You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
454 lines
18 KiB
454 lines
18 KiB
/* -----------------------------------------------------------------------------
|
|
Software License for The Fraunhofer FDK AAC Codec Library for Android
|
|
|
|
© Copyright 1995 - 2018 Fraunhofer-Gesellschaft zur Förderung der angewandten
|
|
Forschung e.V. All rights reserved.
|
|
|
|
1. INTRODUCTION
|
|
The Fraunhofer FDK AAC Codec Library for Android ("FDK AAC Codec") is software
|
|
that implements the MPEG Advanced Audio Coding ("AAC") encoding and decoding
|
|
scheme for digital audio. This FDK AAC Codec software is intended to be used on
|
|
a wide variety of Android devices.
|
|
|
|
AAC's HE-AAC and HE-AAC v2 versions are regarded as today's most efficient
|
|
general perceptual audio codecs. AAC-ELD is considered the best-performing
|
|
full-bandwidth communications codec by independent studies and is widely
|
|
deployed. AAC has been standardized by ISO and IEC as part of the MPEG
|
|
specifications.
|
|
|
|
Patent licenses for necessary patent claims for the FDK AAC Codec (including
|
|
those of Fraunhofer) may be obtained through Via Licensing
|
|
(www.vialicensing.com) or through the respective patent owners individually for
|
|
the purpose of encoding or decoding bit streams in products that are compliant
|
|
with the ISO/IEC MPEG audio standards. Please note that most manufacturers of
|
|
Android devices already license these patent claims through Via Licensing or
|
|
directly from the patent owners, and therefore FDK AAC Codec software may
|
|
already be covered under those patent licenses when it is used for those
|
|
licensed purposes only.
|
|
|
|
Commercially-licensed AAC software libraries, including floating-point versions
|
|
with enhanced sound quality, are also available from Fraunhofer. Users are
|
|
encouraged to check the Fraunhofer website for additional applications
|
|
information and documentation.
|
|
|
|
2. COPYRIGHT LICENSE
|
|
|
|
Redistribution and use in source and binary forms, with or without modification,
|
|
are permitted without payment of copyright license fees provided that you
|
|
satisfy the following conditions:
|
|
|
|
You must retain the complete text of this software license in redistributions of
|
|
the FDK AAC Codec or your modifications thereto in source code form.
|
|
|
|
You must retain the complete text of this software license in the documentation
|
|
and/or other materials provided with redistributions of the FDK AAC Codec or
|
|
your modifications thereto in binary form. You must make available free of
|
|
charge copies of the complete source code of the FDK AAC Codec and your
|
|
modifications thereto to recipients of copies in binary form.
|
|
|
|
The name of Fraunhofer may not be used to endorse or promote products derived
|
|
from this library without prior written permission.
|
|
|
|
You may not charge copyright license fees for anyone to use, copy or distribute
|
|
the FDK AAC Codec software or your modifications thereto.
|
|
|
|
Your modified versions of the FDK AAC Codec must carry prominent notices stating
|
|
that you changed the software and the date of any change. For modified versions
|
|
of the FDK AAC Codec, the term "Fraunhofer FDK AAC Codec Library for Android"
|
|
must be replaced by the term "Third-Party Modified Version of the Fraunhofer FDK
|
|
AAC Codec Library for Android."
|
|
|
|
3. NO PATENT LICENSE
|
|
|
|
NO EXPRESS OR IMPLIED LICENSES TO ANY PATENT CLAIMS, including without
|
|
limitation the patents of Fraunhofer, ARE GRANTED BY THIS SOFTWARE LICENSE.
|
|
Fraunhofer provides no warranty of patent non-infringement with respect to this
|
|
software.
|
|
|
|
You may use this FDK AAC Codec software or modifications thereto only for
|
|
purposes that are authorized by appropriate patent licenses.
|
|
|
|
4. DISCLAIMER
|
|
|
|
This FDK AAC Codec software is provided by Fraunhofer on behalf of the copyright
|
|
holders and contributors "AS IS" and WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES,
|
|
including but not limited to the implied warranties of merchantability and
|
|
fitness for a particular purpose. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
|
|
CONTRIBUTORS BE LIABLE for any direct, indirect, incidental, special, exemplary,
|
|
or consequential damages, including but not limited to procurement of substitute
|
|
goods or services; loss of use, data, or profits, or business interruption,
|
|
however caused and on any theory of liability, whether in contract, strict
|
|
liability, or tort (including negligence), arising in any way out of the use of
|
|
this software, even if advised of the possibility of such damage.
|
|
|
|
5. CONTACT INFORMATION
|
|
|
|
Fraunhofer Institute for Integrated Circuits IIS
|
|
Attention: Audio and Multimedia Departments - FDK AAC LL
|
|
Am Wolfsmantel 33
|
|
91058 Erlangen, Germany
|
|
|
|
www.iis.fraunhofer.de/amm
|
|
amm-info@iis.fraunhofer.de
|
|
----------------------------------------------------------------------------- */
|
|
|
|
/******************* Library for basic calculation routines ********************
|
|
|
|
Author(s): M. Lohwasser, M. Gayer
|
|
|
|
Description: Flexible fixpoint library configuration
|
|
|
|
*******************************************************************************/
|
|
|
|
#ifndef COMMON_FIX_H
|
|
#define COMMON_FIX_H
|
|
|
|
#include "FDK_archdef.h"
|
|
#include "machine_type.h"
|
|
|
|
/* ***** Start of former fix.h ****** */
|
|
|
|
/* Define bit sizes of integer fixpoint fractional data types */
|
|
#define FRACT_BITS 16 /* single precision */
|
|
#define DFRACT_BITS 32 /* double precision */
|
|
#define ACCU_BITS 40 /* double precision plus overflow */
|
|
|
|
/* Fixpoint equivalent type fot PCM audio time domain data. */
|
|
#if defined(SAMPLE_BITS)
|
|
#if (SAMPLE_BITS == DFRACT_BITS)
|
|
#define FIXP_PCM FIXP_DBL
|
|
#define MAXVAL_FIXP_PCM MAXVAL_DBL
|
|
#define MINVAL_FIXP_PCM MINVAL_DBL
|
|
#define FX_PCM2FX_DBL(x) ((FIXP_DBL)(x))
|
|
#define FX_DBL2FX_PCM(x) ((INT_PCM)(x))
|
|
#elif (SAMPLE_BITS == FRACT_BITS)
|
|
#define FIXP_PCM FIXP_SGL
|
|
#define MAXVAL_FIXP_PCM MAXVAL_SGL
|
|
#define MINVAL_FIXP_PCM MINVAL_SGL
|
|
#define FX_PCM2FX_DBL(x) FX_SGL2FX_DBL((FIXP_SGL)(x))
|
|
#define FX_DBL2FX_PCM(x) FX_DBL2FX_SGL(x)
|
|
#else
|
|
#error SAMPLE_BITS different from FRACT_BITS or DFRACT_BITS not implemented!
|
|
#endif
|
|
#endif
|
|
|
|
/* ****** End of former fix.h ****** */
|
|
|
|
#define SGL_MASK ((1UL << FRACT_BITS) - 1) /* 16bit: (2^16)-1 = 0xFFFF */
|
|
|
|
#define MAX_SHIFT_SGL \
|
|
(FRACT_BITS - 1) /* maximum possible shift for FIXP_SGL values */
|
|
#define MAX_SHIFT_DBL \
|
|
(DFRACT_BITS - 1) /* maximum possible shift for FIXP_DBL values */
|
|
|
|
/* Scale factor from/to float/fixpoint values. DO NOT USE THESE VALUES AS
|
|
* SATURATION LIMITS !! */
|
|
#define FRACT_FIX_SCALE ((INT64(1) << (FRACT_BITS - 1)))
|
|
#define DFRACT_FIX_SCALE ((INT64(1) << (DFRACT_BITS - 1)))
|
|
|
|
/* Max and Min values for saturation purposes. DO NOT USE THESE VALUES AS SCALE
|
|
* VALUES !! */
|
|
#define MAXVAL_SGL \
|
|
((signed)0x00007FFF) /* this has to be synchronized to FRACT_BITS */
|
|
#define MINVAL_SGL \
|
|
((signed)0xFFFF8000) /* this has to be synchronized to FRACT_BITS */
|
|
#define MAXVAL_DBL \
|
|
((signed)0x7FFFFFFF) /* this has to be synchronized to DFRACT_BITS */
|
|
#define MINVAL_DBL \
|
|
((signed)0x80000000) /* this has to be synchronized to DFRACT_BITS */
|
|
|
|
#define FX_DBL2FXCONST_SGL(val) \
|
|
((((((val) >> (DFRACT_BITS - FRACT_BITS - 1)) + 1) > \
|
|
(((LONG)1 << FRACT_BITS) - 1)) && \
|
|
((LONG)(val) > 0)) \
|
|
? (FIXP_SGL)(SHORT)(((LONG)1 << (FRACT_BITS - 1)) - 1) \
|
|
: (FIXP_SGL)(SHORT)((((val) >> (DFRACT_BITS - FRACT_BITS - 1)) + 1) >> \
|
|
1))
|
|
|
|
#define shouldBeUnion union /* unions are possible */
|
|
|
|
typedef SHORT FIXP_SGL;
|
|
typedef LONG FIXP_DBL;
|
|
|
|
/* macros for compile-time conversion of constant float values to fixedpoint */
|
|
#define FL2FXCONST_SPC FL2FXCONST_DBL
|
|
|
|
#define MINVAL_DBL_CONST MINVAL_DBL
|
|
#define MINVAL_SGL_CONST MINVAL_SGL
|
|
|
|
#define FL2FXCONST_SGL(val) \
|
|
(FIXP_SGL)( \
|
|
((val) >= 0) \
|
|
? ((((double)(val) * (FRACT_FIX_SCALE) + 0.5) >= \
|
|
(double)(MAXVAL_SGL)) \
|
|
? (SHORT)(MAXVAL_SGL) \
|
|
: (SHORT)((double)(val) * (double)(FRACT_FIX_SCALE) + 0.5)) \
|
|
: ((((double)(val) * (FRACT_FIX_SCALE)-0.5) <= \
|
|
(double)(MINVAL_SGL_CONST)) \
|
|
? (SHORT)(MINVAL_SGL_CONST) \
|
|
: (SHORT)((double)(val) * (double)(FRACT_FIX_SCALE)-0.5)))
|
|
|
|
#define FL2FXCONST_DBL(val) \
|
|
(FIXP_DBL)( \
|
|
((val) >= 0) \
|
|
? ((((double)(val) * (DFRACT_FIX_SCALE) + 0.5) >= \
|
|
(double)(MAXVAL_DBL)) \
|
|
? (LONG)(MAXVAL_DBL) \
|
|
: (LONG)((double)(val) * (double)(DFRACT_FIX_SCALE) + 0.5)) \
|
|
: ((((double)(val) * (DFRACT_FIX_SCALE)-0.5) <= \
|
|
(double)(MINVAL_DBL_CONST)) \
|
|
? (LONG)(MINVAL_DBL_CONST) \
|
|
: (LONG)((double)(val) * (double)(DFRACT_FIX_SCALE)-0.5)))
|
|
|
|
/* macros for runtime conversion of float values to integer fixedpoint. NO
|
|
* OVERFLOW CHECK!!! */
|
|
#define FL2FX_SPC FL2FX_DBL
|
|
#define FL2FX_SGL(val) \
|
|
((val) > 0.0f ? (SHORT)((val) * (float)(FRACT_FIX_SCALE) + 0.5f) \
|
|
: (SHORT)((val) * (float)(FRACT_FIX_SCALE)-0.5f))
|
|
#define FL2FX_DBL(val) \
|
|
((val) > 0.0f ? (LONG)((val) * (float)(DFRACT_FIX_SCALE) + 0.5f) \
|
|
: (LONG)((val) * (float)(DFRACT_FIX_SCALE)-0.5f))
|
|
|
|
/* macros for runtime conversion of fixedpoint values to other fixedpoint. NO
|
|
* ROUNDING!!! */
|
|
#define FX_ACC2FX_SGL(val) ((FIXP_SGL)((val) >> (ACCU_BITS - FRACT_BITS)))
|
|
#define FX_ACC2FX_DBL(val) ((FIXP_DBL)((val) >> (ACCU_BITS - DFRACT_BITS)))
|
|
#define FX_SGL2FX_ACC(val) ((FIXP_ACC)((LONG)(val) << (ACCU_BITS - FRACT_BITS)))
|
|
#define FX_SGL2FX_DBL(val) \
|
|
((FIXP_DBL)((LONG)(val) << (DFRACT_BITS - FRACT_BITS)))
|
|
#define FX_DBL2FX_SGL(val) ((FIXP_SGL)((val) >> (DFRACT_BITS - FRACT_BITS)))
|
|
|
|
/* ############################################################# */
|
|
|
|
/* macros for runtime conversion of integer fixedpoint values to float. */
|
|
|
|
/* #define FX_DBL2FL(val) ((float)(pow(2.,-31.)*(float)val)) */ /* version #1
|
|
*/
|
|
#define FX_DBL2FL(val) \
|
|
((float)((double)(val) / (double)DFRACT_FIX_SCALE)) /* version #2 - \
|
|
identical to class \
|
|
dfract cast from \
|
|
dfract to float */
|
|
#define FX_DBL2DOUBLE(val) (((double)(val) / (double)DFRACT_FIX_SCALE))
|
|
|
|
/* ############################################################# */
|
|
#include "fixmul.h"
|
|
|
|
FDK_INLINE LONG fMult(SHORT a, SHORT b) { return fixmul_SS(a, b); }
|
|
FDK_INLINE LONG fMult(SHORT a, LONG b) { return fixmul_SD(a, b); }
|
|
FDK_INLINE LONG fMult(LONG a, SHORT b) { return fixmul_DS(a, b); }
|
|
FDK_INLINE LONG fMult(LONG a, LONG b) { return fixmul_DD(a, b); }
|
|
FDK_INLINE LONG fPow2(LONG a) { return fixpow2_D(a); }
|
|
FDK_INLINE LONG fPow2(SHORT a) { return fixpow2_S(a); }
|
|
|
|
FDK_INLINE LONG fMultDiv2(SHORT a, SHORT b) { return fixmuldiv2_SS(a, b); }
|
|
FDK_INLINE LONG fMultDiv2(SHORT a, LONG b) { return fixmuldiv2_SD(a, b); }
|
|
FDK_INLINE LONG fMultDiv2(LONG a, SHORT b) { return fixmuldiv2_DS(a, b); }
|
|
FDK_INLINE LONG fMultDiv2(LONG a, LONG b) { return fixmuldiv2_DD(a, b); }
|
|
FDK_INLINE LONG fPow2Div2(LONG a) { return fixpow2div2_D(a); }
|
|
FDK_INLINE LONG fPow2Div2(SHORT a) { return fixpow2div2_S(a); }
|
|
|
|
FDK_INLINE LONG fMultDiv2BitExact(LONG a, LONG b) {
|
|
return fixmuldiv2BitExact_DD(a, b);
|
|
}
|
|
FDK_INLINE LONG fMultDiv2BitExact(SHORT a, LONG b) {
|
|
return fixmuldiv2BitExact_SD(a, b);
|
|
}
|
|
FDK_INLINE LONG fMultDiv2BitExact(LONG a, SHORT b) {
|
|
return fixmuldiv2BitExact_DS(a, b);
|
|
}
|
|
FDK_INLINE LONG fMultBitExact(LONG a, LONG b) {
|
|
return fixmulBitExact_DD(a, b);
|
|
}
|
|
FDK_INLINE LONG fMultBitExact(SHORT a, LONG b) {
|
|
return fixmulBitExact_SD(a, b);
|
|
}
|
|
FDK_INLINE LONG fMultBitExact(LONG a, SHORT b) {
|
|
return fixmulBitExact_DS(a, b);
|
|
}
|
|
|
|
/* ********************************************************************************
|
|
*/
|
|
#include "abs.h"
|
|
|
|
FDK_INLINE FIXP_DBL fAbs(FIXP_DBL x) { return fixabs_D(x); }
|
|
FDK_INLINE FIXP_SGL fAbs(FIXP_SGL x) { return fixabs_S(x); }
|
|
|
|
#if !defined(__LP64__)
|
|
FDK_INLINE INT fAbs(INT x) { return fixabs_I(x); }
|
|
#endif
|
|
|
|
/* ********************************************************************************
|
|
*/
|
|
|
|
#include "clz.h"
|
|
|
|
FDK_INLINE INT fNormz(INT64 x) {
|
|
INT clz = fixnormz_D((INT)(x >> 32));
|
|
if (clz == 32) clz += fixnormz_D((INT)x);
|
|
return clz;
|
|
}
|
|
FDK_INLINE INT fNormz(FIXP_DBL x) { return fixnormz_D(x); }
|
|
FDK_INLINE INT fNormz(FIXP_SGL x) { return fixnormz_S(x); }
|
|
FDK_INLINE INT fNorm(FIXP_DBL x) { return fixnorm_D(x); }
|
|
FDK_INLINE INT fNorm(FIXP_SGL x) { return fixnorm_S(x); }
|
|
|
|
/* ********************************************************************************
|
|
*/
|
|
/* ********************************************************************************
|
|
*/
|
|
/* ********************************************************************************
|
|
*/
|
|
|
|
#include "clz.h"
|
|
#define fixp_abs(x) fAbs(x)
|
|
#define fixMin(a, b) fMin(a, b)
|
|
#define fixMax(a, b) fMax(a, b)
|
|
#define CntLeadingZeros(x) fixnormz_D(x)
|
|
#define CountLeadingBits(x) fixnorm_D(x)
|
|
|
|
#include "fixmadd.h"
|
|
|
|
/* y = (x+0.5*a*b) */
|
|
FDK_INLINE FIXP_DBL fMultAddDiv2(FIXP_DBL x, FIXP_DBL a, FIXP_DBL b) {
|
|
return fixmadddiv2_DD(x, a, b);
|
|
}
|
|
FDK_INLINE FIXP_DBL fMultAddDiv2(FIXP_DBL x, FIXP_SGL a, FIXP_DBL b) {
|
|
return fixmadddiv2_SD(x, a, b);
|
|
}
|
|
FDK_INLINE FIXP_DBL fMultAddDiv2(FIXP_DBL x, FIXP_DBL a, FIXP_SGL b) {
|
|
return fixmadddiv2_DS(x, a, b);
|
|
}
|
|
FDK_INLINE FIXP_DBL fMultAddDiv2(FIXP_DBL x, FIXP_SGL a, FIXP_SGL b) {
|
|
return fixmadddiv2_SS(x, a, b);
|
|
}
|
|
|
|
FDK_INLINE FIXP_DBL fPow2AddDiv2(FIXP_DBL x, FIXP_DBL a) {
|
|
return fixpadddiv2_D(x, a);
|
|
}
|
|
FDK_INLINE FIXP_DBL fPow2AddDiv2(FIXP_DBL x, FIXP_SGL a) {
|
|
return fixpadddiv2_S(x, a);
|
|
}
|
|
|
|
/* y = 2*(x+0.5*a*b) = (2x+a*b) */
|
|
FDK_INLINE FIXP_DBL fMultAdd(FIXP_DBL x, FIXP_DBL a, FIXP_DBL b) {
|
|
return fixmadd_DD(x, a, b);
|
|
}
|
|
inline FIXP_DBL fMultAdd(FIXP_DBL x, FIXP_SGL a, FIXP_DBL b) {
|
|
return fixmadd_SD(x, a, b);
|
|
}
|
|
inline FIXP_DBL fMultAdd(FIXP_DBL x, FIXP_DBL a, FIXP_SGL b) {
|
|
return fixmadd_DS(x, a, b);
|
|
}
|
|
inline FIXP_DBL fMultAdd(FIXP_DBL x, FIXP_SGL a, FIXP_SGL b) {
|
|
return fixmadd_SS(x, a, b);
|
|
}
|
|
|
|
inline FIXP_DBL fPow2Add(FIXP_DBL x, FIXP_DBL a) { return fixpadd_D(x, a); }
|
|
inline FIXP_DBL fPow2Add(FIXP_DBL x, FIXP_SGL a) { return fixpadd_S(x, a); }
|
|
|
|
/* y = (x-0.5*a*b) */
|
|
inline FIXP_DBL fMultSubDiv2(FIXP_DBL x, FIXP_DBL a, FIXP_DBL b) {
|
|
return fixmsubdiv2_DD(x, a, b);
|
|
}
|
|
inline FIXP_DBL fMultSubDiv2(FIXP_DBL x, FIXP_SGL a, FIXP_DBL b) {
|
|
return fixmsubdiv2_SD(x, a, b);
|
|
}
|
|
inline FIXP_DBL fMultSubDiv2(FIXP_DBL x, FIXP_DBL a, FIXP_SGL b) {
|
|
return fixmsubdiv2_DS(x, a, b);
|
|
}
|
|
inline FIXP_DBL fMultSubDiv2(FIXP_DBL x, FIXP_SGL a, FIXP_SGL b) {
|
|
return fixmsubdiv2_SS(x, a, b);
|
|
}
|
|
|
|
/* y = 2*(x-0.5*a*b) = (2*x-a*b) */
|
|
FDK_INLINE FIXP_DBL fMultSub(FIXP_DBL x, FIXP_DBL a, FIXP_DBL b) {
|
|
return fixmsub_DD(x, a, b);
|
|
}
|
|
inline FIXP_DBL fMultSub(FIXP_DBL x, FIXP_SGL a, FIXP_DBL b) {
|
|
return fixmsub_SD(x, a, b);
|
|
}
|
|
inline FIXP_DBL fMultSub(FIXP_DBL x, FIXP_DBL a, FIXP_SGL b) {
|
|
return fixmsub_DS(x, a, b);
|
|
}
|
|
inline FIXP_DBL fMultSub(FIXP_DBL x, FIXP_SGL a, FIXP_SGL b) {
|
|
return fixmsub_SS(x, a, b);
|
|
}
|
|
|
|
FDK_INLINE FIXP_DBL fMultAddDiv2BitExact(FIXP_DBL x, FIXP_DBL a, FIXP_DBL b) {
|
|
return fixmadddiv2BitExact_DD(x, a, b);
|
|
}
|
|
FDK_INLINE FIXP_DBL fMultAddDiv2BitExact(FIXP_DBL x, FIXP_SGL a, FIXP_DBL b) {
|
|
return fixmadddiv2BitExact_SD(x, a, b);
|
|
}
|
|
FDK_INLINE FIXP_DBL fMultAddDiv2BitExact(FIXP_DBL x, FIXP_DBL a, FIXP_SGL b) {
|
|
return fixmadddiv2BitExact_DS(x, a, b);
|
|
}
|
|
FDK_INLINE FIXP_DBL fMultSubDiv2BitExact(FIXP_DBL x, FIXP_DBL a, FIXP_DBL b) {
|
|
return fixmsubdiv2BitExact_DD(x, a, b);
|
|
}
|
|
FDK_INLINE FIXP_DBL fMultSubDiv2BitExact(FIXP_DBL x, FIXP_SGL a, FIXP_DBL b) {
|
|
return fixmsubdiv2BitExact_SD(x, a, b);
|
|
}
|
|
FDK_INLINE FIXP_DBL fMultSubDiv2BitExact(FIXP_DBL x, FIXP_DBL a, FIXP_SGL b) {
|
|
return fixmsubdiv2BitExact_DS(x, a, b);
|
|
}
|
|
|
|
#include "fixminmax.h"
|
|
|
|
FDK_INLINE FIXP_DBL fMin(FIXP_DBL a, FIXP_DBL b) { return fixmin_D(a, b); }
|
|
FDK_INLINE FIXP_DBL fMax(FIXP_DBL a, FIXP_DBL b) { return fixmax_D(a, b); }
|
|
|
|
FDK_INLINE FIXP_SGL fMin(FIXP_SGL a, FIXP_SGL b) { return fixmin_S(a, b); }
|
|
FDK_INLINE FIXP_SGL fMax(FIXP_SGL a, FIXP_SGL b) { return fixmax_S(a, b); }
|
|
|
|
#if !defined(__LP64__)
|
|
FDK_INLINE INT fMax(INT a, INT b) { return fixmax_I(a, b); }
|
|
FDK_INLINE INT fMin(INT a, INT b) { return fixmin_I(a, b); }
|
|
#if !defined(_MSC_VER) && defined(__x86_64__)
|
|
FDK_INLINE SHORT fMax(SHORT a, SHORT b) { return fixmax_S(a, b); }
|
|
FDK_INLINE SHORT fMin(SHORT a, SHORT b) { return fixmin_S(a, b); }
|
|
#endif
|
|
#endif
|
|
|
|
inline UINT fMax(UINT a, UINT b) { return fixmax_UI(a, b); }
|
|
inline UINT fMin(UINT a, UINT b) { return fixmin_UI(a, b); }
|
|
|
|
inline UCHAR fMax(UCHAR a, UCHAR b) {
|
|
return (UCHAR)fixmax_UI((UINT)a, (UINT)b);
|
|
}
|
|
inline UCHAR fMin(UCHAR a, UCHAR b) {
|
|
return (UCHAR)fixmin_UI((UINT)a, (UINT)b);
|
|
}
|
|
|
|
/* Complex data types */
|
|
typedef shouldBeUnion {
|
|
/* vector representation for arithmetic */
|
|
struct {
|
|
FIXP_SGL re;
|
|
FIXP_SGL im;
|
|
} v;
|
|
/* word representation for memory move */
|
|
LONG w;
|
|
}
|
|
FIXP_SPK;
|
|
|
|
typedef shouldBeUnion {
|
|
/* vector representation for arithmetic */
|
|
struct {
|
|
FIXP_DBL re;
|
|
FIXP_DBL im;
|
|
} v;
|
|
/* word representation for memory move */
|
|
INT64 w;
|
|
}
|
|
FIXP_DPK;
|
|
|
|
#include "fixmul.h"
|
|
#include "fixmadd.h"
|
|
#include "cplx_mul.h"
|
|
#include "fixpoint_math.h"
|
|
|
|
#endif
|