--- /dev/null
+/* ----------------------------------------------------------------------\r
+ * Copyright (C) 2010-2011 ARM Limited. All rights reserved.\r
+ *\r
+ * $Date: 15. July 2011\r
+ * $Revision: V1.0.10\r
+ *\r
+ * Project: CMSIS DSP Library\r
+ * Title: arm_math.h\r
+ *\r
+ * Description: Public header file for CMSIS DSP Library\r
+ *\r
+ * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0\r
+ *\r
+ * Version 1.0.10 2011/7/15\r
+ * Big Endian support added and Merged M0 and M3/M4 Source code.\r
+ *\r
+ * Version 1.0.3 2010/11/29\r
+ * Re-organized the CMSIS folders and updated documentation.\r
+ *\r
+ * Version 1.0.2 2010/11/11\r
+ * Documentation updated.\r
+ *\r
+ * Version 1.0.1 2010/10/05\r
+ * Production release and review comments incorporated.\r
+ *\r
+ * Version 1.0.0 2010/09/20\r
+ * Production release and review comments incorporated.\r
+ * -------------------------------------------------------------------- */\r
+\r
+/**\r
+ \mainpage CMSIS DSP Software Library\r
+ *\r
+ * <b>Introduction</b>\r
+ *\r
+ * This user manual describes the CMSIS DSP software library,\r
+ * a suite of common signal processing functions for use on Cortex-M processor based devices.\r
+ *\r
+ * The library is divided into a number of modules each covering a specific category:\r
+ * - Basic math functions\r
+ * - Fast math functions\r
+ * - Complex math functions\r
+ * - Filters\r
+ * - Matrix functions\r
+ * - Transforms\r
+ * - Motor control functions\r
+ * - Statistical functions\r
+ * - Support functions\r
+ * - Interpolation functions\r
+ *\r
+ * The library has separate functions for operating on 8-bit integers, 16-bit integers,\r
+ * 32-bit integer and 32-bit floating-point values.\r
+ *\r
+ * <b>Processor Support</b>\r
+ *\r
+ * The library is completely written in C and is fully CMSIS compliant.\r
+ * High performance is achieved through maximum use of Cortex-M4 intrinsics.\r
+ *\r
+ * The supplied library source code also builds and runs on the Cortex-M3 and Cortex-M0 processor,\r
+ * with the DSP intrinsics being emulated through software.\r
+ *\r
+ *\r
+ * <b>Toolchain Support</b>\r
+ *\r
+ * The library has been developed and tested with MDK-ARM version 4.21.\r
+ * The library is being tested in GCC and IAR toolchains and updates on this activity will be made available shortly.\r
+ *\r
+ * <b>Using the Library</b>\r
+ *\r
+ * The library installer contains prebuilt versions of the libraries in the <code>Lib</code> folder.\r
+ * - arm_cortexM4lf_math.lib (Little endian and Floating Point Unit on Cortex-M4)\r
+ * - arm_cortexM4bf_math.lib (Big endian and Floating Point Unit on Cortex-M4)\r
+ * - arm_cortexM4l_math.lib (Little endian on Cortex-M4)\r
+ * - arm_cortexM4b_math.lib (Big endian on Cortex-M4)\r
+ * - arm_cortexM3l_math.lib (Little endian on Cortex-M3)\r
+ * - arm_cortexM3b_math.lib (Big endian on Cortex-M3)\r
+ * - arm_cortexM0l_math.lib (Little endian on Cortex-M0)\r
+ * - arm_cortexM0b_math.lib (Big endian on Cortex-M3)\r
+ *\r
+ * The library functions are declared in the public file <code>arm_math.h</code> which is placed in the <code>Include</code> folder.\r
+ * Simply include this file and link the appropriate library in the application and begin calling the library functions. The Library supports single\r
+ * public header file <code>arm_math.h</code> for Cortex-M4/M3/M0 with little endian and big endian. Same header file will be used for floating point unit(FPU) variants.\r
+ * Define the appropriate pre processor MACRO ARM_MATH_CM4 or ARM_MATH_CM3 or\r
+ * ARM_MATH_CM0 depending on the target processor in the application.\r
+ *\r
+ * <b>Examples</b>\r
+ *\r
+ * The library ships with a number of examples which demonstrate how to use the library functions.\r
+ *\r
+ * <b>Building the Library</b>\r
+ *\r
+ * The library installer contains project files to re build libraries on MDK Tool chain in the <code>CMSIS\\DSP_Lib\\Source\\ARM</code> folder.\r
+ * - arm_cortexM0b_math.uvproj\r
+ * - arm_cortexM0l_math.uvproj\r
+ * - arm_cortexM3b_math.uvproj\r
+ * - arm_cortexM3l_math.uvproj\r
+ * - arm_cortexM4b_math.uvproj\r
+ * - arm_cortexM4l_math.uvproj\r
+ * - arm_cortexM4bf_math.uvproj\r
+ * - arm_cortexM4lf_math.uvproj\r
+ *\r
+ * Each library project have differant pre-processor macros.\r
+ *\r
+ * <b>ARM_MATH_CMx:</b>\r
+ * Define macro ARM_MATH_CM4 for building the library on Cortex-M4 target, ARM_MATH_CM3 for building library on Cortex-M3 target\r
+ * and ARM_MATH_CM0 for building library on cortex-M0 target.\r
+ *\r
+ * <b>ARM_MATH_BIG_ENDIAN:</b>\r
+ * Define macro ARM_MATH_BIG_ENDIAN to build the library for big endian targets. By default library builds for little endian targets.\r
+ *\r
+ * <b>ARM_MATH_MATRIX_CHECK:</b>\r
+ * Define macro for checking on the input and output sizes of matrices\r
+ *\r
+ * <b>ARM_MATH_ROUNDING:</b>\r
+ * Define macro for rounding on support functions\r
+ *\r
+ * <b>__FPU_PRESENT:</b>\r
+ * Initialize macro __FPU_PRESENT = 1 when building on FPU supported Targets. Enable this macro for M4bf and M4lf libraries\r
+ *\r
+ *\r
+ * The project can be built by opening the appropriate project in MDK-ARM 4.21 chain and defining the optional pre processor MACROs detailed above.\r
+ *\r
+ * <b>Copyright Notice</b>\r
+ *\r
+ * Copyright (C) 2010 ARM Limited. All rights reserved.\r
+ */\r
+\r
+\r
+/**\r
+ * @defgroup groupMath Basic Math Functions\r
+ */\r
+\r
+/**\r
+ * @defgroup groupFastMath Fast Math Functions\r
+ * This set of functions provides a fast approximation to sine, cosine, and square root.\r
+ * As compared to most of the other functions in the CMSIS math library, the fast math functions\r
+ * operate on individual values and not arrays.\r
+ * There are separate functions for Q15, Q31, and floating-point data.\r
+ *\r
+ */\r
+\r
+/**\r
+ * @defgroup groupCmplxMath Complex Math Functions\r
+ * This set of functions operates on complex data vectors.\r
+ * The data in the complex arrays is stored in an interleaved fashion\r
+ * (real, imag, real, imag, ...).\r
+ * In the API functions, the number of samples in a complex array refers\r
+ * to the number of complex values; the array contains twice this number of\r
+ * real values.\r
+ */\r
+\r
+/**\r
+ * @defgroup groupFilters Filtering Functions\r
+ */\r
+\r
+/**\r
+ * @defgroup groupMatrix Matrix Functions\r
+ *\r
+ * This set of functions provides basic matrix math operations.\r
+ * The functions operate on matrix data structures. For example,\r
+ * the type\r
+ * definition for the floating-point matrix structure is shown\r
+ * below:\r
+ * <pre>\r
+ * typedef struct\r
+ * {\r
+ * uint16_t numRows; // number of rows of the matrix.\r
+ * uint16_t numCols; // number of columns of the matrix.\r
+ * float32_t *pData; // points to the data of the matrix.\r
+ * } arm_matrix_instance_f32;\r
+ * </pre>\r
+ * There are similar definitions for Q15 and Q31 data types.\r
+ *\r
+ * The structure specifies the size of the matrix and then points to\r
+ * an array of data. The array is of size <code>numRows X numCols</code>\r
+ * and the values are arranged in row order. That is, the\r
+ * matrix element (i, j) is stored at:\r
+ * <pre>\r
+ * pData[i*numCols + j]\r
+ * </pre>\r
+ *\r
+ * \par Init Functions\r
+ * There is an associated initialization function for each type of matrix\r
+ * data structure.\r
+ * The initialization function sets the values of the internal structure fields.\r
+ * Refer to the function <code>arm_mat_init_f32()</code>, <code>arm_mat_init_q31()</code>\r
+ * and <code>arm_mat_init_q15()</code> for floating-point, Q31 and Q15 types, respectively.\r
+ *\r
+ * \par\r
+ * Use of the initialization function is optional. However, if initialization function is used\r
+ * then the instance structure cannot be placed into a const data section.\r
+ * To place the instance structure in a const data\r
+ * section, manually initialize the data structure. For example:\r
+ * <pre>\r
+ * <code>arm_matrix_instance_f32 S = {nRows, nColumns, pData};</code>\r
+ * <code>arm_matrix_instance_q31 S = {nRows, nColumns, pData};</code>\r
+ * <code>arm_matrix_instance_q15 S = {nRows, nColumns, pData};</code>\r
+ * </pre>\r
+ * where <code>nRows</code> specifies the number of rows, <code>nColumns</code>\r
+ * specifies the number of columns, and <code>pData</code> points to the\r
+ * data array.\r
+ *\r
+ * \par Size Checking\r
+ * By default all of the matrix functions perform size checking on the input and\r
+ * output matrices. For example, the matrix addition function verifies that the\r
+ * two input matrices and the output matrix all have the same number of rows and\r
+ * columns. If the size check fails the functions return:\r
+ * <pre>\r
+ * ARM_MATH_SIZE_MISMATCH\r
+ * </pre>\r
+ * Otherwise the functions return\r
+ * <pre>\r
+ * ARM_MATH_SUCCESS\r
+ * </pre>\r
+ * There is some overhead associated with this matrix size checking.\r
+ * The matrix size checking is enabled via the \#define\r
+ * <pre>\r
+ * ARM_MATH_MATRIX_CHECK\r
+ * </pre>\r
+ * within the library project settings. By default this macro is defined\r
+ * and size checking is enabled. By changing the project settings and\r
+ * undefining this macro size checking is eliminated and the functions\r
+ * run a bit faster. With size checking disabled the functions always\r
+ * return <code>ARM_MATH_SUCCESS</code>.\r
+ */\r
+\r
+/**\r
+ * @defgroup groupTransforms Transform Functions\r
+ */\r
+\r
+/**\r
+ * @defgroup groupController Controller Functions\r
+ */\r
+\r
+/**\r
+ * @defgroup groupStats Statistics Functions\r
+ */\r
+/**\r
+ * @defgroup groupSupport Support Functions\r
+ */\r
+\r
+/**\r
+ * @defgroup groupInterpolation Interpolation Functions\r
+ * These functions perform 1- and 2-dimensional interpolation of data.\r
+ * Linear interpolation is used for 1-dimensional data and\r
+ * bilinear interpolation is used for 2-dimensional data.\r
+ */\r
+\r
+/**\r
+ * @defgroup groupExamples Examples\r
+ */\r
+#ifndef _ARM_MATH_H\r
+#define _ARM_MATH_H\r
+\r
+#define __CMSIS_GENERIC /* disable NVIC and Systick functions */\r
+\r
+#if defined (ARM_MATH_CM4)\r
+ #include "core_cm4.h"\r
+#elif defined (ARM_MATH_CM3)\r
+ #include "core_cm3.h"\r
+#elif defined (ARM_MATH_CM0)\r
+ #include "core_cm0.h"\r
+#else\r
+#include "ARMCM4.h"\r
+#warning "Define either ARM_MATH_CM4 OR ARM_MATH_CM3...By Default building on ARM_MATH_CM4....."\r
+#endif\r
+\r
+#undef __CMSIS_GENERIC /* enable NVIC and Systick functions */\r
+#include "string.h"\r
+ #include "math.h"\r
+#ifdef __cplusplus\r
+extern "C"\r
+{\r
+#endif\r
+\r
+\r
+ /**\r
+ * @brief Macros required for reciprocal calculation in Normalized LMS\r
+ */\r
+\r
+#define DELTA_Q31 (0x100)\r
+#define DELTA_Q15 0x5\r
+#define INDEX_MASK 0x0000003F\r
+#define PI 3.14159265358979f\r
+\r
+ /**\r
+ * @brief Macros required for SINE and COSINE Fast math approximations\r
+ */\r
+\r
+#define TABLE_SIZE 256\r
+#define TABLE_SPACING_Q31 0x800000\r
+#define TABLE_SPACING_Q15 0x80\r
+\r
+ /**\r
+ * @brief Macros required for SINE and COSINE Controller functions\r
+ */\r
+ /* 1.31(q31) Fixed value of 2/360 */\r
+ /* -1 to +1 is divided into 360 values so total spacing is (2/360) */\r
+#define INPUT_SPACING 0xB60B61\r
+\r
+\r
+ /**\r
+ * @brief Error status returned by some functions in the library.\r
+ */\r
+\r
+ typedef enum\r
+ {\r
+ ARM_MATH_SUCCESS = 0, /**< No error */\r
+ ARM_MATH_ARGUMENT_ERROR = -1, /**< One or more arguments are incorrect */\r
+ ARM_MATH_LENGTH_ERROR = -2, /**< Length of data buffer is incorrect */\r
+ ARM_MATH_SIZE_MISMATCH = -3, /**< Size of matrices is not compatible with the operation. */\r
+ ARM_MATH_NANINF = -4, /**< Not-a-number (NaN) or infinity is generated */\r
+ ARM_MATH_SINGULAR = -5, /**< Generated by matrix inversion if the input matrix is singular and cannot be inverted. */\r
+ ARM_MATH_TEST_FAILURE = -6 /**< Test Failed */\r
+ } arm_status;\r
+\r
+ /**\r
+ * @brief 8-bit fractional data type in 1.7 format.\r
+ */\r
+ typedef int8_t q7_t;\r
+\r
+ /**\r
+ * @brief 16-bit fractional data type in 1.15 format.\r
+ */\r
+ typedef int16_t q15_t;\r
+\r
+ /**\r
+ * @brief 32-bit fractional data type in 1.31 format.\r
+ */\r
+ typedef int32_t q31_t;\r
+\r
+ /**\r
+ * @brief 64-bit fractional data type in 1.63 format.\r
+ */\r
+ typedef int64_t q63_t;\r
+\r
+ /**\r
+ * @brief 32-bit floating-point type definition.\r
+ */\r
+ typedef float float32_t;\r
+\r
+ /**\r
+ * @brief 64-bit floating-point type definition.\r
+ */\r
+ typedef double float64_t;\r
+\r
+ /**\r
+ * @brief definition to read/write two 16 bit values.\r
+ */\r
+#define __SIMD32(addr) (*(int32_t **) & (addr))\r
+\r
+#if defined (ARM_MATH_CM3) || defined (ARM_MATH_CM0)\r
+ /**\r
+ * @brief definition to pack two 16 bit values.\r
+ */\r
+#define __PKHBT(ARG1, ARG2, ARG3) ( (((int32_t)(ARG1) << 0) & (int32_t)0x0000FFFF) | \\r
+ (((int32_t)(ARG2) << ARG3) & (int32_t)0xFFFF0000) )\r
+\r
+#endif\r
+\r
+\r
+ /**\r
+ * @brief definition to pack four 8 bit values.\r
+ */\r
+#ifndef ARM_MATH_BIG_ENDIAN\r
+\r
+#define __PACKq7(v0,v1,v2,v3) ( (((int32_t)(v0) << 0) & (int32_t)0x000000FF) | \\r
+ (((int32_t)(v1) << 8) & (int32_t)0x0000FF00) | \\r
+ (((int32_t)(v2) << 16) & (int32_t)0x00FF0000) | \\r
+ (((int32_t)(v3) << 24) & (int32_t)0xFF000000) )\r
+#else\r
+\r
+#define __PACKq7(v0,v1,v2,v3) ( (((int32_t)(v3) << 0) & (int32_t)0x000000FF) | \\r
+ (((int32_t)(v2) << 8) & (int32_t)0x0000FF00) | \\r
+ (((int32_t)(v1) << 16) & (int32_t)0x00FF0000) | \\r
+ (((int32_t)(v0) << 24) & (int32_t)0xFF000000) )\r
+\r
+#endif\r
+\r
+\r
+ /**\r
+ * @brief Clips Q63 to Q31 values.\r
+ */\r
+ __STATIC_INLINE q31_t clip_q63_to_q31(\r
+ q63_t x)\r
+ {\r
+ return ((q31_t) (x >> 32) != ((q31_t) x >> 31)) ?\r
+ ((0x7FFFFFFF ^ ((q31_t) (x >> 63)))) : (q31_t) x;\r
+ }\r
+\r
+ /**\r
+ * @brief Clips Q63 to Q15 values.\r
+ */\r
+ __STATIC_INLINE q15_t clip_q63_to_q15(\r
+ q63_t x)\r
+ {\r
+ return ((q31_t) (x >> 32) != ((q31_t) x >> 31)) ?\r
+ ((0x7FFF ^ ((q15_t) (x >> 63)))) : (q15_t) (x >> 15);\r
+ }\r
+\r
+ /**\r
+ * @brief Clips Q31 to Q7 values.\r
+ */\r
+ __STATIC_INLINE q7_t clip_q31_to_q7(\r
+ q31_t x)\r
+ {\r
+ return ((q31_t) (x >> 24) != ((q31_t) x >> 23)) ?\r
+ ((0x7F ^ ((q7_t) (x >> 31)))) : (q7_t) x;\r
+ }\r
+\r
+ /**\r
+ * @brief Clips Q31 to Q15 values.\r
+ */\r
+ __STATIC_INLINE q15_t clip_q31_to_q15(\r
+ q31_t x)\r
+ {\r
+ return ((q31_t) (x >> 16) != ((q31_t) x >> 15)) ?\r
+ ((0x7FFF ^ ((q15_t) (x >> 31)))) : (q15_t) x;\r
+ }\r
+\r
+ /**\r
+ * @brief Multiplies 32 X 64 and returns 32 bit result in 2.30 format.\r
+ */\r
+\r
+ __STATIC_INLINE q63_t mult32x64(\r
+ q63_t x,\r
+ q31_t y)\r
+ {\r
+ return ((((q63_t) (x & 0x00000000FFFFFFFF) * y) >> 32) +\r
+ (((q63_t) (x >> 32) * y)));\r
+ }\r
+\r
+\r
+#if defined (ARM_MATH_CM0) && defined ( __CC_ARM )\r
+#define __CLZ __clz\r
+#endif\r
+\r
+#if defined (ARM_MATH_CM0) && defined ( __TASKING__ )\r
+/* No need to redefine __CLZ */\r
+#endif\r
+\r
+#if defined (ARM_MATH_CM0) && ((defined (__ICCARM__)) ||(defined (__GNUC__)) )\r
+\r
+ __STATIC_INLINE uint32_t __CLZ(q31_t data);\r
+\r
+\r
+ __STATIC_INLINE uint32_t __CLZ(q31_t data)\r
+ {\r
+ uint32_t count = 0;\r
+ uint32_t mask = 0x80000000;\r
+\r
+ while((data & mask) == 0)\r
+ {\r
+ count += 1u;\r
+ mask = mask >> 1u;\r
+ }\r
+\r
+ return(count);\r
+\r
+ }\r
+\r
+#endif\r
+\r
+ /**\r
+ * @brief Function to Calculates 1/in(reciprocal) value of Q31 Data type.\r
+ */\r
+\r
+ __STATIC_INLINE uint32_t arm_recip_q31(\r
+ q31_t in,\r
+ q31_t * dst,\r
+ q31_t * pRecipTable)\r
+ {\r
+\r
+ uint32_t out, tempVal;\r
+ uint32_t index, i;\r
+ uint32_t signBits;\r
+\r
+ if(in > 0)\r
+ {\r
+ signBits = __CLZ(in) - 1;\r
+ }\r
+ else\r
+ {\r
+ signBits = __CLZ(-in) - 1;\r
+ }\r
+\r
+ /* Convert input sample to 1.31 format */\r
+ in = in << signBits;\r
+\r
+ /* calculation of index for initial approximated Val */\r
+ index = (uint32_t) (in >> 24u);\r
+ index = (index & INDEX_MASK);\r
+\r
+ /* 1.31 with exp 1 */\r
+ out = pRecipTable[index];\r
+\r
+ /* calculation of reciprocal value */\r
+ /* running approximation for two iterations */\r
+ for (i = 0u; i < 2u; i++)\r
+ {\r
+ tempVal = (q31_t) (((q63_t) in * out) >> 31u);\r
+ tempVal = 0x7FFFFFFF - tempVal;\r
+ /* 1.31 with exp 1 */\r
+ //out = (q31_t) (((q63_t) out * tempVal) >> 30u);\r
+ out = (q31_t) clip_q63_to_q31(((q63_t) out * tempVal) >> 30u);\r
+ }\r
+\r
+ /* write output */\r
+ *dst = out;\r
+\r
+ /* return num of signbits of out = 1/in value */\r
+ return (signBits + 1u);\r
+\r
+ }\r
+\r
+ /**\r
+ * @brief Function to Calculates 1/in(reciprocal) value of Q15 Data type.\r
+ */\r
+ __STATIC_INLINE uint32_t arm_recip_q15(\r
+ q15_t in,\r
+ q15_t * dst,\r
+ q15_t * pRecipTable)\r
+ {\r
+\r
+ uint32_t out = 0, tempVal = 0;\r
+ uint32_t index = 0, i = 0;\r
+ uint32_t signBits = 0;\r
+\r
+ if(in > 0)\r
+ {\r
+ signBits = __CLZ(in) - 17;\r
+ }\r
+ else\r
+ {\r
+ signBits = __CLZ(-in) - 17;\r
+ }\r
+\r
+ /* Convert input sample to 1.15 format */\r
+ in = in << signBits;\r
+\r
+ /* calculation of index for initial approximated Val */\r
+ index = in >> 8;\r
+ index = (index & INDEX_MASK);\r
+\r
+ /* 1.15 with exp 1 */\r
+ out = pRecipTable[index];\r
+\r
+ /* calculation of reciprocal value */\r
+ /* running approximation for two iterations */\r
+ for (i = 0; i < 2; i++)\r
+ {\r
+ tempVal = (q15_t) (((q31_t) in * out) >> 15);\r
+ tempVal = 0x7FFF - tempVal;\r
+ /* 1.15 with exp 1 */\r
+ out = (q15_t) (((q31_t) out * tempVal) >> 14);\r
+ }\r
+\r
+ /* write output */\r
+ *dst = out;\r
+\r
+ /* return num of signbits of out = 1/in value */\r
+ return (signBits + 1);\r
+\r
+ }\r
+\r
+\r
+ /*\r
+ * @brief C custom defined intrinisic function for only M0 processors\r
+ */\r
+#if defined(ARM_MATH_CM0)\r
+\r
+ __STATIC_INLINE q31_t __SSAT(\r
+ q31_t x,\r
+ uint32_t y)\r
+ {\r
+ int32_t posMax, negMin;\r
+ uint32_t i;\r
+\r
+ posMax = 1;\r
+ for (i = 0; i < (y - 1); i++)\r
+ {\r
+ posMax = posMax * 2;\r
+ }\r
+\r
+ if(x > 0)\r
+ {\r
+ posMax = (posMax - 1);\r
+\r
+ if(x > posMax)\r
+ {\r
+ x = posMax;\r
+ }\r
+ }\r
+ else\r
+ {\r
+ negMin = -posMax;\r
+\r
+ if(x < negMin)\r
+ {\r
+ x = negMin;\r
+ }\r
+ }\r
+ return (x);\r
+\r
+\r
+ }\r
+\r
+#endif /* end of ARM_MATH_CM0 */\r
+\r
+\r
+\r
+ /*\r
+ * @brief C custom defined intrinsic function for M3 and M0 processors\r
+ */\r
+#if defined (ARM_MATH_CM3) || defined (ARM_MATH_CM0)\r
+\r
+ /*\r
+ * @brief C custom defined QADD8 for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __QADD8(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ q31_t sum;\r
+ q7_t r, s, t, u;\r
+\r
+ r = (char) x;\r
+ s = (char) y;\r
+\r
+ r = __SSAT((q31_t) (r + s), 8);\r
+ s = __SSAT(((q31_t) (((x << 16) >> 24) + ((y << 16) >> 24))), 8);\r
+ t = __SSAT(((q31_t) (((x << 8) >> 24) + ((y << 8) >> 24))), 8);\r
+ u = __SSAT(((q31_t) ((x >> 24) + (y >> 24))), 8);\r
+\r
+ sum = (((q31_t) u << 24) & 0xFF000000) | (((q31_t) t << 16) & 0x00FF0000) |\r
+ (((q31_t) s << 8) & 0x0000FF00) | (r & 0x000000FF);\r
+\r
+ return sum;\r
+\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined QSUB8 for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __QSUB8(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ q31_t sum;\r
+ q31_t r, s, t, u;\r
+\r
+ r = (char) x;\r
+ s = (char) y;\r
+\r
+ r = __SSAT((r - s), 8);\r
+ s = __SSAT(((q31_t) (((x << 16) >> 24) - ((y << 16) >> 24))), 8) << 8;\r
+ t = __SSAT(((q31_t) (((x << 8) >> 24) - ((y << 8) >> 24))), 8) << 16;\r
+ u = __SSAT(((q31_t) ((x >> 24) - (y >> 24))), 8) << 24;\r
+\r
+ sum =\r
+ (u & 0xFF000000) | (t & 0x00FF0000) | (s & 0x0000FF00) | (r & 0x000000FF);\r
+\r
+ return sum;\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined QADD16 for M3 and M0 processors\r
+ */\r
+\r
+ /*\r
+ * @brief C custom defined QADD16 for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __QADD16(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ q31_t sum;\r
+ q31_t r, s;\r
+\r
+ r = (short) x;\r
+ s = (short) y;\r
+\r
+ r = __SSAT(r + s, 16);\r
+ s = __SSAT(((q31_t) ((x >> 16) + (y >> 16))), 16) << 16;\r
+\r
+ sum = (s & 0xFFFF0000) | (r & 0x0000FFFF);\r
+\r
+ return sum;\r
+\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined SHADD16 for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __SHADD16(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ q31_t sum;\r
+ q31_t r, s;\r
+\r
+ r = (short) x;\r
+ s = (short) y;\r
+\r
+ r = ((r >> 1) + (s >> 1));\r
+ s = ((q31_t) ((x >> 17) + (y >> 17))) << 16;\r
+\r
+ sum = (s & 0xFFFF0000) | (r & 0x0000FFFF);\r
+\r
+ return sum;\r
+\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined QSUB16 for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __QSUB16(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ q31_t sum;\r
+ q31_t r, s;\r
+\r
+ r = (short) x;\r
+ s = (short) y;\r
+\r
+ r = __SSAT(r - s, 16);\r
+ s = __SSAT(((q31_t) ((x >> 16) - (y >> 16))), 16) << 16;\r
+\r
+ sum = (s & 0xFFFF0000) | (r & 0x0000FFFF);\r
+\r
+ return sum;\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined SHSUB16 for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __SHSUB16(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ q31_t diff;\r
+ q31_t r, s;\r
+\r
+ r = (short) x;\r
+ s = (short) y;\r
+\r
+ r = ((r >> 1) - (s >> 1));\r
+ s = (((x >> 17) - (y >> 17)) << 16);\r
+\r
+ diff = (s & 0xFFFF0000) | (r & 0x0000FFFF);\r
+\r
+ return diff;\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined QASX for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __QASX(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ q31_t sum = 0;\r
+\r
+ sum = ((sum + clip_q31_to_q15((q31_t) ((short) (x >> 16) + (short) y))) << 16) +\r
+ clip_q31_to_q15((q31_t) ((short) x - (short) (y >> 16)));\r
+\r
+ return sum;\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined SHASX for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __SHASX(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ q31_t sum;\r
+ q31_t r, s;\r
+\r
+ r = (short) x;\r
+ s = (short) y;\r
+\r
+ r = ((r >> 1) - (y >> 17));\r
+ s = (((x >> 17) + (s >> 1)) << 16);\r
+\r
+ sum = (s & 0xFFFF0000) | (r & 0x0000FFFF);\r
+\r
+ return sum;\r
+ }\r
+\r
+\r
+ /*\r
+ * @brief C custom defined QSAX for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __QSAX(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ q31_t sum = 0;\r
+\r
+ sum = ((sum + clip_q31_to_q15((q31_t) ((short) (x >> 16) - (short) y))) << 16) +\r
+ clip_q31_to_q15((q31_t) ((short) x + (short) (y >> 16)));\r
+\r
+ return sum;\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined SHSAX for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __SHSAX(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ q31_t sum;\r
+ q31_t r, s;\r
+\r
+ r = (short) x;\r
+ s = (short) y;\r
+\r
+ r = ((r >> 1) + (y >> 17));\r
+ s = (((x >> 17) - (s >> 1)) << 16);\r
+\r
+ sum = (s & 0xFFFF0000) | (r & 0x0000FFFF);\r
+\r
+ return sum;\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined SMUSDX for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __SMUSDX(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ return ((q31_t)(((short) x * (short) (y >> 16)) -\r
+ ((short) (x >> 16) * (short) y)));\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined SMUADX for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __SMUADX(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ return ((q31_t)(((short) x * (short) (y >> 16)) +\r
+ ((short) (x >> 16) * (short) y)));\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined QADD for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __QADD(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+ return clip_q63_to_q31((q63_t) x + y);\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined QSUB for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __QSUB(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+ return clip_q63_to_q31((q63_t) x - y);\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined SMLAD for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __SMLAD(\r
+ q31_t x,\r
+ q31_t y,\r
+ q31_t sum)\r
+ {\r
+\r
+ return (sum + ((short) (x >> 16) * (short) (y >> 16)) +\r
+ ((short) x * (short) y));\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined SMLADX for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __SMLADX(\r
+ q31_t x,\r
+ q31_t y,\r
+ q31_t sum)\r
+ {\r
+\r
+ return (sum + ((short) (x >> 16) * (short) (y)) +\r
+ ((short) x * (short) (y >> 16)));\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined SMLSDX for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __SMLSDX(\r
+ q31_t x,\r
+ q31_t y,\r
+ q31_t sum)\r
+ {\r
+\r
+ return (sum - ((short) (x >> 16) * (short) (y)) +\r
+ ((short) x * (short) (y >> 16)));\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined SMLALD for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q63_t __SMLALD(\r
+ q31_t x,\r
+ q31_t y,\r
+ q63_t sum)\r
+ {\r
+\r
+ return (sum + ((short) (x >> 16) * (short) (y >> 16)) +\r
+ ((short) x * (short) y));\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined SMLALDX for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q63_t __SMLALDX(\r
+ q31_t x,\r
+ q31_t y,\r
+ q63_t sum)\r
+ {\r
+\r
+ return (sum + ((short) (x >> 16) * (short) y)) +\r
+ ((short) x * (short) (y >> 16));\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined SMUAD for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __SMUAD(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ return (((x >> 16) * (y >> 16)) +\r
+ (((x << 16) >> 16) * ((y << 16) >> 16)));\r
+ }\r
+\r
+ /*\r
+ * @brief C custom defined SMUSD for M3 and M0 processors\r
+ */\r
+ __STATIC_INLINE q31_t __SMUSD(\r
+ q31_t x,\r
+ q31_t y)\r
+ {\r
+\r
+ return (-((x >> 16) * (y >> 16)) +\r
+ (((x << 16) >> 16) * ((y << 16) >> 16)));\r
+ }\r
+\r
+\r
+\r
+\r
+#endif /* (ARM_MATH_CM3) || defined (ARM_MATH_CM0) */\r
+\r
+\r
+ /**\r
+ * @brief Instance structure for the Q7 FIR filter.\r
+ */\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< number of filter coefficients in the filter. */\r
+ q7_t *pState; /**< points to the state variable array. The array is of length numTaps+blockSize-1. */\r
+ q7_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps.*/\r
+ } arm_fir_instance_q7;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 FIR filter.\r
+ */\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< number of filter coefficients in the filter. */\r
+ q15_t *pState; /**< points to the state variable array. The array is of length numTaps+blockSize-1. */\r
+ q15_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps.*/\r
+ } arm_fir_instance_q15;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 FIR filter.\r
+ */\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< number of filter coefficients in the filter. */\r
+ q31_t *pState; /**< points to the state variable array. The array is of length numTaps+blockSize-1. */\r
+ q31_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps. */\r
+ } arm_fir_instance_q31;\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point FIR filter.\r
+ */\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< number of filter coefficients in the filter. */\r
+ float32_t *pState; /**< points to the state variable array. The array is of length numTaps+blockSize-1. */\r
+ float32_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps. */\r
+ } arm_fir_instance_f32;\r
+\r
+\r
+ /**\r
+ * @brief Processing function for the Q7 FIR filter.\r
+ * @param[in] *S points to an instance of the Q7 FIR filter structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+ void arm_fir_q7(\r
+ const arm_fir_instance_q7 * S,\r
+ q7_t * pSrc,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Initialization function for the Q7 FIR filter.\r
+ * @param[in,out] *S points to an instance of the Q7 FIR structure.\r
+ * @param[in] numTaps Number of filter coefficients in the filter.\r
+ * @param[in] *pCoeffs points to the filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] blockSize number of samples that are processed.\r
+ * @return none\r
+ */\r
+ void arm_fir_init_q7(\r
+ arm_fir_instance_q7 * S,\r
+ uint16_t numTaps,\r
+ q7_t * pCoeffs,\r
+ q7_t * pState,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Processing function for the Q15 FIR filter.\r
+ * @param[in] *S points to an instance of the Q15 FIR structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+ void arm_fir_q15(\r
+ const arm_fir_instance_q15 * S,\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Processing function for the fast Q15 FIR filter for Cortex-M3 and Cortex-M4.\r
+ * @param[in] *S points to an instance of the Q15 FIR filter structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+ void arm_fir_fast_q15(\r
+ const arm_fir_instance_q15 * S,\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for the Q15 FIR filter.\r
+ * @param[in,out] *S points to an instance of the Q15 FIR filter structure.\r
+ * @param[in] numTaps Number of filter coefficients in the filter. Must be even and greater than or equal to 4.\r
+ * @param[in] *pCoeffs points to the filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] blockSize number of samples that are processed at a time.\r
+ * @return The function returns ARM_MATH_SUCCESS if initialization was successful or ARM_MATH_ARGUMENT_ERROR if\r
+ * <code>numTaps</code> is not a supported value.\r
+ */\r
+\r
+ arm_status arm_fir_init_q15(\r
+ arm_fir_instance_q15 * S,\r
+ uint16_t numTaps,\r
+ q15_t * pCoeffs,\r
+ q15_t * pState,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Processing function for the Q31 FIR filter.\r
+ * @param[in] *S points to an instance of the Q31 FIR filter structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+ void arm_fir_q31(\r
+ const arm_fir_instance_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Processing function for the fast Q31 FIR filter for Cortex-M3 and Cortex-M4.\r
+ * @param[in] *S points to an instance of the Q31 FIR structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+ void arm_fir_fast_q31(\r
+ const arm_fir_instance_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for the Q31 FIR filter.\r
+ * @param[in,out] *S points to an instance of the Q31 FIR structure.\r
+ * @param[in] numTaps Number of filter coefficients in the filter.\r
+ * @param[in] *pCoeffs points to the filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] blockSize number of samples that are processed at a time.\r
+ * @return none.\r
+ */\r
+ void arm_fir_init_q31(\r
+ arm_fir_instance_q31 * S,\r
+ uint16_t numTaps,\r
+ q31_t * pCoeffs,\r
+ q31_t * pState,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Processing function for the floating-point FIR filter.\r
+ * @param[in] *S points to an instance of the floating-point FIR structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+ void arm_fir_f32(\r
+ const arm_fir_instance_f32 * S,\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for the floating-point FIR filter.\r
+ * @param[in,out] *S points to an instance of the floating-point FIR filter structure.\r
+ * @param[in] numTaps Number of filter coefficients in the filter.\r
+ * @param[in] *pCoeffs points to the filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] blockSize number of samples that are processed at a time.\r
+ * @return none.\r
+ */\r
+ void arm_fir_init_f32(\r
+ arm_fir_instance_f32 * S,\r
+ uint16_t numTaps,\r
+ float32_t * pCoeffs,\r
+ float32_t * pState,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 Biquad cascade filter.\r
+ */\r
+ typedef struct\r
+ {\r
+ int8_t numStages; /**< number of 2nd order stages in the filter. Overall order is 2*numStages. */\r
+ q15_t *pState; /**< Points to the array of state coefficients. The array is of length 4*numStages. */\r
+ q15_t *pCoeffs; /**< Points to the array of coefficients. The array is of length 5*numStages. */\r
+ int8_t postShift; /**< Additional shift, in bits, applied to each output sample. */\r
+\r
+ } arm_biquad_casd_df1_inst_q15;\r
+\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 Biquad cascade filter.\r
+ */\r
+ typedef struct\r
+ {\r
+ uint32_t numStages; /**< number of 2nd order stages in the filter. Overall order is 2*numStages. */\r
+ q31_t *pState; /**< Points to the array of state coefficients. The array is of length 4*numStages. */\r
+ q31_t *pCoeffs; /**< Points to the array of coefficients. The array is of length 5*numStages. */\r
+ uint8_t postShift; /**< Additional shift, in bits, applied to each output sample. */\r
+\r
+ } arm_biquad_casd_df1_inst_q31;\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point Biquad cascade filter.\r
+ */\r
+ typedef struct\r
+ {\r
+ uint32_t numStages; /**< number of 2nd order stages in the filter. Overall order is 2*numStages. */\r
+ float32_t *pState; /**< Points to the array of state coefficients. The array is of length 4*numStages. */\r
+ float32_t *pCoeffs; /**< Points to the array of coefficients. The array is of length 5*numStages. */\r
+\r
+\r
+ } arm_biquad_casd_df1_inst_f32;\r
+\r
+\r
+\r
+ /**\r
+ * @brief Processing function for the Q15 Biquad cascade filter.\r
+ * @param[in] *S points to an instance of the Q15 Biquad cascade structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_biquad_cascade_df1_q15(\r
+ const arm_biquad_casd_df1_inst_q15 * S,\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for the Q15 Biquad cascade filter.\r
+ * @param[in,out] *S points to an instance of the Q15 Biquad cascade structure.\r
+ * @param[in] numStages number of 2nd order stages in the filter.\r
+ * @param[in] *pCoeffs points to the filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] postShift Shift to be applied to the output. Varies according to the coefficients format\r
+ * @return none\r
+ */\r
+\r
+ void arm_biquad_cascade_df1_init_q15(\r
+ arm_biquad_casd_df1_inst_q15 * S,\r
+ uint8_t numStages,\r
+ q15_t * pCoeffs,\r
+ q15_t * pState,\r
+ int8_t postShift);\r
+\r
+\r
+ /**\r
+ * @brief Fast but less precise processing function for the Q15 Biquad cascade filter for Cortex-M3 and Cortex-M4.\r
+ * @param[in] *S points to an instance of the Q15 Biquad cascade structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_biquad_cascade_df1_fast_q15(\r
+ const arm_biquad_casd_df1_inst_q15 * S,\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Processing function for the Q31 Biquad cascade filter\r
+ * @param[in] *S points to an instance of the Q31 Biquad cascade structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_biquad_cascade_df1_q31(\r
+ const arm_biquad_casd_df1_inst_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Fast but less precise processing function for the Q31 Biquad cascade filter for Cortex-M3 and Cortex-M4.\r
+ * @param[in] *S points to an instance of the Q31 Biquad cascade structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_biquad_cascade_df1_fast_q31(\r
+ const arm_biquad_casd_df1_inst_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for the Q31 Biquad cascade filter.\r
+ * @param[in,out] *S points to an instance of the Q31 Biquad cascade structure.\r
+ * @param[in] numStages number of 2nd order stages in the filter.\r
+ * @param[in] *pCoeffs points to the filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] postShift Shift to be applied to the output. Varies according to the coefficients format\r
+ * @return none\r
+ */\r
+\r
+ void arm_biquad_cascade_df1_init_q31(\r
+ arm_biquad_casd_df1_inst_q31 * S,\r
+ uint8_t numStages,\r
+ q31_t * pCoeffs,\r
+ q31_t * pState,\r
+ int8_t postShift);\r
+\r
+ /**\r
+ * @brief Processing function for the floating-point Biquad cascade filter.\r
+ * @param[in] *S points to an instance of the floating-point Biquad cascade structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_biquad_cascade_df1_f32(\r
+ const arm_biquad_casd_df1_inst_f32 * S,\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for the floating-point Biquad cascade filter.\r
+ * @param[in,out] *S points to an instance of the floating-point Biquad cascade structure.\r
+ * @param[in] numStages number of 2nd order stages in the filter.\r
+ * @param[in] *pCoeffs points to the filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @return none\r
+ */\r
+\r
+ void arm_biquad_cascade_df1_init_f32(\r
+ arm_biquad_casd_df1_inst_f32 * S,\r
+ uint8_t numStages,\r
+ float32_t * pCoeffs,\r
+ float32_t * pState);\r
+\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point matrix structure.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numRows; /**< number of rows of the matrix. */\r
+ uint16_t numCols; /**< number of columns of the matrix. */\r
+ float32_t *pData; /**< points to the data of the matrix. */\r
+ } arm_matrix_instance_f32;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 matrix structure.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numRows; /**< number of rows of the matrix. */\r
+ uint16_t numCols; /**< number of columns of the matrix. */\r
+ q15_t *pData; /**< points to the data of the matrix. */\r
+\r
+ } arm_matrix_instance_q15;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 matrix structure.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numRows; /**< number of rows of the matrix. */\r
+ uint16_t numCols; /**< number of columns of the matrix. */\r
+ q31_t *pData; /**< points to the data of the matrix. */\r
+\r
+ } arm_matrix_instance_q31;\r
+\r
+\r
+\r
+ /**\r
+ * @brief Floating-point matrix addition.\r
+ * @param[in] *pSrcA points to the first input matrix structure\r
+ * @param[in] *pSrcB points to the second input matrix structure\r
+ * @param[out] *pDst points to output matrix structure\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_add_f32(\r
+ const arm_matrix_instance_f32 * pSrcA,\r
+ const arm_matrix_instance_f32 * pSrcB,\r
+ arm_matrix_instance_f32 * pDst);\r
+\r
+ /**\r
+ * @brief Q15 matrix addition.\r
+ * @param[in] *pSrcA points to the first input matrix structure\r
+ * @param[in] *pSrcB points to the second input matrix structure\r
+ * @param[out] *pDst points to output matrix structure\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_add_q15(\r
+ const arm_matrix_instance_q15 * pSrcA,\r
+ const arm_matrix_instance_q15 * pSrcB,\r
+ arm_matrix_instance_q15 * pDst);\r
+\r
+ /**\r
+ * @brief Q31 matrix addition.\r
+ * @param[in] *pSrcA points to the first input matrix structure\r
+ * @param[in] *pSrcB points to the second input matrix structure\r
+ * @param[out] *pDst points to output matrix structure\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_add_q31(\r
+ const arm_matrix_instance_q31 * pSrcA,\r
+ const arm_matrix_instance_q31 * pSrcB,\r
+ arm_matrix_instance_q31 * pDst);\r
+\r
+\r
+ /**\r
+ * @brief Floating-point matrix transpose.\r
+ * @param[in] *pSrc points to the input matrix\r
+ * @param[out] *pDst points to the output matrix\r
+ * @return The function returns either <code>ARM_MATH_SIZE_MISMATCH</code>\r
+ * or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_trans_f32(\r
+ const arm_matrix_instance_f32 * pSrc,\r
+ arm_matrix_instance_f32 * pDst);\r
+\r
+\r
+ /**\r
+ * @brief Q15 matrix transpose.\r
+ * @param[in] *pSrc points to the input matrix\r
+ * @param[out] *pDst points to the output matrix\r
+ * @return The function returns either <code>ARM_MATH_SIZE_MISMATCH</code>\r
+ * or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_trans_q15(\r
+ const arm_matrix_instance_q15 * pSrc,\r
+ arm_matrix_instance_q15 * pDst);\r
+\r
+ /**\r
+ * @brief Q31 matrix transpose.\r
+ * @param[in] *pSrc points to the input matrix\r
+ * @param[out] *pDst points to the output matrix\r
+ * @return The function returns either <code>ARM_MATH_SIZE_MISMATCH</code>\r
+ * or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_trans_q31(\r
+ const arm_matrix_instance_q31 * pSrc,\r
+ arm_matrix_instance_q31 * pDst);\r
+\r
+\r
+ /**\r
+ * @brief Floating-point matrix multiplication\r
+ * @param[in] *pSrcA points to the first input matrix structure\r
+ * @param[in] *pSrcB points to the second input matrix structure\r
+ * @param[out] *pDst points to output matrix structure\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_mult_f32(\r
+ const arm_matrix_instance_f32 * pSrcA,\r
+ const arm_matrix_instance_f32 * pSrcB,\r
+ arm_matrix_instance_f32 * pDst);\r
+\r
+ /**\r
+ * @brief Q15 matrix multiplication\r
+ * @param[in] *pSrcA points to the first input matrix structure\r
+ * @param[in] *pSrcB points to the second input matrix structure\r
+ * @param[out] *pDst points to output matrix structure\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_mult_q15(\r
+ const arm_matrix_instance_q15 * pSrcA,\r
+ const arm_matrix_instance_q15 * pSrcB,\r
+ arm_matrix_instance_q15 * pDst,\r
+ q15_t * pState);\r
+\r
+ /**\r
+ * @brief Q15 matrix multiplication (fast variant) for Cortex-M3 and Cortex-M4\r
+ * @param[in] *pSrcA points to the first input matrix structure\r
+ * @param[in] *pSrcB points to the second input matrix structure\r
+ * @param[out] *pDst points to output matrix structure\r
+ * @param[in] *pState points to the array for storing intermediate results\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_mult_fast_q15(\r
+ const arm_matrix_instance_q15 * pSrcA,\r
+ const arm_matrix_instance_q15 * pSrcB,\r
+ arm_matrix_instance_q15 * pDst,\r
+ q15_t * pState);\r
+\r
+ /**\r
+ * @brief Q31 matrix multiplication\r
+ * @param[in] *pSrcA points to the first input matrix structure\r
+ * @param[in] *pSrcB points to the second input matrix structure\r
+ * @param[out] *pDst points to output matrix structure\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_mult_q31(\r
+ const arm_matrix_instance_q31 * pSrcA,\r
+ const arm_matrix_instance_q31 * pSrcB,\r
+ arm_matrix_instance_q31 * pDst);\r
+\r
+ /**\r
+ * @brief Q31 matrix multiplication (fast variant) for Cortex-M3 and Cortex-M4\r
+ * @param[in] *pSrcA points to the first input matrix structure\r
+ * @param[in] *pSrcB points to the second input matrix structure\r
+ * @param[out] *pDst points to output matrix structure\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_mult_fast_q31(\r
+ const arm_matrix_instance_q31 * pSrcA,\r
+ const arm_matrix_instance_q31 * pSrcB,\r
+ arm_matrix_instance_q31 * pDst);\r
+\r
+\r
+ /**\r
+ * @brief Floating-point matrix subtraction\r
+ * @param[in] *pSrcA points to the first input matrix structure\r
+ * @param[in] *pSrcB points to the second input matrix structure\r
+ * @param[out] *pDst points to output matrix structure\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_sub_f32(\r
+ const arm_matrix_instance_f32 * pSrcA,\r
+ const arm_matrix_instance_f32 * pSrcB,\r
+ arm_matrix_instance_f32 * pDst);\r
+\r
+ /**\r
+ * @brief Q15 matrix subtraction\r
+ * @param[in] *pSrcA points to the first input matrix structure\r
+ * @param[in] *pSrcB points to the second input matrix structure\r
+ * @param[out] *pDst points to output matrix structure\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_sub_q15(\r
+ const arm_matrix_instance_q15 * pSrcA,\r
+ const arm_matrix_instance_q15 * pSrcB,\r
+ arm_matrix_instance_q15 * pDst);\r
+\r
+ /**\r
+ * @brief Q31 matrix subtraction\r
+ * @param[in] *pSrcA points to the first input matrix structure\r
+ * @param[in] *pSrcB points to the second input matrix structure\r
+ * @param[out] *pDst points to output matrix structure\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_sub_q31(\r
+ const arm_matrix_instance_q31 * pSrcA,\r
+ const arm_matrix_instance_q31 * pSrcB,\r
+ arm_matrix_instance_q31 * pDst);\r
+\r
+ /**\r
+ * @brief Floating-point matrix scaling.\r
+ * @param[in] *pSrc points to the input matrix\r
+ * @param[in] scale scale factor\r
+ * @param[out] *pDst points to the output matrix\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_scale_f32(\r
+ const arm_matrix_instance_f32 * pSrc,\r
+ float32_t scale,\r
+ arm_matrix_instance_f32 * pDst);\r
+\r
+ /**\r
+ * @brief Q15 matrix scaling.\r
+ * @param[in] *pSrc points to input matrix\r
+ * @param[in] scaleFract fractional portion of the scale factor\r
+ * @param[in] shift number of bits to shift the result by\r
+ * @param[out] *pDst points to output matrix\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_scale_q15(\r
+ const arm_matrix_instance_q15 * pSrc,\r
+ q15_t scaleFract,\r
+ int32_t shift,\r
+ arm_matrix_instance_q15 * pDst);\r
+\r
+ /**\r
+ * @brief Q31 matrix scaling.\r
+ * @param[in] *pSrc points to input matrix\r
+ * @param[in] scaleFract fractional portion of the scale factor\r
+ * @param[in] shift number of bits to shift the result by\r
+ * @param[out] *pDst points to output matrix structure\r
+ * @return The function returns either\r
+ * <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.\r
+ */\r
+\r
+ arm_status arm_mat_scale_q31(\r
+ const arm_matrix_instance_q31 * pSrc,\r
+ q31_t scaleFract,\r
+ int32_t shift,\r
+ arm_matrix_instance_q31 * pDst);\r
+\r
+\r
+ /**\r
+ * @brief Q31 matrix initialization.\r
+ * @param[in,out] *S points to an instance of the floating-point matrix structure.\r
+ * @param[in] nRows number of rows in the matrix.\r
+ * @param[in] nColumns number of columns in the matrix.\r
+ * @param[in] *pData points to the matrix data array.\r
+ * @return none\r
+ */\r
+\r
+ void arm_mat_init_q31(\r
+ arm_matrix_instance_q31 * S,\r
+ uint16_t nRows,\r
+ uint16_t nColumns,\r
+ q31_t *pData);\r
+\r
+ /**\r
+ * @brief Q15 matrix initialization.\r
+ * @param[in,out] *S points to an instance of the floating-point matrix structure.\r
+ * @param[in] nRows number of rows in the matrix.\r
+ * @param[in] nColumns number of columns in the matrix.\r
+ * @param[in] *pData points to the matrix data array.\r
+ * @return none\r
+ */\r
+\r
+ void arm_mat_init_q15(\r
+ arm_matrix_instance_q15 * S,\r
+ uint16_t nRows,\r
+ uint16_t nColumns,\r
+ q15_t *pData);\r
+\r
+ /**\r
+ * @brief Floating-point matrix initialization.\r
+ * @param[in,out] *S points to an instance of the floating-point matrix structure.\r
+ * @param[in] nRows number of rows in the matrix.\r
+ * @param[in] nColumns number of columns in the matrix.\r
+ * @param[in] *pData points to the matrix data array.\r
+ * @return none\r
+ */\r
+\r
+ void arm_mat_init_f32(\r
+ arm_matrix_instance_f32 * S,\r
+ uint16_t nRows,\r
+ uint16_t nColumns,\r
+ float32_t *pData);\r
+\r
+\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 PID Control.\r
+ */\r
+ typedef struct\r
+ {\r
+ q15_t A0; /**< The derived gain, A0 = Kp + Ki + Kd . */\r
+ #ifdef ARM_MATH_CM0\r
+ q15_t A1;\r
+ q15_t A2;\r
+ #else\r
+ q31_t A1; /**< The derived gain A1 = -Kp - 2Kd | Kd.*/\r
+ #endif\r
+ q15_t state[3]; /**< The state array of length 3. */\r
+ q15_t Kp; /**< The proportional gain. */\r
+ q15_t Ki; /**< The integral gain. */\r
+ q15_t Kd; /**< The derivative gain. */\r
+ } arm_pid_instance_q15;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 PID Control.\r
+ */\r
+ typedef struct\r
+ {\r
+ q31_t A0; /**< The derived gain, A0 = Kp + Ki + Kd . */\r
+ q31_t A1; /**< The derived gain, A1 = -Kp - 2Kd. */\r
+ q31_t A2; /**< The derived gain, A2 = Kd . */\r
+ q31_t state[3]; /**< The state array of length 3. */\r
+ q31_t Kp; /**< The proportional gain. */\r
+ q31_t Ki; /**< The integral gain. */\r
+ q31_t Kd; /**< The derivative gain. */\r
+\r
+ } arm_pid_instance_q31;\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point PID Control.\r
+ */\r
+ typedef struct\r
+ {\r
+ float32_t A0; /**< The derived gain, A0 = Kp + Ki + Kd . */\r
+ float32_t A1; /**< The derived gain, A1 = -Kp - 2Kd. */\r
+ float32_t A2; /**< The derived gain, A2 = Kd . */\r
+ float32_t state[3]; /**< The state array of length 3. */\r
+ float32_t Kp; /**< The proportional gain. */\r
+ float32_t Ki; /**< The integral gain. */\r
+ float32_t Kd; /**< The derivative gain. */\r
+ } arm_pid_instance_f32;\r
+\r
+\r
+\r
+ /**\r
+ * @brief Initialization function for the floating-point PID Control.\r
+ * @param[in,out] *S points to an instance of the PID structure.\r
+ * @param[in] resetStateFlag flag to reset the state. 0 = no change in state 1 = reset the state.\r
+ * @return none.\r
+ */\r
+ void arm_pid_init_f32(\r
+ arm_pid_instance_f32 * S,\r
+ int32_t resetStateFlag);\r
+\r
+ /**\r
+ * @brief Reset function for the floating-point PID Control.\r
+ * @param[in,out] *S is an instance of the floating-point PID Control structure\r
+ * @return none\r
+ */\r
+ void arm_pid_reset_f32(\r
+ arm_pid_instance_f32 * S);\r
+\r
+\r
+ /**\r
+ * @brief Initialization function for the Q31 PID Control.\r
+ * @param[in,out] *S points to an instance of the Q15 PID structure.\r
+ * @param[in] resetStateFlag flag to reset the state. 0 = no change in state 1 = reset the state.\r
+ * @return none.\r
+ */\r
+ void arm_pid_init_q31(\r
+ arm_pid_instance_q31 * S,\r
+ int32_t resetStateFlag);\r
+\r
+\r
+ /**\r
+ * @brief Reset function for the Q31 PID Control.\r
+ * @param[in,out] *S points to an instance of the Q31 PID Control structure\r
+ * @return none\r
+ */\r
+\r
+ void arm_pid_reset_q31(\r
+ arm_pid_instance_q31 * S);\r
+\r
+ /**\r
+ * @brief Initialization function for the Q15 PID Control.\r
+ * @param[in,out] *S points to an instance of the Q15 PID structure.\r
+ * @param[in] resetStateFlag flag to reset the state. 0 = no change in state 1 = reset the state.\r
+ * @return none.\r
+ */\r
+ void arm_pid_init_q15(\r
+ arm_pid_instance_q15 * S,\r
+ int32_t resetStateFlag);\r
+\r
+ /**\r
+ * @brief Reset function for the Q15 PID Control.\r
+ * @param[in,out] *S points to an instance of the q15 PID Control structure\r
+ * @return none\r
+ */\r
+ void arm_pid_reset_q15(\r
+ arm_pid_instance_q15 * S);\r
+\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point Linear Interpolate function.\r
+ */\r
+ typedef struct\r
+ {\r
+ uint32_t nValues; /**< nValues */\r
+ float32_t x1; /**< x1 */\r
+ float32_t xSpacing; /**< xSpacing */\r
+ float32_t *pYData; /**< pointer to the table of Y values */\r
+ } arm_linear_interp_instance_f32;\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point bilinear interpolation function.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numRows; /**< number of rows in the data table. */\r
+ uint16_t numCols; /**< number of columns in the data table. */\r
+ float32_t *pData; /**< points to the data table. */\r
+ } arm_bilinear_interp_instance_f32;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 bilinear interpolation function.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numRows; /**< number of rows in the data table. */\r
+ uint16_t numCols; /**< number of columns in the data table. */\r
+ q31_t *pData; /**< points to the data table. */\r
+ } arm_bilinear_interp_instance_q31;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 bilinear interpolation function.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numRows; /**< number of rows in the data table. */\r
+ uint16_t numCols; /**< number of columns in the data table. */\r
+ q15_t *pData; /**< points to the data table. */\r
+ } arm_bilinear_interp_instance_q15;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 bilinear interpolation function.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numRows; /**< number of rows in the data table. */\r
+ uint16_t numCols; /**< number of columns in the data table. */\r
+ q7_t *pData; /**< points to the data table. */\r
+ } arm_bilinear_interp_instance_q7;\r
+\r
+\r
+ /**\r
+ * @brief Q7 vector multiplication.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_mult_q7(\r
+ q7_t * pSrcA,\r
+ q7_t * pSrcB,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Q15 vector multiplication.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_mult_q15(\r
+ q15_t * pSrcA,\r
+ q15_t * pSrcB,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Q31 vector multiplication.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_mult_q31(\r
+ q31_t * pSrcA,\r
+ q31_t * pSrcB,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Floating-point vector multiplication.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_mult_f32(\r
+ float32_t * pSrcA,\r
+ float32_t * pSrcB,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 CFFT/CIFFT function.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t fftLen; /**< length of the FFT. */\r
+ uint8_t ifftFlag; /**< flag that selects forward (ifftFlag=0) or inverse (ifftFlag=1) transform. */\r
+ uint8_t bitReverseFlag; /**< flag that enables (bitReverseFlag=1) or disables (bitReverseFlag=0) bit reversal of output. */\r
+ q15_t *pTwiddle; /**< points to the twiddle factor table. */\r
+ uint16_t *pBitRevTable; /**< points to the bit reversal table. */\r
+ uint16_t twidCoefModifier; /**< twiddle coefficient modifier that supports different size FFTs with the same twiddle factor table. */\r
+ uint16_t bitRevFactor; /**< bit reversal modifier that supports different size FFTs with the same bit reversal table. */\r
+ } arm_cfft_radix4_instance_q15;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 CFFT/CIFFT function.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t fftLen; /**< length of the FFT. */\r
+ uint8_t ifftFlag; /**< flag that selects forward (ifftFlag=0) or inverse (ifftFlag=1) transform. */\r
+ uint8_t bitReverseFlag; /**< flag that enables (bitReverseFlag=1) or disables (bitReverseFlag=0) bit reversal of output. */\r
+ q31_t *pTwiddle; /**< points to the twiddle factor table. */\r
+ uint16_t *pBitRevTable; /**< points to the bit reversal table. */\r
+ uint16_t twidCoefModifier; /**< twiddle coefficient modifier that supports different size FFTs with the same twiddle factor table. */\r
+ uint16_t bitRevFactor; /**< bit reversal modifier that supports different size FFTs with the same bit reversal table. */\r
+ } arm_cfft_radix4_instance_q31;\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point CFFT/CIFFT function.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t fftLen; /**< length of the FFT. */\r
+ uint8_t ifftFlag; /**< flag that selects forward (ifftFlag=0) or inverse (ifftFlag=1) transform. */\r
+ uint8_t bitReverseFlag; /**< flag that enables (bitReverseFlag=1) or disables (bitReverseFlag=0) bit reversal of output. */\r
+ float32_t *pTwiddle; /**< points to the twiddle factor table. */\r
+ uint16_t *pBitRevTable; /**< points to the bit reversal table. */\r
+ uint16_t twidCoefModifier; /**< twiddle coefficient modifier that supports different size FFTs with the same twiddle factor table. */\r
+ uint16_t bitRevFactor; /**< bit reversal modifier that supports different size FFTs with the same bit reversal table. */\r
+ float32_t onebyfftLen; /**< value of 1/fftLen. */\r
+ } arm_cfft_radix4_instance_f32;\r
+\r
+ /**\r
+ * @brief Processing function for the Q15 CFFT/CIFFT.\r
+ * @param[in] *S points to an instance of the Q15 CFFT/CIFFT structure.\r
+ * @param[in, out] *pSrc points to the complex data buffer. Processing occurs in-place.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cfft_radix4_q15(\r
+ const arm_cfft_radix4_instance_q15 * S,\r
+ q15_t * pSrc);\r
+\r
+ /**\r
+ * @brief Initialization function for the Q15 CFFT/CIFFT.\r
+ * @param[in,out] *S points to an instance of the Q15 CFFT/CIFFT structure.\r
+ * @param[in] fftLen length of the FFT.\r
+ * @param[in] ifftFlag flag that selects forward (ifftFlag=0) or inverse (ifftFlag=1) transform.\r
+ * @param[in] bitReverseFlag flag that enables (bitReverseFlag=1) or disables (bitReverseFlag=0) bit reversal of output.\r
+ * @return arm_status function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_ARGUMENT_ERROR if <code>fftLen</code> is not a supported value.\r
+ */\r
+\r
+ arm_status arm_cfft_radix4_init_q15(\r
+ arm_cfft_radix4_instance_q15 * S,\r
+ uint16_t fftLen,\r
+ uint8_t ifftFlag,\r
+ uint8_t bitReverseFlag);\r
+\r
+ /**\r
+ * @brief Processing function for the Q31 CFFT/CIFFT.\r
+ * @param[in] *S points to an instance of the Q31 CFFT/CIFFT structure.\r
+ * @param[in, out] *pSrc points to the complex data buffer. Processing occurs in-place.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cfft_radix4_q31(\r
+ const arm_cfft_radix4_instance_q31 * S,\r
+ q31_t * pSrc);\r
+\r
+ /**\r
+ * @brief Initialization function for the Q31 CFFT/CIFFT.\r
+ * @param[in,out] *S points to an instance of the Q31 CFFT/CIFFT structure.\r
+ * @param[in] fftLen length of the FFT.\r
+ * @param[in] ifftFlag flag that selects forward (ifftFlag=0) or inverse (ifftFlag=1) transform.\r
+ * @param[in] bitReverseFlag flag that enables (bitReverseFlag=1) or disables (bitReverseFlag=0) bit reversal of output.\r
+ * @return arm_status function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_ARGUMENT_ERROR if <code>fftLen</code> is not a supported value.\r
+ */\r
+\r
+ arm_status arm_cfft_radix4_init_q31(\r
+ arm_cfft_radix4_instance_q31 * S,\r
+ uint16_t fftLen,\r
+ uint8_t ifftFlag,\r
+ uint8_t bitReverseFlag);\r
+\r
+ /**\r
+ * @brief Processing function for the floating-point CFFT/CIFFT.\r
+ * @param[in] *S points to an instance of the floating-point CFFT/CIFFT structure.\r
+ * @param[in, out] *pSrc points to the complex data buffer. Processing occurs in-place.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cfft_radix4_f32(\r
+ const arm_cfft_radix4_instance_f32 * S,\r
+ float32_t * pSrc);\r
+\r
+ /**\r
+ * @brief Initialization function for the floating-point CFFT/CIFFT.\r
+ * @param[in,out] *S points to an instance of the floating-point CFFT/CIFFT structure.\r
+ * @param[in] fftLen length of the FFT.\r
+ * @param[in] ifftFlag flag that selects forward (ifftFlag=0) or inverse (ifftFlag=1) transform.\r
+ * @param[in] bitReverseFlag flag that enables (bitReverseFlag=1) or disables (bitReverseFlag=0) bit reversal of output.\r
+ * @return The function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_ARGUMENT_ERROR if <code>fftLen</code> is not a supported value.\r
+ */\r
+\r
+ arm_status arm_cfft_radix4_init_f32(\r
+ arm_cfft_radix4_instance_f32 * S,\r
+ uint16_t fftLen,\r
+ uint8_t ifftFlag,\r
+ uint8_t bitReverseFlag);\r
+\r
+\r
+\r
+ /*----------------------------------------------------------------------\r
+ * Internal functions prototypes FFT function\r
+ ----------------------------------------------------------------------*/\r
+\r
+ /**\r
+ * @brief Core function for the floating-point CFFT butterfly process.\r
+ * @param[in, out] *pSrc points to the in-place buffer of floating-point data type.\r
+ * @param[in] fftLen length of the FFT.\r
+ * @param[in] *pCoef points to the twiddle coefficient buffer.\r
+ * @param[in] twidCoefModifier twiddle coefficient modifier that supports different size FFTs with the same twiddle factor table.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_radix4_butterfly_f32(\r
+ float32_t * pSrc,\r
+ uint16_t fftLen,\r
+ float32_t * pCoef,\r
+ uint16_t twidCoefModifier);\r
+\r
+ /**\r
+ * @brief Core function for the floating-point CIFFT butterfly process.\r
+ * @param[in, out] *pSrc points to the in-place buffer of floating-point data type.\r
+ * @param[in] fftLen length of the FFT.\r
+ * @param[in] *pCoef points to twiddle coefficient buffer.\r
+ * @param[in] twidCoefModifier twiddle coefficient modifier that supports different size FFTs with the same twiddle factor table.\r
+ * @param[in] onebyfftLen value of 1/fftLen.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_radix4_butterfly_inverse_f32(\r
+ float32_t * pSrc,\r
+ uint16_t fftLen,\r
+ float32_t * pCoef,\r
+ uint16_t twidCoefModifier,\r
+ float32_t onebyfftLen);\r
+\r
+ /**\r
+ * @brief In-place bit reversal function.\r
+ * @param[in, out] *pSrc points to the in-place buffer of floating-point data type.\r
+ * @param[in] fftSize length of the FFT.\r
+ * @param[in] bitRevFactor bit reversal modifier that supports different size FFTs with the same bit reversal table.\r
+ * @param[in] *pBitRevTab points to the bit reversal table.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_bitreversal_f32(\r
+ float32_t *pSrc,\r
+ uint16_t fftSize,\r
+ uint16_t bitRevFactor,\r
+ uint16_t *pBitRevTab);\r
+\r
+ /**\r
+ * @brief Core function for the Q31 CFFT butterfly process.\r
+ * @param[in, out] *pSrc points to the in-place buffer of Q31 data type.\r
+ * @param[in] fftLen length of the FFT.\r
+ * @param[in] *pCoef points to twiddle coefficient buffer.\r
+ * @param[in] twidCoefModifier twiddle coefficient modifier that supports different size FFTs with the same twiddle factor table.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_radix4_butterfly_q31(\r
+ q31_t *pSrc,\r
+ uint32_t fftLen,\r
+ q31_t *pCoef,\r
+ uint32_t twidCoefModifier);\r
+\r
+ /**\r
+ * @brief Core function for the Q31 CIFFT butterfly process.\r
+ * @param[in, out] *pSrc points to the in-place buffer of Q31 data type.\r
+ * @param[in] fftLen length of the FFT.\r
+ * @param[in] *pCoef points to twiddle coefficient buffer.\r
+ * @param[in] twidCoefModifier twiddle coefficient modifier that supports different size FFTs with the same twiddle factor table.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_radix4_butterfly_inverse_q31(\r
+ q31_t * pSrc,\r
+ uint32_t fftLen,\r
+ q31_t * pCoef,\r
+ uint32_t twidCoefModifier);\r
+\r
+ /**\r
+ * @brief In-place bit reversal function.\r
+ * @param[in, out] *pSrc points to the in-place buffer of Q31 data type.\r
+ * @param[in] fftLen length of the FFT.\r
+ * @param[in] bitRevFactor bit reversal modifier that supports different size FFTs with the same bit reversal table\r
+ * @param[in] *pBitRevTab points to bit reversal table.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_bitreversal_q31(\r
+ q31_t * pSrc,\r
+ uint32_t fftLen,\r
+ uint16_t bitRevFactor,\r
+ uint16_t *pBitRevTab);\r
+\r
+ /**\r
+ * @brief Core function for the Q15 CFFT butterfly process.\r
+ * @param[in, out] *pSrc16 points to the in-place buffer of Q15 data type.\r
+ * @param[in] fftLen length of the FFT.\r
+ * @param[in] *pCoef16 points to twiddle coefficient buffer.\r
+ * @param[in] twidCoefModifier twiddle coefficient modifier that supports different size FFTs with the same twiddle factor table.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_radix4_butterfly_q15(\r
+ q15_t *pSrc16,\r
+ uint32_t fftLen,\r
+ q15_t *pCoef16,\r
+ uint32_t twidCoefModifier);\r
+\r
+ /**\r
+ * @brief Core function for the Q15 CIFFT butterfly process.\r
+ * @param[in, out] *pSrc16 points to the in-place buffer of Q15 data type.\r
+ * @param[in] fftLen length of the FFT.\r
+ * @param[in] *pCoef16 points to twiddle coefficient buffer.\r
+ * @param[in] twidCoefModifier twiddle coefficient modifier that supports different size FFTs with the same twiddle factor table.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_radix4_butterfly_inverse_q15(\r
+ q15_t *pSrc16,\r
+ uint32_t fftLen,\r
+ q15_t *pCoef16,\r
+ uint32_t twidCoefModifier);\r
+\r
+ /**\r
+ * @brief In-place bit reversal function.\r
+ * @param[in, out] *pSrc points to the in-place buffer of Q15 data type.\r
+ * @param[in] fftLen length of the FFT.\r
+ * @param[in] bitRevFactor bit reversal modifier that supports different size FFTs with the same bit reversal table\r
+ * @param[in] *pBitRevTab points to bit reversal table.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_bitreversal_q15(\r
+ q15_t * pSrc,\r
+ uint32_t fftLen,\r
+ uint16_t bitRevFactor,\r
+ uint16_t *pBitRevTab);\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 RFFT/RIFFT function.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint32_t fftLenReal; /**< length of the real FFT. */\r
+ uint32_t fftLenBy2; /**< length of the complex FFT. */\r
+ uint8_t ifftFlagR; /**< flag that selects forward (ifftFlagR=0) or inverse (ifftFlagR=1) transform. */\r
+ uint8_t bitReverseFlagR; /**< flag that enables (bitReverseFlagR=1) or disables (bitReverseFlagR=0) bit reversal of output. */\r
+ uint32_t twidCoefRModifier; /**< twiddle coefficient modifier that supports different size FFTs with the same twiddle factor table. */\r
+ q15_t *pTwiddleAReal; /**< points to the real twiddle factor table. */\r
+ q15_t *pTwiddleBReal; /**< points to the imag twiddle factor table. */\r
+ arm_cfft_radix4_instance_q15 *pCfft; /**< points to the complex FFT instance. */\r
+ } arm_rfft_instance_q15;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 RFFT/RIFFT function.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint32_t fftLenReal; /**< length of the real FFT. */\r
+ uint32_t fftLenBy2; /**< length of the complex FFT. */\r
+ uint8_t ifftFlagR; /**< flag that selects forward (ifftFlagR=0) or inverse (ifftFlagR=1) transform. */\r
+ uint8_t bitReverseFlagR; /**< flag that enables (bitReverseFlagR=1) or disables (bitReverseFlagR=0) bit reversal of output. */\r
+ uint32_t twidCoefRModifier; /**< twiddle coefficient modifier that supports different size FFTs with the same twiddle factor table. */\r
+ q31_t *pTwiddleAReal; /**< points to the real twiddle factor table. */\r
+ q31_t *pTwiddleBReal; /**< points to the imag twiddle factor table. */\r
+ arm_cfft_radix4_instance_q31 *pCfft; /**< points to the complex FFT instance. */\r
+ } arm_rfft_instance_q31;\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point RFFT/RIFFT function.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint32_t fftLenReal; /**< length of the real FFT. */\r
+ uint16_t fftLenBy2; /**< length of the complex FFT. */\r
+ uint8_t ifftFlagR; /**< flag that selects forward (ifftFlagR=0) or inverse (ifftFlagR=1) transform. */\r
+ uint8_t bitReverseFlagR; /**< flag that enables (bitReverseFlagR=1) or disables (bitReverseFlagR=0) bit reversal of output. */\r
+ uint32_t twidCoefRModifier; /**< twiddle coefficient modifier that supports different size FFTs with the same twiddle factor table. */\r
+ float32_t *pTwiddleAReal; /**< points to the real twiddle factor table. */\r
+ float32_t *pTwiddleBReal; /**< points to the imag twiddle factor table. */\r
+ arm_cfft_radix4_instance_f32 *pCfft; /**< points to the complex FFT instance. */\r
+ } arm_rfft_instance_f32;\r
+\r
+ /**\r
+ * @brief Processing function for the Q15 RFFT/RIFFT.\r
+ * @param[in] *S points to an instance of the Q15 RFFT/RIFFT structure.\r
+ * @param[in] *pSrc points to the input buffer.\r
+ * @param[out] *pDst points to the output buffer.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_rfft_q15(\r
+ const arm_rfft_instance_q15 * S,\r
+ q15_t * pSrc,\r
+ q15_t * pDst);\r
+\r
+ /**\r
+ * @brief Initialization function for the Q15 RFFT/RIFFT.\r
+ * @param[in, out] *S points to an instance of the Q15 RFFT/RIFFT structure.\r
+ * @param[in] *S_CFFT points to an instance of the Q15 CFFT/CIFFT structure.\r
+ * @param[in] fftLenReal length of the FFT.\r
+ * @param[in] ifftFlagR flag that selects forward (ifftFlagR=0) or inverse (ifftFlagR=1) transform.\r
+ * @param[in] bitReverseFlag flag that enables (bitReverseFlag=1) or disables (bitReverseFlag=0) bit reversal of output.\r
+ * @return The function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_ARGUMENT_ERROR if <code>fftLenReal</code> is not a supported value.\r
+ */\r
+\r
+ arm_status arm_rfft_init_q15(\r
+ arm_rfft_instance_q15 * S,\r
+ arm_cfft_radix4_instance_q15 * S_CFFT,\r
+ uint32_t fftLenReal,\r
+ uint32_t ifftFlagR,\r
+ uint32_t bitReverseFlag);\r
+\r
+ /**\r
+ * @brief Processing function for the Q31 RFFT/RIFFT.\r
+ * @param[in] *S points to an instance of the Q31 RFFT/RIFFT structure.\r
+ * @param[in] *pSrc points to the input buffer.\r
+ * @param[out] *pDst points to the output buffer.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_rfft_q31(\r
+ const arm_rfft_instance_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pDst);\r
+\r
+ /**\r
+ * @brief Initialization function for the Q31 RFFT/RIFFT.\r
+ * @param[in, out] *S points to an instance of the Q31 RFFT/RIFFT structure.\r
+ * @param[in, out] *S_CFFT points to an instance of the Q31 CFFT/CIFFT structure.\r
+ * @param[in] fftLenReal length of the FFT.\r
+ * @param[in] ifftFlagR flag that selects forward (ifftFlagR=0) or inverse (ifftFlagR=1) transform.\r
+ * @param[in] bitReverseFlag flag that enables (bitReverseFlag=1) or disables (bitReverseFlag=0) bit reversal of output.\r
+ * @return The function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_ARGUMENT_ERROR if <code>fftLenReal</code> is not a supported value.\r
+ */\r
+\r
+ arm_status arm_rfft_init_q31(\r
+ arm_rfft_instance_q31 * S,\r
+ arm_cfft_radix4_instance_q31 * S_CFFT,\r
+ uint32_t fftLenReal,\r
+ uint32_t ifftFlagR,\r
+ uint32_t bitReverseFlag);\r
+\r
+ /**\r
+ * @brief Initialization function for the floating-point RFFT/RIFFT.\r
+ * @param[in,out] *S points to an instance of the floating-point RFFT/RIFFT structure.\r
+ * @param[in,out] *S_CFFT points to an instance of the floating-point CFFT/CIFFT structure.\r
+ * @param[in] fftLenReal length of the FFT.\r
+ * @param[in] ifftFlagR flag that selects forward (ifftFlagR=0) or inverse (ifftFlagR=1) transform.\r
+ * @param[in] bitReverseFlag flag that enables (bitReverseFlag=1) or disables (bitReverseFlag=0) bit reversal of output.\r
+ * @return The function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_ARGUMENT_ERROR if <code>fftLenReal</code> is not a supported value.\r
+ */\r
+\r
+ arm_status arm_rfft_init_f32(\r
+ arm_rfft_instance_f32 * S,\r
+ arm_cfft_radix4_instance_f32 * S_CFFT,\r
+ uint32_t fftLenReal,\r
+ uint32_t ifftFlagR,\r
+ uint32_t bitReverseFlag);\r
+\r
+ /**\r
+ * @brief Processing function for the floating-point RFFT/RIFFT.\r
+ * @param[in] *S points to an instance of the floating-point RFFT/RIFFT structure.\r
+ * @param[in] *pSrc points to the input buffer.\r
+ * @param[out] *pDst points to the output buffer.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_rfft_f32(\r
+ const arm_rfft_instance_f32 * S,\r
+ float32_t * pSrc,\r
+ float32_t * pDst);\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point DCT4/IDCT4 function.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t N; /**< length of the DCT4. */\r
+ uint16_t Nby2; /**< half of the length of the DCT4. */\r
+ float32_t normalize; /**< normalizing factor. */\r
+ float32_t *pTwiddle; /**< points to the twiddle factor table. */\r
+ float32_t *pCosFactor; /**< points to the cosFactor table. */\r
+ arm_rfft_instance_f32 *pRfft; /**< points to the real FFT instance. */\r
+ arm_cfft_radix4_instance_f32 *pCfft; /**< points to the complex FFT instance. */\r
+ } arm_dct4_instance_f32;\r
+\r
+ /**\r
+ * @brief Initialization function for the floating-point DCT4/IDCT4.\r
+ * @param[in,out] *S points to an instance of floating-point DCT4/IDCT4 structure.\r
+ * @param[in] *S_RFFT points to an instance of floating-point RFFT/RIFFT structure.\r
+ * @param[in] *S_CFFT points to an instance of floating-point CFFT/CIFFT structure.\r
+ * @param[in] N length of the DCT4.\r
+ * @param[in] Nby2 half of the length of the DCT4.\r
+ * @param[in] normalize normalizing factor.\r
+ * @return arm_status function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_ARGUMENT_ERROR if <code>fftLenReal</code> is not a supported transform length.\r
+ */\r
+\r
+ arm_status arm_dct4_init_f32(\r
+ arm_dct4_instance_f32 * S,\r
+ arm_rfft_instance_f32 * S_RFFT,\r
+ arm_cfft_radix4_instance_f32 * S_CFFT,\r
+ uint16_t N,\r
+ uint16_t Nby2,\r
+ float32_t normalize);\r
+\r
+ /**\r
+ * @brief Processing function for the floating-point DCT4/IDCT4.\r
+ * @param[in] *S points to an instance of the floating-point DCT4/IDCT4 structure.\r
+ * @param[in] *pState points to state buffer.\r
+ * @param[in,out] *pInlineBuffer points to the in-place input and output buffer.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_dct4_f32(\r
+ const arm_dct4_instance_f32 * S,\r
+ float32_t * pState,\r
+ float32_t * pInlineBuffer);\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 DCT4/IDCT4 function.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t N; /**< length of the DCT4. */\r
+ uint16_t Nby2; /**< half of the length of the DCT4. */\r
+ q31_t normalize; /**< normalizing factor. */\r
+ q31_t *pTwiddle; /**< points to the twiddle factor table. */\r
+ q31_t *pCosFactor; /**< points to the cosFactor table. */\r
+ arm_rfft_instance_q31 *pRfft; /**< points to the real FFT instance. */\r
+ arm_cfft_radix4_instance_q31 *pCfft; /**< points to the complex FFT instance. */\r
+ } arm_dct4_instance_q31;\r
+\r
+ /**\r
+ * @brief Initialization function for the Q31 DCT4/IDCT4.\r
+ * @param[in,out] *S points to an instance of Q31 DCT4/IDCT4 structure.\r
+ * @param[in] *S_RFFT points to an instance of Q31 RFFT/RIFFT structure\r
+ * @param[in] *S_CFFT points to an instance of Q31 CFFT/CIFFT structure\r
+ * @param[in] N length of the DCT4.\r
+ * @param[in] Nby2 half of the length of the DCT4.\r
+ * @param[in] normalize normalizing factor.\r
+ * @return arm_status function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_ARGUMENT_ERROR if <code>N</code> is not a supported transform length.\r
+ */\r
+\r
+ arm_status arm_dct4_init_q31(\r
+ arm_dct4_instance_q31 * S,\r
+ arm_rfft_instance_q31 * S_RFFT,\r
+ arm_cfft_radix4_instance_q31 * S_CFFT,\r
+ uint16_t N,\r
+ uint16_t Nby2,\r
+ q31_t normalize);\r
+\r
+ /**\r
+ * @brief Processing function for the Q31 DCT4/IDCT4.\r
+ * @param[in] *S points to an instance of the Q31 DCT4 structure.\r
+ * @param[in] *pState points to state buffer.\r
+ * @param[in,out] *pInlineBuffer points to the in-place input and output buffer.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_dct4_q31(\r
+ const arm_dct4_instance_q31 * S,\r
+ q31_t * pState,\r
+ q31_t * pInlineBuffer);\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 DCT4/IDCT4 function.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t N; /**< length of the DCT4. */\r
+ uint16_t Nby2; /**< half of the length of the DCT4. */\r
+ q15_t normalize; /**< normalizing factor. */\r
+ q15_t *pTwiddle; /**< points to the twiddle factor table. */\r
+ q15_t *pCosFactor; /**< points to the cosFactor table. */\r
+ arm_rfft_instance_q15 *pRfft; /**< points to the real FFT instance. */\r
+ arm_cfft_radix4_instance_q15 *pCfft; /**< points to the complex FFT instance. */\r
+ } arm_dct4_instance_q15;\r
+\r
+ /**\r
+ * @brief Initialization function for the Q15 DCT4/IDCT4.\r
+ * @param[in,out] *S points to an instance of Q15 DCT4/IDCT4 structure.\r
+ * @param[in] *S_RFFT points to an instance of Q15 RFFT/RIFFT structure.\r
+ * @param[in] *S_CFFT points to an instance of Q15 CFFT/CIFFT structure.\r
+ * @param[in] N length of the DCT4.\r
+ * @param[in] Nby2 half of the length of the DCT4.\r
+ * @param[in] normalize normalizing factor.\r
+ * @return arm_status function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_ARGUMENT_ERROR if <code>N</code> is not a supported transform length.\r
+ */\r
+\r
+ arm_status arm_dct4_init_q15(\r
+ arm_dct4_instance_q15 * S,\r
+ arm_rfft_instance_q15 * S_RFFT,\r
+ arm_cfft_radix4_instance_q15 * S_CFFT,\r
+ uint16_t N,\r
+ uint16_t Nby2,\r
+ q15_t normalize);\r
+\r
+ /**\r
+ * @brief Processing function for the Q15 DCT4/IDCT4.\r
+ * @param[in] *S points to an instance of the Q15 DCT4 structure.\r
+ * @param[in] *pState points to state buffer.\r
+ * @param[in,out] *pInlineBuffer points to the in-place input and output buffer.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_dct4_q15(\r
+ const arm_dct4_instance_q15 * S,\r
+ q15_t * pState,\r
+ q15_t * pInlineBuffer);\r
+\r
+ /**\r
+ * @brief Floating-point vector addition.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_add_f32(\r
+ float32_t * pSrcA,\r
+ float32_t * pSrcB,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Q7 vector addition.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_add_q7(\r
+ q7_t * pSrcA,\r
+ q7_t * pSrcB,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Q15 vector addition.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_add_q15(\r
+ q15_t * pSrcA,\r
+ q15_t * pSrcB,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Q31 vector addition.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_add_q31(\r
+ q31_t * pSrcA,\r
+ q31_t * pSrcB,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Floating-point vector subtraction.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_sub_f32(\r
+ float32_t * pSrcA,\r
+ float32_t * pSrcB,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Q7 vector subtraction.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_sub_q7(\r
+ q7_t * pSrcA,\r
+ q7_t * pSrcB,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Q15 vector subtraction.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_sub_q15(\r
+ q15_t * pSrcA,\r
+ q15_t * pSrcB,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Q31 vector subtraction.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_sub_q31(\r
+ q31_t * pSrcA,\r
+ q31_t * pSrcB,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Multiplies a floating-point vector by a scalar.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[in] scale scale factor to be applied\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_scale_f32(\r
+ float32_t * pSrc,\r
+ float32_t scale,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Multiplies a Q7 vector by a scalar.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[in] scaleFract fractional portion of the scale value\r
+ * @param[in] shift number of bits to shift the result by\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_scale_q7(\r
+ q7_t * pSrc,\r
+ q7_t scaleFract,\r
+ int8_t shift,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Multiplies a Q15 vector by a scalar.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[in] scaleFract fractional portion of the scale value\r
+ * @param[in] shift number of bits to shift the result by\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_scale_q15(\r
+ q15_t * pSrc,\r
+ q15_t scaleFract,\r
+ int8_t shift,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Multiplies a Q31 vector by a scalar.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[in] scaleFract fractional portion of the scale value\r
+ * @param[in] shift number of bits to shift the result by\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_scale_q31(\r
+ q31_t * pSrc,\r
+ q31_t scaleFract,\r
+ int8_t shift,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Q7 vector absolute value.\r
+ * @param[in] *pSrc points to the input buffer\r
+ * @param[out] *pDst points to the output buffer\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_abs_q7(\r
+ q7_t * pSrc,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Floating-point vector absolute value.\r
+ * @param[in] *pSrc points to the input buffer\r
+ * @param[out] *pDst points to the output buffer\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_abs_f32(\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Q15 vector absolute value.\r
+ * @param[in] *pSrc points to the input buffer\r
+ * @param[out] *pDst points to the output buffer\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_abs_q15(\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Q31 vector absolute value.\r
+ * @param[in] *pSrc points to the input buffer\r
+ * @param[out] *pDst points to the output buffer\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_abs_q31(\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Dot product of floating-point vectors.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @param[out] *result output result returned here\r
+ * @return none.\r
+ */\r
+\r
+ void arm_dot_prod_f32(\r
+ float32_t * pSrcA,\r
+ float32_t * pSrcB,\r
+ uint32_t blockSize,\r
+ float32_t * result);\r
+\r
+ /**\r
+ * @brief Dot product of Q7 vectors.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @param[out] *result output result returned here\r
+ * @return none.\r
+ */\r
+\r
+ void arm_dot_prod_q7(\r
+ q7_t * pSrcA,\r
+ q7_t * pSrcB,\r
+ uint32_t blockSize,\r
+ q31_t * result);\r
+\r
+ /**\r
+ * @brief Dot product of Q15 vectors.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @param[out] *result output result returned here\r
+ * @return none.\r
+ */\r
+\r
+ void arm_dot_prod_q15(\r
+ q15_t * pSrcA,\r
+ q15_t * pSrcB,\r
+ uint32_t blockSize,\r
+ q63_t * result);\r
+\r
+ /**\r
+ * @brief Dot product of Q31 vectors.\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[in] blockSize number of samples in each vector\r
+ * @param[out] *result output result returned here\r
+ * @return none.\r
+ */\r
+\r
+ void arm_dot_prod_q31(\r
+ q31_t * pSrcA,\r
+ q31_t * pSrcB,\r
+ uint32_t blockSize,\r
+ q63_t * result);\r
+\r
+ /**\r
+ * @brief Shifts the elements of a Q7 vector a specified number of bits.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[in] shiftBits number of bits to shift. A positive value shifts left; a negative value shifts right.\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_shift_q7(\r
+ q7_t * pSrc,\r
+ int8_t shiftBits,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Shifts the elements of a Q15 vector a specified number of bits.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[in] shiftBits number of bits to shift. A positive value shifts left; a negative value shifts right.\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_shift_q15(\r
+ q15_t * pSrc,\r
+ int8_t shiftBits,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Shifts the elements of a Q31 vector a specified number of bits.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[in] shiftBits number of bits to shift. A positive value shifts left; a negative value shifts right.\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_shift_q31(\r
+ q31_t * pSrc,\r
+ int8_t shiftBits,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Adds a constant offset to a floating-point vector.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[in] offset is the offset to be added\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_offset_f32(\r
+ float32_t * pSrc,\r
+ float32_t offset,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Adds a constant offset to a Q7 vector.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[in] offset is the offset to be added\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_offset_q7(\r
+ q7_t * pSrc,\r
+ q7_t offset,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Adds a constant offset to a Q15 vector.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[in] offset is the offset to be added\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_offset_q15(\r
+ q15_t * pSrc,\r
+ q15_t offset,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Adds a constant offset to a Q31 vector.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[in] offset is the offset to be added\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_offset_q31(\r
+ q31_t * pSrc,\r
+ q31_t offset,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Negates the elements of a floating-point vector.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_negate_f32(\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Negates the elements of a Q7 vector.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_negate_q7(\r
+ q7_t * pSrc,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Negates the elements of a Q15 vector.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_negate_q15(\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Negates the elements of a Q31 vector.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] blockSize number of samples in the vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_negate_q31(\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+ /**\r
+ * @brief Copies the elements of a floating-point vector.\r
+ * @param[in] *pSrc input pointer\r
+ * @param[out] *pDst output pointer\r
+ * @param[in] blockSize number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_copy_f32(\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Copies the elements of a Q7 vector.\r
+ * @param[in] *pSrc input pointer\r
+ * @param[out] *pDst output pointer\r
+ * @param[in] blockSize number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_copy_q7(\r
+ q7_t * pSrc,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Copies the elements of a Q15 vector.\r
+ * @param[in] *pSrc input pointer\r
+ * @param[out] *pDst output pointer\r
+ * @param[in] blockSize number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_copy_q15(\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Copies the elements of a Q31 vector.\r
+ * @param[in] *pSrc input pointer\r
+ * @param[out] *pDst output pointer\r
+ * @param[in] blockSize number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_copy_q31(\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+ /**\r
+ * @brief Fills a constant value into a floating-point vector.\r
+ * @param[in] value input value to be filled\r
+ * @param[out] *pDst output pointer\r
+ * @param[in] blockSize number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_fill_f32(\r
+ float32_t value,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Fills a constant value into a Q7 vector.\r
+ * @param[in] value input value to be filled\r
+ * @param[out] *pDst output pointer\r
+ * @param[in] blockSize number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_fill_q7(\r
+ q7_t value,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Fills a constant value into a Q15 vector.\r
+ * @param[in] value input value to be filled\r
+ * @param[out] *pDst output pointer\r
+ * @param[in] blockSize number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_fill_q15(\r
+ q15_t value,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Fills a constant value into a Q31 vector.\r
+ * @param[in] value input value to be filled\r
+ * @param[out] *pDst output pointer\r
+ * @param[in] blockSize number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_fill_q31(\r
+ q31_t value,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+/**\r
+ * @brief Convolution of floating-point sequences.\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the location where the output result is written. Length srcALen+srcBLen-1.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_conv_f32(\r
+ float32_t * pSrcA,\r
+ uint32_t srcALen,\r
+ float32_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ float32_t * pDst);\r
+\r
+/**\r
+ * @brief Convolution of Q15 sequences.\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the location where the output result is written. Length srcALen+srcBLen-1.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_conv_q15(\r
+ q15_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q15_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q15_t * pDst);\r
+\r
+ /**\r
+ * @brief Convolution of Q15 sequences (fast version) for Cortex-M3 and Cortex-M4\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data Length srcALen+srcBLen-1.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_conv_fast_q15(\r
+ q15_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q15_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q15_t * pDst);\r
+\r
+ /**\r
+ * @brief Convolution of Q31 sequences.\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data Length srcALen+srcBLen-1.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_conv_q31(\r
+ q31_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q31_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q31_t * pDst);\r
+\r
+ /**\r
+ * @brief Convolution of Q31 sequences (fast version) for Cortex-M3 and Cortex-M4\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data Length srcALen+srcBLen-1.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_conv_fast_q31(\r
+ q31_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q31_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q31_t * pDst);\r
+\r
+ /**\r
+ * @brief Convolution of Q7 sequences.\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data Length srcALen+srcBLen-1.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_conv_q7(\r
+ q7_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q7_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q7_t * pDst);\r
+\r
+ /**\r
+ * @brief Partial convolution of floating-point sequences.\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] firstIndex is the first output sample to start with.\r
+ * @param[in] numPoints is the number of output points to be computed.\r
+ * @return Returns either ARM_MATH_SUCCESS if the function completed correctly or ARM_MATH_ARGUMENT_ERROR if the requested subset is not in the range [0 srcALen+srcBLen-2].\r
+ */\r
+\r
+ arm_status arm_conv_partial_f32(\r
+ float32_t * pSrcA,\r
+ uint32_t srcALen,\r
+ float32_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ float32_t * pDst,\r
+ uint32_t firstIndex,\r
+ uint32_t numPoints);\r
+\r
+ /**\r
+ * @brief Partial convolution of Q15 sequences.\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] firstIndex is the first output sample to start with.\r
+ * @param[in] numPoints is the number of output points to be computed.\r
+ * @return Returns either ARM_MATH_SUCCESS if the function completed correctly or ARM_MATH_ARGUMENT_ERROR if the requested subset is not in the range [0 srcALen+srcBLen-2].\r
+ */\r
+\r
+ arm_status arm_conv_partial_q15(\r
+ q15_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q15_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q15_t * pDst,\r
+ uint32_t firstIndex,\r
+ uint32_t numPoints);\r
+\r
+ /**\r
+ * @brief Partial convolution of Q15 sequences (fast version) for Cortex-M3 and Cortex-M4\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] firstIndex is the first output sample to start with.\r
+ * @param[in] numPoints is the number of output points to be computed.\r
+ * @return Returns either ARM_MATH_SUCCESS if the function completed correctly or ARM_MATH_ARGUMENT_ERROR if the requested subset is not in the range [0 srcALen+srcBLen-2].\r
+ */\r
+\r
+ arm_status arm_conv_partial_fast_q15(\r
+ q15_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q15_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q15_t * pDst,\r
+ uint32_t firstIndex,\r
+ uint32_t numPoints);\r
+\r
+ /**\r
+ * @brief Partial convolution of Q31 sequences.\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] firstIndex is the first output sample to start with.\r
+ * @param[in] numPoints is the number of output points to be computed.\r
+ * @return Returns either ARM_MATH_SUCCESS if the function completed correctly or ARM_MATH_ARGUMENT_ERROR if the requested subset is not in the range [0 srcALen+srcBLen-2].\r
+ */\r
+\r
+ arm_status arm_conv_partial_q31(\r
+ q31_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q31_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q31_t * pDst,\r
+ uint32_t firstIndex,\r
+ uint32_t numPoints);\r
+\r
+\r
+ /**\r
+ * @brief Partial convolution of Q31 sequences (fast version) for Cortex-M3 and Cortex-M4\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] firstIndex is the first output sample to start with.\r
+ * @param[in] numPoints is the number of output points to be computed.\r
+ * @return Returns either ARM_MATH_SUCCESS if the function completed correctly or ARM_MATH_ARGUMENT_ERROR if the requested subset is not in the range [0 srcALen+srcBLen-2].\r
+ */\r
+\r
+ arm_status arm_conv_partial_fast_q31(\r
+ q31_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q31_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q31_t * pDst,\r
+ uint32_t firstIndex,\r
+ uint32_t numPoints);\r
+\r
+ /**\r
+ * @brief Partial convolution of Q7 sequences.\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] firstIndex is the first output sample to start with.\r
+ * @param[in] numPoints is the number of output points to be computed.\r
+ * @return Returns either ARM_MATH_SUCCESS if the function completed correctly or ARM_MATH_ARGUMENT_ERROR if the requested subset is not in the range [0 srcALen+srcBLen-2].\r
+ */\r
+\r
+ arm_status arm_conv_partial_q7(\r
+ q7_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q7_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q7_t * pDst,\r
+ uint32_t firstIndex,\r
+ uint32_t numPoints);\r
+\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 FIR decimator.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint8_t M; /**< decimation factor. */\r
+ uint16_t numTaps; /**< number of coefficients in the filter. */\r
+ q15_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps.*/\r
+ q15_t *pState; /**< points to the state variable array. The array is of length numTaps+blockSize-1. */\r
+ } arm_fir_decimate_instance_q15;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 FIR decimator.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint8_t M; /**< decimation factor. */\r
+ uint16_t numTaps; /**< number of coefficients in the filter. */\r
+ q31_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps.*/\r
+ q31_t *pState; /**< points to the state variable array. The array is of length numTaps+blockSize-1. */\r
+\r
+ } arm_fir_decimate_instance_q31;\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point FIR decimator.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint8_t M; /**< decimation factor. */\r
+ uint16_t numTaps; /**< number of coefficients in the filter. */\r
+ float32_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps.*/\r
+ float32_t *pState; /**< points to the state variable array. The array is of length numTaps+blockSize-1. */\r
+\r
+ } arm_fir_decimate_instance_f32;\r
+\r
+\r
+\r
+ /**\r
+ * @brief Processing function for the floating-point FIR decimator.\r
+ * @param[in] *S points to an instance of the floating-point FIR decimator structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return none\r
+ */\r
+\r
+ void arm_fir_decimate_f32(\r
+ const arm_fir_decimate_instance_f32 * S,\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Initialization function for the floating-point FIR decimator.\r
+ * @param[in,out] *S points to an instance of the floating-point FIR decimator structure.\r
+ * @param[in] numTaps number of coefficients in the filter.\r
+ * @param[in] M decimation factor.\r
+ * @param[in] *pCoeffs points to the filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return The function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_LENGTH_ERROR if\r
+ * <code>blockSize</code> is not a multiple of <code>M</code>.\r
+ */\r
+\r
+ arm_status arm_fir_decimate_init_f32(\r
+ arm_fir_decimate_instance_f32 * S,\r
+ uint16_t numTaps,\r
+ uint8_t M,\r
+ float32_t * pCoeffs,\r
+ float32_t * pState,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Processing function for the Q15 FIR decimator.\r
+ * @param[in] *S points to an instance of the Q15 FIR decimator structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return none\r
+ */\r
+\r
+ void arm_fir_decimate_q15(\r
+ const arm_fir_decimate_instance_q15 * S,\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Processing function for the Q15 FIR decimator (fast variant) for Cortex-M3 and Cortex-M4.\r
+ * @param[in] *S points to an instance of the Q15 FIR decimator structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return none\r
+ */\r
+\r
+ void arm_fir_decimate_fast_q15(\r
+ const arm_fir_decimate_instance_q15 * S,\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+\r
+ /**\r
+ * @brief Initialization function for the Q15 FIR decimator.\r
+ * @param[in,out] *S points to an instance of the Q15 FIR decimator structure.\r
+ * @param[in] numTaps number of coefficients in the filter.\r
+ * @param[in] M decimation factor.\r
+ * @param[in] *pCoeffs points to the filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return The function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_LENGTH_ERROR if\r
+ * <code>blockSize</code> is not a multiple of <code>M</code>.\r
+ */\r
+\r
+ arm_status arm_fir_decimate_init_q15(\r
+ arm_fir_decimate_instance_q15 * S,\r
+ uint16_t numTaps,\r
+ uint8_t M,\r
+ q15_t * pCoeffs,\r
+ q15_t * pState,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Processing function for the Q31 FIR decimator.\r
+ * @param[in] *S points to an instance of the Q31 FIR decimator structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return none\r
+ */\r
+\r
+ void arm_fir_decimate_q31(\r
+ const arm_fir_decimate_instance_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Processing function for the Q31 FIR decimator (fast variant) for Cortex-M3 and Cortex-M4.\r
+ * @param[in] *S points to an instance of the Q31 FIR decimator structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return none\r
+ */\r
+\r
+ void arm_fir_decimate_fast_q31(\r
+ arm_fir_decimate_instance_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Initialization function for the Q31 FIR decimator.\r
+ * @param[in,out] *S points to an instance of the Q31 FIR decimator structure.\r
+ * @param[in] numTaps number of coefficients in the filter.\r
+ * @param[in] M decimation factor.\r
+ * @param[in] *pCoeffs points to the filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return The function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_LENGTH_ERROR if\r
+ * <code>blockSize</code> is not a multiple of <code>M</code>.\r
+ */\r
+\r
+ arm_status arm_fir_decimate_init_q31(\r
+ arm_fir_decimate_instance_q31 * S,\r
+ uint16_t numTaps,\r
+ uint8_t M,\r
+ q31_t * pCoeffs,\r
+ q31_t * pState,\r
+ uint32_t blockSize);\r
+\r
+\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 FIR interpolator.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint8_t L; /**< upsample factor. */\r
+ uint16_t phaseLength; /**< length of each polyphase filter component. */\r
+ q15_t *pCoeffs; /**< points to the coefficient array. The array is of length L*phaseLength. */\r
+ q15_t *pState; /**< points to the state variable array. The array is of length blockSize+phaseLength-1. */\r
+ } arm_fir_interpolate_instance_q15;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 FIR interpolator.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint8_t L; /**< upsample factor. */\r
+ uint16_t phaseLength; /**< length of each polyphase filter component. */\r
+ q31_t *pCoeffs; /**< points to the coefficient array. The array is of length L*phaseLength. */\r
+ q31_t *pState; /**< points to the state variable array. The array is of length blockSize+phaseLength-1. */\r
+ } arm_fir_interpolate_instance_q31;\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point FIR interpolator.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint8_t L; /**< upsample factor. */\r
+ uint16_t phaseLength; /**< length of each polyphase filter component. */\r
+ float32_t *pCoeffs; /**< points to the coefficient array. The array is of length L*phaseLength. */\r
+ float32_t *pState; /**< points to the state variable array. The array is of length phaseLength+numTaps-1. */\r
+ } arm_fir_interpolate_instance_f32;\r
+\r
+\r
+ /**\r
+ * @brief Processing function for the Q15 FIR interpolator.\r
+ * @param[in] *S points to an instance of the Q15 FIR interpolator structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_fir_interpolate_q15(\r
+ const arm_fir_interpolate_instance_q15 * S,\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Initialization function for the Q15 FIR interpolator.\r
+ * @param[in,out] *S points to an instance of the Q15 FIR interpolator structure.\r
+ * @param[in] L upsample factor.\r
+ * @param[in] numTaps number of filter coefficients in the filter.\r
+ * @param[in] *pCoeffs points to the filter coefficient buffer.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return The function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_LENGTH_ERROR if\r
+ * the filter length <code>numTaps</code> is not a multiple of the interpolation factor <code>L</code>.\r
+ */\r
+\r
+ arm_status arm_fir_interpolate_init_q15(\r
+ arm_fir_interpolate_instance_q15 * S,\r
+ uint8_t L,\r
+ uint16_t numTaps,\r
+ q15_t * pCoeffs,\r
+ q15_t * pState,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Processing function for the Q31 FIR interpolator.\r
+ * @param[in] *S points to an instance of the Q15 FIR interpolator structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_fir_interpolate_q31(\r
+ const arm_fir_interpolate_instance_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for the Q31 FIR interpolator.\r
+ * @param[in,out] *S points to an instance of the Q31 FIR interpolator structure.\r
+ * @param[in] L upsample factor.\r
+ * @param[in] numTaps number of filter coefficients in the filter.\r
+ * @param[in] *pCoeffs points to the filter coefficient buffer.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return The function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_LENGTH_ERROR if\r
+ * the filter length <code>numTaps</code> is not a multiple of the interpolation factor <code>L</code>.\r
+ */\r
+\r
+ arm_status arm_fir_interpolate_init_q31(\r
+ arm_fir_interpolate_instance_q31 * S,\r
+ uint8_t L,\r
+ uint16_t numTaps,\r
+ q31_t * pCoeffs,\r
+ q31_t * pState,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Processing function for the floating-point FIR interpolator.\r
+ * @param[in] *S points to an instance of the floating-point FIR interpolator structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_fir_interpolate_f32(\r
+ const arm_fir_interpolate_instance_f32 * S,\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for the floating-point FIR interpolator.\r
+ * @param[in,out] *S points to an instance of the floating-point FIR interpolator structure.\r
+ * @param[in] L upsample factor.\r
+ * @param[in] numTaps number of filter coefficients in the filter.\r
+ * @param[in] *pCoeffs points to the filter coefficient buffer.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return The function returns ARM_MATH_SUCCESS if initialization is successful or ARM_MATH_LENGTH_ERROR if\r
+ * the filter length <code>numTaps</code> is not a multiple of the interpolation factor <code>L</code>.\r
+ */\r
+\r
+ arm_status arm_fir_interpolate_init_f32(\r
+ arm_fir_interpolate_instance_f32 * S,\r
+ uint8_t L,\r
+ uint16_t numTaps,\r
+ float32_t * pCoeffs,\r
+ float32_t * pState,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Instance structure for the high precision Q31 Biquad cascade filter.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint8_t numStages; /**< number of 2nd order stages in the filter. Overall order is 2*numStages. */\r
+ q63_t *pState; /**< points to the array of state coefficients. The array is of length 4*numStages. */\r
+ q31_t *pCoeffs; /**< points to the array of coefficients. The array is of length 5*numStages. */\r
+ uint8_t postShift; /**< additional shift, in bits, applied to each output sample. */\r
+\r
+ } arm_biquad_cas_df1_32x64_ins_q31;\r
+\r
+\r
+ /**\r
+ * @param[in] *S points to an instance of the high precision Q31 Biquad cascade filter structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_biquad_cas_df1_32x64_q31(\r
+ const arm_biquad_cas_df1_32x64_ins_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @param[in,out] *S points to an instance of the high precision Q31 Biquad cascade filter structure.\r
+ * @param[in] numStages number of 2nd order stages in the filter.\r
+ * @param[in] *pCoeffs points to the filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] postShift shift to be applied to the output. Varies according to the coefficients format\r
+ * @return none\r
+ */\r
+\r
+ void arm_biquad_cas_df1_32x64_init_q31(\r
+ arm_biquad_cas_df1_32x64_ins_q31 * S,\r
+ uint8_t numStages,\r
+ q31_t * pCoeffs,\r
+ q63_t * pState,\r
+ uint8_t postShift);\r
+\r
+\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point transposed direct form II Biquad cascade filter.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint8_t numStages; /**< number of 2nd order stages in the filter. Overall order is 2*numStages. */\r
+ float32_t *pState; /**< points to the array of state coefficients. The array is of length 2*numStages. */\r
+ float32_t *pCoeffs; /**< points to the array of coefficients. The array is of length 5*numStages. */\r
+ } arm_biquad_cascade_df2T_instance_f32;\r
+\r
+\r
+ /**\r
+ * @brief Processing function for the floating-point transposed direct form II Biquad cascade filter.\r
+ * @param[in] *S points to an instance of the filter data structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_biquad_cascade_df2T_f32(\r
+ const arm_biquad_cascade_df2T_instance_f32 * S,\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Initialization function for the floating-point transposed direct form II Biquad cascade filter.\r
+ * @param[in,out] *S points to an instance of the filter data structure.\r
+ * @param[in] numStages number of 2nd order stages in the filter.\r
+ * @param[in] *pCoeffs points to the filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @return none\r
+ */\r
+\r
+ void arm_biquad_cascade_df2T_init_f32(\r
+ arm_biquad_cascade_df2T_instance_f32 * S,\r
+ uint8_t numStages,\r
+ float32_t * pCoeffs,\r
+ float32_t * pState);\r
+\r
+\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 FIR lattice filter.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numStages; /**< number of filter stages. */\r
+ q15_t *pState; /**< points to the state variable array. The array is of length numStages. */\r
+ q15_t *pCoeffs; /**< points to the coefficient array. The array is of length numStages. */\r
+ } arm_fir_lattice_instance_q15;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 FIR lattice filter.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numStages; /**< number of filter stages. */\r
+ q31_t *pState; /**< points to the state variable array. The array is of length numStages. */\r
+ q31_t *pCoeffs; /**< points to the coefficient array. The array is of length numStages. */\r
+ } arm_fir_lattice_instance_q31;\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point FIR lattice filter.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numStages; /**< number of filter stages. */\r
+ float32_t *pState; /**< points to the state variable array. The array is of length numStages. */\r
+ float32_t *pCoeffs; /**< points to the coefficient array. The array is of length numStages. */\r
+ } arm_fir_lattice_instance_f32;\r
+\r
+ /**\r
+ * @brief Initialization function for the Q15 FIR lattice filter.\r
+ * @param[in] *S points to an instance of the Q15 FIR lattice structure.\r
+ * @param[in] numStages number of filter stages.\r
+ * @param[in] *pCoeffs points to the coefficient buffer. The array is of length numStages.\r
+ * @param[in] *pState points to the state buffer. The array is of length numStages.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_fir_lattice_init_q15(\r
+ arm_fir_lattice_instance_q15 * S,\r
+ uint16_t numStages,\r
+ q15_t * pCoeffs,\r
+ q15_t * pState);\r
+\r
+\r
+ /**\r
+ * @brief Processing function for the Q15 FIR lattice filter.\r
+ * @param[in] *S points to an instance of the Q15 FIR lattice structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+ void arm_fir_lattice_q15(\r
+ const arm_fir_lattice_instance_q15 * S,\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for the Q31 FIR lattice filter.\r
+ * @param[in] *S points to an instance of the Q31 FIR lattice structure.\r
+ * @param[in] numStages number of filter stages.\r
+ * @param[in] *pCoeffs points to the coefficient buffer. The array is of length numStages.\r
+ * @param[in] *pState points to the state buffer. The array is of length numStages.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_fir_lattice_init_q31(\r
+ arm_fir_lattice_instance_q31 * S,\r
+ uint16_t numStages,\r
+ q31_t * pCoeffs,\r
+ q31_t * pState);\r
+\r
+\r
+ /**\r
+ * @brief Processing function for the Q31 FIR lattice filter.\r
+ * @param[in] *S points to an instance of the Q31 FIR lattice structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_fir_lattice_q31(\r
+ const arm_fir_lattice_instance_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+/**\r
+ * @brief Initialization function for the floating-point FIR lattice filter.\r
+ * @param[in] *S points to an instance of the floating-point FIR lattice structure.\r
+ * @param[in] numStages number of filter stages.\r
+ * @param[in] *pCoeffs points to the coefficient buffer. The array is of length numStages.\r
+ * @param[in] *pState points to the state buffer. The array is of length numStages.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_fir_lattice_init_f32(\r
+ arm_fir_lattice_instance_f32 * S,\r
+ uint16_t numStages,\r
+ float32_t * pCoeffs,\r
+ float32_t * pState);\r
+\r
+ /**\r
+ * @brief Processing function for the floating-point FIR lattice filter.\r
+ * @param[in] *S points to an instance of the floating-point FIR lattice structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_fir_lattice_f32(\r
+ const arm_fir_lattice_instance_f32 * S,\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 IIR lattice filter.\r
+ */\r
+ typedef struct\r
+ {\r
+ uint16_t numStages; /**< number of stages in the filter. */\r
+ q15_t *pState; /**< points to the state variable array. The array is of length numStages+blockSize. */\r
+ q15_t *pkCoeffs; /**< points to the reflection coefficient array. The array is of length numStages. */\r
+ q15_t *pvCoeffs; /**< points to the ladder coefficient array. The array is of length numStages+1. */\r
+ } arm_iir_lattice_instance_q15;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 IIR lattice filter.\r
+ */\r
+ typedef struct\r
+ {\r
+ uint16_t numStages; /**< number of stages in the filter. */\r
+ q31_t *pState; /**< points to the state variable array. The array is of length numStages+blockSize. */\r
+ q31_t *pkCoeffs; /**< points to the reflection coefficient array. The array is of length numStages. */\r
+ q31_t *pvCoeffs; /**< points to the ladder coefficient array. The array is of length numStages+1. */\r
+ } arm_iir_lattice_instance_q31;\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point IIR lattice filter.\r
+ */\r
+ typedef struct\r
+ {\r
+ uint16_t numStages; /**< number of stages in the filter. */\r
+ float32_t *pState; /**< points to the state variable array. The array is of length numStages+blockSize. */\r
+ float32_t *pkCoeffs; /**< points to the reflection coefficient array. The array is of length numStages. */\r
+ float32_t *pvCoeffs; /**< points to the ladder coefficient array. The array is of length numStages+1. */\r
+ } arm_iir_lattice_instance_f32;\r
+\r
+ /**\r
+ * @brief Processing function for the floating-point IIR lattice filter.\r
+ * @param[in] *S points to an instance of the floating-point IIR lattice structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_iir_lattice_f32(\r
+ const arm_iir_lattice_instance_f32 * S,\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for the floating-point IIR lattice filter.\r
+ * @param[in] *S points to an instance of the floating-point IIR lattice structure.\r
+ * @param[in] numStages number of stages in the filter.\r
+ * @param[in] *pkCoeffs points to the reflection coefficient buffer. The array is of length numStages.\r
+ * @param[in] *pvCoeffs points to the ladder coefficient buffer. The array is of length numStages+1.\r
+ * @param[in] *pState points to the state buffer. The array is of length numStages+blockSize-1.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_iir_lattice_init_f32(\r
+ arm_iir_lattice_instance_f32 * S,\r
+ uint16_t numStages,\r
+ float32_t *pkCoeffs,\r
+ float32_t *pvCoeffs,\r
+ float32_t *pState,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Processing function for the Q31 IIR lattice filter.\r
+ * @param[in] *S points to an instance of the Q31 IIR lattice structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_iir_lattice_q31(\r
+ const arm_iir_lattice_instance_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Initialization function for the Q31 IIR lattice filter.\r
+ * @param[in] *S points to an instance of the Q31 IIR lattice structure.\r
+ * @param[in] numStages number of stages in the filter.\r
+ * @param[in] *pkCoeffs points to the reflection coefficient buffer. The array is of length numStages.\r
+ * @param[in] *pvCoeffs points to the ladder coefficient buffer. The array is of length numStages+1.\r
+ * @param[in] *pState points to the state buffer. The array is of length numStages+blockSize.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_iir_lattice_init_q31(\r
+ arm_iir_lattice_instance_q31 * S,\r
+ uint16_t numStages,\r
+ q31_t *pkCoeffs,\r
+ q31_t *pvCoeffs,\r
+ q31_t *pState,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Processing function for the Q15 IIR lattice filter.\r
+ * @param[in] *S points to an instance of the Q15 IIR lattice structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_iir_lattice_q15(\r
+ const arm_iir_lattice_instance_q15 * S,\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+/**\r
+ * @brief Initialization function for the Q15 IIR lattice filter.\r
+ * @param[in] *S points to an instance of the fixed-point Q15 IIR lattice structure.\r
+ * @param[in] numStages number of stages in the filter.\r
+ * @param[in] *pkCoeffs points to reflection coefficient buffer. The array is of length numStages.\r
+ * @param[in] *pvCoeffs points to ladder coefficient buffer. The array is of length numStages+1.\r
+ * @param[in] *pState points to state buffer. The array is of length numStages+blockSize.\r
+ * @param[in] blockSize number of samples to process per call.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_iir_lattice_init_q15(\r
+ arm_iir_lattice_instance_q15 * S,\r
+ uint16_t numStages,\r
+ q15_t *pkCoeffs,\r
+ q15_t *pvCoeffs,\r
+ q15_t *pState,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point LMS filter.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< number of coefficients in the filter. */\r
+ float32_t *pState; /**< points to the state variable array. The array is of length numTaps+blockSize-1. */\r
+ float32_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps. */\r
+ float32_t mu; /**< step size that controls filter coefficient updates. */\r
+ } arm_lms_instance_f32;\r
+\r
+ /**\r
+ * @brief Processing function for floating-point LMS filter.\r
+ * @param[in] *S points to an instance of the floating-point LMS filter structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[in] *pRef points to the block of reference data.\r
+ * @param[out] *pOut points to the block of output data.\r
+ * @param[out] *pErr points to the block of error data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_lms_f32(\r
+ const arm_lms_instance_f32 * S,\r
+ float32_t * pSrc,\r
+ float32_t * pRef,\r
+ float32_t * pOut,\r
+ float32_t * pErr,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for floating-point LMS filter.\r
+ * @param[in] *S points to an instance of the floating-point LMS filter structure.\r
+ * @param[in] numTaps number of filter coefficients.\r
+ * @param[in] *pCoeffs points to the coefficient buffer.\r
+ * @param[in] *pState points to state buffer.\r
+ * @param[in] mu step size that controls filter coefficient updates.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_lms_init_f32(\r
+ arm_lms_instance_f32 * S,\r
+ uint16_t numTaps,\r
+ float32_t * pCoeffs,\r
+ float32_t * pState,\r
+ float32_t mu,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 LMS filter.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< number of coefficients in the filter. */\r
+ q15_t *pState; /**< points to the state variable array. The array is of length numTaps+blockSize-1. */\r
+ q15_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps. */\r
+ q15_t mu; /**< step size that controls filter coefficient updates. */\r
+ uint32_t postShift; /**< bit shift applied to coefficients. */\r
+ } arm_lms_instance_q15;\r
+\r
+\r
+ /**\r
+ * @brief Initialization function for the Q15 LMS filter.\r
+ * @param[in] *S points to an instance of the Q15 LMS filter structure.\r
+ * @param[in] numTaps number of filter coefficients.\r
+ * @param[in] *pCoeffs points to the coefficient buffer.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] mu step size that controls filter coefficient updates.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @param[in] postShift bit shift applied to coefficients.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_lms_init_q15(\r
+ arm_lms_instance_q15 * S,\r
+ uint16_t numTaps,\r
+ q15_t * pCoeffs,\r
+ q15_t * pState,\r
+ q15_t mu,\r
+ uint32_t blockSize,\r
+ uint32_t postShift);\r
+\r
+ /**\r
+ * @brief Processing function for Q15 LMS filter.\r
+ * @param[in] *S points to an instance of the Q15 LMS filter structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[in] *pRef points to the block of reference data.\r
+ * @param[out] *pOut points to the block of output data.\r
+ * @param[out] *pErr points to the block of error data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_lms_q15(\r
+ const arm_lms_instance_q15 * S,\r
+ q15_t * pSrc,\r
+ q15_t * pRef,\r
+ q15_t * pOut,\r
+ q15_t * pErr,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 LMS filter.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< number of coefficients in the filter. */\r
+ q31_t *pState; /**< points to the state variable array. The array is of length numTaps+blockSize-1. */\r
+ q31_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps. */\r
+ q31_t mu; /**< step size that controls filter coefficient updates. */\r
+ uint32_t postShift; /**< bit shift applied to coefficients. */\r
+\r
+ } arm_lms_instance_q31;\r
+\r
+ /**\r
+ * @brief Processing function for Q31 LMS filter.\r
+ * @param[in] *S points to an instance of the Q15 LMS filter structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[in] *pRef points to the block of reference data.\r
+ * @param[out] *pOut points to the block of output data.\r
+ * @param[out] *pErr points to the block of error data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_lms_q31(\r
+ const arm_lms_instance_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pRef,\r
+ q31_t * pOut,\r
+ q31_t * pErr,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for Q31 LMS filter.\r
+ * @param[in] *S points to an instance of the Q31 LMS filter structure.\r
+ * @param[in] numTaps number of filter coefficients.\r
+ * @param[in] *pCoeffs points to coefficient buffer.\r
+ * @param[in] *pState points to state buffer.\r
+ * @param[in] mu step size that controls filter coefficient updates.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @param[in] postShift bit shift applied to coefficients.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_lms_init_q31(\r
+ arm_lms_instance_q31 * S,\r
+ uint16_t numTaps,\r
+ q31_t *pCoeffs,\r
+ q31_t *pState,\r
+ q31_t mu,\r
+ uint32_t blockSize,\r
+ uint32_t postShift);\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point normalized LMS filter.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< number of coefficients in the filter. */\r
+ float32_t *pState; /**< points to the state variable array. The array is of length numTaps+blockSize-1. */\r
+ float32_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps. */\r
+ float32_t mu; /**< step size that control filter coefficient updates. */\r
+ float32_t energy; /**< saves previous frame energy. */\r
+ float32_t x0; /**< saves previous input sample. */\r
+ } arm_lms_norm_instance_f32;\r
+\r
+ /**\r
+ * @brief Processing function for floating-point normalized LMS filter.\r
+ * @param[in] *S points to an instance of the floating-point normalized LMS filter structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[in] *pRef points to the block of reference data.\r
+ * @param[out] *pOut points to the block of output data.\r
+ * @param[out] *pErr points to the block of error data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_lms_norm_f32(\r
+ arm_lms_norm_instance_f32 * S,\r
+ float32_t * pSrc,\r
+ float32_t * pRef,\r
+ float32_t * pOut,\r
+ float32_t * pErr,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for floating-point normalized LMS filter.\r
+ * @param[in] *S points to an instance of the floating-point LMS filter structure.\r
+ * @param[in] numTaps number of filter coefficients.\r
+ * @param[in] *pCoeffs points to coefficient buffer.\r
+ * @param[in] *pState points to state buffer.\r
+ * @param[in] mu step size that controls filter coefficient updates.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_lms_norm_init_f32(\r
+ arm_lms_norm_instance_f32 * S,\r
+ uint16_t numTaps,\r
+ float32_t * pCoeffs,\r
+ float32_t * pState,\r
+ float32_t mu,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 normalized LMS filter.\r
+ */\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< number of coefficients in the filter. */\r
+ q31_t *pState; /**< points to the state variable array. The array is of length numTaps+blockSize-1. */\r
+ q31_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps. */\r
+ q31_t mu; /**< step size that controls filter coefficient updates. */\r
+ uint8_t postShift; /**< bit shift applied to coefficients. */\r
+ q31_t *recipTable; /**< points to the reciprocal initial value table. */\r
+ q31_t energy; /**< saves previous frame energy. */\r
+ q31_t x0; /**< saves previous input sample. */\r
+ } arm_lms_norm_instance_q31;\r
+\r
+ /**\r
+ * @brief Processing function for Q31 normalized LMS filter.\r
+ * @param[in] *S points to an instance of the Q31 normalized LMS filter structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[in] *pRef points to the block of reference data.\r
+ * @param[out] *pOut points to the block of output data.\r
+ * @param[out] *pErr points to the block of error data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_lms_norm_q31(\r
+ arm_lms_norm_instance_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pRef,\r
+ q31_t * pOut,\r
+ q31_t * pErr,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for Q31 normalized LMS filter.\r
+ * @param[in] *S points to an instance of the Q31 normalized LMS filter structure.\r
+ * @param[in] numTaps number of filter coefficients.\r
+ * @param[in] *pCoeffs points to coefficient buffer.\r
+ * @param[in] *pState points to state buffer.\r
+ * @param[in] mu step size that controls filter coefficient updates.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @param[in] postShift bit shift applied to coefficients.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_lms_norm_init_q31(\r
+ arm_lms_norm_instance_q31 * S,\r
+ uint16_t numTaps,\r
+ q31_t * pCoeffs,\r
+ q31_t * pState,\r
+ q31_t mu,\r
+ uint32_t blockSize,\r
+ uint8_t postShift);\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 normalized LMS filter.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< Number of coefficients in the filter. */\r
+ q15_t *pState; /**< points to the state variable array. The array is of length numTaps+blockSize-1. */\r
+ q15_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps. */\r
+ q15_t mu; /**< step size that controls filter coefficient updates. */\r
+ uint8_t postShift; /**< bit shift applied to coefficients. */\r
+ q15_t *recipTable; /**< Points to the reciprocal initial value table. */\r
+ q15_t energy; /**< saves previous frame energy. */\r
+ q15_t x0; /**< saves previous input sample. */\r
+ } arm_lms_norm_instance_q15;\r
+\r
+ /**\r
+ * @brief Processing function for Q15 normalized LMS filter.\r
+ * @param[in] *S points to an instance of the Q15 normalized LMS filter structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[in] *pRef points to the block of reference data.\r
+ * @param[out] *pOut points to the block of output data.\r
+ * @param[out] *pErr points to the block of error data.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_lms_norm_q15(\r
+ arm_lms_norm_instance_q15 * S,\r
+ q15_t * pSrc,\r
+ q15_t * pRef,\r
+ q15_t * pOut,\r
+ q15_t * pErr,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Initialization function for Q15 normalized LMS filter.\r
+ * @param[in] *S points to an instance of the Q15 normalized LMS filter structure.\r
+ * @param[in] numTaps number of filter coefficients.\r
+ * @param[in] *pCoeffs points to coefficient buffer.\r
+ * @param[in] *pState points to state buffer.\r
+ * @param[in] mu step size that controls filter coefficient updates.\r
+ * @param[in] blockSize number of samples to process.\r
+ * @param[in] postShift bit shift applied to coefficients.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_lms_norm_init_q15(\r
+ arm_lms_norm_instance_q15 * S,\r
+ uint16_t numTaps,\r
+ q15_t * pCoeffs,\r
+ q15_t * pState,\r
+ q15_t mu,\r
+ uint32_t blockSize,\r
+ uint8_t postShift);\r
+\r
+ /**\r
+ * @brief Correlation of floating-point sequences.\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data Length 2 * max(srcALen, srcBLen) - 1.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_correlate_f32(\r
+ float32_t * pSrcA,\r
+ uint32_t srcALen,\r
+ float32_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ float32_t * pDst);\r
+\r
+ /**\r
+ * @brief Correlation of Q15 sequences.\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data Length 2 * max(srcALen, srcBLen) - 1.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_correlate_q15(\r
+ q15_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q15_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q15_t * pDst);\r
+\r
+ /**\r
+ * @brief Correlation of Q15 sequences (fast version) for Cortex-M3 and Cortex-M4.\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data Length 2 * max(srcALen, srcBLen) - 1.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_correlate_fast_q15(\r
+ q15_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q15_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q15_t * pDst);\r
+\r
+ /**\r
+ * @brief Correlation of Q31 sequences.\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data Length 2 * max(srcALen, srcBLen) - 1.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_correlate_q31(\r
+ q31_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q31_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q31_t * pDst);\r
+\r
+ /**\r
+ * @brief Correlation of Q31 sequences (fast version) for Cortex-M3 and Cortex-M4\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data Length 2 * max(srcALen, srcBLen) - 1.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_correlate_fast_q31(\r
+ q31_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q31_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q31_t * pDst);\r
+\r
+ /**\r
+ * @brief Correlation of Q7 sequences.\r
+ * @param[in] *pSrcA points to the first input sequence.\r
+ * @param[in] srcALen length of the first input sequence.\r
+ * @param[in] *pSrcB points to the second input sequence.\r
+ * @param[in] srcBLen length of the second input sequence.\r
+ * @param[out] *pDst points to the block of output data Length 2 * max(srcALen, srcBLen) - 1.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_correlate_q7(\r
+ q7_t * pSrcA,\r
+ uint32_t srcALen,\r
+ q7_t * pSrcB,\r
+ uint32_t srcBLen,\r
+ q7_t * pDst);\r
+\r
+ /**\r
+ * @brief Instance structure for the floating-point sparse FIR filter.\r
+ */\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< number of coefficients in the filter. */\r
+ uint16_t stateIndex; /**< state buffer index. Points to the oldest sample in the state buffer. */\r
+ float32_t *pState; /**< points to the state buffer array. The array is of length maxDelay+blockSize-1. */\r
+ float32_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps.*/\r
+ uint16_t maxDelay; /**< maximum offset specified by the pTapDelay array. */\r
+ int32_t *pTapDelay; /**< points to the array of delay values. The array is of length numTaps. */\r
+ } arm_fir_sparse_instance_f32;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q31 sparse FIR filter.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< number of coefficients in the filter. */\r
+ uint16_t stateIndex; /**< state buffer index. Points to the oldest sample in the state buffer. */\r
+ q31_t *pState; /**< points to the state buffer array. The array is of length maxDelay+blockSize-1. */\r
+ q31_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps.*/\r
+ uint16_t maxDelay; /**< maximum offset specified by the pTapDelay array. */\r
+ int32_t *pTapDelay; /**< points to the array of delay values. The array is of length numTaps. */\r
+ } arm_fir_sparse_instance_q31;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q15 sparse FIR filter.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< number of coefficients in the filter. */\r
+ uint16_t stateIndex; /**< state buffer index. Points to the oldest sample in the state buffer. */\r
+ q15_t *pState; /**< points to the state buffer array. The array is of length maxDelay+blockSize-1. */\r
+ q15_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps.*/\r
+ uint16_t maxDelay; /**< maximum offset specified by the pTapDelay array. */\r
+ int32_t *pTapDelay; /**< points to the array of delay values. The array is of length numTaps. */\r
+ } arm_fir_sparse_instance_q15;\r
+\r
+ /**\r
+ * @brief Instance structure for the Q7 sparse FIR filter.\r
+ */\r
+\r
+ typedef struct\r
+ {\r
+ uint16_t numTaps; /**< number of coefficients in the filter. */\r
+ uint16_t stateIndex; /**< state buffer index. Points to the oldest sample in the state buffer. */\r
+ q7_t *pState; /**< points to the state buffer array. The array is of length maxDelay+blockSize-1. */\r
+ q7_t *pCoeffs; /**< points to the coefficient array. The array is of length numTaps.*/\r
+ uint16_t maxDelay; /**< maximum offset specified by the pTapDelay array. */\r
+ int32_t *pTapDelay; /**< points to the array of delay values. The array is of length numTaps. */\r
+ } arm_fir_sparse_instance_q7;\r
+\r
+ /**\r
+ * @brief Processing function for the floating-point sparse FIR filter.\r
+ * @param[in] *S points to an instance of the floating-point sparse FIR structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] *pScratchIn points to a temporary buffer of size blockSize.\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_fir_sparse_f32(\r
+ arm_fir_sparse_instance_f32 * S,\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ float32_t * pScratchIn,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for the floating-point sparse FIR filter.\r
+ * @param[in,out] *S points to an instance of the floating-point sparse FIR structure.\r
+ * @param[in] numTaps number of nonzero coefficients in the filter.\r
+ * @param[in] *pCoeffs points to the array of filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] *pTapDelay points to the array of offset times.\r
+ * @param[in] maxDelay maximum offset time supported.\r
+ * @param[in] blockSize number of samples that will be processed per block.\r
+ * @return none\r
+ */\r
+\r
+ void arm_fir_sparse_init_f32(\r
+ arm_fir_sparse_instance_f32 * S,\r
+ uint16_t numTaps,\r
+ float32_t * pCoeffs,\r
+ float32_t * pState,\r
+ int32_t * pTapDelay,\r
+ uint16_t maxDelay,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Processing function for the Q31 sparse FIR filter.\r
+ * @param[in] *S points to an instance of the Q31 sparse FIR structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] *pScratchIn points to a temporary buffer of size blockSize.\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_fir_sparse_q31(\r
+ arm_fir_sparse_instance_q31 * S,\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ q31_t * pScratchIn,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for the Q31 sparse FIR filter.\r
+ * @param[in,out] *S points to an instance of the Q31 sparse FIR structure.\r
+ * @param[in] numTaps number of nonzero coefficients in the filter.\r
+ * @param[in] *pCoeffs points to the array of filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] *pTapDelay points to the array of offset times.\r
+ * @param[in] maxDelay maximum offset time supported.\r
+ * @param[in] blockSize number of samples that will be processed per block.\r
+ * @return none\r
+ */\r
+\r
+ void arm_fir_sparse_init_q31(\r
+ arm_fir_sparse_instance_q31 * S,\r
+ uint16_t numTaps,\r
+ q31_t * pCoeffs,\r
+ q31_t * pState,\r
+ int32_t * pTapDelay,\r
+ uint16_t maxDelay,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Processing function for the Q15 sparse FIR filter.\r
+ * @param[in] *S points to an instance of the Q15 sparse FIR structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] *pScratchIn points to a temporary buffer of size blockSize.\r
+ * @param[in] *pScratchOut points to a temporary buffer of size blockSize.\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_fir_sparse_q15(\r
+ arm_fir_sparse_instance_q15 * S,\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ q15_t * pScratchIn,\r
+ q31_t * pScratchOut,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Initialization function for the Q15 sparse FIR filter.\r
+ * @param[in,out] *S points to an instance of the Q15 sparse FIR structure.\r
+ * @param[in] numTaps number of nonzero coefficients in the filter.\r
+ * @param[in] *pCoeffs points to the array of filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] *pTapDelay points to the array of offset times.\r
+ * @param[in] maxDelay maximum offset time supported.\r
+ * @param[in] blockSize number of samples that will be processed per block.\r
+ * @return none\r
+ */\r
+\r
+ void arm_fir_sparse_init_q15(\r
+ arm_fir_sparse_instance_q15 * S,\r
+ uint16_t numTaps,\r
+ q15_t * pCoeffs,\r
+ q15_t * pState,\r
+ int32_t * pTapDelay,\r
+ uint16_t maxDelay,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Processing function for the Q7 sparse FIR filter.\r
+ * @param[in] *S points to an instance of the Q7 sparse FIR structure.\r
+ * @param[in] *pSrc points to the block of input data.\r
+ * @param[out] *pDst points to the block of output data\r
+ * @param[in] *pScratchIn points to a temporary buffer of size blockSize.\r
+ * @param[in] *pScratchOut points to a temporary buffer of size blockSize.\r
+ * @param[in] blockSize number of input samples to process per call.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_fir_sparse_q7(\r
+ arm_fir_sparse_instance_q7 * S,\r
+ q7_t * pSrc,\r
+ q7_t * pDst,\r
+ q7_t * pScratchIn,\r
+ q31_t * pScratchOut,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Initialization function for the Q7 sparse FIR filter.\r
+ * @param[in,out] *S points to an instance of the Q7 sparse FIR structure.\r
+ * @param[in] numTaps number of nonzero coefficients in the filter.\r
+ * @param[in] *pCoeffs points to the array of filter coefficients.\r
+ * @param[in] *pState points to the state buffer.\r
+ * @param[in] *pTapDelay points to the array of offset times.\r
+ * @param[in] maxDelay maximum offset time supported.\r
+ * @param[in] blockSize number of samples that will be processed per block.\r
+ * @return none\r
+ */\r
+\r
+ void arm_fir_sparse_init_q7(\r
+ arm_fir_sparse_instance_q7 * S,\r
+ uint16_t numTaps,\r
+ q7_t * pCoeffs,\r
+ q7_t * pState,\r
+ int32_t *pTapDelay,\r
+ uint16_t maxDelay,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /*\r
+ * @brief Floating-point sin_cos function.\r
+ * @param[in] theta input value in degrees\r
+ * @param[out] *pSinVal points to the processed sine output.\r
+ * @param[out] *pCosVal points to the processed cos output.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_sin_cos_f32(\r
+ float32_t theta,\r
+ float32_t *pSinVal,\r
+ float32_t *pCcosVal);\r
+\r
+ /*\r
+ * @brief Q31 sin_cos function.\r
+ * @param[in] theta scaled input value in degrees\r
+ * @param[out] *pSinVal points to the processed sine output.\r
+ * @param[out] *pCosVal points to the processed cosine output.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_sin_cos_q31(\r
+ q31_t theta,\r
+ q31_t *pSinVal,\r
+ q31_t *pCosVal);\r
+\r
+\r
+ /**\r
+ * @brief Floating-point complex conjugate.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] numSamples number of complex samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_conj_f32(\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t numSamples);\r
+\r
+ /**\r
+ * @brief Q31 complex conjugate.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] numSamples number of complex samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_conj_q31(\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t numSamples);\r
+\r
+ /**\r
+ * @brief Q15 complex conjugate.\r
+ * @param[in] *pSrc points to the input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] numSamples number of complex samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_conj_q15(\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t numSamples);\r
+\r
+\r
+\r
+ /**\r
+ * @brief Floating-point complex magnitude squared\r
+ * @param[in] *pSrc points to the complex input vector\r
+ * @param[out] *pDst points to the real output vector\r
+ * @param[in] numSamples number of complex samples in the input vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_mag_squared_f32(\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t numSamples);\r
+\r
+ /**\r
+ * @brief Q31 complex magnitude squared\r
+ * @param[in] *pSrc points to the complex input vector\r
+ * @param[out] *pDst points to the real output vector\r
+ * @param[in] numSamples number of complex samples in the input vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_mag_squared_q31(\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t numSamples);\r
+\r
+ /**\r
+ * @brief Q15 complex magnitude squared\r
+ * @param[in] *pSrc points to the complex input vector\r
+ * @param[out] *pDst points to the real output vector\r
+ * @param[in] numSamples number of complex samples in the input vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_mag_squared_q15(\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t numSamples);\r
+\r
+\r
+ /**\r
+ * @ingroup groupController\r
+ */\r
+\r
+ /**\r
+ * @defgroup PID PID Motor Control\r
+ *\r
+ * A Proportional Integral Derivative (PID) controller is a generic feedback control\r
+ * loop mechanism widely used in industrial control systems.\r
+ * A PID controller is the most commonly used type of feedback controller.\r
+ *\r
+ * This set of functions implements (PID) controllers\r
+ * for Q15, Q31, and floating-point data types. The functions operate on a single sample\r
+ * of data and each call to the function returns a single processed value.\r
+ * <code>S</code> points to an instance of the PID control data structure. <code>in</code>\r
+ * is the input sample value. The functions return the output value.\r
+ *\r
+ * \par Algorithm:\r
+ * <pre>\r
+ * y[n] = y[n-1] + A0 * x[n] + A1 * x[n-1] + A2 * x[n-2]\r
+ * A0 = Kp + Ki + Kd\r
+ * A1 = (-Kp ) - (2 * Kd )\r
+ * A2 = Kd </pre>\r
+ *\r
+ * \par\r
+ * where \c Kp is proportional constant, \c Ki is Integral constant and \c Kd is Derivative constant\r
+ *\r
+ * \par\r
+ * \image html PID.gif "Proportional Integral Derivative Controller"\r
+ *\r
+ * \par\r
+ * The PID controller calculates an "error" value as the difference between\r
+ * the measured output and the reference input.\r
+ * The controller attempts to minimize the error by adjusting the process control inputs.\r
+ * The proportional value determines the reaction to the current error,\r
+ * the integral value determines the reaction based on the sum of recent errors,\r
+ * and the derivative value determines the reaction based on the rate at which the error has been changing.\r
+ *\r
+ * \par Instance Structure\r
+ * The Gains A0, A1, A2 and state variables for a PID controller are stored together in an instance data structure.\r
+ * A separate instance structure must be defined for each PID Controller.\r
+ * There are separate instance structure declarations for each of the 3 supported data types.\r
+ *\r
+ * \par Reset Functions\r
+ * There is also an associated reset function for each data type which clears the state array.\r
+ *\r
+ * \par Initialization Functions\r
+ * There is also an associated initialization function for each data type.\r
+ * The initialization function performs the following operations:\r
+ * - Initializes the Gains A0, A1, A2 from Kp,Ki, Kd gains.\r
+ * - Zeros out the values in the state buffer.\r
+ *\r
+ * \par\r
+ * Instance structure cannot be placed into a const data section and it is recommended to use the initialization function.\r
+ *\r
+ * \par Fixed-Point Behavior\r
+ * Care must be taken when using the fixed-point versions of the PID Controller functions.\r
+ * In particular, the overflow and saturation behavior of the accumulator used in each function must be considered.\r
+ * Refer to the function specific documentation below for usage guidelines.\r
+ */\r
+\r
+ /**\r
+ * @addtogroup PID\r
+ * @{\r
+ */\r
+\r
+ /**\r
+ * @brief Process function for the floating-point PID Control.\r
+ * @param[in,out] *S is an instance of the floating-point PID Control structure\r
+ * @param[in] in input sample to process\r
+ * @return out processed output sample.\r
+ */\r
+\r
+\r
+ __STATIC_INLINE float32_t arm_pid_f32(\r
+ arm_pid_instance_f32 * S,\r
+ float32_t in)\r
+ {\r
+ float32_t out;\r
+\r
+ /* y[n] = y[n-1] + A0 * x[n] + A1 * x[n-1] + A2 * x[n-2] */\r
+ out = (S->A0 * in) +\r
+ (S->A1 * S->state[0]) + (S->A2 * S->state[1]) + (S->state[2]);\r
+\r
+ /* Update state */\r
+ S->state[1] = S->state[0];\r
+ S->state[0] = in;\r
+ S->state[2] = out;\r
+\r
+ /* return to application */\r
+ return (out);\r
+\r
+ }\r
+\r
+ /**\r
+ * @brief Process function for the Q31 PID Control.\r
+ * @param[in,out] *S points to an instance of the Q31 PID Control structure\r
+ * @param[in] in input sample to process\r
+ * @return out processed output sample.\r
+ *\r
+ * <b>Scaling and Overflow Behavior:</b>\r
+ * \par\r
+ * The function is implemented using an internal 64-bit accumulator.\r
+ * The accumulator has a 2.62 format and maintains full precision of the intermediate multiplication results but provides only a single guard bit.\r
+ * Thus, if the accumulator result overflows it wraps around rather than clip.\r
+ * In order to avoid overflows completely the input signal must be scaled down by 2 bits as there are four additions.\r
+ * After all multiply-accumulates are performed, the 2.62 accumulator is truncated to 1.32 format and then saturated to 1.31 format.\r
+ */\r
+\r
+ __STATIC_INLINE q31_t arm_pid_q31(\r
+ arm_pid_instance_q31 * S,\r
+ q31_t in)\r
+ {\r
+ q63_t acc;\r
+ q31_t out;\r
+\r
+ /* acc = A0 * x[n] */\r
+ acc = (q63_t) S->A0 * in;\r
+\r
+ /* acc += A1 * x[n-1] */\r
+ acc += (q63_t) S->A1 * S->state[0];\r
+\r
+ /* acc += A2 * x[n-2] */\r
+ acc += (q63_t) S->A2 * S->state[1];\r
+\r
+ /* convert output to 1.31 format to add y[n-1] */\r
+ out = (q31_t) (acc >> 31u);\r
+\r
+ /* out += y[n-1] */\r
+ out += S->state[2];\r
+\r
+ /* Update state */\r
+ S->state[1] = S->state[0];\r
+ S->state[0] = in;\r
+ S->state[2] = out;\r
+\r
+ /* return to application */\r
+ return (out);\r
+\r
+ }\r
+\r
+ /**\r
+ * @brief Process function for the Q15 PID Control.\r
+ * @param[in,out] *S points to an instance of the Q15 PID Control structure\r
+ * @param[in] in input sample to process\r
+ * @return out processed output sample.\r
+ *\r
+ * <b>Scaling and Overflow Behavior:</b>\r
+ * \par\r
+ * The function is implemented using a 64-bit internal accumulator.\r
+ * Both Gains and state variables are represented in 1.15 format and multiplications yield a 2.30 result.\r
+ * The 2.30 intermediate results are accumulated in a 64-bit accumulator in 34.30 format.\r
+ * There is no risk of internal overflow with this approach and the full precision of intermediate multiplications is preserved.\r
+ * After all additions have been performed, the accumulator is truncated to 34.15 format by discarding low 15 bits.\r
+ * Lastly, the accumulator is saturated to yield a result in 1.15 format.\r
+ */\r
+\r
+ __STATIC_INLINE q15_t arm_pid_q15(\r
+ arm_pid_instance_q15 * S,\r
+ q15_t in)\r
+ {\r
+ q63_t acc;\r
+ q15_t out;\r
+\r
+ /* Implementation of PID controller */\r
+\r
+ #ifdef ARM_MATH_CM0\r
+\r
+ /* acc = A0 * x[n] */\r
+ acc = ((q31_t) S->A0 )* in ;\r
+\r
+ #else\r
+\r
+ /* acc = A0 * x[n] */\r
+ acc = (q31_t) __SMUAD(S->A0, in);\r
+\r
+ #endif\r
+\r
+ #ifdef ARM_MATH_CM0\r
+\r
+ /* acc += A1 * x[n-1] + A2 * x[n-2] */\r
+ acc += (q31_t) S->A1 * S->state[0] ;\r
+ acc += (q31_t) S->A2 * S->state[1] ;\r
+\r
+ #else\r
+\r
+ /* acc += A1 * x[n-1] + A2 * x[n-2] */\r
+ acc = __SMLALD(S->A1, (q31_t)__SIMD32(S->state), acc);\r
+\r
+ #endif\r
+\r
+ /* acc += y[n-1] */\r
+ acc += (q31_t) S->state[2] << 15;\r
+\r
+ /* saturate the output */\r
+ out = (q15_t) (__SSAT((acc >> 15), 16));\r
+\r
+ /* Update state */\r
+ S->state[1] = S->state[0];\r
+ S->state[0] = in;\r
+ S->state[2] = out;\r
+\r
+ /* return to application */\r
+ return (out);\r
+\r
+ }\r
+\r
+ /**\r
+ * @} end of PID group\r
+ */\r
+\r
+\r
+ /**\r
+ * @brief Floating-point matrix inverse.\r
+ * @param[in] *src points to the instance of the input floating-point matrix structure.\r
+ * @param[out] *dst points to the instance of the output floating-point matrix structure.\r
+ * @return The function returns ARM_MATH_SIZE_MISMATCH, if the dimensions do not match.\r
+ * If the input matrix is singular (does not have an inverse), then the algorithm terminates and returns error status ARM_MATH_SINGULAR.\r
+ */\r
+\r
+ arm_status arm_mat_inverse_f32(\r
+ const arm_matrix_instance_f32 * src,\r
+ arm_matrix_instance_f32 * dst);\r
+\r
+\r
+\r
+ /**\r
+ * @ingroup groupController\r
+ */\r
+\r
+\r
+ /**\r
+ * @defgroup clarke Vector Clarke Transform\r
+ * Forward Clarke transform converts the instantaneous stator phases into a two-coordinate time invariant vector.\r
+ * Generally the Clarke transform uses three-phase currents <code>Ia, Ib and Ic</code> to calculate currents\r
+ * in the two-phase orthogonal stator axis <code>Ialpha</code> and <code>Ibeta</code>.\r
+ * When <code>Ialpha</code> is superposed with <code>Ia</code> as shown in the figure below\r
+ * \image html clarke.gif Stator current space vector and its components in (a,b).\r
+ * and <code>Ia + Ib + Ic = 0</code>, in this condition <code>Ialpha</code> and <code>Ibeta</code>\r
+ * can be calculated using only <code>Ia</code> and <code>Ib</code>.\r
+ *\r
+ * The function operates on a single sample of data and each call to the function returns the processed output.\r
+ * The library provides separate functions for Q31 and floating-point data types.\r
+ * \par Algorithm\r
+ * \image html clarkeFormula.gif\r
+ * where <code>Ia</code> and <code>Ib</code> are the instantaneous stator phases and\r
+ * <code>pIalpha</code> and <code>pIbeta</code> are the two coordinates of time invariant vector.\r
+ * \par Fixed-Point Behavior\r
+ * Care must be taken when using the Q31 version of the Clarke transform.\r
+ * In particular, the overflow and saturation behavior of the accumulator used must be considered.\r
+ * Refer to the function specific documentation below for usage guidelines.\r
+ */\r
+\r
+ /**\r
+ * @addtogroup clarke\r
+ * @{\r
+ */\r
+\r
+ /**\r
+ *\r
+ * @brief Floating-point Clarke transform\r
+ * @param[in] Ia input three-phase coordinate <code>a</code>\r
+ * @param[in] Ib input three-phase coordinate <code>b</code>\r
+ * @param[out] *pIalpha points to output two-phase orthogonal vector axis alpha\r
+ * @param[out] *pIbeta points to output two-phase orthogonal vector axis beta\r
+ * @return none.\r
+ */\r
+\r
+ __STATIC_INLINE void arm_clarke_f32(\r
+ float32_t Ia,\r
+ float32_t Ib,\r
+ float32_t * pIalpha,\r
+ float32_t * pIbeta)\r
+ {\r
+ /* Calculate pIalpha using the equation, pIalpha = Ia */\r
+ *pIalpha = Ia;\r
+\r
+ /* Calculate pIbeta using the equation, pIbeta = (1/sqrt(3)) * Ia + (2/sqrt(3)) * Ib */\r
+ *pIbeta = ((float32_t) 0.57735026919 * Ia + (float32_t) 1.15470053838 * Ib);\r
+\r
+ }\r
+\r
+ /**\r
+ * @brief Clarke transform for Q31 version\r
+ * @param[in] Ia input three-phase coordinate <code>a</code>\r
+ * @param[in] Ib input three-phase coordinate <code>b</code>\r
+ * @param[out] *pIalpha points to output two-phase orthogonal vector axis alpha\r
+ * @param[out] *pIbeta points to output two-phase orthogonal vector axis beta\r
+ * @return none.\r
+ *\r
+ * <b>Scaling and Overflow Behavior:</b>\r
+ * \par\r
+ * The function is implemented using an internal 32-bit accumulator.\r
+ * The accumulator maintains 1.31 format by truncating lower 31 bits of the intermediate multiplication in 2.62 format.\r
+ * There is saturation on the addition, hence there is no risk of overflow.\r
+ */\r
+\r
+ __STATIC_INLINE void arm_clarke_q31(\r
+ q31_t Ia,\r
+ q31_t Ib,\r
+ q31_t * pIalpha,\r
+ q31_t * pIbeta)\r
+ {\r
+ q31_t product1, product2; /* Temporary variables used to store intermediate results */\r
+\r
+ /* Calculating pIalpha from Ia by equation pIalpha = Ia */\r
+ *pIalpha = Ia;\r
+\r
+ /* Intermediate product is calculated by (1/(sqrt(3)) * Ia) */\r
+ product1 = (q31_t) (((q63_t) Ia * 0x24F34E8B) >> 30);\r
+\r
+ /* Intermediate product is calculated by (2/sqrt(3) * Ib) */\r
+ product2 = (q31_t) (((q63_t) Ib * 0x49E69D16) >> 30);\r
+\r
+ /* pIbeta is calculated by adding the intermediate products */\r
+ *pIbeta = __QADD(product1, product2);\r
+ }\r
+\r
+ /**\r
+ * @} end of clarke group\r
+ */\r
+\r
+ /**\r
+ * @brief Converts the elements of the Q7 vector to Q31 vector.\r
+ * @param[in] *pSrc input pointer\r
+ * @param[out] *pDst output pointer\r
+ * @param[in] blockSize number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_q7_to_q31(\r
+ q7_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+\r
+\r
+ /**\r
+ * @ingroup groupController\r
+ */\r
+\r
+ /**\r
+ * @defgroup inv_clarke Vector Inverse Clarke Transform\r
+ * Inverse Clarke transform converts the two-coordinate time invariant vector into instantaneous stator phases.\r
+ *\r
+ * The function operates on a single sample of data and each call to the function returns the processed output.\r
+ * The library provides separate functions for Q31 and floating-point data types.\r
+ * \par Algorithm\r
+ * \image html clarkeInvFormula.gif\r
+ * where <code>pIa</code> and <code>pIb</code> are the instantaneous stator phases and\r
+ * <code>Ialpha</code> and <code>Ibeta</code> are the two coordinates of time invariant vector.\r
+ * \par Fixed-Point Behavior\r
+ * Care must be taken when using the Q31 version of the Clarke transform.\r
+ * In particular, the overflow and saturation behavior of the accumulator used must be considered.\r
+ * Refer to the function specific documentation below for usage guidelines.\r
+ */\r
+\r
+ /**\r
+ * @addtogroup inv_clarke\r
+ * @{\r
+ */\r
+\r
+ /**\r
+ * @brief Floating-point Inverse Clarke transform\r
+ * @param[in] Ialpha input two-phase orthogonal vector axis alpha\r
+ * @param[in] Ibeta input two-phase orthogonal vector axis beta\r
+ * @param[out] *pIa points to output three-phase coordinate <code>a</code>\r
+ * @param[out] *pIb points to output three-phase coordinate <code>b</code>\r
+ * @return none.\r
+ */\r
+\r
+\r
+ __STATIC_INLINE void arm_inv_clarke_f32(\r
+ float32_t Ialpha,\r
+ float32_t Ibeta,\r
+ float32_t * pIa,\r
+ float32_t * pIb)\r
+ {\r
+ /* Calculating pIa from Ialpha by equation pIa = Ialpha */\r
+ *pIa = Ialpha;\r
+\r
+ /* Calculating pIb from Ialpha and Ibeta by equation pIb = -(1/2) * Ialpha + (sqrt(3)/2) * Ibeta */\r
+ *pIb = -0.5 * Ialpha + (float32_t) 0.8660254039 *Ibeta;\r
+\r
+ }\r
+\r
+ /**\r
+ * @brief Inverse Clarke transform for Q31 version\r
+ * @param[in] Ialpha input two-phase orthogonal vector axis alpha\r
+ * @param[in] Ibeta input two-phase orthogonal vector axis beta\r
+ * @param[out] *pIa points to output three-phase coordinate <code>a</code>\r
+ * @param[out] *pIb points to output three-phase coordinate <code>b</code>\r
+ * @return none.\r
+ *\r
+ * <b>Scaling and Overflow Behavior:</b>\r
+ * \par\r
+ * The function is implemented using an internal 32-bit accumulator.\r
+ * The accumulator maintains 1.31 format by truncating lower 31 bits of the intermediate multiplication in 2.62 format.\r
+ * There is saturation on the subtraction, hence there is no risk of overflow.\r
+ */\r
+\r
+ __STATIC_INLINE void arm_inv_clarke_q31(\r
+ q31_t Ialpha,\r
+ q31_t Ibeta,\r
+ q31_t * pIa,\r
+ q31_t * pIb)\r
+ {\r
+ q31_t product1, product2; /* Temporary variables used to store intermediate results */\r
+\r
+ /* Calculating pIa from Ialpha by equation pIa = Ialpha */\r
+ *pIa = Ialpha;\r
+\r
+ /* Intermediate product is calculated by (1/(2*sqrt(3)) * Ia) */\r
+ product1 = (q31_t) (((q63_t) (Ialpha) * (0x40000000)) >> 31);\r
+\r
+ /* Intermediate product is calculated by (1/sqrt(3) * pIb) */\r
+ product2 = (q31_t) (((q63_t) (Ibeta) * (0x6ED9EBA1)) >> 31);\r
+\r
+ /* pIb is calculated by subtracting the products */\r
+ *pIb = __QSUB(product2, product1);\r
+\r
+ }\r
+\r
+ /**\r
+ * @} end of inv_clarke group\r
+ */\r
+\r
+ /**\r
+ * @brief Converts the elements of the Q7 vector to Q15 vector.\r
+ * @param[in] *pSrc input pointer\r
+ * @param[out] *pDst output pointer\r
+ * @param[in] blockSize number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_q7_to_q15(\r
+ q7_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+\r
+ /**\r
+ * @ingroup groupController\r
+ */\r
+\r
+ /**\r
+ * @defgroup park Vector Park Transform\r
+ *\r
+ * Forward Park transform converts the input two-coordinate vector to flux and torque components.\r
+ * The Park transform can be used to realize the transformation of the <code>Ialpha</code> and the <code>Ibeta</code> currents\r
+ * from the stationary to the moving reference frame and control the spatial relationship between\r
+ * the stator vector current and rotor flux vector.\r
+ * If we consider the d axis aligned with the rotor flux, the diagram below shows the\r
+ * current vector and the relationship from the two reference frames:\r
+ * \image html park.gif "Stator current space vector and its component in (a,b) and in the d,q rotating reference frame"\r
+ *\r
+ * The function operates on a single sample of data and each call to the function returns the processed output.\r
+ * The library provides separate functions for Q31 and floating-point data types.\r
+ * \par Algorithm\r
+ * \image html parkFormula.gif\r
+ * where <code>Ialpha</code> and <code>Ibeta</code> are the stator vector components,\r
+ * <code>pId</code> and <code>pIq</code> are rotor vector components and <code>cosVal</code> and <code>sinVal</code> are the\r
+ * cosine and sine values of theta (rotor flux position).\r
+ * \par Fixed-Point Behavior\r
+ * Care must be taken when using the Q31 version of the Park transform.\r
+ * In particular, the overflow and saturation behavior of the accumulator used must be considered.\r
+ * Refer to the function specific documentation below for usage guidelines.\r
+ */\r
+\r
+ /**\r
+ * @addtogroup park\r
+ * @{\r
+ */\r
+\r
+ /**\r
+ * @brief Floating-point Park transform\r
+ * @param[in] Ialpha input two-phase vector coordinate alpha\r
+ * @param[in] Ibeta input two-phase vector coordinate beta\r
+ * @param[out] *pId points to output rotor reference frame d\r
+ * @param[out] *pIq points to output rotor reference frame q\r
+ * @param[in] sinVal sine value of rotation angle theta\r
+ * @param[in] cosVal cosine value of rotation angle theta\r
+ * @return none.\r
+ *\r
+ * The function implements the forward Park transform.\r
+ *\r
+ */\r
+\r
+ __STATIC_INLINE void arm_park_f32(\r
+ float32_t Ialpha,\r
+ float32_t Ibeta,\r
+ float32_t * pId,\r
+ float32_t * pIq,\r
+ float32_t sinVal,\r
+ float32_t cosVal)\r
+ {\r
+ /* Calculate pId using the equation, pId = Ialpha * cosVal + Ibeta * sinVal */\r
+ *pId = Ialpha * cosVal + Ibeta * sinVal;\r
+\r
+ /* Calculate pIq using the equation, pIq = - Ialpha * sinVal + Ibeta * cosVal */\r
+ *pIq = -Ialpha * sinVal + Ibeta * cosVal;\r
+\r
+ }\r
+\r
+ /**\r
+ * @brief Park transform for Q31 version\r
+ * @param[in] Ialpha input two-phase vector coordinate alpha\r
+ * @param[in] Ibeta input two-phase vector coordinate beta\r
+ * @param[out] *pId points to output rotor reference frame d\r
+ * @param[out] *pIq points to output rotor reference frame q\r
+ * @param[in] sinVal sine value of rotation angle theta\r
+ * @param[in] cosVal cosine value of rotation angle theta\r
+ * @return none.\r
+ *\r
+ * <b>Scaling and Overflow Behavior:</b>\r
+ * \par\r
+ * The function is implemented using an internal 32-bit accumulator.\r
+ * The accumulator maintains 1.31 format by truncating lower 31 bits of the intermediate multiplication in 2.62 format.\r
+ * There is saturation on the addition and subtraction, hence there is no risk of overflow.\r
+ */\r
+\r
+\r
+ __STATIC_INLINE void arm_park_q31(\r
+ q31_t Ialpha,\r
+ q31_t Ibeta,\r
+ q31_t * pId,\r
+ q31_t * pIq,\r
+ q31_t sinVal,\r
+ q31_t cosVal)\r
+ {\r
+ q31_t product1, product2; /* Temporary variables used to store intermediate results */\r
+ q31_t product3, product4; /* Temporary variables used to store intermediate results */\r
+\r
+ /* Intermediate product is calculated by (Ialpha * cosVal) */\r
+ product1 = (q31_t) (((q63_t) (Ialpha) * (cosVal)) >> 31);\r
+\r
+ /* Intermediate product is calculated by (Ibeta * sinVal) */\r
+ product2 = (q31_t) (((q63_t) (Ibeta) * (sinVal)) >> 31);\r
+\r
+\r
+ /* Intermediate product is calculated by (Ialpha * sinVal) */\r
+ product3 = (q31_t) (((q63_t) (Ialpha) * (sinVal)) >> 31);\r
+\r
+ /* Intermediate product is calculated by (Ibeta * cosVal) */\r
+ product4 = (q31_t) (((q63_t) (Ibeta) * (cosVal)) >> 31);\r
+\r
+ /* Calculate pId by adding the two intermediate products 1 and 2 */\r
+ *pId = __QADD(product1, product2);\r
+\r
+ /* Calculate pIq by subtracting the two intermediate products 3 from 4 */\r
+ *pIq = __QSUB(product4, product3);\r
+ }\r
+\r
+ /**\r
+ * @} end of park group\r
+ */\r
+\r
+ /**\r
+ * @brief Converts the elements of the Q7 vector to floating-point vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[out] *pDst is output pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_q7_to_float(\r
+ q7_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @ingroup groupController\r
+ */\r
+\r
+ /**\r
+ * @defgroup inv_park Vector Inverse Park transform\r
+ * Inverse Park transform converts the input flux and torque components to two-coordinate vector.\r
+ *\r
+ * The function operates on a single sample of data and each call to the function returns the processed output.\r
+ * The library provides separate functions for Q31 and floating-point data types.\r
+ * \par Algorithm\r
+ * \image html parkInvFormula.gif\r
+ * where <code>pIalpha</code> and <code>pIbeta</code> are the stator vector components,\r
+ * <code>Id</code> and <code>Iq</code> are rotor vector components and <code>cosVal</code> and <code>sinVal</code> are the\r
+ * cosine and sine values of theta (rotor flux position).\r
+ * \par Fixed-Point Behavior\r
+ * Care must be taken when using the Q31 version of the Park transform.\r
+ * In particular, the overflow and saturation behavior of the accumulator used must be considered.\r
+ * Refer to the function specific documentation below for usage guidelines.\r
+ */\r
+\r
+ /**\r
+ * @addtogroup inv_park\r
+ * @{\r
+ */\r
+\r
+ /**\r
+ * @brief Floating-point Inverse Park transform\r
+ * @param[in] Id input coordinate of rotor reference frame d\r
+ * @param[in] Iq input coordinate of rotor reference frame q\r
+ * @param[out] *pIalpha points to output two-phase orthogonal vector axis alpha\r
+ * @param[out] *pIbeta points to output two-phase orthogonal vector axis beta\r
+ * @param[in] sinVal sine value of rotation angle theta\r
+ * @param[in] cosVal cosine value of rotation angle theta\r
+ * @return none.\r
+ */\r
+\r
+ __STATIC_INLINE void arm_inv_park_f32(\r
+ float32_t Id,\r
+ float32_t Iq,\r
+ float32_t * pIalpha,\r
+ float32_t * pIbeta,\r
+ float32_t sinVal,\r
+ float32_t cosVal)\r
+ {\r
+ /* Calculate pIalpha using the equation, pIalpha = Id * cosVal - Iq * sinVal */\r
+ *pIalpha = Id * cosVal - Iq * sinVal;\r
+\r
+ /* Calculate pIbeta using the equation, pIbeta = Id * sinVal + Iq * cosVal */\r
+ *pIbeta = Id * sinVal + Iq * cosVal;\r
+\r
+ }\r
+\r
+\r
+ /**\r
+ * @brief Inverse Park transform for Q31 version\r
+ * @param[in] Id input coordinate of rotor reference frame d\r
+ * @param[in] Iq input coordinate of rotor reference frame q\r
+ * @param[out] *pIalpha points to output two-phase orthogonal vector axis alpha\r
+ * @param[out] *pIbeta points to output two-phase orthogonal vector axis beta\r
+ * @param[in] sinVal sine value of rotation angle theta\r
+ * @param[in] cosVal cosine value of rotation angle theta\r
+ * @return none.\r
+ *\r
+ * <b>Scaling and Overflow Behavior:</b>\r
+ * \par\r
+ * The function is implemented using an internal 32-bit accumulator.\r
+ * The accumulator maintains 1.31 format by truncating lower 31 bits of the intermediate multiplication in 2.62 format.\r
+ * There is saturation on the addition, hence there is no risk of overflow.\r
+ */\r
+\r
+\r
+ __STATIC_INLINE void arm_inv_park_q31(\r
+ q31_t Id,\r
+ q31_t Iq,\r
+ q31_t * pIalpha,\r
+ q31_t * pIbeta,\r
+ q31_t sinVal,\r
+ q31_t cosVal)\r
+ {\r
+ q31_t product1, product2; /* Temporary variables used to store intermediate results */\r
+ q31_t product3, product4; /* Temporary variables used to store intermediate results */\r
+\r
+ /* Intermediate product is calculated by (Id * cosVal) */\r
+ product1 = (q31_t) (((q63_t) (Id) * (cosVal)) >> 31);\r
+\r
+ /* Intermediate product is calculated by (Iq * sinVal) */\r
+ product2 = (q31_t) (((q63_t) (Iq) * (sinVal)) >> 31);\r
+\r
+\r
+ /* Intermediate product is calculated by (Id * sinVal) */\r
+ product3 = (q31_t) (((q63_t) (Id) * (sinVal)) >> 31);\r
+\r
+ /* Intermediate product is calculated by (Iq * cosVal) */\r
+ product4 = (q31_t) (((q63_t) (Iq) * (cosVal)) >> 31);\r
+\r
+ /* Calculate pIalpha by using the two intermediate products 1 and 2 */\r
+ *pIalpha = __QSUB(product1, product2);\r
+\r
+ /* Calculate pIbeta by using the two intermediate products 3 and 4 */\r
+ *pIbeta = __QADD(product4, product3);\r
+\r
+ }\r
+\r
+ /**\r
+ * @} end of Inverse park group\r
+ */\r
+\r
+\r
+ /**\r
+ * @brief Converts the elements of the Q31 vector to floating-point vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[out] *pDst is output pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_q31_to_float(\r
+ q31_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @ingroup groupInterpolation\r
+ */\r
+\r
+ /**\r
+ * @defgroup LinearInterpolate Linear Interpolation\r
+ *\r
+ * Linear interpolation is a method of curve fitting using linear polynomials.\r
+ * Linear interpolation works by effectively drawing a straight line between two neighboring samples and returning the appropriate point along that line\r
+ *\r
+ * \par\r
+ * \image html LinearInterp.gif "Linear interpolation"\r
+ *\r
+ * \par\r
+ * A Linear Interpolate function calculates an output value(y), for the input(x)\r
+ * using linear interpolation of the input values x0, x1( nearest input values) and the output values y0 and y1(nearest output values)\r
+ *\r
+ * \par Algorithm:\r
+ * <pre>\r
+ * y = y0 + (x - x0) * ((y1 - y0)/(x1-x0))\r
+ * where x0, x1 are nearest values of input x\r
+ * y0, y1 are nearest values to output y\r
+ * </pre>\r
+ *\r
+ * \par\r
+ * This set of functions implements Linear interpolation process\r
+ * for Q7, Q15, Q31, and floating-point data types. The functions operate on a single\r
+ * sample of data and each call to the function returns a single processed value.\r
+ * <code>S</code> points to an instance of the Linear Interpolate function data structure.\r
+ * <code>x</code> is the input sample value. The functions returns the output value.\r
+ *\r
+ * \par\r
+ * if x is outside of the table boundary, Linear interpolation returns first value of the table\r
+ * if x is below input range and returns last value of table if x is above range.\r
+ */\r
+\r
+ /**\r
+ * @addtogroup LinearInterpolate\r
+ * @{\r
+ */\r
+\r
+ /**\r
+ * @brief Process function for the floating-point Linear Interpolation Function.\r
+ * @param[in,out] *S is an instance of the floating-point Linear Interpolation structure\r
+ * @param[in] x input sample to process\r
+ * @return y processed output sample.\r
+ *\r
+ */\r
+\r
+ __STATIC_INLINE float32_t arm_linear_interp_f32(\r
+ arm_linear_interp_instance_f32 * S,\r
+ float32_t x)\r
+ {\r
+\r
+ float32_t y;\r
+ float32_t x0, x1; /* Nearest input values */\r
+ float32_t y0, y1; /* Nearest output values */\r
+ float32_t xSpacing = S->xSpacing; /* spacing between input values */\r
+ int32_t i; /* Index variable */\r
+ float32_t *pYData = S->pYData; /* pointer to output table */\r
+\r
+ /* Calculation of index */\r
+ i = (x - S->x1) / xSpacing;\r
+\r
+ if(i < 0)\r
+ {\r
+ /* Iniatilize output for below specified range as least output value of table */\r
+ y = pYData[0];\r
+ }\r
+ else if(i >= S->nValues)\r
+ {\r
+ /* Iniatilize output for above specified range as last output value of table */\r
+ y = pYData[S->nValues-1];\r
+ }\r
+ else\r
+ {\r
+ /* Calculation of nearest input values */\r
+ x0 = S->x1 + i * xSpacing;\r
+ x1 = S->x1 + (i +1) * xSpacing;\r
+\r
+ /* Read of nearest output values */\r
+ y0 = pYData[i];\r
+ y1 = pYData[i + 1];\r
+\r
+ /* Calculation of output */\r
+ y = y0 + (x - x0) * ((y1 - y0)/(x1-x0));\r
+\r
+ }\r
+\r
+ /* returns output value */\r
+ return (y);\r
+ }\r
+\r
+ /**\r
+ *\r
+ * @brief Process function for the Q31 Linear Interpolation Function.\r
+ * @param[in] *pYData pointer to Q31 Linear Interpolation table\r
+ * @param[in] x input sample to process\r
+ * @param[in] nValues number of table values\r
+ * @return y processed output sample.\r
+ *\r
+ * \par\r
+ * Input sample <code>x</code> is in 12.20 format which contains 12 bits for table index and 20 bits for fractional part.\r
+ * This function can support maximum of table size 2^12.\r
+ *\r
+ */\r
+\r
+\r
+ __STATIC_INLINE q31_t arm_linear_interp_q31(q31_t *pYData,\r
+ q31_t x, uint32_t nValues)\r
+ {\r
+ q31_t y; /* output */\r
+ q31_t y0, y1; /* Nearest output values */\r
+ q31_t fract; /* fractional part */\r
+ int32_t index; /* Index to read nearest output values */\r
+\r
+ /* Input is in 12.20 format */\r
+ /* 12 bits for the table index */\r
+ /* Index value calculation */\r
+ index = ((x & 0xFFF00000) >> 20);\r
+\r
+ if(index >= (nValues - 1))\r
+ {\r
+ return(pYData[nValues - 1]);\r
+ }\r
+ else if(index < 0)\r
+ {\r
+ return(pYData[0]);\r
+ }\r
+ else\r
+ {\r
+\r
+ /* 20 bits for the fractional part */\r
+ /* shift left by 11 to keep fract in 1.31 format */\r
+ fract = (x & 0x000FFFFF) << 11;\r
+\r
+ /* Read two nearest output values from the index in 1.31(q31) format */\r
+ y0 = pYData[index];\r
+ y1 = pYData[index + 1u];\r
+\r
+ /* Calculation of y0 * (1-fract) and y is in 2.30 format */\r
+ y = ((q31_t) ((q63_t) y0 * (0x7FFFFFFF - fract) >> 32));\r
+\r
+ /* Calculation of y0 * (1-fract) + y1 *fract and y is in 2.30 format */\r
+ y += ((q31_t) (((q63_t) y1 * fract) >> 32));\r
+\r
+ /* Convert y to 1.31 format */\r
+ return (y << 1u);\r
+\r
+ }\r
+\r
+ }\r
+\r
+ /**\r
+ *\r
+ * @brief Process function for the Q15 Linear Interpolation Function.\r
+ * @param[in] *pYData pointer to Q15 Linear Interpolation table\r
+ * @param[in] x input sample to process\r
+ * @param[in] nValues number of table values\r
+ * @return y processed output sample.\r
+ *\r
+ * \par\r
+ * Input sample <code>x</code> is in 12.20 format which contains 12 bits for table index and 20 bits for fractional part.\r
+ * This function can support maximum of table size 2^12.\r
+ *\r
+ */\r
+\r
+\r
+ __STATIC_INLINE q15_t arm_linear_interp_q15(q15_t *pYData, q31_t x, uint32_t nValues)\r
+ {\r
+ q63_t y; /* output */\r
+ q15_t y0, y1; /* Nearest output values */\r
+ q31_t fract; /* fractional part */\r
+ int32_t index; /* Index to read nearest output values */\r
+\r
+ /* Input is in 12.20 format */\r
+ /* 12 bits for the table index */\r
+ /* Index value calculation */\r
+ index = ((x & 0xFFF00000) >> 20u);\r
+\r
+ if(index >= (nValues - 1))\r
+ {\r
+ return(pYData[nValues - 1]);\r
+ }\r
+ else if(index < 0)\r
+ {\r
+ return(pYData[0]);\r
+ }\r
+ else\r
+ {\r
+ /* 20 bits for the fractional part */\r
+ /* fract is in 12.20 format */\r
+ fract = (x & 0x000FFFFF);\r
+\r
+ /* Read two nearest output values from the index */\r
+ y0 = pYData[index];\r
+ y1 = pYData[index + 1u];\r
+\r
+ /* Calculation of y0 * (1-fract) and y is in 13.35 format */\r
+ y = ((q63_t) y0 * (0xFFFFF - fract));\r
+\r
+ /* Calculation of (y0 * (1-fract) + y1 * fract) and y is in 13.35 format */\r
+ y += ((q63_t) y1 * (fract));\r
+\r
+ /* convert y to 1.15 format */\r
+ return (y >> 20);\r
+ }\r
+\r
+\r
+ }\r
+\r
+ /**\r
+ *\r
+ * @brief Process function for the Q7 Linear Interpolation Function.\r
+ * @param[in] *pYData pointer to Q7 Linear Interpolation table\r
+ * @param[in] x input sample to process\r
+ * @param[in] nValues number of table values\r
+ * @return y processed output sample.\r
+ *\r
+ * \par\r
+ * Input sample <code>x</code> is in 12.20 format which contains 12 bits for table index and 20 bits for fractional part.\r
+ * This function can support maximum of table size 2^12.\r
+ */\r
+\r
+\r
+ __STATIC_INLINE q7_t arm_linear_interp_q7(q7_t *pYData, q31_t x, uint32_t nValues)\r
+ {\r
+ q31_t y; /* output */\r
+ q7_t y0, y1; /* Nearest output values */\r
+ q31_t fract; /* fractional part */\r
+ int32_t index; /* Index to read nearest output values */\r
+\r
+ /* Input is in 12.20 format */\r
+ /* 12 bits for the table index */\r
+ /* Index value calculation */\r
+ index = ((x & 0xFFF00000) >> 20u);\r
+\r
+\r
+ if(index >= (nValues - 1))\r
+ {\r
+ return(pYData[nValues - 1]);\r
+ }\r
+ else if(index < 0)\r
+ {\r
+ return(pYData[0]);\r
+ }\r
+ else\r
+ {\r
+\r
+ /* 20 bits for the fractional part */\r
+ /* fract is in 12.20 format */\r
+ fract = (x & 0x000FFFFF);\r
+\r
+ /* Read two nearest output values from the index and are in 1.7(q7) format */\r
+ y0 = pYData[index];\r
+ y1 = pYData[index + 1u];\r
+\r
+ /* Calculation of y0 * (1-fract ) and y is in 13.27(q27) format */\r
+ y = ((y0 * (0xFFFFF - fract)));\r
+\r
+ /* Calculation of y1 * fract + y0 * (1-fract) and y is in 13.27(q27) format */\r
+ y += (y1 * fract);\r
+\r
+ /* convert y to 1.7(q7) format */\r
+ return (y >> 20u);\r
+\r
+ }\r
+\r
+ }\r
+ /**\r
+ * @} end of LinearInterpolate group\r
+ */\r
+\r
+ /**\r
+ * @brief Fast approximation to the trigonometric sine function for floating-point data.\r
+ * @param[in] x input value in radians.\r
+ * @return sin(x).\r
+ */\r
+\r
+ float32_t arm_sin_f32(\r
+ float32_t x);\r
+\r
+ /**\r
+ * @brief Fast approximation to the trigonometric sine function for Q31 data.\r
+ * @param[in] x Scaled input value in radians.\r
+ * @return sin(x).\r
+ */\r
+\r
+ q31_t arm_sin_q31(\r
+ q31_t x);\r
+\r
+ /**\r
+ * @brief Fast approximation to the trigonometric sine function for Q15 data.\r
+ * @param[in] x Scaled input value in radians.\r
+ * @return sin(x).\r
+ */\r
+\r
+ q15_t arm_sin_q15(\r
+ q15_t x);\r
+\r
+ /**\r
+ * @brief Fast approximation to the trigonometric cosine function for floating-point data.\r
+ * @param[in] x input value in radians.\r
+ * @return cos(x).\r
+ */\r
+\r
+ float32_t arm_cos_f32(\r
+ float32_t x);\r
+\r
+ /**\r
+ * @brief Fast approximation to the trigonometric cosine function for Q31 data.\r
+ * @param[in] x Scaled input value in radians.\r
+ * @return cos(x).\r
+ */\r
+\r
+ q31_t arm_cos_q31(\r
+ q31_t x);\r
+\r
+ /**\r
+ * @brief Fast approximation to the trigonometric cosine function for Q15 data.\r
+ * @param[in] x Scaled input value in radians.\r
+ * @return cos(x).\r
+ */\r
+\r
+ q15_t arm_cos_q15(\r
+ q15_t x);\r
+\r
+\r
+ /**\r
+ * @ingroup groupFastMath\r
+ */\r
+\r
+\r
+ /**\r
+ * @defgroup SQRT Square Root\r
+ *\r
+ * Computes the square root of a number.\r
+ * There are separate functions for Q15, Q31, and floating-point data types.\r
+ * The square root function is computed using the Newton-Raphson algorithm.\r
+ * This is an iterative algorithm of the form:\r
+ * <pre>\r
+ * x1 = x0 - f(x0)/f'(x0)\r
+ * </pre>\r
+ * where <code>x1</code> is the current estimate,\r
+ * <code>x0</code> is the previous estimate and\r
+ * <code>f'(x0)</code> is the derivative of <code>f()</code> evaluated at <code>x0</code>.\r
+ * For the square root function, the algorithm reduces to:\r
+ * <pre>\r
+ * x0 = in/2 [initial guess]\r
+ * x1 = 1/2 * ( x0 + in / x0) [each iteration]\r
+ * </pre>\r
+ */\r
+\r
+\r
+ /**\r
+ * @addtogroup SQRT\r
+ * @{\r
+ */\r
+\r
+ /**\r
+ * @brief Floating-point square root function.\r
+ * @param[in] in input value.\r
+ * @param[out] *pOut square root of input value.\r
+ * @return The function returns ARM_MATH_SUCCESS if input value is positive value or ARM_MATH_ARGUMENT_ERROR if\r
+ * <code>in</code> is negative value and returns zero output for negative values.\r
+ */\r
+\r
+ __STATIC_INLINE arm_status arm_sqrt_f32(\r
+ float32_t in, float32_t *pOut)\r
+ {\r
+ if(in > 0)\r
+ {\r
+\r
+// #if __FPU_USED\r
+ #if (__FPU_USED == 1) && defined ( __CC_ARM )\r
+ *pOut = __sqrtf(in);\r
+ #elif (__FPU_USED == 1) && defined ( __TMS_740 )\r
+ *pOut = __builtin_sqrtf(in);\r
+ #else\r
+ *pOut = sqrtf(in);\r
+ #endif\r
+\r
+ return (ARM_MATH_SUCCESS);\r
+ }\r
+ else\r
+ {\r
+ *pOut = 0.0f;\r
+ return (ARM_MATH_ARGUMENT_ERROR);\r
+ }\r
+\r
+ }\r
+\r
+\r
+ /**\r
+ * @brief Q31 square root function.\r
+ * @param[in] in input value. The range of the input value is [0 +1) or 0x00000000 to 0x7FFFFFFF.\r
+ * @param[out] *pOut square root of input value.\r
+ * @return The function returns ARM_MATH_SUCCESS if input value is positive value or ARM_MATH_ARGUMENT_ERROR if\r
+ * <code>in</code> is negative value and returns zero output for negative values.\r
+ */\r
+ arm_status arm_sqrt_q31(\r
+ q31_t in, q31_t *pOut);\r
+\r
+ /**\r
+ * @brief Q15 square root function.\r
+ * @param[in] in input value. The range of the input value is [0 +1) or 0x0000 to 0x7FFF.\r
+ * @param[out] *pOut square root of input value.\r
+ * @return The function returns ARM_MATH_SUCCESS if input value is positive value or ARM_MATH_ARGUMENT_ERROR if\r
+ * <code>in</code> is negative value and returns zero output for negative values.\r
+ */\r
+ arm_status arm_sqrt_q15(\r
+ q15_t in, q15_t *pOut);\r
+\r
+ /**\r
+ * @} end of SQRT group\r
+ */\r
+\r
+\r
+\r
+\r
+\r
+\r
+ /**\r
+ * @brief floating-point Circular write function.\r
+ */\r
+\r
+ __STATIC_INLINE void arm_circularWrite_f32(\r
+ int32_t * circBuffer,\r
+ int32_t L,\r
+ uint16_t * writeOffset,\r
+ int32_t bufferInc,\r
+ const int32_t * src,\r
+ int32_t srcInc,\r
+ uint32_t blockSize)\r
+ {\r
+ uint32_t i = 0u;\r
+ int32_t wOffset;\r
+\r
+ /* Copy the value of Index pointer that points\r
+ * to the current location where the input samples to be copied */\r
+ wOffset = *writeOffset;\r
+\r
+ /* Loop over the blockSize */\r
+ i = blockSize;\r
+\r
+ while(i > 0u)\r
+ {\r
+ /* copy the input sample to the circular buffer */\r
+ circBuffer[wOffset] = *src;\r
+\r
+ /* Update the input pointer */\r
+ src += srcInc;\r
+\r
+ /* Circularly update wOffset. Watch out for positive and negative value */\r
+ wOffset += bufferInc;\r
+ if(wOffset >= L)\r
+ wOffset -= L;\r
+\r
+ /* Decrement the loop counter */\r
+ i--;\r
+ }\r
+\r
+ /* Update the index pointer */\r
+ *writeOffset = wOffset;\r
+ }\r
+\r
+\r
+\r
+ /**\r
+ * @brief floating-point Circular Read function.\r
+ */\r
+ __STATIC_INLINE void arm_circularRead_f32(\r
+ int32_t * circBuffer,\r
+ int32_t L,\r
+ int32_t * readOffset,\r
+ int32_t bufferInc,\r
+ int32_t * dst,\r
+ int32_t * dst_base,\r
+ int32_t dst_length,\r
+ int32_t dstInc,\r
+ uint32_t blockSize)\r
+ {\r
+ uint32_t i = 0u;\r
+ int32_t rOffset, dst_end;\r
+\r
+ /* Copy the value of Index pointer that points\r
+ * to the current location from where the input samples to be read */\r
+ rOffset = *readOffset;\r
+ dst_end = (int32_t) (dst_base + dst_length);\r
+\r
+ /* Loop over the blockSize */\r
+ i = blockSize;\r
+\r
+ while(i > 0u)\r
+ {\r
+ /* copy the sample from the circular buffer to the destination buffer */\r
+ *dst = circBuffer[rOffset];\r
+\r
+ /* Update the input pointer */\r
+ dst += dstInc;\r
+\r
+ if(dst == (int32_t *) dst_end)\r
+ {\r
+ dst = dst_base;\r
+ }\r
+\r
+ /* Circularly update rOffset. Watch out for positive and negative value */\r
+ rOffset += bufferInc;\r
+\r
+ if(rOffset >= L)\r
+ {\r
+ rOffset -= L;\r
+ }\r
+\r
+ /* Decrement the loop counter */\r
+ i--;\r
+ }\r
+\r
+ /* Update the index pointer */\r
+ *readOffset = rOffset;\r
+ }\r
+\r
+ /**\r
+ * @brief Q15 Circular write function.\r
+ */\r
+\r
+ __STATIC_INLINE void arm_circularWrite_q15(\r
+ q15_t * circBuffer,\r
+ int32_t L,\r
+ uint16_t * writeOffset,\r
+ int32_t bufferInc,\r
+ const q15_t * src,\r
+ int32_t srcInc,\r
+ uint32_t blockSize)\r
+ {\r
+ uint32_t i = 0u;\r
+ int32_t wOffset;\r
+\r
+ /* Copy the value of Index pointer that points\r
+ * to the current location where the input samples to be copied */\r
+ wOffset = *writeOffset;\r
+\r
+ /* Loop over the blockSize */\r
+ i = blockSize;\r
+\r
+ while(i > 0u)\r
+ {\r
+ /* copy the input sample to the circular buffer */\r
+ circBuffer[wOffset] = *src;\r
+\r
+ /* Update the input pointer */\r
+ src += srcInc;\r
+\r
+ /* Circularly update wOffset. Watch out for positive and negative value */\r
+ wOffset += bufferInc;\r
+ if(wOffset >= L)\r
+ wOffset -= L;\r
+\r
+ /* Decrement the loop counter */\r
+ i--;\r
+ }\r
+\r
+ /* Update the index pointer */\r
+ *writeOffset = wOffset;\r
+ }\r
+\r
+\r
+\r
+ /**\r
+ * @brief Q15 Circular Read function.\r
+ */\r
+ __STATIC_INLINE void arm_circularRead_q15(\r
+ q15_t * circBuffer,\r
+ int32_t L,\r
+ int32_t * readOffset,\r
+ int32_t bufferInc,\r
+ q15_t * dst,\r
+ q15_t * dst_base,\r
+ int32_t dst_length,\r
+ int32_t dstInc,\r
+ uint32_t blockSize)\r
+ {\r
+ uint32_t i = 0;\r
+ int32_t rOffset, dst_end;\r
+\r
+ /* Copy the value of Index pointer that points\r
+ * to the current location from where the input samples to be read */\r
+ rOffset = *readOffset;\r
+\r
+ dst_end = (int32_t) (dst_base + dst_length);\r
+\r
+ /* Loop over the blockSize */\r
+ i = blockSize;\r
+\r
+ while(i > 0u)\r
+ {\r
+ /* copy the sample from the circular buffer to the destination buffer */\r
+ *dst = circBuffer[rOffset];\r
+\r
+ /* Update the input pointer */\r
+ dst += dstInc;\r
+\r
+ if(dst == (q15_t *) dst_end)\r
+ {\r
+ dst = dst_base;\r
+ }\r
+\r
+ /* Circularly update wOffset. Watch out for positive and negative value */\r
+ rOffset += bufferInc;\r
+\r
+ if(rOffset >= L)\r
+ {\r
+ rOffset -= L;\r
+ }\r
+\r
+ /* Decrement the loop counter */\r
+ i--;\r
+ }\r
+\r
+ /* Update the index pointer */\r
+ *readOffset = rOffset;\r
+ }\r
+\r
+\r
+ /**\r
+ * @brief Q7 Circular write function.\r
+ */\r
+\r
+ __STATIC_INLINE void arm_circularWrite_q7(\r
+ q7_t * circBuffer,\r
+ int32_t L,\r
+ uint16_t * writeOffset,\r
+ int32_t bufferInc,\r
+ const q7_t * src,\r
+ int32_t srcInc,\r
+ uint32_t blockSize)\r
+ {\r
+ uint32_t i = 0u;\r
+ int32_t wOffset;\r
+\r
+ /* Copy the value of Index pointer that points\r
+ * to the current location where the input samples to be copied */\r
+ wOffset = *writeOffset;\r
+\r
+ /* Loop over the blockSize */\r
+ i = blockSize;\r
+\r
+ while(i > 0u)\r
+ {\r
+ /* copy the input sample to the circular buffer */\r
+ circBuffer[wOffset] = *src;\r
+\r
+ /* Update the input pointer */\r
+ src += srcInc;\r
+\r
+ /* Circularly update wOffset. Watch out for positive and negative value */\r
+ wOffset += bufferInc;\r
+ if(wOffset >= L)\r
+ wOffset -= L;\r
+\r
+ /* Decrement the loop counter */\r
+ i--;\r
+ }\r
+\r
+ /* Update the index pointer */\r
+ *writeOffset = wOffset;\r
+ }\r
+\r
+\r
+\r
+ /**\r
+ * @brief Q7 Circular Read function.\r
+ */\r
+ __STATIC_INLINE void arm_circularRead_q7(\r
+ q7_t * circBuffer,\r
+ int32_t L,\r
+ int32_t * readOffset,\r
+ int32_t bufferInc,\r
+ q7_t * dst,\r
+ q7_t * dst_base,\r
+ int32_t dst_length,\r
+ int32_t dstInc,\r
+ uint32_t blockSize)\r
+ {\r
+ uint32_t i = 0;\r
+ int32_t rOffset, dst_end;\r
+\r
+ /* Copy the value of Index pointer that points\r
+ * to the current location from where the input samples to be read */\r
+ rOffset = *readOffset;\r
+\r
+ dst_end = (int32_t) (dst_base + dst_length);\r
+\r
+ /* Loop over the blockSize */\r
+ i = blockSize;\r
+\r
+ while(i > 0u)\r
+ {\r
+ /* copy the sample from the circular buffer to the destination buffer */\r
+ *dst = circBuffer[rOffset];\r
+\r
+ /* Update the input pointer */\r
+ dst += dstInc;\r
+\r
+ if(dst == (q7_t *) dst_end)\r
+ {\r
+ dst = dst_base;\r
+ }\r
+\r
+ /* Circularly update rOffset. Watch out for positive and negative value */\r
+ rOffset += bufferInc;\r
+\r
+ if(rOffset >= L)\r
+ {\r
+ rOffset -= L;\r
+ }\r
+\r
+ /* Decrement the loop counter */\r
+ i--;\r
+ }\r
+\r
+ /* Update the index pointer */\r
+ *readOffset = rOffset;\r
+ }\r
+\r
+\r
+ /**\r
+ * @brief Sum of the squares of the elements of a Q31 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_power_q31(\r
+ q31_t * pSrc,\r
+ uint32_t blockSize,\r
+ q63_t * pResult);\r
+\r
+ /**\r
+ * @brief Sum of the squares of the elements of a floating-point vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_power_f32(\r
+ float32_t * pSrc,\r
+ uint32_t blockSize,\r
+ float32_t * pResult);\r
+\r
+ /**\r
+ * @brief Sum of the squares of the elements of a Q15 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_power_q15(\r
+ q15_t * pSrc,\r
+ uint32_t blockSize,\r
+ q63_t * pResult);\r
+\r
+ /**\r
+ * @brief Sum of the squares of the elements of a Q7 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_power_q7(\r
+ q7_t * pSrc,\r
+ uint32_t blockSize,\r
+ q31_t * pResult);\r
+\r
+ /**\r
+ * @brief Mean value of a Q7 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_mean_q7(\r
+ q7_t * pSrc,\r
+ uint32_t blockSize,\r
+ q7_t * pResult);\r
+\r
+ /**\r
+ * @brief Mean value of a Q15 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+ void arm_mean_q15(\r
+ q15_t * pSrc,\r
+ uint32_t blockSize,\r
+ q15_t * pResult);\r
+\r
+ /**\r
+ * @brief Mean value of a Q31 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+ void arm_mean_q31(\r
+ q31_t * pSrc,\r
+ uint32_t blockSize,\r
+ q31_t * pResult);\r
+\r
+ /**\r
+ * @brief Mean value of a floating-point vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+ void arm_mean_f32(\r
+ float32_t * pSrc,\r
+ uint32_t blockSize,\r
+ float32_t * pResult);\r
+\r
+ /**\r
+ * @brief Variance of the elements of a floating-point vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_var_f32(\r
+ float32_t * pSrc,\r
+ uint32_t blockSize,\r
+ float32_t * pResult);\r
+\r
+ /**\r
+ * @brief Variance of the elements of a Q31 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_var_q31(\r
+ q31_t * pSrc,\r
+ uint32_t blockSize,\r
+ q63_t * pResult);\r
+\r
+ /**\r
+ * @brief Variance of the elements of a Q15 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_var_q15(\r
+ q15_t * pSrc,\r
+ uint32_t blockSize,\r
+ q31_t * pResult);\r
+\r
+ /**\r
+ * @brief Root Mean Square of the elements of a floating-point vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_rms_f32(\r
+ float32_t * pSrc,\r
+ uint32_t blockSize,\r
+ float32_t * pResult);\r
+\r
+ /**\r
+ * @brief Root Mean Square of the elements of a Q31 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_rms_q31(\r
+ q31_t * pSrc,\r
+ uint32_t blockSize,\r
+ q31_t * pResult);\r
+\r
+ /**\r
+ * @brief Root Mean Square of the elements of a Q15 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_rms_q15(\r
+ q15_t * pSrc,\r
+ uint32_t blockSize,\r
+ q15_t * pResult);\r
+\r
+ /**\r
+ * @brief Standard deviation of the elements of a floating-point vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_std_f32(\r
+ float32_t * pSrc,\r
+ uint32_t blockSize,\r
+ float32_t * pResult);\r
+\r
+ /**\r
+ * @brief Standard deviation of the elements of a Q31 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_std_q31(\r
+ q31_t * pSrc,\r
+ uint32_t blockSize,\r
+ q31_t * pResult);\r
+\r
+ /**\r
+ * @brief Standard deviation of the elements of a Q15 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output value.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_std_q15(\r
+ q15_t * pSrc,\r
+ uint32_t blockSize,\r
+ q15_t * pResult);\r
+\r
+ /**\r
+ * @brief Floating-point complex magnitude\r
+ * @param[in] *pSrc points to the complex input vector\r
+ * @param[out] *pDst points to the real output vector\r
+ * @param[in] numSamples number of complex samples in the input vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_mag_f32(\r
+ float32_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t numSamples);\r
+\r
+ /**\r
+ * @brief Q31 complex magnitude\r
+ * @param[in] *pSrc points to the complex input vector\r
+ * @param[out] *pDst points to the real output vector\r
+ * @param[in] numSamples number of complex samples in the input vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_mag_q31(\r
+ q31_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t numSamples);\r
+\r
+ /**\r
+ * @brief Q15 complex magnitude\r
+ * @param[in] *pSrc points to the complex input vector\r
+ * @param[out] *pDst points to the real output vector\r
+ * @param[in] numSamples number of complex samples in the input vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_mag_q15(\r
+ q15_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t numSamples);\r
+\r
+ /**\r
+ * @brief Q15 complex dot product\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[in] numSamples number of complex samples in each vector\r
+ * @param[out] *realResult real part of the result returned here\r
+ * @param[out] *imagResult imaginary part of the result returned here\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_dot_prod_q15(\r
+ q15_t * pSrcA,\r
+ q15_t * pSrcB,\r
+ uint32_t numSamples,\r
+ q31_t * realResult,\r
+ q31_t * imagResult);\r
+\r
+ /**\r
+ * @brief Q31 complex dot product\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[in] numSamples number of complex samples in each vector\r
+ * @param[out] *realResult real part of the result returned here\r
+ * @param[out] *imagResult imaginary part of the result returned here\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_dot_prod_q31(\r
+ q31_t * pSrcA,\r
+ q31_t * pSrcB,\r
+ uint32_t numSamples,\r
+ q63_t * realResult,\r
+ q63_t * imagResult);\r
+\r
+ /**\r
+ * @brief Floating-point complex dot product\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[in] numSamples number of complex samples in each vector\r
+ * @param[out] *realResult real part of the result returned here\r
+ * @param[out] *imagResult imaginary part of the result returned here\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_dot_prod_f32(\r
+ float32_t * pSrcA,\r
+ float32_t * pSrcB,\r
+ uint32_t numSamples,\r
+ float32_t * realResult,\r
+ float32_t * imagResult);\r
+\r
+ /**\r
+ * @brief Q15 complex-by-real multiplication\r
+ * @param[in] *pSrcCmplx points to the complex input vector\r
+ * @param[in] *pSrcReal points to the real input vector\r
+ * @param[out] *pCmplxDst points to the complex output vector\r
+ * @param[in] numSamples number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_mult_real_q15(\r
+ q15_t * pSrcCmplx,\r
+ q15_t * pSrcReal,\r
+ q15_t * pCmplxDst,\r
+ uint32_t numSamples);\r
+\r
+ /**\r
+ * @brief Q31 complex-by-real multiplication\r
+ * @param[in] *pSrcCmplx points to the complex input vector\r
+ * @param[in] *pSrcReal points to the real input vector\r
+ * @param[out] *pCmplxDst points to the complex output vector\r
+ * @param[in] numSamples number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_mult_real_q31(\r
+ q31_t * pSrcCmplx,\r
+ q31_t * pSrcReal,\r
+ q31_t * pCmplxDst,\r
+ uint32_t numSamples);\r
+\r
+ /**\r
+ * @brief Floating-point complex-by-real multiplication\r
+ * @param[in] *pSrcCmplx points to the complex input vector\r
+ * @param[in] *pSrcReal points to the real input vector\r
+ * @param[out] *pCmplxDst points to the complex output vector\r
+ * @param[in] numSamples number of samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_mult_real_f32(\r
+ float32_t * pSrcCmplx,\r
+ float32_t * pSrcReal,\r
+ float32_t * pCmplxDst,\r
+ uint32_t numSamples);\r
+\r
+ /**\r
+ * @brief Minimum value of a Q7 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *result is output pointer\r
+ * @param[in] index is the array index of the minimum value in the input buffer.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_min_q7(\r
+ q7_t * pSrc,\r
+ uint32_t blockSize,\r
+ q7_t * result,\r
+ uint32_t * index);\r
+\r
+ /**\r
+ * @brief Minimum value of a Q15 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output pointer\r
+ * @param[in] *pIndex is the array index of the minimum value in the input buffer.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_min_q15(\r
+ q15_t * pSrc,\r
+ uint32_t blockSize,\r
+ q15_t * pResult,\r
+ uint32_t * pIndex);\r
+\r
+ /**\r
+ * @brief Minimum value of a Q31 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output pointer\r
+ * @param[out] *pIndex is the array index of the minimum value in the input buffer.\r
+ * @return none.\r
+ */\r
+ void arm_min_q31(\r
+ q31_t * pSrc,\r
+ uint32_t blockSize,\r
+ q31_t * pResult,\r
+ uint32_t * pIndex);\r
+\r
+ /**\r
+ * @brief Minimum value of a floating-point vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @param[out] *pResult is output pointer\r
+ * @param[out] *pIndex is the array index of the minimum value in the input buffer.\r
+ * @return none.\r
+ */\r
+\r
+ void arm_min_f32(\r
+ float32_t * pSrc,\r
+ uint32_t blockSize,\r
+ float32_t * pResult,\r
+ uint32_t * pIndex);\r
+\r
+/**\r
+ * @brief Maximum value of a Q7 vector.\r
+ * @param[in] *pSrc points to the input buffer\r
+ * @param[in] blockSize length of the input vector\r
+ * @param[out] *pResult maximum value returned here\r
+ * @param[out] *pIndex index of maximum value returned here\r
+ * @return none.\r
+ */\r
+\r
+ void arm_max_q7(\r
+ q7_t * pSrc,\r
+ uint32_t blockSize,\r
+ q7_t * pResult,\r
+ uint32_t * pIndex);\r
+\r
+/**\r
+ * @brief Maximum value of a Q15 vector.\r
+ * @param[in] *pSrc points to the input buffer\r
+ * @param[in] blockSize length of the input vector\r
+ * @param[out] *pResult maximum value returned here\r
+ * @param[out] *pIndex index of maximum value returned here\r
+ * @return none.\r
+ */\r
+\r
+ void arm_max_q15(\r
+ q15_t * pSrc,\r
+ uint32_t blockSize,\r
+ q15_t * pResult,\r
+ uint32_t * pIndex);\r
+\r
+/**\r
+ * @brief Maximum value of a Q31 vector.\r
+ * @param[in] *pSrc points to the input buffer\r
+ * @param[in] blockSize length of the input vector\r
+ * @param[out] *pResult maximum value returned here\r
+ * @param[out] *pIndex index of maximum value returned here\r
+ * @return none.\r
+ */\r
+\r
+ void arm_max_q31(\r
+ q31_t * pSrc,\r
+ uint32_t blockSize,\r
+ q31_t * pResult,\r
+ uint32_t * pIndex);\r
+\r
+/**\r
+ * @brief Maximum value of a floating-point vector.\r
+ * @param[in] *pSrc points to the input buffer\r
+ * @param[in] blockSize length of the input vector\r
+ * @param[out] *pResult maximum value returned here\r
+ * @param[out] *pIndex index of maximum value returned here\r
+ * @return none.\r
+ */\r
+\r
+ void arm_max_f32(\r
+ float32_t * pSrc,\r
+ uint32_t blockSize,\r
+ float32_t * pResult,\r
+ uint32_t * pIndex);\r
+\r
+ /**\r
+ * @brief Q15 complex-by-complex multiplication\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] numSamples number of complex samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_mult_cmplx_q15(\r
+ q15_t * pSrcA,\r
+ q15_t * pSrcB,\r
+ q15_t * pDst,\r
+ uint32_t numSamples);\r
+\r
+ /**\r
+ * @brief Q31 complex-by-complex multiplication\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] numSamples number of complex samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_mult_cmplx_q31(\r
+ q31_t * pSrcA,\r
+ q31_t * pSrcB,\r
+ q31_t * pDst,\r
+ uint32_t numSamples);\r
+\r
+ /**\r
+ * @brief Floating-point complex-by-complex multiplication\r
+ * @param[in] *pSrcA points to the first input vector\r
+ * @param[in] *pSrcB points to the second input vector\r
+ * @param[out] *pDst points to the output vector\r
+ * @param[in] numSamples number of complex samples in each vector\r
+ * @return none.\r
+ */\r
+\r
+ void arm_cmplx_mult_cmplx_f32(\r
+ float32_t * pSrcA,\r
+ float32_t * pSrcB,\r
+ float32_t * pDst,\r
+ uint32_t numSamples);\r
+\r
+ /**\r
+ * @brief Converts the elements of the floating-point vector to Q31 vector.\r
+ * @param[in] *pSrc points to the floating-point input vector\r
+ * @param[out] *pDst points to the Q31 output vector\r
+ * @param[in] blockSize length of the input vector\r
+ * @return none.\r
+ */\r
+ void arm_float_to_q31(\r
+ float32_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Converts the elements of the floating-point vector to Q15 vector.\r
+ * @param[in] *pSrc points to the floating-point input vector\r
+ * @param[out] *pDst points to the Q15 output vector\r
+ * @param[in] blockSize length of the input vector\r
+ * @return none\r
+ */\r
+ void arm_float_to_q15(\r
+ float32_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Converts the elements of the floating-point vector to Q7 vector.\r
+ * @param[in] *pSrc points to the floating-point input vector\r
+ * @param[out] *pDst points to the Q7 output vector\r
+ * @param[in] blockSize length of the input vector\r
+ * @return none\r
+ */\r
+ void arm_float_to_q7(\r
+ float32_t * pSrc,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Converts the elements of the Q31 vector to Q15 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[out] *pDst is output pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_q31_to_q15(\r
+ q31_t * pSrc,\r
+ q15_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Converts the elements of the Q31 vector to Q7 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[out] *pDst is output pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_q31_to_q7(\r
+ q31_t * pSrc,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+ /**\r
+ * @brief Converts the elements of the Q15 vector to floating-point vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[out] *pDst is output pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_q15_to_float(\r
+ q15_t * pSrc,\r
+ float32_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Converts the elements of the Q15 vector to Q31 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[out] *pDst is output pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_q15_to_q31(\r
+ q15_t * pSrc,\r
+ q31_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @brief Converts the elements of the Q15 vector to Q7 vector.\r
+ * @param[in] *pSrc is input pointer\r
+ * @param[out] *pDst is output pointer\r
+ * @param[in] blockSize is the number of samples to process\r
+ * @return none.\r
+ */\r
+ void arm_q15_to_q7(\r
+ q15_t * pSrc,\r
+ q7_t * pDst,\r
+ uint32_t blockSize);\r
+\r
+\r
+ /**\r
+ * @ingroup groupInterpolation\r
+ */\r
+\r
+ /**\r
+ * @defgroup BilinearInterpolate Bilinear Interpolation\r
+ *\r
+ * Bilinear interpolation is an extension of linear interpolation applied to a two dimensional grid.\r
+ * The underlying function <code>f(x, y)</code> is sampled on a regular grid and the interpolation process\r
+ * determines values between the grid points.\r
+ * Bilinear interpolation is equivalent to two step linear interpolation, first in the x-dimension and then in the y-dimension.\r
+ * Bilinear interpolation is often used in image processing to rescale images.\r
+ * The CMSIS DSP library provides bilinear interpolation functions for Q7, Q15, Q31, and floating-point data types.\r
+ *\r
+ * <b>Algorithm</b>\r
+ * \par\r
+ * The instance structure used by the bilinear interpolation functions describes a two dimensional data table.\r
+ * For floating-point, the instance structure is defined as:\r
+ * <pre>\r
+ * typedef struct\r
+ * {\r
+ * uint16_t numRows;\r
+ * uint16_t numCols;\r
+ * float32_t *pData;\r
+ * } arm_bilinear_interp_instance_f32;\r
+ * </pre>\r
+ *\r
+ * \par\r
+ * where <code>numRows</code> specifies the number of rows in the table;\r
+ * <code>numCols</code> specifies the number of columns in the table;\r
+ * and <code>pData</code> points to an array of size <code>numRows*numCols</code> values.\r
+ * The data table <code>pTable</code> is organized in row order and the supplied data values fall on integer indexes.\r
+ * That is, table element (x,y) is located at <code>pTable[x + y*numCols]</code> where x and y are integers.\r
+ *\r
+ * \par\r
+ * Let <code>(x, y)</code> specify the desired interpolation point. Then define:\r
+ * <pre>\r
+ * XF = floor(x)\r
+ * YF = floor(y)\r
+ * </pre>\r
+ * \par\r
+ * The interpolated output point is computed as:\r
+ * <pre>\r
+ * f(x, y) = f(XF, YF) * (1-(x-XF)) * (1-(y-YF))\r
+ * + f(XF+1, YF) * (x-XF)*(1-(y-YF))\r
+ * + f(XF, YF+1) * (1-(x-XF))*(y-YF)\r
+ * + f(XF+1, YF+1) * (x-XF)*(y-YF)\r
+ * </pre>\r
+ * Note that the coordinates (x, y) contain integer and fractional components.\r
+ * The integer components specify which portion of the table to use while the\r
+ * fractional components control the interpolation processor.\r
+ *\r
+ * \par\r
+ * if (x,y) are outside of the table boundary, Bilinear interpolation returns zero output.\r
+ */\r
+\r
+ /**\r
+ * @addtogroup BilinearInterpolate\r
+ * @{\r
+ */\r
+\r
+ /**\r
+ *\r
+ * @brief Floating-point bilinear interpolation.\r
+ * @param[in,out] *S points to an instance of the interpolation structure.\r
+ * @param[in] X interpolation coordinate.\r
+ * @param[in] Y interpolation coordinate.\r
+ * @return out interpolated value.\r
+ */\r
+\r
+\r
+ __STATIC_INLINE float32_t arm_bilinear_interp_f32(\r
+ const arm_bilinear_interp_instance_f32 * S,\r
+ float32_t X,\r
+ float32_t Y)\r
+ {\r
+ float32_t out;\r
+ float32_t f00, f01, f10, f11;\r
+ float32_t *pData = S->pData;\r
+ int32_t xIndex, yIndex, index;\r
+ float32_t xdiff, ydiff;\r
+ float32_t b1, b2, b3, b4;\r
+\r
+ xIndex = (int32_t) X;\r
+ yIndex = (int32_t) Y;\r
+\r
+ /* Care taken for table outside boundary */\r
+ /* Returns zero output when values are outside table boundary */\r
+ if(xIndex < 0 || xIndex > (S->numRows-1) || yIndex < 0 || yIndex > ( S->numCols-1))\r
+ {\r
+ return(0);\r
+ }\r
+\r
+ /* Calculation of index for two nearest points in X-direction */\r
+ index = (xIndex - 1) + (yIndex-1) * S->numCols ;\r
+\r
+\r
+ /* Read two nearest points in X-direction */\r
+ f00 = pData[index];\r
+ f01 = pData[index + 1];\r
+\r
+ /* Calculation of index for two nearest points in Y-direction */\r
+ index = (xIndex-1) + (yIndex) * S->numCols;\r
+\r
+\r
+ /* Read two nearest points in Y-direction */\r
+ f10 = pData[index];\r
+ f11 = pData[index + 1];\r
+\r
+ /* Calculation of intermediate values */\r
+ b1 = f00;\r
+ b2 = f01 - f00;\r
+ b3 = f10 - f00;\r
+ b4 = f00 - f01 - f10 + f11;\r
+\r
+ /* Calculation of fractional part in X */\r
+ xdiff = X - xIndex;\r
+\r
+ /* Calculation of fractional part in Y */\r
+ ydiff = Y - yIndex;\r
+\r
+ /* Calculation of bi-linear interpolated output */\r
+ out = b1 + b2 * xdiff + b3 * ydiff + b4 * xdiff * ydiff;\r
+\r
+ /* return to application */\r
+ return (out);\r
+\r
+ }\r
+\r
+ /**\r
+ *\r
+ * @brief Q31 bilinear interpolation.\r
+ * @param[in,out] *S points to an instance of the interpolation structure.\r
+ * @param[in] X interpolation coordinate in 12.20 format.\r
+ * @param[in] Y interpolation coordinate in 12.20 format.\r
+ * @return out interpolated value.\r
+ */\r
+\r
+ __STATIC_INLINE q31_t arm_bilinear_interp_q31(\r
+ arm_bilinear_interp_instance_q31 * S,\r
+ q31_t X,\r
+ q31_t Y)\r
+ {\r
+ q31_t out; /* Temporary output */\r
+ q31_t acc = 0; /* output */\r
+ q31_t xfract, yfract; /* X, Y fractional parts */\r
+ q31_t x1, x2, y1, y2; /* Nearest output values */\r
+ int32_t rI, cI; /* Row and column indices */\r
+ q31_t *pYData = S->pData; /* pointer to output table values */\r
+ uint32_t nCols = S->numCols; /* num of rows */\r
+\r
+\r
+ /* Input is in 12.20 format */\r
+ /* 12 bits for the table index */\r
+ /* Index value calculation */\r
+ rI = ((X & 0xFFF00000) >> 20u);\r
+\r
+ /* Input is in 12.20 format */\r
+ /* 12 bits for the table index */\r
+ /* Index value calculation */\r
+ cI = ((Y & 0xFFF00000) >> 20u);\r
+\r
+ /* Care taken for table outside boundary */\r
+ /* Returns zero output when values are outside table boundary */\r
+ if(rI < 0 || rI > (S->numRows-1) || cI < 0 || cI > ( S->numCols-1))\r
+ {\r
+ return(0);\r
+ }\r
+\r
+ /* 20 bits for the fractional part */\r
+ /* shift left xfract by 11 to keep 1.31 format */\r
+ xfract = (X & 0x000FFFFF) << 11u;\r
+\r
+ /* Read two nearest output values from the index */\r
+ x1 = pYData[(rI) + nCols * (cI)];\r
+ x2 = pYData[(rI) + nCols * (cI) + 1u];\r
+\r
+ /* 20 bits for the fractional part */\r
+ /* shift left yfract by 11 to keep 1.31 format */\r
+ yfract = (Y & 0x000FFFFF) << 11u;\r
+\r
+ /* Read two nearest output values from the index */\r
+ y1 = pYData[(rI) + nCols * (cI + 1)];\r
+ y2 = pYData[(rI) + nCols * (cI + 1) + 1u];\r
+\r
+ /* Calculation of x1 * (1-xfract ) * (1-yfract) and acc is in 3.29(q29) format */\r
+ out = ((q31_t) (((q63_t) x1 * (0x7FFFFFFF - xfract)) >> 32));\r
+ acc = ((q31_t) (((q63_t) out * (0x7FFFFFFF - yfract)) >> 32));\r
+\r
+ /* x2 * (xfract) * (1-yfract) in 3.29(q29) and adding to acc */\r
+ out = ((q31_t) ((q63_t) x2 * (0x7FFFFFFF - yfract) >> 32));\r
+ acc += ((q31_t) ((q63_t) out * (xfract) >> 32));\r
+\r
+ /* y1 * (1 - xfract) * (yfract) in 3.29(q29) and adding to acc */\r
+ out = ((q31_t) ((q63_t) y1 * (0x7FFFFFFF - xfract) >> 32));\r
+ acc += ((q31_t) ((q63_t) out * (yfract) >> 32));\r
+\r
+ /* y2 * (xfract) * (yfract) in 3.29(q29) and adding to acc */\r
+ out = ((q31_t) ((q63_t) y2 * (xfract) >> 32));\r
+ acc += ((q31_t) ((q63_t) out * (yfract) >> 32));\r
+\r
+ /* Convert acc to 1.31(q31) format */\r
+ return (acc << 2u);\r
+\r
+ }\r
+\r
+ /**\r
+ * @brief Q15 bilinear interpolation.\r
+ * @param[in,out] *S points to an instance of the interpolation structure.\r
+ * @param[in] X interpolation coordinate in 12.20 format.\r
+ * @param[in] Y interpolation coordinate in 12.20 format.\r
+ * @return out interpolated value.\r
+ */\r
+\r
+ __STATIC_INLINE q15_t arm_bilinear_interp_q15(\r
+ arm_bilinear_interp_instance_q15 * S,\r
+ q31_t X,\r
+ q31_t Y)\r
+ {\r
+ q63_t acc = 0; /* output */\r
+ q31_t out; /* Temporary output */\r
+ q15_t x1, x2, y1, y2; /* Nearest output values */\r
+ q31_t xfract, yfract; /* X, Y fractional parts */\r
+ int32_t rI, cI; /* Row and column indices */\r
+ q15_t *pYData = S->pData; /* pointer to output table values */\r
+ uint32_t nCols = S->numCols; /* num of rows */\r
+\r
+ /* Input is in 12.20 format */\r
+ /* 12 bits for the table index */\r
+ /* Index value calculation */\r
+ rI = ((X & 0xFFF00000) >> 20);\r
+\r
+ /* Input is in 12.20 format */\r
+ /* 12 bits for the table index */\r
+ /* Index value calculation */\r
+ cI = ((Y & 0xFFF00000) >> 20);\r
+\r
+ /* Care taken for table outside boundary */\r
+ /* Returns zero output when values are outside table boundary */\r
+ if(rI < 0 || rI > (S->numRows-1) || cI < 0 || cI > ( S->numCols-1))\r
+ {\r
+ return(0);\r
+ }\r
+\r
+ /* 20 bits for the fractional part */\r
+ /* xfract should be in 12.20 format */\r
+ xfract = (X & 0x000FFFFF);\r
+\r
+ /* Read two nearest output values from the index */\r
+ x1 = pYData[(rI) + nCols * (cI)];\r
+ x2 = pYData[(rI) + nCols * (cI) + 1u];\r
+\r
+\r
+ /* 20 bits for the fractional part */\r
+ /* yfract should be in 12.20 format */\r
+ yfract = (Y & 0x000FFFFF);\r
+\r
+ /* Read two nearest output values from the index */\r
+ y1 = pYData[(rI) + nCols * (cI + 1)];\r
+ y2 = pYData[(rI) + nCols * (cI + 1) + 1u];\r
+\r
+ /* Calculation of x1 * (1-xfract ) * (1-yfract) and acc is in 13.51 format */\r
+\r
+ /* x1 is in 1.15(q15), xfract in 12.20 format and out is in 13.35 format */\r
+ /* convert 13.35 to 13.31 by right shifting and out is in 1.31 */\r
+ out = (q31_t) (((q63_t) x1 * (0xFFFFF - xfract)) >> 4u);\r
+ acc = ((q63_t) out * (0xFFFFF - yfract));\r
+\r
+ /* x2 * (xfract) * (1-yfract) in 1.51 and adding to acc */\r
+ out = (q31_t) (((q63_t) x2 * (0xFFFFF - yfract)) >> 4u);\r
+ acc += ((q63_t) out * (xfract));\r
+\r
+ /* y1 * (1 - xfract) * (yfract) in 1.51 and adding to acc */\r
+ out = (q31_t) (((q63_t) y1 * (0xFFFFF - xfract)) >> 4u);\r
+ acc += ((q63_t) out * (yfract));\r
+\r
+ /* y2 * (xfract) * (yfract) in 1.51 and adding to acc */\r
+ out = (q31_t) (((q63_t) y2 * (xfract)) >> 4u);\r
+ acc += ((q63_t) out * (yfract));\r
+\r
+ /* acc is in 13.51 format and down shift acc by 36 times */\r
+ /* Convert out to 1.15 format */\r
+ return (acc >> 36);\r
+\r
+ }\r
+\r
+ /**\r
+ * @brief Q7 bilinear interpolation.\r
+ * @param[in,out] *S points to an instance of the interpolation structure.\r
+ * @param[in] X interpolation coordinate in 12.20 format.\r
+ * @param[in] Y interpolation coordinate in 12.20 format.\r
+ * @return out interpolated value.\r
+ */\r
+\r
+ __STATIC_INLINE q7_t arm_bilinear_interp_q7(\r
+ arm_bilinear_interp_instance_q7 * S,\r
+ q31_t X,\r
+ q31_t Y)\r
+ {\r
+ q63_t acc = 0; /* output */\r
+ q31_t out; /* Temporary output */\r
+ q31_t xfract, yfract; /* X, Y fractional parts */\r
+ q7_t x1, x2, y1, y2; /* Nearest output values */\r
+ int32_t rI, cI; /* Row and column indices */\r
+ q7_t *pYData = S->pData; /* pointer to output table values */\r
+ uint32_t nCols = S->numCols; /* num of rows */\r
+\r
+ /* Input is in 12.20 format */\r
+ /* 12 bits for the table index */\r
+ /* Index value calculation */\r
+ rI = ((X & 0xFFF00000) >> 20);\r
+\r
+ /* Input is in 12.20 format */\r
+ /* 12 bits for the table index */\r
+ /* Index value calculation */\r
+ cI = ((Y & 0xFFF00000) >> 20);\r
+\r
+ /* Care taken for table outside boundary */\r
+ /* Returns zero output when values are outside table boundary */\r
+ if(rI < 0 || rI > (S->numRows-1) || cI < 0 || cI > ( S->numCols-1))\r
+ {\r
+ return(0);\r
+ }\r
+\r
+ /* 20 bits for the fractional part */\r
+ /* xfract should be in 12.20 format */\r
+ xfract = (X & 0x000FFFFF);\r
+\r
+ /* Read two nearest output values from the index */\r
+ x1 = pYData[(rI) + nCols * (cI)];\r
+ x2 = pYData[(rI) + nCols * (cI) + 1u];\r
+\r
+\r
+ /* 20 bits for the fractional part */\r
+ /* yfract should be in 12.20 format */\r
+ yfract = (Y & 0x000FFFFF);\r
+\r
+ /* Read two nearest output values from the index */\r
+ y1 = pYData[(rI) + nCols * (cI + 1)];\r
+ y2 = pYData[(rI) + nCols * (cI + 1) + 1u];\r
+\r
+ /* Calculation of x1 * (1-xfract ) * (1-yfract) and acc is in 16.47 format */\r
+ out = ((x1 * (0xFFFFF - xfract)));\r
+ acc = (((q63_t) out * (0xFFFFF - yfract)));\r
+\r
+ /* x2 * (xfract) * (1-yfract) in 2.22 and adding to acc */\r
+ out = ((x2 * (0xFFFFF - yfract)));\r
+ acc += (((q63_t) out * (xfract)));\r
+\r
+ /* y1 * (1 - xfract) * (yfract) in 2.22 and adding to acc */\r
+ out = ((y1 * (0xFFFFF - xfract)));\r
+ acc += (((q63_t) out * (yfract)));\r
+\r
+ /* y2 * (xfract) * (yfract) in 2.22 and adding to acc */\r
+ out = ((y2 * (yfract)));\r
+ acc += (((q63_t) out * (xfract)));\r
+\r
+ /* acc in 16.47 format and down shift by 40 to convert to 1.7 format */\r
+ return (acc >> 40);\r
+\r
+ }\r
+\r
+ /**\r
+ * @} end of BilinearInterpolate group\r
+ */\r
+\r
+\r
+\r
+\r
+\r
+\r
+#ifdef __cplusplus\r
+}\r
+#endif\r
+\r
+\r
+#endif /* _ARM_MATH_H */\r
+\r
+\r
+/**\r
+ *\r
+ * End of file.\r
+ */\r
projects / freertos / blobdiff