmlaq_f32(accI,vec3.val[1],vec4.val[0]);
accI = vmlaq_f32(accI,vec3.val[0],vec4.val[1]);
/* Decrement the loop counter */
blkCnt--;
}
accum = vpadd_f32(vget_low_f32(accR), vget_high_f32(accR));
real_sum += accum[0] + accum[1];
accum = vpadd_f32(vget_low_f32(accI), vget_high_f32(accI));
imag_sum += accum[0] + accum[1];
/* Tail */
blkCnt = numSamples & 0x7;
#else
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = numSamples >> 2U;
while (blkCnt > 0U)
{
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += a0 * c0;
imag_sum += a0 * d0;
real_sum -= b0 * d0;
imag_sum += b0 * c0;
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += a0 * c0;
imag_sum += a0 * d0;
real_sum -= b0 * d0;
imag_sum += b0 * c0;
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += a0 * c0;
imag_sum += a0 * d0;
real_sum -= b0 * d0;
imag_sum += b0 * c0;
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += a0 * c0;
imag_sum += a0 * d0;
real_sum -= b0 * d0;
imag_sum += b0 * c0;
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = numSamples % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = numSamples;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
#endif /* #if defined(ARM_MATH_NEON) */
while (blkCnt > 0U)
{
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += a0 * c0;
imag_sum += a0 * d0;
real_sum -= b0 * d0;
imag_sum += b0 * c0;
/* Decrement loop counter */
blkCnt--;
}
/* Store real and imaginary result in destination buffer. */
*realResult = real_sum;
*imagResult = imag_sum;
}
/**
@} end of cmplx_dot_prod group
*/

154
Drivers/CMSIS/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_q15.c Normal file
View File

@@ -0,0 +1,154 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_cmplx_dot_prod_q15.c
* Description: Processing function for the Q15 Complex Dot product
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupCmplxMath
*/
/**
@addtogroup cmplx_dot_prod
@{
*/
/**
@brief Q15 complex dot product.
@param[in] pSrcA points to the first input vector
@param[in] pSrcB points to the second input vector
@param[in] numSamples number of samples in each vector
@param[out] realResult real part of the result returned here
@param[out] imagResult imaginary part of the result returned her
@return none
@par Scaling and Overflow Behavior
The function is implemented using an internal 64-bit accumulator.
The intermediate 1.15 by 1.15 multiplications are performed with full precision and yield a 2.30 result.
These are accumulated in a 64-bit accumulator with 34.30 precision.
As a final step, the accumulators are converted to 8.24 format.
The return results <code>realResult</code> and <code>imagResult</code> are in 8.24 format.
*/
void arm_cmplx_dot_prod_q15(
const q15_t * pSrcA,
const q15_t * pSrcB,
uint32_t numSamples,
q31_t * realResult,
q31_t * imagResult)
{
uint32_t blkCnt; /* Loop counter */
q63_t real_sum = 0, imag_sum = 0; /* Temporary result variables */
q15_t a0,b0,c0,d0;
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = numSamples >> 2U;
while (blkCnt > 0U)
{
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += (q31_t)a0 * c0;
imag_sum += (q31_t)a0 * d0;
real_sum -= (q31_t)b0 * d0;
imag_sum += (q31_t)b0 * c0;
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += (q31_t)a0 * c0;
imag_sum += (q31_t)a0 * d0;
real_sum -= (q31_t)b0 * d0;
imag_sum += (q31_t)b0 * c0;
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += (q31_t)a0 * c0;
imag_sum += (q31_t)a0 * d0;
real_sum -= (q31_t)b0 * d0;
imag_sum += (q31_t)b0 * c0;
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += (q31_t)a0 * c0;
imag_sum += (q31_t)a0 * d0;
real_sum -= (q31_t)b0 * d0;
imag_sum += (q31_t)b0 * c0;
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = numSamples % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = numSamples;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += (q31_t)a0 * c0;
imag_sum += (q31_t)a0 * d0;
real_sum -= (q31_t)b0 * d0;
imag_sum += (q31_t)b0 * c0;
/* Decrement loop counter */
blkCnt--;
}
/* Store real and imaginary result in 8.24 format */
/* Convert real data in 34.30 to 8.24 by 6 right shifts */
*realResult = (q31_t) (real_sum >> 6);
/* Convert imaginary data in 34.30 to 8.24 by 6 right shifts */
*imagResult = (q31_t) (imag_sum >> 6);
}
/**
@} end of cmplx_dot_prod group
*/

153
Drivers/CMSIS/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_q31.c Normal file
View File

@@ -0,0 +1,153 @@
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_cmplx_dot_prod_q31.c
* Description: Q31 complex dot product
*
* $Date: 18. March 2019
* $Revision: V1.6.0
*
* Target Processor: Cortex-M cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "arm_math.h"
/**
@ingroup groupCmplxMath
*/
/**
@addtogroup cmplx_dot_prod
@{
*/
/**
@brief Q31 complex dot product.
@param[in] pSrcA points to the first input vector
@param[in] pSrcB points to the second input vector
@param[in] numSamples number of samples in each vector
@param[out] realResult real part of the result returned here
@param[out] imagResult imaginary part of the result returned here
@return none
@par Scaling and Overflow Behavior
The function is implemented using an internal 64-bit accumulator.
The intermediate 1.31 by 1.31 multiplications are performed with 64-bit precision and then shifted to 16.48 format.
The internal real and imaginary accumulators are in 16.48 format and provide 15 guard bits.
Additions are nonsaturating and no overflow will occur as long as <code>numSamples</code> is less than 32768.
The return results <code>realResult</code> and <code>imagResult</code> are in 16.48 format.
Input down scaling is not required.
*/
void arm_cmplx_dot_prod_q31(
const q31_t * pSrcA,
const q31_t * pSrcB,
uint32_t numSamples,
q63_t * realResult,
q63_t * imagResult)
{
uint32_t blkCnt; /* Loop counter */
q63_t real_sum = 0, imag_sum = 0; /* Temporary result variables */
q31_t a0,b0,c0,d0;
#if defined (ARM_MATH_LOOPUNROLL)
/* Loop unrolling: Compute 4 outputs at a time */
blkCnt = numSamples >> 2U;
while (blkCnt > 0U)
{
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += ((q63_t)a0 * c0) >> 14;
imag_sum += ((q63_t)a0 * d0) >> 14;
real_sum -= ((q63_t)b0 * d0) >> 14;
imag_sum += ((q63_t)b0 * c0) >> 14;
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += ((q63_t)a0 * c0) >> 14;
imag_sum += ((q63_t)a0 * d0) >> 14;
real_sum -= ((q63_t)b0 * d0) >> 14;
imag_sum += ((q63_t)b0 * c0) >> 14;
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += ((q63_t)a0 * c0) >> 14;
imag_sum += ((q63_t)a0 * d0) >> 14;
real_sum -= ((q63_t)b0 * d0) >> 14;
imag_sum += ((q63_t)b0 * c0) >> 14;
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += ((q63_t)a0 * c0) >> 14;
imag_sum += ((q63_t)a0 * d0) >> 14;
real_sum -= ((q63_t)b0 * d0) >> 14;
imag_sum += ((q63_t)b0 * c0) >> 14;
/* Decrement loop counter */
blkCnt--;
}
/* Loop unrolling: Compute remaining outputs */
blkCnt = numSamples % 0x4U;
#else
/* Initialize blkCnt with number of samples */
blkCnt = numSamples;
#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
while (blkCnt > 0U)
{
a0 = *pSrcA++;
b0 = *pSrcA++;
c0 = *pSrcB++;
d0 = *pSrcB++;
real_sum += ((q63_t)a0 * c0) >> 14;
imag_sum += ((q63_t)a0 * d0) >> 14;
real_sum -= ((q63_t)b0 * d0) >> 14;
imag_sum += ((q63_t)b0 * c0) >> 14;
/* Decrement loop counter */
blkCnt--;
}
/* Store real and imaginary result in 16.48 format */
*realResult = real_sum;
*imagResult = imag_sum;
}
/**
@} end of cmplx_dot_prod group
*/

Some files were not shown because too many files have changed in this diff Show More