Math Library
User Reference Manual
56800E, 56800Ex
Digital Signal Controller
56800Ex_MLIB
Rev. 0
02/2014
freescale.com
Chapter 1 License Agreement
FREESCALE SEMICONDUCTOR SOFTWARE LICENSE AGREEMENT.
This is a legal agreement between you (either as an individual or as an authorized
representative of your employer) and Freescale Semiconductor, Inc.
("Freescale"). It concerns your rights to use this file and any accompanying
written materials (the "Software"). In consideration for Freescale allowing you to
access the Software, you are agreeing to be bound by the terms of this Agreement.
If you do not agree to all of the terms of this Agreement, do not download the
Software. If you change your mind later, stop using the Software and delete all
copies of the Software in your possession or control. Any copies of the Software
that you have already distributed, where permitted, and do not destroy will
continue to be governed by this Agreement. Your prior use will also continue to
be governed by this Agreement.
OBJECT PROVIDED, OBJECT REDISTRIBUTION LICENSE GRANT.
Freescale grants to you, free of charge, the non-exclusive, non-transferable right
(1) to reproduce the Software, (2) to distribute the Software, and (3) to sublicense
to others the right to use the distributed Software. The Software is provided to you
only in object (machine-readable) form. You may exercise the rights above only
with respect to such object form. You may not translate, reverse engineer,
decompile, or disassemble the Software except to the extent applicable law
specifically prohibits such restriction. In addition, you must prohibit your
sublicensees from doing the same. If you violate any of the terms or restrictions
of this Agreement, Freescale may immediately terminate this Agreement, and
require that you stop using and delete all copies of the Software in your
possession or control.
COPYRIGHT. The Software is licensed to you, not sold. Freescale owns the
Software, and United States copyright laws and international treaty provisions
protect the Software. Therefore, you must treat the Software like any other
copyrighted material (e.g. a book or musical recording). You may not use or copy
the Software for any other purpose than what is described in this Agreement.
Except as expressly provided herein, Freescale does not grant to you any express
or implied rights under any Freescale or third-party patents, copyrights,
trademarks, or trade secrets. Additionally, you must reproduce and apply any
copyright or other proprietary rights notices included on or embedded in the
Software to any copies or derivative works made thereof, in whole or in part, if
any.
SUPPORT. Freescale is NOT obligated to provide any support, upgrades or new
releases of the Software. If you wish, you may contact Freescale and report
problems and provide suggestions regarding the Software. Freescale has no
obligation whatsoever to respond in any way to such a problem report or
suggestion. Freescale may make changes to the Software at any time, without any
obligation to notify or provide updated versions of the Software to you.
Math Library, Rev. 0
1-2
Freescale Semiconductor
NO WARRANTY. TO THE MAXIMUM EXTENT PERMITTED BY LAW,
FREESCALE EXPRESSLY DISCLAIMS ANY WARRANTY FOR THE
SOFTWARE. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT
WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED,
INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR
NON-INFRINGEMENT. YOU ASSUME THE ENTIRE RISK ARISING
OUT OF THE USE OR PERFORMANCE OF THE SOFTWARE, OR ANY
SYSTEMS YOU DESIGN USING THE SOFTWARE (IF ANY). NOTHING
IN THIS AGREEMENT MAY BE CONSTRUED AS A WARRANTY OR
REPRESENTATION BY FREESCALE THAT THE SOFTWARE OR ANY
DERIVATIVE WORK DEVELOPED WITH OR INCORPORATING THE
SOFTWARE WILL BE FREE FROM INFRINGEMENT OF THE
INTELLECTUAL PROPERTY RIGHTS OF THIRD PARTIES.
INDEMNITY. You agree to fully defend and indemnify Freescale from any and
all claims, liabilities, and costs (including reasonable attorney's fees) related to
(1) your use (including your sublicensee's use, if permitted) of the Software or
(2) your violation of the terms and conditions of this Agreement.
LIMITATION OF LIABILITY. IN NO EVENT WILL FREESCALE BE
LIABLE, WHETHER IN CONTRACT, TORT, OR OTHERWISE, FOR ANY
INCIDENTAL, SPECIAL, INDIRECT, CONSEQUENTIAL OR PUNITIVE
DAMAGES, INCLUDING, BUT NOT LIMITED TO, DAMAGES FOR ANY
LOSS OF USE, LOSS OF TIME, INCONVENIENCE, COMMERCIAL
LOSS, OR LOST PROFITS, SAVINGS, OR REVENUES TO THE FULL
EXTENT SUCH MAY BE DISCLAIMED BY LAW.
COMPLIANCE WITH LAWS; EXPORT RESTRICTIONS. You must use the
Software in accordance with all applicable U.S. laws, regulations and statutes.
You agree that neither you nor your licensees (if any) intend to or will, directly
or indirectly, export or transmit the Software to any country in violation of U.S.
export restrictions.
GOVERNMENT USE. Use of the Software and any corresponding
documentation, if any, is provided with RESTRICTED RIGHTS. Use,
duplication or disclosure by the Government is subject to restrictions as set forth
in subparagraph (c)(1)(ii) of The Rights in Technical Data and Computer
Software clause at DFARS 252.227-7013 or subparagraphs (c)(l) and (2) of the
Commercial Computer Software--Restricted Rights at 48 CFR 52.227-19, as
applicable. Manufacturer is Freescale Semiconductor, Inc., 6501 William
Cannon Drive West, Austin, TX, 78735.
HIGH RISK ACTIVITIES. You acknowledge that the Software is not fault
tolerant and is not designed, manufactured or intended by Freescale for
incorporation into products intended for use or resale in on-line control
Math Library, Rev. 0
Freescale Semiconductor
1-3
equipment in hazardous, dangerous to life or potentially life-threatening
environments requiring fail-safe performance, such as in the operation of nuclear
facilities, aircraft navigation or communication systems, air traffic control, direct
life support machines or weapons systems, in which the failure of products could
lead directly to death, personal injury or severe physical or environmental
damage ("High Risk Activities"). You specifically represent and warrant that you
will not use the Software or any derivative work of the Software for High Risk
Activities.
CHOICE OF LAW; VENUE; LIMITATIONS. You agree that the statutes and
laws of the United States and the State of Texas, USA, without regard to conflicts
of laws principles, will apply to all matters relating to this Agreement or the
Software, and you agree that any litigation will be subject to the exclusive
jurisdiction of the state or federal courts in Texas, USA. You agree that regardless
of any statute or law to the contrary, any claim or cause of action arising out of or
related to this Agreement or the Software must be filed within one (1) year after
such claim or cause of action arose or be forever barred.
PRODUCT LABELING. You are not authorized to use any Freescale trademarks,
brand names, or logos.
ENTIRE AGREEMENT. This Agreement constitutes the entire agreement
between you and Freescale regarding the subject matter of this Agreement, and
supersedes all prior communications, negotiations, understandings, agreements
or representations, either written or oral, if any. This Agreement may only be
amended in written form, executed by you and Freescale.
SEVERABILITY. If any provision of this Agreement is held for any reason to be
invalid or unenforceable, then the remaining provisions of this Agreement will be
unimpaired and, unless a modification or replacement of the invalid or
unenforceable provision is further held to deprive you or Freescale of a material
benefit, in which case the Agreement will immediately terminate, the invalid or
unenforceable provision will be replaced with a provision that is valid and
enforceable and that comes closest to the intention underlying the invalid or
unenforceable provision.
NO WAIVER. The waiver by Freescale of any breach of any provision of this
Agreement will not operate or be construed as a waiver of any other or a
subsequent breach of the same or a different provision.
Math Library, Rev. 0
1-4
Freescale Semiconductor
Overview
Chapter 2 INTRODUCTION
2.1
Overview
This reference manual describes the Math Library (MLIB) for the Freescale
56F800E(X) family of Digital Signal Controllers. This library contains optimized
functions.
2.2
Supported Compilers
Math Library (MLIB) is written in assembly language with C-callable interface.
The library was built and tested using the CodeWarrior™ Development Studio
version 10.3.
The library is delivered in library module 56800Ex_MLIB.lib and is intended for
use in small data memory model projects. The interfaces to the algorithms
included in this library have been combined into a single public interface include
file, gdflib.h. This was done to simplify the number of files required for inclusion
by application programs. Refer to the specific algorithm sections of this
document for details on the software Application Programming Interface (API),
defined and functionality provided for the individual algorithms.
2.3
Installation
If user wants to fully use this library, the CodeWarrior™ Development Studio
should be installed prior to the Math Library. In case that Math Library is
installed while CodeWarrior™ Development Studio is not present, users can only
browse the installed software package, but will not be able to build, download and
run code. The installation itself consists of copying the required files to the
destination hard drive, checking the presence of CodeWarrior and creating the
shortcut under the Start->Programs menu.
The Math Library release is installed in its own folder named 56800Ex_MLIB.
To start the installation process, perform the following steps:
1. Execute 56800Ex_FSLESL_rXX.exe.
2. Follow the FSLESL software installation instructions on your screen.
2.4
Library Integration
The library integration is described in AN4586 which can be downloaded
from www.freescale.com.
Math Library, Rev. 0
Freescale Semiconductor
2-5
2.5
API Definition
The description of each function described in this Math Library user reference
manual consists of a number of subsections:
Synopsis
This subsection gives the header files that should be included within
a source file that references the function or macro. It also shows an
appropriate declaration for the function or for a function that can be
substituted by a macro. This declaration is not included in your
program; only the header file(s) should be included.
Prototype
This subsection shows the original function prototype declaration
with all its arguments.
Arguments
This optional subsection describes input arguments to a function or
macro.
Description
This subsection is a description of the function or macro. It explains
algorithms being used by functions or macros.
Return
This optional subsection describes the return value (if any) of the
function or macro.
Range Issues
This optional subsection specifies the ranges of input variables.
Special Issues
This optional subsection specifies special assumptions that are
mandatory for correct function calculation; for example saturation,
rounding, and so on.
Implementation
This optional subsection specifies, whether a call of the function
generates a library function call or a macro expansion.
This subsection also consists of one or more examples of the use of
the function. The examples are often fragments of code (not
completed programs) for illustration purposes.
See Also
This optional subsection provides a list of related functions or
macros.
Performance
Math Library, Rev. 0
Freescale Semiconductor
2-6
Data Types
This section specifies the actual requirements of the function or macro
in terms of required code memory, data memory, and number of clock
cycles to execute.
2.6
Data Types
The 16-bit DSC core supports four types of two’s-complement data formats:
• Signed integer
• Unsigned integer
• Signed fractional
• Unsigned fractional
Signed and unsigned integer data types are useful for general-purpose
computation; they are familiar with the microprocessor and microcontroller
programmers. Fractional data types allow powerful numeric and
digital-signal-processing algorithms to be implemented.
2.6.1
Signed Integer (SI)
This format is used for processing data as integers. In this format, the N-bit
operand is represented using the N.0 format (N integer bits). The signed integer
numbers lie in the following range:
– 2 N – 1 SI 2 N – 1 – 1
Eqn. 2-1
This data format is available for bytes, words, and longs. The most negative,
signed word that can be represented is –32,768 ($8000), and the most negative,
signed long word is –2,147,483,648 ($80000000).
The most positive, signed word is 32,767 ($7FFF), and the most positive signed
long word is 2,147,483,647 ($7FFFFFFF).
2.6.2
Unsigned Integer (UI)
The unsigned integer numbers are positive only, and they have nearly twice the
magnitude of a signed number of the same size. The unsigned integer numbers lie
in the following range:
0 UI 2 N – 1 – 1
Eqn. 2-2
The binary word is interpreted as having a binary point immediately to the right
of the integer’s least significant bit. This data format is available for bytes, words,
and long words. The most positive, 16-bit, unsigned integer is 65,535 ($FFFF),
and the most positive, 32-bit, unsigned integer is 4,294,967,295 ($FFFFFFFF).
The smallest unsigned integer number is zero ($0000), regardless of size.
Math Library, Rev. 0
Freescale Semiconductor
2-7
User Common Types
2.6.3
Signed Fractional (SF)
In this format, the N-bit operand is represented using the 1.[N–1] format (one sign
bit, N–1 fractional bits). The signed fractional numbers lie in the following range:
– 1.0 SF 1.0 – 2 – N – 1
Eqn. 2-3
This data format is available for words and long words. For both word and
long-word signed fractions, the most negative number that can be represented
is –1.0; its internal representation is $8000 (word) or $80000000 (long word).
The most positive word is $7FFF (1.0 – 2–15); its most positive long word
is $7FFFFFFF (1.0 – 2–31).
2.6.4
Unsigned Fractional (UF)
The unsigned fractional numbers can be positive only, and they have nearly twice
the magnitude of a signed number with the same number of bits. The unsigned
fractional numbers lie in the following range:
0.0 UF 2.0 – 2 – N – 1
Eqn. 2-4
The binary word is interpreted as having a binary point after the MSB. This data
format is available for words and longs. The most positive, 16-bit, unsigned
number is $FFFF, or {1.0 + (1.0 – 2–[N–1])} = 1.99997. The smallest unsigned
fractional number is zero ($0000).
2.7
User Common Types
Table 2-1. User-Defined Typedefs in 56800E_types.h
Mnemonics
Size — bits
Description
Word8
8
To represent 8-bit signed variable/value.
UWord8
8
To represent 16-bit unsigned variable/value.
Word16
16
To represent 16-bit signed variable/value.
UWord16
16
To represent 16-bit unsigned variable/value.
Word32
32
To represent 32-bit signed variable/value.
UWord32
32
To represent 16-bit unsigned variable/value.
Int8
8
To represent 8-bit signed variable/value.
UInt8
8
To represent 16-bit unsigned variable/value.
Int16
16
To represent 16-bit signed variable/value.
UInt16
16
To represent 16-bit unsigned variable/value.
Int32
32
To represent 32-bit signed variable/value.
Math Library, Rev. 0
2-8
Freescale Semiconductor
V2 and V3 Core Support
Table 2-1. User-Defined Typedefs in 56800E_types.h (continued)
2.8
UInt32
32
To represent 16-bit unsigned variable/value.
Frac16
16
To represent 16-bit signed variable/value.
Frac32
32
To represent 32-bit signed variable/value.
NULL
constant
Represents NULL pointer.
bool
16
Boolean variable.
false
constant
Represents false value.
true
constant
Represents true value.
FRAC16()
macro
Transforms float value from <–1, 1) range into fractional
representation <–32768, 32767>.
FRAC32()
macro
Transforms float value from <–1, 1) range into fractional
representation <–2147483648, 2147483648>.
V2 and V3 Core Support
The library has been written to support both 56800E (V2) and 56800Ex (V3)
cores. The V3 core offers new set of math instructions which can simplify and
accelarete the algorithm runtime. Therefore certain algorithms can have two
prototypes.
2
If the library is used on the 56800Ex core, the V3 algorithms use is recommended
because:
• the code is shorter
• the execution is faster
• the precision of 32-bit calculation is higher
The final algorithm is selected by a define. To select the correct algorithm
implementation the user has to set up a define: OPTION_CORE_V3. If this
define is not defined, it is automatically set up as 0. If its value is 0, the V2
algorithms are used. If its value is 1, the V3 algorithms are used.
The best way is to define this define is in the project properties (see Figure 2-1):
1. In the left hand tree, expand the C/C++ Build node
2. Click on the Settings node
3. Under the Tool Settings tab, click on the DSC Compiler/Input node
4. In the Defined Macros dialog box click on the first icon (+) and type the
following: OPTION_CORE_V3=1
5. Click OK
6. Click OK on the Properties dialog box
Math Library, Rev. 0
Freescale Semiconductor
2-9
Special Issues
Figure 2-1. V2/V3 core option
2.9
Special Issues
All functions in the Math Library are implemented without storing any of the
volatile registers (refer to the compiler manual) used by the respective routine.
Only non-volatile registers (C10, D10, R5) are saved by pushing the registers on
the stack. Therefore, if the particular registers initialized before the library
function call are to be used after the function call, it is necessary to save them
manually.
Math Library, Rev. 0
2-10
Freescale Semiconductor
Special Issues
Chapter 3 FUNCTION API
3.0.1
API Summary
Table 3-1. Function API summary
Name
Arguments
Output
Description
MLIB_Abs16
Frac16 f16In
Frac16
Absolute value: MLIB_Abs16 a = a
MLIB_Abs16Sat
Frac16 f16In
Frac16
Absolute value with saturation: MLIB_Abs16Sat a = a
MLIB_Abs32
Frac32 f32In
Frac32
Absolute value: MLIB_Abs32 a = a
MLIB_Abs32Sat
Frac32 f32In
Frac32
Absolute value with saturation: MLIB_Abs32Sat a = a
MLIB_Rnd16
Frac16 f16In
Frac16
Round to the nearest.
MLIB_Rnd16Sat
Frac16 f16In
Frac16
Round to the nearest with saturation.
MLIB_Rnd32
Frac32 f32In
Frac32
Round upper 16 bits to the nearest.
MLIB_Rnd32Sat
Frac32 f32In
Frac32
Round upper 16 bits to the nearest with saturation.
MLIB_Neg16
Frac16 f16In
Frac16
Negative value: MLIB_Neg16 a = – a
MLIB_Neg16Sat
Frac16 f16In
Frac16
Negative value with saturation: MLIB_Neg16Sat a = – a
MLIB_Neg32
Frac32 f32In
Frac32
Negative value: MLIB_Neg32 a = – a
MLIB_Neg32Sat
Frac32 f32In
Frac32
Negative value with saturation: MLIB_Neg32Sat a = – a
MLIB_Add16
Frac16 f16In
Frac16 f16In
Frac16
Addition: MLIB_Add16 a b = a + b
MLIB_Add16Sat
Frac16 f16In
Frac16 f16In
Frac16
Addition with saturation: MLIB_Add16Sat a b = a + b
MLIB_Add32
Frac32 f32In
Frac32 f32In
Frac32
Addition: MLIB_Add32 a b = a + b
MLIB_Add32Sat
Frac32 f32In
Frac32 f32In
Frac32
Addition with saturation: MLIB_Add32Sat a b = a + b
MLIB_Sub16
Frac16 f16In
Frac16 f16In
Frac16
Subtraction: MLIB_Sub16 a b = a – b
MLIB_Sub16Sat
Frac16 f16In
Frac16 f16In
Frac16
Subtraction with saturation:
MLIB_Sub32
Frac32 f32In
Frac32 f32In
Frac32
Subtraction: MLIB_Sub32 a b = a – b
MLIB_Sub32Sat
Frac32 f32In
Frac32 f32In
Frac32
Subtraction with saturation: MLIB_Sub32Sat a b = a – b
MLIB_Sh1L16
Frac16 f16In
Frac16
1-bit left shift: MLIB_Sh1L16 a = a « 1
MLIB_Sh1L16Sat
Frac16 f16In
Frac16
1-bit left shift with saturation: MLIB_Sh1L16Sat a = a « 1
Math Library, Rev. 0
Freescale Semiconductor
3-11
Special Issues
Table 3-1. Function API summary
MLIB_Sh1R16
Frac16 f16In
Frac16
1-bit right shift: MLIB_ShR16 a b = a » b
MLIB_Sh1L32
Frac32 f32In
Frac32
1-bit left shift: MLIB_Sh1L32 a = a « 1
MLIB_Sh1L32Sat
Frac32 f32In
Frac32
1-bit left shift with saturation: MLIB_Sh1L32Sat a = a « 1
MLIB_Sh1R32
Frac32 f32In
Frac32
1-bit right shift: MLIB_Sh1R32 a = a » 1
MLIB_ShL16
Frac16 f16In
Word16 w16N
Frac16
Multi-bit left shift: MLIB_ShL16 a b = a « b
MLIB_ShL16Sat
Frac16 f16In
Word16 w16N
Frac16
Multi-bit left shift with saturation:
MLIB_ShL16Sat a b = a « b
MLIB_ShR16
Frac16 f16In
Word16 w16N
Frac16
Multi-bit right shift: MLIB_ShR16 a b = a » b
MLIB_ShR16Sat
Frac16 f16In
Word16 w16N
Frac16
Multi-bit right shift with saturation:
MLIB_ShR16Sat a b = a » b
MLIB_ShL32
Frac32 f32In
Word16 w16N
Frac32
Multi-bit left shift: MLIB_ShL32 a b = a « b
MLIB_ShL32Sat
Frac32 f32In
Word16 w16N
Frac32
Multi-bit left shift with saturation:
MLIB_ShL32Sat a b = a « b
MLIB_ShR32
Frac32 f32In
Word16 w16N
Frac32
Multi-bit right shift: MLIB_ShR32 a b = a » b
MLIB_ShR32Sat
Frac32 f32In
Word16 w16N
Frac32
Multi-bit right shift with saturation:
MLIB_ShR32Sat a b = a » b
MLIB_Mul16SS
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication:
MLIB_Mul16SS a b = a b » 16
MLIB_Mul16SSSat
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication with saturation:
MLIB_Mul16SSSat a b = a b » 16
MLIB_MulNeg16SS
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication with negation:
MLIB_MulNeg16SS a b = – a b » 16
MLIB_Mul32SS
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication: MLIB_Mul32SS a b = a b
MLIB_Mul32SSSat
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication with saturation:
MLIB_Mul32SSSat a b = a b
MLIB_MulNeg32SS
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication with negation:
MLIB_MulNeg32SS a b = – a b
MLIB_MulRnd16SS
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication with rounding to the nearest:
ab
MLIB_MulRnd16SS a b = round ---------------
65536
MLIB_MulRnd16SSSat
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication with rounding to the nearest with
ab
saturation: MLIB_MulRnd16SSSat a b = round ---------------
65536
MLIB_MulNegRnd16SS
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication with negation with rounding to the
–a b
nearest: MLIB_MulRnd16SS a b = round ---------------
65536
Math Library, Rev. 0
3-12
Freescale Semiconductor
Special Issues
Table 3-1. Function API summary
MLIB_MulNegRnd16SSS
at
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication with negation with rounding to the
nearest with saturation:
–a b
MLIB_MulRnd16SSSat a b = round ---------------
65536
Fractional multiplication with the upper 16 bits rounding to
the nearest:
ab
MLIB_MulRnd32SS a b = round --------------- « 16
65536
Fractional multiplication with the upper 16 bits rounding to
the nearest with saturation:
ab
MLIB_MulRnd32SSSat a b = round --------------- « 16
65536
MLIB_MulRnd32SS
Frac16 f16In
Frac16 f16In
Frac32
MLIB_MulRnd32SSSat
Frac16 f16In
Frac16 f16In
Frac32
MLIB_MulNegRnd32SS
Frac16 f16In
Frac16 f16In
Frac32
MLIB_MulNegRnd32SSS
at
Frac16 f16In
Frac16 f16In
Frac32
MLIB_Mul32LS
Frac32 f32In
Frac16 f16In
Frac32
MLIB_Mul32LSSat
Frac32 f32In
Frac16 f16In
Frac32
Fractional multiplication with saturation:
MLIB_Mul32LSSat a b = a b » 16
MLIB_MulNeg32LS
Frac32 f32In
Frac16 f16In
Frac32
Fractional multiplication with negation:
MLIB_MulNeg32LS a b = – a b » 16
MLIB_Mul32LL
Frac32 f32In
Frac32 f32In
Frac32
Fractional multiplication:
MLIB_Mul32LL a b = a b » 32
MLIB_Mul32LLSat
Frac32 f32In
Frac32 f32In
Frac32
Fractional multiplication with saturation:
MLIB_Mul32LLSat a b = a b » 32
MLIB_MulNeg32LL
Frac32 f32In
Frac32 f32In
Frac32
Fractional multiplication with negation:
MLIB_Mul32NegLL a b = – a b » 32
MLIB_Mac16SSS
Frac16 f16In
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication-accumulation:
MLIB_Mac16SSS a b c = a + b c » 16
MLIB_Mac16SSSSat
Frac16 f16In
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication-accumulation with saturation:
MLIB_Mac16SSSSat a b c = a + b c » 16
MLIB_Msu16SSS
Frac16 f16In
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication-subtraction:
MLIB_Msu16SSS a b c = a – b c » 16
MLIB_Msu16SSSSat
Frac16 f16In
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication-subtraction with saturation:
MLIB_Msu16SSSSat a b c = a – b c » 16
Fractional multiplication with negation with the upper 16 bits
rounding to the nearest:
–a b
MLIB_MulNegRnd32SS a b = round --------------- « 16
65536
Fractional multiplication with negation with the upper 16 bits
rounding to the nearest with saturation:
–a b
MLIB_MulNegRnd32SSSat a b = round --------------- « 16
65536
Fractional multiplication:
MLIB_Mul32LS a b = a b » 16
Math Library, Rev. 0
Freescale Semiconductor
3-13
Special Issues
Table 3-1. Function API summary
MLIB_Mac32LSS
Frac32 f32In
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication-accumulation:
MLIB_Mac32LSS a b c = a + b c
MLIB_Mac32LSSSat
Frac32 f32In
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication-accumulation with saturation:
MLIB_Mac32LSSSat a b c = a + b c
MLIB_Msu32LSS
Frac32 f32In
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication-subtraction:
MLIB_Msu32LSS a b c = a – b c
MLIB_Msu32LSSSat
Frac32 f32In
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication-subtraction with saturation:
MLIB_Msu32LSSSat a b c = a – b c
MLIB_MacRnd16SSS
Frac16 f16In
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication-accumulation with rounding to the
nearest:
MLIB_MacRnd16SSS a b c = round a + b c » 16
MLIB_MacRnd16SSSSat
Frac16 f16In
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication-accumulation with rounding to the
nearest with saturation:
MLIB_MacRnd16SSSSat a b c = round a + b c » 16
MLIB_MsuRnd16SSS
Frac16 f16In
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication-subtraction with rounding to the
nearest:
MLIB_MsuRnd16SSS a b c = round a – b c » 16
MLIB_MsuRnd16SSSSat
Frac16 f16In
Frac16 f16In
Frac16 f16In
Frac16
Fractional multiplication-subtraction with roudning to the
nearest saturation:
MLIB_MsuRnd16SSS a b c = round a – b c » 16
MLIB_MacRnd32LSS
Frac32 f32In
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication-accumulation with rounding to the
nearest:
a+bc
MLIB_MacRnd32LSS a b c = round ------------------- « 16
65536
MLIB_MacRnd32LSSSat
Frac32 f32In
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication-accumulation with rounding to the
nearest with saturation:
a+bc
MLIB_MacRnd32LSSSat a b c = round ------------------- « 16
65536
MLIB_MsuRnd32LSS
Frac32 f32In
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication-subtraction with rounding to the
nearest:
a–bc
MLIB_MsuRnd32LSS a b c = round ------------------- « 16
65536
MLIB_MsuRnd32LSSSat
Frac32 f32In
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication-subtraction with roudning to the
nearest saturation:
a–bc
MLIB_MsuRnd32LSSSat a b c = round ------------------- « 16
65536
MLIB_MacRnd32LSS
Frac32 f32In
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication-accumulation with rounding to the
nearest:
a+bc
MLIB_MacRnd32LSS a b c = round ------------------- « 16
65536
Math Library, Rev. 0
3-14
Freescale Semiconductor
Special Issues
Table 3-1. Function API summary
MLIB_MacRnd32LSSSat
Frac32 f32In
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication-accumulation with rounding to the
nearest with saturation:
a+bc
MLIB_MacRnd32LSSSat a b c = round ------------------- « 16
65536
MLIB_MsuRnd32LSS
Frac32 f32In
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication-subtraction with rounding to the
nearest:
a–bc
MLIB_MsuRnd32LSS a b c = round ------------------- « 16
65536
MLIB_MsuRnd32LSSSat
Frac32 f32In
Frac16 f16In
Frac16 f16In
Frac32
Fractional multiplication-subtraction with roudning to the
nearest saturation:
a–bc
MLIB_MsuRnd32LSSSat a b c = round ------------------- « 16
65536
MLIB_Mac32LLS
Frac32 f32In
Frac32 f32In
Frac16 f16In
Frac32
Fractional multiplication-accumulation with rounding to the
nearest: MLIB_Mac32LLS a b c = a + b c » 16
MLIB_Mac32LLSSat
Frac32 f32In
Frac32 f32In
Frac16 f16In
Frac32
Fractional multiplication-accumulation with rounding to the
nearest with saturation:
MLIB_Mac32LLSSat a b c = a + b c » 16
MLIB_Msu32LLS
Frac32 f32In
Frac32 f32In
Frac16 f16In
Frac32
Fractional multiplication-subtraction with rounding to the
nearest: MLIB_Msu32LLS a b c = a – b c » 16
MLIB_Msu32LLSSat
Frac32 f32In
Frac32 f32In
Frac16 f16In
Frac32
Fractional multiplication-subtraction with roudning to the
nearest saturation:
MLIB_Msu32LLSSat a b c = a – b c » 16
MLIB_Mac32LLL
Frac32 f32In
Frac32 f32In
Frac32 f32In
Frac32
Fractional multiplication-accumulation with rounding to the
nearest: MLIB_Mac32LLL a b c = a + b c » 32
MLIB_Mac32LLLSat
Frac32 f32In
Frac32 f32In
Frac32 f32In
Frac32
Fractional multiplication-accumulation with rounding to the
nearest with saturation:
MLIB_Mac32LLLSat a b c = a + b c » 32
MLIB_Msu32LLL
Frac32 f32In
Frac32 f32In
Frac32 f32In
Frac32
Fractional multiplication-subtraction with rounding to the
nearest: MLIB_Msu32LLL a b c = a – b c » 32
MLIB_Msu32LLLSat
Frac32 f32In
Frac32 f32In
Frac32 f32In
Frac32
Fractional multiplication-subtraction with roudning to the
nearest saturation:
MLIB_Msu32LLLSat a b c = a – b c » 32
MLIB_Div1Q16SS
Frac16 f16In
Frac16 f16In
Frac16
Single-quadrant fractional division:
a « 16
MLIB_Div1Q16SS a b = --------------b
MLIB_Div4Q16SS
Frac16 f16In
Frac16 f16In
Frac16
Four-quadrant fractional division:
a « 16
MLIB_Div4Q16SS a b = --------------b
MLIB_Div1Q16LS
Frac32 f32In
Frac16 f16In
Frac16
Single-quadrant fractional division:
a
MLIB_Div1Q16LS a b = --b
Math Library, Rev. 0
Freescale Semiconductor
3-15
Special Issues
Table 3-1. Function API summary
MLIB_Div4Q16LS
Frac32 f32In
Frac16 f16In
Frac16
Four-quadrant fractional division:
a
MLIB_Div4Q16LS a b = --b
MLIB_Div1Q32LS
Frac32 f32In
Frac16 f16In
Frac32
Single-quadrant fractional division:
a
MLIB_Div1Q32LS a b = --- « 16
b
MLIB_Div4Q32LS
Frac32 f32In
Frac16 f16In
Frac32
Four-quadrant fractional division:
a
MLIB_Div4Q32LS a b = --- « 16
b
MLIB_Rcp161Q
Frac16 f16In
Frac32
MLIB_Rcp164Q
Frac16 f16In
Frac32
MLIB_Rcp321Q
Frac16 f16In
Frac32
MLIB_Rcp324Q
Frac16 f16In
Frac32
Single-quadrant 16-bit precision reciprocal: function:
1
MLIB_Rcp161Q a = --a
Four-quadrant 16-bit precision reciprocal function:
1
MLIB_Rcp324Q a = --a
Single-quadrant 32-bit precision reciprocal: function:
1
MLIB_Rcp161Q a = --a
Four-quadrant 32-bit precision reciprocal function:
1
MLIB_Rcp324Q a = --a
Math Library, Rev. 0
3-16
Freescale Semiconductor
MLIB_Abs16
3.1
MLIB_Abs16
This function performs an absolute value of the 16-bit argument.
3.1.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Abs16(Frac16 f16In)
3.1.2
Prototype
asm inline Frac16 MLIB_Abs16FAsmi(register Frac16 f16In)
3.1.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-2. Function Arguments
Name
f16In
3.1.4
In/Out
In
Format
SF16
Range
0x8000...
0x7FFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.1.5
Dependencies
The dependent files are:
• MLIB_AbsAsm.h
• MLIB_types.h
3.1.6
Description
The MLIB_Abs16 function returns the absolute value of the argument.
MLIB_Abs16 a = a
Eqn. 3-1
where:
• result - Frac16
• a - Frac16
Math Library, Rev. 0
Freescale Semiconductor
3-17
MLIB_Abs16
3.1.7
Returns
The function returns the 16-bit absolute value of the input f16In.
3.1.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.1.9
Special Issues
If the input is 0x8000 the output is 0x8000 if the saturation mode is turned off.
The function MLIB_Abs16 requires the saturation mode to be turned on for
full-range correct operation or the MLIB_Abs16Sat has to be used instead.
3.1.10
Implementation
The MLIB_Abs16 function is implemented as an inline function.
Example 3-1. Implementation Code
#include "mlib.h"
static Frac16 mf16In;
static Frac16 mf16Out;
void main(void)
{
mf16In = FRAC16(-0.5);
/* Absolute value */
mf16Out = MLIB_Abs16(mf16In);
}
3.1.11
See Also
See MLIB_Abs16Sat, MLIB_Abs32 and MLIB_Abs32Sat for more
information.
3.1.12
Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-18
Freescale Semiconductor
MLIB_Abs16
Table 3-3. Performance of the MLIB_Abs16 Function
Code Size (words)
1
Data Size (words)
0
Execution Clock
Min
5 cycles
Max
5 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-19
MLIB_Abs16Sat
3.2
MLIB_Abs16Sat
This function returns an absolute value of the 16-bit argument. The function
saturates the output if necessary.
3.2.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Abs16Sat(Frac16 f16In)
3.2.2
Prototype
inline Frac16 MLIB_Abs16SatFAsmi(register Frac16 f16In)
3.2.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-4. Function Arguments
Name
f16In
3.2.4
In/Out
In
Format
SF16
Range
0x8000
0x7FFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.2.5
Dependencies
The dependent files are:
• MLIB_AbsAsm.h
• MLIB_types.h
3.2.6
Description
The MLIB_Abs16Sat function returns the absolute value of the argument.
MLIB_Abs16Sat a = a
Eqn. 3-2
where:
• result - Frac16
Math Library, Rev. 0
Freescale Semiconductor
3-21
MLIB_Abs16Sat
a - Frac16
3.2.7
Returns
The function returns the 16-bit absolute value of the 16-bit input f16In. The
function saturates the output if necessary.
3.2.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.2.9
Special Issues
If the input is 0x8000 the output is 0x7FFF.
The function MLIB_Abs16Sat does not require the saturation mode to be turned
on.
3.2.10
Implementation
The MLIB_Abs16Sat function is implemented as an inline function.
Example 3-2. Implementation Code
#include "mlib.h"
static Frac16 mf16In;
static Frac16 mf16Out;
void main(void)
{
mf16In = FRAC16(-0.5);
/* Absolute value */
mf16Out = MLIB_Abs16Sat(mf16In);
}
3.2.11
See Also
See MLIB_Abs16, MLIB_Abs32 and MLIB_Abs32Sat for more information.
3.2.12
Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-22
Freescale Semiconductor
MLIB_Abs16Sat
Table 3-5. Performance of the MLIB_Abs16Sat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
7 cycles
Max
7 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-23
MLIB_Abs32
3.3
MLIB_Abs32
This function performs an absolute value of the 32-bit argument.
3.3.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Abs32(Frac32 f32In)
3.3.2
Prototype
inline Frac32 MLIB_Abs32FAsmi(register Frac32 f32In)
3.3.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-6. Function Arguments
Name
f32In
3.3.4
In/Out
In
Format
SF32
Range
0x8000 0000...
0x7FFF FFFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.3.5
Dependencies
The dependent files are:
• MLIB_AbsAsm.h
• MLIB_types.h
3.3.6
Description
The MLIB_Abs32 function returns the absolute value of the argument.
MLIB_Abs32 a = a
Eqn. 3-3
where:
• result - Frac32
• a - Frac32
Math Library, Rev. 0
Freescale Semiconductor
3-25
MLIB_Abs32
3.3.7
Returns
The function returns the 32-bit absolute value of the 32-bit input f32In.
3.3.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.3.9
Special Issues
If the input is 0x8000 0000 the output is 0x8000 0000 if the saturation mode is
turned off.
The function MLIB_Abs32 requires the saturation mode to be turned on for
full-range correct operation or the MLIB_Abs32Sat has to be used instead.
3.3.10
Implementation
The MLIB_Abs32 function is implemented as an inline function.
Example 3-3. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac32 mf32Out;
void main(void)
{
mf32In = FRAC32(-0.5);
/* Absolute value */
mf32Out = MLIB_Abs32(mf32In);
}
3.3.11
See Also
See MLIB_Abs16, MLIB_Abs16Sat and MLIB_Abs32Sat for more
information.
3.3.12
Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-26
Freescale Semiconductor
MLIB_Abs32
Table 3-7. Performance of the MLIB_Abs32 Function
Code Size (words)
1
Data Size (words)
0
Execution Clock
Min
5 cycles
Max
5 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-27
MLIB_Abs32Sat
3.4
MLIB_Abs32Sat
This function returns an absolute value of the 32-bit argument. The function
saturates the output if necessary.
3.4.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Abs32Sat(Frac32 f32In)
3.4.2
Prototype
inline Frac32 MLIB_Abs32SatFAsmi(register Frac32 f32In)
3.4.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-8. Function Arguments
Name
f32In
3.4.4
In/Out
In
Format
SF32
Range
0x8000 0000...
0x7FFF FFFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.4.5
Dependencies
The dependent files are:
• MLIB_AbsAsm.h
• MLIB_types.h
3.4.6
Description
The MLIB_Abs32Sat function returns the absolute value of the argument.
MLIB_Abs32Sat a = a
Eqn. 3-4
where:
• result - Frac32
• a - Frac32
Math Library, Rev. 0
Freescale Semiconductor
3-29
MLIB_Abs32Sat
3.4.7
Returns
The function returns the 32-bit absolute value of the 32-bit input f32In. The
function saturates the output if necessary.
3.4.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.4.9
Special Issues
If the input is 0x8000 0000 the output is 0x7FFF FFFF.
The function MLIB_Abs32Sat does not require the saturation mode to be turned
on.
3.4.10
Implementation
The MLIB_Abs32Sat function is implemented as an inline function.
Example 3-4. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac32 mf32Out;
void main(void)
{
mf32In = FRAC32(-0.5);
/* Absolute value */
mf32Out = MLIB_Abs32Sat(mf32In);
}
3.4.11
See Also
See MLIB_Abs16, MLIB_Abs16Sat and MLIB_Abs32 for more information.
3.4.12
Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-30
Freescale Semiconductor
MLIB_Abs32Sat
Table 3-9. Performance of the MLIB_Abs32Sat Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
8 cycles
Max
8 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-31
MLIB_Rnd16
3.5
MLIB_Rnd16
This function performs rounds the 16-bit argument.
3.5.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Rnd16(Frac32 f32In)
3.5.2
Prototype
asm inline Frac16 MLIB_Rnd16FAsmi(register Frac32 f32In)
3.5.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-10. Function Arguments
Name
f32In
3.5.4
In/Out
In
Format
SF32
Range
0x8000 0000...
0x7FFF FFFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.5.5
Dependencies
The dependent files are:
• MLIB_RoundAsm.h
• MLIB_types.h
3.5.6
Description
The MLIB_Rnd16 function returns the rounded argument. The rounding is to the
nearest.
3.5.7
Returns
The function returns the 16-bit rounded input f16In.
Math Library, Rev. 0
Freescale Semiconductor
3-33
MLIB_Rnd16
3.5.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.5.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the input is 0x7FFF 8000 to 0x7FFF FFF the output is 0x8000 if the saturation
mode is turned off.
The function MLIB_Rnd16 requires the saturation mode to be turned on for
full-range correct operation or the MLIB_Rnd16Sat has to be used instead.
3.5.10
Implementation
The MLIB_Rnd16 function is implemented as an inline function.
Example 3-5. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac16 mf16Out;
void main(void)
{
mf32In = FRAC32(-0.5);
/* Rounding */
mf16Out = MLIB_Rnd16(mf32In);
}
3.5.11
See Also
See MLIB_Rnd16Sat, MLIB_Rnd32 and MLIB_Rnd32Sat for more
information.
3.5.12
Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-34
Freescale Semiconductor
MLIB_Rnd16
Table 3-11. Performance of the MLIB_Rnd16 Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
7 cycles
Max
7 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-35
MLIB_Rnd16Sat
3.6
MLIB_Rnd16Sat
This function performs rounds the 16-bit argument with saturation.
3.6.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Rnd16Sat(Frac32 f32In)
3.6.2
Prototype
asm inline Frac16 MLIB_Rnd16SatFAsmi(register Frac32 f32In)
3.6.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-12. Function Arguments
Name
f32In
3.6.4
In/Out
In
Format
SF32
Range
0x8000 0000...
0x7FFF FFFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.6.5
Dependencies
The dependent files are:
• MLIB_RoundAsm.h
• MLIB_types.h
3.6.6
Description
The MLIB_Rnd16Sat function returns the rounded argument. The rounding is
to the nearest.
3.6.7
Returns
The function returns the 16-bit rounded input f16In. The function saturates the
output if necessary.
Math Library, Rev. 0
Freescale Semiconductor
3-37
MLIB_Rnd16Sat
3.6.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.6.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the input is 0x7FFF 8000 to 0x7FFF FFF the output is 0x7FFF.
The function MLIB_Rnd16Sat does not require the saturation mode to be turned
on.
3.6.10
Implementation
The MLIB_Rnd16Sat function is implemented as an inline function.
Example 3-6. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac16 mf16Out;
void main(void)
{
mf32In = FRAC32(-0.5);
/* Rounding */
mf16Out = MLIB_Rnd16Sat(mf32In);
}
3.6.11
See Also
See MLIB_Rnd16, MLIB_Rnd32 and MLIB_Rnd32Sat for more information.
3.6.12
Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-13. Performance of the MLIB_Rnd16Sat Function
Code Size (words)
2
Data Size (words)
0
Math Library, Rev. 0
3-38
Freescale Semiconductor
MLIB_Rnd16Sat
Table 3-13. Performance of the MLIB_Rnd16Sat Function
Execution Clock
Min
8 cycles
Max
8 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-39
MLIB_Rnd32
3.7
MLIB_Rnd32
This function rounds the upper 32 bits of the argument.
3.7.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Rnd32(Frac32 f32In)
3.7.2
Prototype
asm inline Frac16 MLIB_Rnd32FAsmi(register Frac32 f32In)
3.7.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-14. Function Arguments
Name
f32In
3.7.4
In/Out
In
Format
SF32
Range
0x8000 0000...
0x7FFF FFFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.7.5
Dependencies
The dependent files are:
• MLIB_RoundAsm.h
• MLIB_types.h
3.7.6
Description
The MLIB_Rnd32 function returns the rounded argument. The rounding is to the
nearest.
3.7.7
Returns
The function returns the 32-bit rounded argument (f32In) to the upper 16 bits.
Math Library, Rev. 0
Freescale Semiconductor
3-41
MLIB_Rnd32
3.7.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.7.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the input is 0x7FFF 8000 to 0x7FFF FFF the output is 0x8000 0000 if the
saturation mode is turned off.
The function MLIB_Rnd32 requires the saturation mode to be turned on for
full-range correct operation or the MLIB_Rnd32Sat has to be used instead.
3.7.10
Implementation
The MLIB_Rnd32 function is implemented as an inline function.
Example 3-7. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac32 mf32Out;
void main(void)
{
mf32In = FRAC32(-0.5);
/* Rounding */
mf16Out = MLIB_Rnd32(mf32In);
}
3.7.11
See Also
See MLIB_Rnd16, MLIB_Rnd16Sat and MLIB_Rnd32Sat for more
information.
3.7.12
Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-42
Freescale Semiconductor
MLIB_Rnd32
Table 3-15. Performance of the MLIB_Rnd32 Function
Code Size (words)
1
Data Size (words)
0
Execution Clock
Min
6 cycles
Max
6 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-43
MLIB_Rnd32Sat
3.8
MLIB_Rnd32Sat
This function rounds the upper 32 bits of the argument with saturation.
3.8.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Rnd32Sat(Frac32 f32In)
3.8.2
Prototype
asm inline Frac16 MLIB_Rnd32SatFAsmi(register Frac32 f32In)
3.8.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-16. Function Arguments
Name
f32In
3.8.4
In/Out
In
Format
SF32
Range
0x8000 0000...
0x7FFF FFFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.8.5
Dependencies
The dependent files are:
• MLIB_RoundAsm.h
• MLIB_types.h
3.8.6
Description
The MLIB_Rnd32Sat function returns the rounded argument. The rounding is
to the nearest.
3.8.7
Returns
The function returns 32-bit argument (f32In) rounded to the upper 16 bits. The
function saturates the output if necessary.
Math Library, Rev. 0
Freescale Semiconductor
3-45
MLIB_Rnd32Sat
3.8.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.8.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the input is 0x7FFF 8000 to 0x7FFF FFF the output is 0x7FFF FFFF if the
saturation mode is turned off.
The function MLIB_Rnd32Sat does not require the saturation mode to be turned
on.
3.8.10
Implementation
The MLIB_Rnd32Sat function is implemented as an inline function.
Example 3-8. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac32 mf32Out;
void main(void)
{
mf32In = FRAC32(-0.5);
/* Rounding */
mf16Out = MLIB_Rnd32Sat(mf32In);
}
3.8.11
See Also
See MLIB_Rnd16, MLIB_Rnd16Sat and MLIB_Rnd32 for more information.
3.8.12
Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-46
Freescale Semiconductor
MLIB_Rnd32Sat
Table 3-17. Performance of the MLIB_Rnd32Sat Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
8 cycles
Max
8 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-47
MLIB_Neg16
3.9
MLIB_Neg16
This function negates the 16-bit argument.
3.9.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Neg16(Frac16 f16In)
3.9.2
Prototype
inline Frac16 MLIB_Neg16FAsmi(register Frac16 f16In)
3.9.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-18. Function Arguments
Name
f16In
3.9.4
In/Out
In
Format
SF16
Range
0x8000...
0x7FFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.9.5
Dependencies
The dependent files are:
• MLIB_NegAsm.h
• MLIB_types.h
3.9.6
Description
The MLIB_Neg16 function returns the negated argument.
MLIB_Neg16 a = – a
Eqn. 3-5
where:
• result - Frac16
• a - Frac16
Math Library, Rev. 0
Freescale Semiconductor
3-49
MLIB_Neg16
3.9.7
Returns
The function returns the 16-bit negative value of the 16-bit input f16In.
3.9.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.9.9
Special Issues
If the input is 0x8000 the output is 0x8000 if the saturation mode is turned off.
The function MLIB_Neg16 requires the saturation mode to be turned on for
full-range correct operation or the MLIB_Neg16Sat has to be used instead.
3.9.10
Implementation
The MLIB_Neg16 function is implemented as an inline function.
Example 3-9. Implementation Code
#include "mlib.h"
static Frac16 mf16In;
static Frac16 mf16Out;
void main(void)
{
mf16In = FRAC16(-0.5);
/* Negative value */
mf16Out = MLIB_Neg16(mf16In);
}
3.9.11
See Also
See MLIB_Neg16Sat, MLIB_Neg32 and MLIB_Neg32Sat for more
information.
3.9.12
Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-50
Freescale Semiconductor
MLIB_Neg16
Table 3-19. Performance of the MLIB_Neg16 Function
Code Size (words)
1
Data Size (words)
0
Execution Clock
Min
5 cycles
Max
5 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-51
MLIB_Neg16Sat
3.10
MLIB_Neg16Sat
This function negates the 16-bit argument. The function saturates the output if
necessary.
3.10.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Neg16Sat(Frac16 f16In)
3.10.2
Prototype
inline Frac16 MLIB_Neg16SatFAsmi(register Frac16 f16In)
3.10.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-20. Function Arguments
Name
f16In
3.10.4
In/Out
In
Format
SF16
Range
0x8000...
0x7FFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.10.5
Dependencies
The dependent files are:
• MLIB_NegAsm.h
• MLIB_types.h
3.10.6
Description
The MLIB_Neg16Sat function returns the negated argument.
MLIB_Neg16Sat a = – a
Eqn. 3-6
where:
• result - Frac16
• a - Frac16
Math Library, Rev. 0
Freescale Semiconductor
3-53
MLIB_Neg16Sat
3.10.7
Returns
The function returns the 16-bit negative value of the 16-bit input f16In. The
function saturates the output if necessary.
3.10.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.10.9
Special Issues
If the input is 0x8000 the output is 0x7FFF.
The function MLIB_Neg16Sat does not require the saturation mode to be turned
on.
3.10.10 Implementation
The MLIB_Neg16Sat function is implemented as an inline function.
Example 3-10. Implementation Code
#include "mlib.h"
static Frac16 mf16In;
static Frac16 mf16Out;
void main(void)
{
mf16In = FRAC16(-0.5);
/* Negative value */
mf16Out = MLIB_Neg16Sat(mf16In);
}
3.10.11 See Also
See MLIB_Neg16, MLIB_Neg32 and MLIB_Neg32Sat for more information.
3.10.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-54
Freescale Semiconductor
MLIB_Neg16Sat
Table 3-21. Performance of the MLIB_Neg16Sat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
7 cycles
Max
7 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-55
MLIB_Neg32
3.11
MLIB_Neg32
This function negates the 32-bit argument.
3.11.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Neg32(Frac32 f32In)
3.11.2
Prototype
inline Frac32 MLIB_Neg32FAsmi(register Frac32 f32In)
3.11.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-22. Function Arguments
Name
f32In
3.11.4
In/Out
In
Format
SF32
Range
0x8000 0000...
0x7FFF FFFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.11.5
Dependencies
The dependent files are:
• MLIB_NegAsm.h
• MLIB_types.h
3.11.6
Description
The MLIB_Neg32 function returns the negated argument.
MLIB_Neg32 a = – a
Eqn. 3-7
where:
• result - Frac32
• a - Frac32
Math Library, Rev. 0
Freescale Semiconductor
3-57
MLIB_Neg32
3.11.7
Returns
The function returns the 32-bit negative value of 32-bit the input f32In.
3.11.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.11.9
Special Issues
If the input is 0x8000 0000 the output is 0x8000 0000 if the saturation mode is
turned off.
The function MLIB_Neg32 requires the saturation mode to be turned on for
full-range correct operation or the MLIB_Neg32Sat has to be used instead.
3.11.10 Implementation
The MLIB_Neg32 function is implemented as an inline function.
Example 3-11. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac32 mf32Out;
void main(void)
{
mf32In = FRAC32(-0.5);
/* Negative value */
mf32Out = MLIB_Neg32(mf32In);
}
3.11.11 See Also
See MLIB_Neg16, MLIB_Neg16Sat and MLIB_Neg32Sat for more
information.
3.11.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-58
Freescale Semiconductor
MLIB_Neg32
Table 3-23. Performance of the MLIB_Neg32 Function
Code Size (words)
1
Data Size (words)
0
Execution Clock
Min
6 cycles
Max
6 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-59
MLIB_Neg32Sat
3.12
MLIB_Neg32Sat
This function negates the 32-bit argument. The function saturates the output if
necessary.
3.12.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Neg32Sat(Frac32 f32In)
3.12.2
Prototype
inline Frac32 MLIB_Neg32SatFAsmi(register Frac32 f32In)
3.12.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-24. Function Arguments
Name
f32In
3.12.4
In/Out
In
Format
SF32
Range
0x8000 0000...
0x7FFF FFFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.12.5
Dependencies
The dependent files are:
• MLIB_NegAsm.h
• MLIB_types.h
3.12.6
Description
The MLIB_Neg32Sat function returns the negative value of the argument.
MLIB_Neg32Sat a = – a
Eqn. 3-8
where:
• result - Frac32
• a - Frac32
Math Library, Rev. 0
Freescale Semiconductor
3-61
MLIB_Neg32Sat
3.12.7
Returns
The function returns the 32-bit negated value of the 32-bit input f32In. The
function saturates the output if necessary.
3.12.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.12.9
Special Issues
If the input is 0x8000 0000 the output is 0x7FFF FFFF.
The function MLIB_Neg32Sat does not require the saturation mode to be turned
on.
3.12.10 Implementation
The MLIB_Neg32Sat function is implemented as an inline function.
Example 3-12. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac32 mf32Out;
void main(void)
{
mf32In = FRAC32(-0.5);
/* Negative value */
mf32Out = MLIB_Neg32Sat(mf32In);
}
3.12.11 See Also
See MLIB_Neg16, MLIB_Neg16Sat and MLIB_Neg32 for more information.
3.12.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-62
Freescale Semiconductor
MLIB_Neg32Sat
Table 3-25. Performance of the MLIB_Neg32Sat Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
8 cycles
Max
8 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-63
MLIB_Add16
3.13
MLIB_Add16
This function returns the 16-bit sum of two 16-bit inputs.
3.13.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Add16(Frac16 f16In1, Frac16 f16In2)
3.13.2
Prototype
inline Frac16 MLIB_Add16FAsmi(register Frac16 f16In1, register Frac16
f16In2)
3.13.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-26. Function Arguments
Name
3.13.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.13.5
Dependencies
The dependent files are:
• MLIB_AddAsm.h
• MLIB_types.h
3.13.6
Description
The MLIB_Add16 function returns the sum of two arguments. The function does
not saturate the output if the saturation mode is turned off.
MLIB_Add16 a b = a + b
Eqn. 3-9
Math Library, Rev. 0
Freescale Semiconductor
3-65
MLIB_Add16
where:
• result - Frac16
• a - Frac16
• b - Frac16
3.13.7
Returns
The function returns the 16-bit sum of the f16In1 and f16In2 inputs.
3.13.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.13.9
Special Issues
If the sum is greater than 0x7FFF or smaller than 0x8000, the output overflows if
the saturation mode is turned off.
In case of desired saturation, the MLIB_Add16 function requires the saturation
mode to be turned on or the MLIB_Add16Sat has to be used instead.
3.13.10 Implementation
The MLIB_Add16 function is implemented as an inline function.
Example 3-13. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac16 mf16Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16In1 + mf16In2 */
mf16Out = MLIB_Add16(mf16In1, mf16In2);
}
3.13.11 See Also
See MLIB_Add16Sat, MLIB_Add32 and MLIB_Add32Sat for more
information.
Math Library, Rev. 0
3-66
Freescale Semiconductor
MLIB_Add16
3.13.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-27. Performance of the MLIB_Add16 Function
Code Size (words)
1
Data Size (words)
0
Execution Clock
Min
8 cycles
Max
8 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-67
MLIB_Add16Sat
3.14
MLIB_Add16Sat
This function returns the 16-bit sum of two 16-bit inputs with saturation.
3.14.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Add16Sat(Frac16 f16In1, Frac16 f16In2)
3.14.2
Prototype
inline Frac16 MLIB_Add16SatFAsmi(register Frac16 f16In1, register Frac16
f16In2)
3.14.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-28. Function Arguments
Name
3.14.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.14.5
Dependencies
The dependent files are:
• MLIB_AddAsm.h
• MLIB_types.h
3.14.6
Description
The MLIB_Add16Sat function returns the sum of two arguments. The function
saturates the output if necessary.
MLIB_Add16Sat a b = a + b
Eqn. 3-10
Math Library, Rev. 0
Freescale Semiconductor
3-69
MLIB_Add16Sat
where:
• result - Frac16
• a - Frac16
• b - Frac16
3.14.7
Returns
The function returns the 16-bit sum of thef16In1 and f16In2 inputs.
3.14.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.14.9
Special Issues
If the sum is greater than 0x7FFF the output is 0x7FFF. If the sum is smaller than
0x8000 the output is 0x8000.
The function MLIB_Add16Sat does not require the saturation mode to be turned
on.
3.14.10 Implementation
The MLIB_Add16Sat function is implemented as an inline function.
Example 3-14. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac16 mf16Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16In1 + mf16In2 */
mf16Out = MLIB_Add16Sat(mf16In1, mf16In2);
}
3.14.11 See Also
See MLIB_Add16, MLIB_Add32 and MLIB_Add32Sat for more information.
Math Library, Rev. 0
3-70
Freescale Semiconductor
MLIB_Add16Sat
3.14.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-29. Performance of the MLIB_Add16Sat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
11 cycles
Max
11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-71
MLIB_Add32
3.15
MLIB_Add32
This function returns the 32-bit sum of two 32-bit inputs.
3.15.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Add32(Frac32 f32In1, Frac32 f32In2)
3.15.2
Prototype
inline Frac32 MLIB_Add32FAsmi(register Frac32 f32In1, register Frac32
f32In2)
3.15.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-30. Function Arguments
Name
3.15.4
In/Out
Format
Range
Description
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f32In2
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.15.5
Dependencies
The dependent files are:
• MLIB_AddAsm.h
• MLIB_types.h
3.15.6
Description
The MLIB_Add32 function returns the sum of two arguments. The function does
not saturate the output if the saturation mode is turned off.
MLIB_Add32 a b = a + b
Eqn. 3-11
Math Library, Rev. 0
Freescale Semiconductor
3-73
MLIB_Add32
where:
• result - Frac32
• a - Frac32
• b - Frac32
3.15.7
Returns
The function returns the 32-bit sum of the f32In1 and f32In2 inputs.
3.15.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.15.9
Special Issues
If the sum is greater than 0x7FFF FFFF or smaller than 0x8000 0000, the output
overflows if the saturation mode is turned off.
In case of desired saturation, the MLIB_Add32 function requires the saturation
mode to be turned on.
3.15.10 Implementation
The MLIB_Add32 function is implemented as an inline function.
Example 3-15. Implementation Code
#include "mlib.h"
static Frac32 mf32In1;
static Frac32 mf32In2;
static Frac32 mf32Out;
void main(void)
{
mf32In1 = FRAC32(0.1);
mf32In2 = FRAC32(-0.2);
/* mf32Out = mf32In1 + mf32In2 */
mf32Out = MLIB_Add32(mf32In1, mf32In2);
}
3.15.11 See Also
See MLIB_Add16, MLIB_Add16Sat and MLIB_Add32Sat for more
information.
Math Library, Rev. 0
3-74
Freescale Semiconductor
MLIB_Add32
3.15.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-31. Performance of the MLIB_Add32 Function
Code Size (words)
1
Data Size (words)
0
Execution Clock
Min
9 cycles
Max
9 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-75
MLIB_Add32Sat
3.16
MLIB_Add32Sat
This function returns the 32-bit sum of two 32-bit inputs with saturation.
3.16.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Add32Sat(Frac32 f32In1, Frac32 f32In2)
3.16.2
Prototype
inline Frac32 MLIB_Add32SatFAsmi(register Frac32 f32In1, register Frac32
f32In2)
3.16.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-32. Function Arguments
Name
3.16.4
In/Out
Format
Range
Description
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f32In2
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.16.5
Dependencies
The dependent files are:
• MLIB_AddAsm.h
• MLIB_types.h
3.16.6
Description
The MLIB_Add32Sat function returns the sum of two arguments. The function
saturates the output if necessary.
MLIB_Add32Sat a b = a + b
Eqn. 3-12
Math Library, Rev. 0
Freescale Semiconductor
3-77
MLIB_Add32Sat
where:
• result - Frac32
• a - Frac32
• b - Frac32
3.16.7
Returns
The function returns the 32-bit sum of the f32In1 and f32In2 inputs.
3.16.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.16.9
Special Issues
If the sum is greater than 0x7FFF FFFF the output is 0x7FFF FFFF. If the sum is
smaller than 0x8000 0000 the output is 0x8000 0000.
The function MLIB_Add32Sat does not require the saturation mode to be turned
on.
3.16.10 Implementation
The MLIB_Add32Sat function is implemented as an inline function.
Example 3-16. Implementation Code
#include "mlib.h"
static Frac32 mf32In1;
static Frac32 mf32In2;
static Frac32 mf32Out;
void main(void)
{
mf32In1 = FRAC32(0.1);
mf32In2 = FRAC32(-0.2);
/* mf32Out = mf32In1 + mf32In2 */
mf32Out = MLIB_Add32(mf32In1, mf32In2);
}
3.16.11 See Also
See MLIB_Add16, MLIB_Add16Sat and MLIB_Add32 for more information.
Math Library, Rev. 0
3-78
Freescale Semiconductor
MLIB_Add32Sat
3.16.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-33. Performance of the MLIB_Add32Sat Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
10 cycles
Max
10 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-79
MLIB_Sub16
3.17
MLIB_Sub16
This function returns the 16-bit difference of two 16-bit inputs.
3.17.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Sub16(Frac16 f16In1, Frac16 f16In2)
3.17.2
Prototype
inline Frac16 MLIB_Sub16FAsmi(register Frac16 f16In1, register Frac16
f16In2)
3.17.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-34. Function Arguments
Name
3.17.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.17.5
Dependencies
The dependent files are:
• MLIB_SubAsm.h
• MLIB_types.h
3.17.6
Description
The MLIB_Sub16 function returns the difference of two arguments. The
function does not saturate the output if the saturation mode is turned off.
MLIB_Sub16 a b = a – b
Eqn. 3-13
Math Library, Rev. 0
Freescale Semiconductor
3-81
MLIB_Sub16
where:
• result - Frac16
• a - Frac16
• b - Frac16
3.17.7
Returns
The function returns the 16-bit difference of the f16In1 and f16In2 inputs.
3.17.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.17.9
Special Issues
If the result is greater than 0x7FFF or smaller than 0x8000, the output overflows
if the saturation mode is turned off.
In case of desired saturation, the MLIB_Sub16 function requires the saturation
mode to be turned on or the MLIB_Sub16Sat has to be used instead.
3.17.10 Implementation
The MLIB_Sub16 function is implemented as an inline function.
Example 3-17. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac16 mf16Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16In1 - mf16In2 */
mf16Out = MLIB_Sub16(mf16In1, mf16In2);
}
3.17.11 See Also
See MLIB_Sub16Sat, MLIB_Sub32 and MLIB_Sub32Sat for more
information.
Math Library, Rev. 0
3-82
Freescale Semiconductor
MLIB_Sub16
3.17.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-35. Performance of the MLIB_Sub16 Function
Code Size (words)
1
Data Size (words)
0
Execution Clock
Min
9 cycles
Max
9 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-83
MLIB_Sub16Sat
3.18
MLIB_Sub16Sat
This function returns the 16-bit difference of two 16-bit inputs with saturation.
3.18.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Sub16Sat(Frac16 f16In1, Frac16 f16In2)
3.18.2
Prototype
inline Frac16 MLIB_Sub16SatFAsmi(register Frac16 f16In1, register Frac16
f16In2)
3.18.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-36. Function Arguments
Name
3.18.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.18.5
Dependencies
The dependent files are:
• MLIB_SubAsm.h
• MLIB_types.h
3.18.6
Description
The MLIB_Sub16Sat function returns the difference of two arguments. The
function saturates the output if necessary.
MLIB_Sub16Sat a b = a – b
Eqn. 3-14
Math Library, Rev. 0
Freescale Semiconductor
3-85
MLIB_Sub16Sat
where:
• result - Frac16
• a - Frac16
• b - Frac16
3.18.7
Returns
The function returns the 16-bit difference of the f16In1 and f16In2 inputs.
3.18.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.18.9
Special Issues
If the result is greater than 0x7FFF the output is 0x7FFF. If the result is smaller
than 0x8000 the output is 0x8000.
The MLIB_Sub16Sat does not require the saturation mode to be turned on.
3.18.10 Implementation
The MLIB_Sub16Sat function is implemented as an inline function.
Example 3-18. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac16 mf16Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16In1 - mf16In2 */
mf16Out = MLIB_Sub16Sat(mf16In1, mf16In2);
}
3.18.11 See Also
See MLIB_Sub16, MLIB_Sub32 and MLIB_Sub32Sat for more information.
Math Library, Rev. 0
3-86
Freescale Semiconductor
MLIB_Sub16Sat
3.18.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-37. Performance of the MLIB_Sub16Sat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
11 cycles
Max
11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-87
MLIB_Sub32
3.19
MLIB_Sub32
This function returns the 32-bit difference of two 32-bit inputs.
3.19.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Sub32(Frac32 f32In1, Frac32 f32In2)
3.19.2
Prototype
inline Frac16 MLIB_Sub32FAsmi(register Frac32 f32In1, register Frac32
f32In2)
3.19.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-38. Function Arguments
Name
3.19.4
In/Out
Format
Range
Description
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f32In2
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.19.5
Dependencies
The dependent files are:
• MLIB_SubAsm.h
• MLIB_types.h
3.19.6
Description
The MLIB_Sub32 function returns the difference of two arguments. The
function does not saturate the output if the saturation mode is turned off.
MLIB_Sub32 a b = a – b
Eqn. 3-15
where:
Math Library, Rev. 0
Freescale Semiconductor
3-89
MLIB_Sub32
•
•
•
3.19.7
result - Frac32
a - Frac32
b - Frac32
Returns
The function returns the 32-bit difference of the f32In1 and f32In2 inputs.
3.19.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.19.9
Special Issues
If the result is greater than 0x7FFF FFFF or smaller than 0x8000 0000, the output
overflows if the saturation mode is turned off.
In case of desired saturation, the MLIB_Sub32 function requires the saturation
mode to be turned on or the MLIB_Sub32Sat has to be used instead.
3.19.10 Implementation
The MLIB_Sub32 function is implemented as an inline function.
Example 3-19. Implementation Code
#include "mlib.h"
static Frac32 mf32In1;
static Frac32 mf32In2;
static Frac32 mf32Out;
void main(void)
{
mf32In1 = FRAC32(0.1);
mf32In2 = FRAC32(-0.2);
/* mf32Out = mf32In1 - mf32In2 */
mf32Out = MLIB_Sub32(mf32In1, mf32In2);
}
3.19.11 See Also
See MLIB_Sub16, MLIB_Sub16Sat and MLIB_Sub32Sat for more
information.
Math Library, Rev. 0
3-90
Freescale Semiconductor
MLIB_Sub32
3.19.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-39. Performance of the MLIB_Sub32 Function
Code Size (words)
1
Data Size (words)
0
Execution Clock
Min
8 cycles
Max
8 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-91
MLIB_Sub32Sat
3.20
MLIB_Sub32Sat
This function returns the 32-bit difference of two 32-bit inputs with saturation.
3.20.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Sub32Sat(Frac32 f32In1, Frac32 f32In2)
3.20.2
Prototype
inline Frac16 MLIB_Sub32SatFAsmi(register Frac32 f32In1, register Frac32
f32In2)
3.20.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
s
Table 3-40. Function Arguments
Name
3.20.4
In/Out
Format
Range
Description
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f32In2
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.20.5
Dependencies
The dependent files are:
• MLIB_SubAsm.h
• MLIB_types.h
3.20.6
Description
The MLIB_Sub32Sat function returns the difference of two arguments. The
function saturates the output if necessary.
MLIB_Sub32Sat a b = a – b
Eqn. 3-16
where:
Math Library, Rev. 0
Freescale Semiconductor
3-93
MLIB_Sub32Sat
•
•
•
3.20.7
result - Frac32
a - Frac32
b - Frac32
Returns
The function returns the 32-bit difference of the f32In1 and f32In2 inputs.
3.20.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.20.9
Special Issues
If the result is greater than 0x7FFF FFFF the output is 0x7FFF FFFF. If the result
is smaller than 0x8000 0000 the output is 0x8000 0000.
The MLIB_Sub32Sat does not require the saturation mode to be turned on..
3.20.10 Implementation
The MLIB_Sub32Sat function is implemented as an inline function.
Example 3-20. Implementation Code
#include "mlib.h"
static Frac32 mf32In1;
static Frac32 mf32In2;
static Frac32 mf32Out;
void main(void)
{
mf32In1 = FRAC32(0.1);
mf32In2 = FRAC32(-0.2);
/* mf32Out = mf32In1 - mf32In2 */
mf32Out = MLIB_Sub32Sat(mf32In1, mf32In2);
}
3.20.11 See Also
See MLIB_Sub16, MLIB_Sub16Sat and MLIB_Sub32 for more information.
Math Library, Rev. 0
3-94
Freescale Semiconductor
MLIB_Sub32Sat
3.20.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-41. Performance of the MLIB_Sub32Sat Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
10 cycles
Max
10 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-95
MLIB_Sh1L16
3.21
MLIB_Sh1L16
This function performs one bit left shift of the 16-bit argument without saturation.
3.21.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Sh1L16(Frac16 f16In)
3.21.2
Prototype
inline Frac16 MLIB_Sh1L16FAsmi(register Frac16 f16In)
3.21.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-42. Function Arguments
Name
f16In
3.21.4
In/Out
In
Format
Range
0x8000...
0x7FFF
SF16
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.21.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.21.6
Description
The MLIB_Sh1L16 function returns the value of the argument shifted one bit to
the left without saturation.
MLIB_Sh1L16 a = a « 1
Eqn. 3-17
where:
• result - Frac16
• a - Frac16
Math Library, Rev. 0
Freescale Semiconductor
3-97
MLIB_Sh1L16
3.21.7
Returns
The function returns the 16-bit value of the f16In input shifted one bit to the left.
3.21.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.21.9
Special Issues
If the input is greater than 0x4000 or smaller than 0xC000, the result overflows.
The function MLIB_Sh1L16 does not saturate the output. If the saturation is
required the MLIB_Sub16Sat has to be used instead.
3.21.10 Implementation
The MLIB_Sh1L16 function is implemented as an inline function.
Example 3-21. Implementation Code
#include "mlib.h"
static Frac16 mf16In;
static Frac16 mf16Out;
void main(void)
{
mf16In = FRAC16(0.25); /* 0x2000 */
/* Left shift by one bit */
mf16Out = MLIB_Sh1L16(mf16In); /* 0x4000 */
}
3.21.11 See Also
See MLIB_Sh1L16Sat, MLIB_Sh1R16, MLIB_Sh1L32, MLIB_Sh1L32Sat
and MLIB_Sh1R32 for more information.
3.21.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-98
Freescale Semiconductor
MLIB_Sh1L16
Table 3-43. Performance of the MLIB_Sh1L16 Function
Code Size (words)
1
Data Size (words)
0
Execution Clock
Min
5 cycles
Max
5 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-99
MLIB_Sh1L16Sat
3.22
MLIB_Sh1L16Sat
This function performs one bit left shift of the 16-bit argument. The function
saturates the output if necessary.
3.22.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Sh1L16Sat(Frac16 f16In)
3.22.2
Prototype
inline Frac16 MLIB_Sh1L16SatFAsmi(register Frac16 f16In)
3.22.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-44. Function Arguments
Name
f16In
3.22.4
In/Out
In
Format
SF16
Range
0x8000...
0x7FFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.22.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.22.6
Description
The MLIB_Sh1L16Sat function returns the value of the argument shifted one bit
to the left with saturation.
MLIB_Sh1L16Sat a = a « 1
Eqn. 3-18
where:
• result - Frac16
Math Library, Rev. 0
Freescale Semiconductor
3-101
MLIB_Sh1L16Sat
•
3.22.7
a - Frac16
Returns
The function returns the 16-bit value of the f16In input shifted one bit to the left.
3.22.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.22.9
Special Issues
If the input is greater than 0x4000, the result is 0x7FFF. If the input is smaller than
0xC000, the result is 0x8000.
The function MLIB_Sh1L16Sat saturates the output if necessary.
3.22.10 Implementation
The MLIB_Sh1L16Sat function is implemented as an inline function.
Example 3-22. Implementation Code
#include "mlib.h"
static Frac16 mf16In;
static Frac16 mf16Out;
void main(void)
{
mf16In = FRAC16(0.25); /* 0x2000 */
/* Left shift by one bit */
mf16Out = MLIB_Sh1L16Sat(mf16In); /* 0x4000 */
}
3.22.11 See Also
See MLIB_Sh1L16, MLIB_Sh1R16, MLIB_Sh1L32, MLIB_Sh1L32Sat and
MLIB_Sh1R32 for more information.
3.22.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-102
Freescale Semiconductor
MLIB_Sh1L16Sat
Table 3-45. Performance of the MLIB_Sh1L16Sat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
7 cycles
Max
7 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-103
MLIB_Sh1R16
3.23
MLIB_Sh1R16
This function performs one bit right shift of the 16-bit argument.
3.23.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Sh1R16(Frac16 f16In)
3.23.2
Prototype
inline Frac16 MLIB_Sh1R16FAsmi(register Frac16 f16In)
3.23.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-46. Function Arguments
Name
f16In
3.23.4
In/Out
In
Format
Range
0x8000...
0x7FFF
SF16
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.23.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.23.6
Description
The MLIB_Sh1R16 function returns the value of the argument shifted one bit the
right.
MLIB_ShR16 a b = a » b
Eqn. 3-19
where:
• result - Frac16
• a - Frac16
Math Library, Rev. 0
Freescale Semiconductor
3-105
MLIB_Sh1R16
•
3.23.7
b - Word16
Returns
The function returns the 16-bit value of the f16In input shifted one bit to the right.
3.23.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.23.9
Special Issues
None.
3.23.10 Implementation
The MLIB_Sh1R16 function is implemented as an inline function.
Example 3-23. Implementation Code
#include "mlib.h"
static Frac16 mf16In;
static Frac16 mf16Out;
void main(void)
{
mf16In = FRAC16(0.5); /* 0x4000 */
/* Right shift by one bit */
mf16Out = MLIB_Sh1R16(mf16In); /* 0x2000 */
}
3.23.11 See Also
See MLIB_Sh1L16, MLIB_Sh1L16Sat, MLIB_Sh1L32, MLIB_Sh1L32Sat
and MLIB_Sh1R32 for more information.
3.23.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-106
Freescale Semiconductor
MLIB_Sh1R16
Table 3-47. Performance of the MLIB_Sh1R16 Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
7 cycles
Max
7 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-107
MLIB_Sh1L32
3.24
MLIB_Sh1L32
This function performs one bit left shift of the 32-bit argument without saturation.
3.24.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Sh1L32(Frac32 f32In)
3.24.2
Prototype
inline Frac32 MLIB_Sh1L32FAsmi(register Frac32 f32In)
3.24.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-48. Function Arguments
Name
f32In
3.24.4
In/Out
In
Format
Range
0x8000 0000...
0x7FFF FFFF
SF32
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.24.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.24.6
Description
The MLIB_Sh1L32 function returns the value of the argument shifted one bit to
the left without saturation.
MLIB_Sh1L32 a = a « 1
Eqn. 3-20
where:
• result - Frac32
• a - Frac32
Math Library, Rev. 0
Freescale Semiconductor
3-109
MLIB_Sh1L32
3.24.7
Returns
The function returns the 32-bit value of the f32In input shifted one bit to the left.
3.24.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.24.9
Special Issues
If the input is greater than 0x4000 0000 or smaller than 0xC000 0000, the result
overflows.
The function MLIB_Sh1L32 does not saturate output. If the saturation is
required the MLIB_Sub32Sat has to be used instead.
3.24.10 Implementation
The MLIB_Sh1L32 function is implemented as an inline function.
Example 3-24. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac32 mf32Out;
void main(void)
{
mf32In = FRAC32(0.25); /* 0x20000000 */
/* Left shift by one bit */
mf32Out = MLIB_Sh1L32(mf32In); /* 0x40000000 */
}
3.24.11 See Also
See MLIB_Sh1L16, MLIB_Sh1L16Sat, MLIB_Sh1R16, MLIB_Sh1L32Sat
and MLIB_Sh1R32 for more information.
3.24.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-110
Freescale Semiconductor
MLIB_Sh1L32
Table 3-49. Performance of the MLIB_Sh1L32 Function
Code Size (words)
1
Data Size (words)
0
Execution Clock
Min
6 cycles
Max
6 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-111
MLIB_Sh1L32Sat
3.25
MLIB_Sh1L32Sat
This function performs one bit left shift of the 32-bit argument with saturation.
3.25.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Sh1L32Sat(Frac32 f32In)
3.25.2
Prototype
inline Frac32 MLIB_Sh1L32SatFAsmi(register Frac32 f32In)
3.25.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-50. Function Arguments
Name
f32In
3.25.4
In/Out
In
Format
SF32
Range
0x8000 0000...
0x7FFF FFFF
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.25.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.25.6
Description
The MLIB_Sh1L32Sat function returns the value of the argument shifted one bit
to the left with saturation.
MLIB_Sh1L32Sat a = a « 1
Eqn. 3-21
where:
• result - Frac32
• a - Frac32
Math Library, Rev. 0
Freescale Semiconductor
3-113
MLIB_Sh1L32Sat
3.25.7
Returns
The function returns the 32-bit value of the f32In input shifted one bit to the
leftwith saturation.
3.25.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.25.9
Special Issues
If the input is greater than 0x4000 0000, the result is 0x7FFF FFFF. If the input
is smaller than 0xC000 0000, the result is 0x8000 0000.
The function MLIB_Sh1L32Sat saturates the output if necessar
3.25.10 Implementation
The MLIB_Sh1L32Sat function is implemented as an inline function.
Example 3-25. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac32 mf32Out;
void main(void)
{
mf32In = FRAC32(0.25); /* 0x20000000 */
/* Left shift by one bit */
mf32Out = MLIB_Sh1L32Sat(mf32In); /* 0x40000000 */
}
3.25.11 See Also
See MLIB_Sh1L16, MLIB_Sh1L16Sat, MLIB_Sh1R16, MLIB_Sh1L32 and
MLIB_Sh1R32 for more information.
3.25.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-114
Freescale Semiconductor
MLIB_Sh1L32Sat
Table 3-51. Performance of the MLIB_Sh1L32Sat Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
7 cycles
Max
7 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-115
MLIB_Sh1R32
3.26
MLIB_Sh1R32
This function performs one bit right shift of the 32-bit argument.
3.26.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Sh1R32(Frac32 f32In)
3.26.2
Prototype
inline Frac32 MLIB_Sh1R32FAsmi(register Frac32 f32In)
3.26.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-52. Function Arguments
Name
f32In
3.26.4
In/Out
In
Format
Range
0x8000 0000...
0x7FFF FFFF
SF32
Description
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.26.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.26.6
Description
The MLIB_Sh1R32 function returns the value of the argument shifted one bit to
the right.
MLIB_Sh1R32 a = a » 1
Eqn. 3-22
where:
• result - Frac32
Math Library, Rev. 0
Freescale Semiconductor
3-117
MLIB_Sh1R32
a - Frac32
3.26.7
Returns
The function returns the 32-bit value of the f32In input shifted one bit to the right.
3.26.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.26.9
Special Issues
None
3.26.10 Implementation
The MLIB_Sh1R32 function is implemented as an inline function.
Example 3-26. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac32 mf32Out;
void main(void)
{
mf32In = FRAC32(0.5); /* 0x40000000 */
/* Right shift by one bit */
mf32Out = MLIB_Sh1R32(mf32In); /* 0x20000000 */
}
3.26.11 See Also
See MLIB_Sh1L16, MLIB_Sh1L16Sat, MLIB_Sh1R16, MLIB_Sh1L32 and
MLIB_Sh1L32Sat for more information.
3.26.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Math Library, Rev. 0
3-118
Freescale Semiconductor
MLIB_Sh1R32
Table 3-53. Performance of the MLIB_Sh1R32 Function
Code Size (words)
1
Data Size (words)
0
Execution Clock
Min
7 cycles
Max
7 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-119
MLIB_ShL16
3.27
MLIB_ShL16
This function performs multi-bit left shift of the 16-bit argument without
saturation.
3.27.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_ShL16(Frac16 f16In, Word16 w16N)
3.27.2
Prototype
inline Frac16 MLIB_ShL16FAsmi(register Frac16 f16In, register Word16
w16N)
3.27.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-54. Function Arguments
Name
3.27.4
In/Out
Format
Range
Description
f16In
In
SF16
0x8000...
0x7FFF
input value
w16N
In
SI16
-15...15
number of shifts to perform
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.27.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.27.6
Description
The MLIB_ShL16 returns the argument shifted left by the specified number of
bits. The function does not saturate the output. If the number of shifts is negative,
the value is shifted to the right.
MLIB_ShL16 a b = a « b
Math Library, Rev. 0
Freescale Semiconductor
Eqn. 3-23
3-121
MLIB_ShL16
where:
• result - Frac16
• a - Frac16
• b - Word16
3.27.7
Returns
This function returns the f16In argument shifted to the left by the number of bits
specified by the w16N argument.
3.27.8
Range Issues
The input data value is in the range of <-1,1); the shift is in the range <-15,15>.
The output data value is in the range <-1, 1).
3.27.9
Special Issues
If the number of shifts is greater than the number of the input’s leading bits, the
result overflows.
The function MLIB_ShL16 does not saturate the output. If the saturation is
required the MLIB_ShL16Sat has to be used instead.
3.27.10 Implementation
The MLIB_ShL16 function is implemented as an inline function.
Example 3-27. Implementation Code
#include "mlib.h"
static Frac16 mf16In;
static Frac16 mf16Out;
static Word16 mw16N;
void main(void)
{
mf16In = FRAC16(0.125); /* 0x1000 */
mw16N = 2;
/* Left shift by the mw16N bits */
mf16Out = MLIB_ShL16(mf16In, mw16N); /* 0x4000 */
}
Math Library, Rev. 0
3-122
Freescale Semiconductor
MLIB_ShL16
3.27.11 See Also
See MLIB_ShL16Sat, MLIB_ShR16, MLIB_ShR16Sat, MLIB_ShL32,
MLIB_ShL32Sat, MLIB_ShR32 and MLIB_ShR32Sat for more information.
3.27.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-55. Performance of the MLIB_ShL16 Function
Code Size (words)
4
Data Size (words)
0
Execution Clock
Min
12 cycles
Max
12 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-123
MLIB_ShL16Sat
3.28
MLIB_ShL16Sat
This function performs mutli-bit left shift of the 16-bit argument with saturation.
3.28.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_ShL16Sat(Frac16 f16In, Word16 w16N)
3.28.2
Prototype
inline Frac16 MLIB_ShL16SatFAsmi(register Frac16 f16In, register Word16
w16N)
3.28.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-56. Function Arguments
Name
3.28.4
In/Out
Format
Range
Description
f16In
In
SF16
0x8000...
0x7FFF
input value
w16N
In
SI16
-15...15
number of shifts to perform
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.28.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.28.6
Description
The MLIB_ShL16Sat returns the argument shifted to the left by the specified
number of bits. The function saturates the output if necessary. If the number of
shifts is negative, the value is shifted to the right. The function saturates the
output if necessary.
MLIB_ShL16Sat a b = a « b
Math Library, Rev. 0
Freescale Semiconductor
Eqn. 3-24
3-125
MLIB_ShL16Sat
where:
• result - Frac16
• a - Frac16
• b - Word16
3.28.7
Returns
This function returns the f16In argument shifted to the left by the number of bits
specified by the w16N argument.
3.28.8
Range Issues
The input data value is in the range of <-1,1); the shift is in the range <-15,15>.
The output data value is in the range <-1, 1).
3.28.9
Special Issues
If the number of shifts is greater than the number of the input’s leading bits, the
result saturates to 0x7FFF (positive input) or to 0x8000 (negative input).
The function MLIB_ShL16Sat saturates the output if necessar
3.28.10 Implementation
The MLIB_ShL16Sat function is implemented as an inline function.
Example 3-28. Implementation Code
#include "mlib.h"
static Frac16 mf16In;
static Frac16 mf16Out;
static Word16 mw16N;
void main(void)
{
mf16In = FRAC16(0.125); /*0x1000*/
mw16N = 2;
/* Left shift by the mw16N bits */
mf16Out = MLIB_ShL16Sat(mf16In, mw16N); /*0x4000*/
}
3.28.11 See Also
See MLIB_ShL16, MLIB_ShR16, MLIB_ShR16Sat, MLIB_ShL32,
MLIB_ShL32Sat, MLIB_ShR32 and MLIB_ShR32Sat for more information.
Math Library, Rev. 0
3-126
Freescale Semiconductor
MLIB_ShL16Sat
3.28.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-57. Performance of the MLIB_ShL16Sat Function
Code Size (words)
9
Data Size (words)
0
Execution Clock
Min
18 cycles
Max
18 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-127
MLIB_ShR16
3.29
MLIB_ShR16
This function performs multi-bit right shift of the 16-bit argument without
saturation.
3.29.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_ShR16(Frac16 f16In, Word16 w16N)
3.29.2
Prototype
inline Frac16 MLIB_ShR16FAsmi(register Frac16 f16In, register Word16
w16N)
3.29.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-58. Function Arguments
Name
3.29.4
In/Out
Format
Range
Description
f16In
In
SF16
0x8000...
0x7FFF
input value
w16N
In
SI16
-15...15
number of shifts to perform
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.29.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.29.6
Description
The MLIB_ShR16 returns the argument shifted to the right by the specified
number of bits. The function does not saturate the output. If the number of shifts
is negative, the value is shifted to the left.
MLIB_ShR16 a b = a » b
Math Library, Rev. 0
Freescale Semiconductor
Eqn. 3-25
3-129
MLIB_ShR16
where:
• result - Frac16
• a - Frac16
• b - Word16
3.29.7
Returns
This function returns the f16In argument shifted to the right by the number of bits
specified by the w16N argument.
3.29.8
Range Issues
The input data value is in the range of <-1,1); the shift is in the range <-15,15>.
The output data value is in the range <-1, 1).
3.29.9
Special Issues
If the negative value of the number of shifts is greater than the number of the
input’s leading bits, the result overflows.
The function MLIB_ShR16 does not saturate the output. If the saturation is
required the MLIB_ShR16Sat has to be used instead.
3.29.10 Implementation
The MLIB_ShR16 function is implemented as an inline function.
Example 3-29. Implementation Code
#include "mlib.h"
static Frac16 mf16In;
static Frac16 mf16Out;
static Word16 mw16N;
void main(void)
{
mf16In = FRAC16(0.5); /*0x4000*/
mw16N = 2;
/* Right shift by the mw16N bits */
mf16Out = MLIB_ShR16(mf16In, mw16N); /*0x1000*/
}
Math Library, Rev. 0
3-130
Freescale Semiconductor
MLIB_ShR16
3.29.11 See Also
See MLIB_ShL16, MLIB_ShL16Sat, MLIB_ShR16Sat, MLIB_ShL32,
MLIB_ShL32Sat, MLIB_ShR32 and MLIB_ShR32Sat for more information.
3.29.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-59. Performance of the MLIB_ShR16 Function
Code Size (words)
4
Data Size (words)
0
Execution Clock
Min
12 cycles
Max
12 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-131
MLIB_ShR16Sat
3.30
MLIB_ShR16Sat
This function performs mulit-bit right shift of the 16-bit argument with
saturation.
3.30.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_ShR16Sat(Frac16 f16In, Word16 w16N)
3.30.2
Prototype
inline Frac16 MLIB_ShR16SatFAsmi(register Frac16 f16In, register Word16
w16N)
3.30.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-60. Function Arguments
Name
3.30.4
In/Out
Format
Range
Description
f16In
In
SF16
0x8000...
0x7FFF
input value
w16N
In
SI16
-15...15
number of shifts to perform
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.30.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.30.6
Description
The MLIB_ShR16Sat returns the argument shifted to the right by the specified
number of bits. The function saturates the output if necessary. If the number of
shifts is negative, the value is shifted to the left.
MLIB_ShR16Sat a b = a » b
Math Library, Rev. 0
Freescale Semiconductor
Eqn. 3-26
3-133
MLIB_ShR16Sat
where:
• result - Frac16
• a - Frac16
• b - Word16
3.30.7
Returns
This function returns the 16-bit argument f16In shifted to the right by the number
of bits specified by the w16N argument.
3.30.8
Range Issues
The input data value is in the range of <-1,1); the shift is in the range <-15,15>.
The output data value is in the range <-1, 1).
3.30.9
Special Issues
If the negative value of the number of shifts is greater than the number of the
input’s leading bits, the result saturates to 0x7FFF (positive input) or to 0x8000
(negative input).
The function MLIB_ShR16Sat saturates the output if necessary.
3.30.10 Implementation
The MLIB_ShR16Sat function is implemented as an inline function.
Example 3-30. Implementation Code
#include "mlib.h"
static Frac16 mf16In;
static Frac16 mf16Out;
static Word16 mw16N;
void main(void)
{
mf16In = FRAC16(0.5); /*0x4000*/
mw16N = 2;
/* Right shift by the mw16N bits */
mf16Out = MLIB_ShR16Sat(mf16In, mw16N); /*0x1000*/
}
Math Library, Rev. 0
3-134
Freescale Semiconductor
MLIB_ShR16Sat
3.30.11 See Also
See MLIB_ShL16, MLIB_ShL16Sat, MLIB_ShR16, MLIB_ShL32,
MLIB_ShL32Sat, MLIB_ShR32 and MLIB_ShR32Sat for more information.
3.30.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-61. Performance of the MLIB_ShR16Sat Function
Code Size (words)
9
Data Size (words)
0
Execution Clock
Min
17 cycles
Max
17 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-135
MLIB_ShL32
3.31
MLIB_ShL32
This function performs multi-bit left shift of the 32-bit argument without
saturation.
3.31.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_ShL32(Frac32 f32In, Word16 w16N)
3.31.2
Prototype
inline Frac32 MLIB_ShL32FAsmi(register Frac32 f32In, register Word16
w16N)
3.31.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-62. Function Arguments
Name
3.31.4
In/Out
Format
Range
Description
f32In
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
w16N
In
SI16
-31...31
number of shifts to perform
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.31.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.31.6
Description
The MLIB_ShL32 returns the argument shifted to the left by the specified
number of bits. The function does not saturate the output. If the number of shifts
is negative, the value is shifted to the right.
MLIB_ShL32 a b = a « b
Math Library, Rev. 0
Freescale Semiconductor
Eqn. 3-27
3-137
MLIB_ShL32
where:
• result - Frac32
• a - Frac32
• b - Word16
3.31.7
Returns
This function returns the f32In argument shifted to the left by the number of bits
specified by the w16N argument.
3.31.8
Range Issues
The input data value is in the range of <-1,1); the shift is in the range <-31,31>.
The output data value is in the range <-1, 1).
3.31.9
Special Issues
If the number of shifts is greater than the number of the input’s leading bits, the
result overflows.
The function MLIB_ShL32 does not saturate the output. If the saturation is
required the MLIB_ShL32Sat has to be used instead.
3.31.10 Implementation
The MLIB_ShL32 function is implemented as an inline function.
Example 3-31. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac32 mf32Out;
static Word16 mw16N;
void main(void)
{
mf32In = FRAC32(0.125); /* 0x10000000 */
mw16N = 2;
/* Left shift by the mw16N bits */
mf32Out = MLIB_ShL32(mf32In, mw16N); /* 0x40000000 */
}
Math Library, Rev. 0
3-138
Freescale Semiconductor
MLIB_ShL32
3.31.11 See Also
See MLIB_ShL16, MLIB_ShL16Sat, MLIB_ShR16, MLIB_ShR16Sat,
MLIB_ShL32Sat, MLIB_ShR32 and MLIB_ShR32Sat for more information.
3.31.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-63. Performance of the MLIB_ShL32 Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
10 cycles
Max
10 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-139
MLIB_ShL32Sat
3.32
MLIB_ShL32Sat
This function performs multi-bit left shift of the 32-bit argument with saturation.
3.32.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_ShL32Sat(Frac32 f32In, Word16 w16N)
3.32.2
Prototype
inline Frac32 MLIB_ShL32SatFAsmi(register Frac32 f32In, register Word16
w16N)
3.32.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-64. Function Arguments
Name
3.32.4
In/Out
Format
Range
Description
f32In
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
w16N
In
SI16
-31...31
number of shifts to perform
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.32.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.32.6
Description
The MLIB_ShL32Sat returns the argument shifted to the left by the specified
number of bits. The function saturates the output if necessary. If the number of
shifts is negative, the value is shifted to the right.
MLIB_ShL32Sat a b = a « b
Eqn. 3-28
Math Library, Rev. 0
Freescale Semiconductor
3-141
MLIB_ShL32Sat
where:
• result - Frac32
• a - Frac32
• b - Word16
3.32.7
Returns
This function returns the f32In argument shifted to the left by the number of bits
specified by the w16N argument.
3.32.8
Range Issues
The input data value is in the range of <-1,1); the shift is in the range <-31,31>.
The output data value is in the range <-1, 1).
3.32.9
Special Issues
If the number of shifts is greater than the number of the input’s leading bits, the
result saturates to 0x7FFF FFFF (positive input) or to 0x8000 0000 (negative
input).
The function MLIB_ShL32Sat saturates the output if necessary.
3.32.10 Implementation
The MLIB_ShL32Sat function is implemented as an inline function.
Example 3-32. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac32 mf32Out;
static Word16 mw16N;
void main(void)
{
mf32In = FRAC32(0.125); /* 0x10000000 */
mw16N = 2;
/* Left shift by the mw16N bits */
mf32Out = MLIB_ShL32Sat(mf32In, mw16N); /* 0x40000000 */
}
Math Library, Rev. 0
3-142
Freescale Semiconductor
MLIB_ShL32Sat
3.32.11 See Also
See MLIB_ShL16, MLIB_ShL16Sat, MLIB_ShR16, MLIB_ShR16Sat,
MLIB_ShL32, MLIB_ShR32 and MLIB_ShR32Sat for more information.
3.32.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-65. Performance of the MLIB_ShL32Sat Function
Code Size (words)
8
Data Size (words)
0
Execution Clock
Min
16 cycles
Max
16 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-143
MLIB_ShR32
3.33
MLIB_ShR32
This function performs multi-bit right shift of the 16-bit argument without
saturation.
3.33.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_ShR32(Frac32 f32In, Word16 w16N)
3.33.2
Prototype
inline Frac32 MLIB_ShR32FAsmi(register Frac32 f32In, register Word16
w16N)
3.33.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-66. Function Arguments
Name
3.33.4
In/Out
Format
Range
Description
f32In
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
w16N
In
SI16
-31...31
number of shifts to perform
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.33.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.33.6
Description
The MLIB_ShR32 returns the argument shifted to the right by the specified
number of bits. The function does not saturate the output. If the number of shifts
is negative, the value is shifted to the left.
MLIB_ShR32 a b = a » b
Math Library, Rev. 0
Freescale Semiconductor
Eqn. 3-29
3-145
MLIB_ShR32
where:
• result - Frac32
• a - Frac32
• b - Word16
3.33.7
Returns
This function returns the f32In argument shifted to the right by the number of bits
specified by the w16N argument.
3.33.8
Range Issues
The input data value is in the range of <-1,1); the shift is in the range <-31,31>.
The output data value is in the range <-1, 1).
3.33.9
Special Issues
If the negative value of the number of shifts is greater than the number of the
input’s leading bits, the result overflows.
The function MLIB_ShR32 does not saturate the output. If the saturation is
required the MLIB_ShR32Sat has to be used instead.
3.33.10 Implementation
The MLIB_ShR32 function is implemented as an inline function.
Example 3-33. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac32 mf32Out;
static Word16 mw16N;
void main(void)
{
mf32In = FRAC32(0.5); /* 0x40000000 */
mw16N = 2;
/* Right shift by the mw16N bits */
mf32Out = MLIB_ShR32(mf32In, mw16N); /* 0x10000000 */
}
Math Library, Rev. 0
3-146
Freescale Semiconductor
MLIB_ShR32
3.33.11 See Also
See MLIB_ShL16, MLIB_ShL16Sat, MLIB_ShR16, MLIB_ShR16Sat,
MLIB_ShL32, MLIB_ShL32Sat and MLIB_ShR32Sat for more information.
3.33.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-67. Performance of the MLIB_ShR32 Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
10 cycles
Max
10 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-147
MLIB_ShR32Sat
3.34
MLIB_ShR32Sat
This function performs multi-bit right shift of the 16-bit argument with
saturation.
3.34.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_ShR32Sat(Frac32 f32In, Word16 w16N)
3.34.2
Prototype
inline Frac32 MLIB_ShR32SatFAsmi(register Frac32 f32In, register Word16
w16N)
3.34.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-68. Function Arguments
Name
3.34.4
In/Out
Format
Range
Description
f32In
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
w16N
In
SI16
-31...31
number of shifts to perform
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.34.5
Dependencies
The dependent files are:
• MLIB_ShiftAsm.h
• MLIB_types.h
3.34.6
Description
The MLIB_ShR32Sat returns the argument shifted to the right by the specified
number of bits. The function saturates the output if necessary. If the number of
shifts is negative, the value is shifted to the left.
MLIB_ShR32Sat a b = a » b
Math Library, Rev. 0
Freescale Semiconductor
Eqn. 3-30
3-149
MLIB_ShR32Sat
where:
• result - Frac32
• a - Frac32
• b - Word16
3.34.7
Returns
This function returns the f32In argument shifted to the right by the number of bits
specified by the w16N argument. The function saturates the output if necessary.
If the number of shifts is negative, the value is shifted to the left shift.
3.34.8
Range Issues
The input data value is in the range of <-1,1); the shift is in the range <-31,31>.
The output data value is in the range <-1, 1).
3.34.9
Special Issues
If the negative value of the number of shifts is greater than the number of the
input’s leading bits, the result saturates to 0x7FFF FFFF (positive input) or to
0x8000 0000 (negative input).
The function MLIB_ShR32Sat saturates the output if necessary.
3.34.10 Implementation
The MLIB_ShR32Sat function is implemented as an inline function.
Example 3-34. Implementation Code
#include "mlib.h"
static Frac32 mf32In;
static Frac32 mf32Out;
static Word16 mw16N;
void main(void)
{
mf32In = FRAC32(0.5); /* 0x40000000 */
mw16N = 2;
/* Right shift by the mw16N bits */
mf32Out = MLIB_ShR32Sat(mf32In, mw16N); /* 0x10000000 */
}
Math Library, Rev. 0
3-150
Freescale Semiconductor
MLIB_ShR32Sat
3.34.11 See Also
See MLIB_ShL16, MLIB_ShL16Sat, MLIB_ShR16, MLIB_ShR16Sat,
MLIB_ShL32, MLIB_ShL32Sat and MLIB_ShR32 for more information.
3.34.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-69. Performance of the MLIB_ShR32Sat Function
Code Size (words)
8
Data Size (words)
0
Execution Clock
Min
17 cycles
Max
17 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-151
MLIB_Mul16SS
3.35
MLIB_Mul16SS
This function returns the 16-bit fractional product of two 16-bit fractional inputs.
3.35.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Mul16SS(Frac16 f16In1, Frac16 f16In2)
3.35.2
Prototype
inline Frac16 MLIB_Mul16SSFAsmi(register Frac16 f16In1, register Frac16
f16In2)
3.35.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-70. Function Arguments
Name
3.35.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.35.5
Dependencies
The dependent files are:
• MLIB_Mul16Asm.h
• MLIB_types.h
3.35.6
Description
The MLIB_Mul16SS function returns the fractional product of two fractional
inputs. The result is the upper 16 bits of the resulted 32-bit product. The function
does not saturate the output if the saturation mode is turned off.
MLIB_Mul16SS a b = a b » 16
Eqn. 3-31
Math Library, Rev. 0
Freescale Semiconductor
3-153
MLIB_Mul16SS
where:
• result - Frac16
• a - Frac16
• b - Frac16
3.35.7
Returns
The function returns the 16-bit fractional product of two 16-bit fractional inputs
f16In1 and f16In2.
3.35.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.35.9
Special Issues
If the both inputs are 0x8000, the output is 0x8000 if the saturation mode is turned
off.
In case of desired saturation, the MLIB_Mul16SS function requires the
saturation mode to be turned on or the MLIB_Mul16SSSat has to be used
instead.
3.35.10 Implementation
The MLIB_Mul16SS function is implemented as an inline function.
Example 3-35. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac16 mf16Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16In1*mf16In2 */
mf16Out = MLIB_Mul16SS(mf16In1, mf16In2);
}
Math Library, Rev. 0
3-154
Freescale Semiconductor
MLIB_Mul16SS
3.35.11 See Also
See MLIB_Mul16SSSat, MLIB_MulNeg16SS, MLIB_Mul32SS,
MLIB_Mul32SSSat and MLIB_MulNeg32SS for more information.
3.35.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-71. Performance of the MLIB_Mul16SS Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
11 cycles
Max
11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-155
MLIB_Mul16SSSat
3.36
MLIB_Mul16SSSat
This function returns the 16-bit fractional product of two 16-bit fractional inputs
with saturation.
3.36.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Mul16SSSat(Frac16 f16In1, Frac16 f16In2)
3.36.2
Prototype
inline Frac16 MLIB_Mul16SSSatFAsmi(register Frac16 f16In1, register
Frac16 f16In2)
3.36.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-72. Function Arguments
Name
3.36.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.36.5
Dependencies
The dependent files are:
• MLIB_Mul16Asm.h
• MLIB_types.h
3.36.6
Description
The MLIB_Mul16SSSat function returns the fractional product of two fractional
inputs. The function saturates the output if necessary.
MLIB_Mul16SSSat a b = a b » 16
Eqn. 3-32
Math Library, Rev. 0
Freescale Semiconductor
3-157
MLIB_Mul16SSSat
where:
• result - Frac16
• a - Frac16
• b - Frac16
3.36.7
Returns
The function returns the 16-bit fractional product of two 16-bit fractional inputs
f16In1 and f16In2.
3.36.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.36.9
Special Issues
If the both inputs are 0x8000, the output is 0x7FFF.
The function MLIB_Mul16SSSat does not require the saturation mode to be
turned on.
3.36.10 Implementation
The MLIB_Mul16SSSat function is implemented as an inline function.
Example 3-36. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac16 mf16Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16In1*mf16In2 */
mf16Out = MLIB_Mul16SSSat(mf16In1, mf16In2);
}
3.36.11 See Also
See MLIB_Mul16SS, MLIB_MulNeg16SS, MLIB_Mul32SS,
MLIB_Mul32SSSat and MLIB_MulNeg32SS for more information.
Math Library, Rev. 0
3-158
Freescale Semiconductor
MLIB_Mul16SSSat
3.36.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-73. Performance of the MLIB_Mul16SSSat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
11 cycles
Max
11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-159
MLIB_MulNeg16SS
3.37
MLIB_MulNeg16SS
This function returns the 16-bit negative fractional product of two 16-bit
fractional inputs.
3.37.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_MulNeg16SS(Frac16 f16In1, Frac16 f16In2)
3.37.2
Prototype
inline Frac16 MLIB_MulNeg16SSFAsmi(register Frac16 f16In1, register
Frac16 f16In2)
3.37.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-74. Function Arguments
Name
3.37.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.37.5
Dependencies
The dependent files are:
• MLIB_Mul16Asm.h
• MLIB_types.h
3.37.6
Description
The MLIB_MulNeg16SS function returns the fractioal negative product of two
fractional inputs.
MLIB_MulNeg16SS a b = – a b » 16
Eqn. 3-33
Math Library, Rev. 0
Freescale Semiconductor
3-161
MLIB_MulNeg16SS
where:
• result - Frac16
• a - Frac16
• b - Frac16
3.37.7
Returns
The function returns the 16-bit fractional negative product two 16-bit fractional
inputs f16In1 and f16In2.
3.37.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.37.9
Special Issues
The function MLIB_MulNeg16SS is saration mode independent.
3.37.10 Implementation
The MLIB_MulNeg16SS function is implemented as an inline function.
Example 3-37. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac16 mf16Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = -mf16In1*mf16In2 */
mf16Out = MLIB_MulNeg16SS(mf16In1, mf16In2);
}
3.37.11 See Also
See MLIB_Mul16SS, MLIB_Mul16SSSat, MLIB_Mul32SS,
MLIB_Mul32SSSat and MLIB_MulNeg32SS for more information.
Math Library, Rev. 0
3-162
Freescale Semiconductor
MLIB_MulNeg16SS
3.37.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-75. Performance of the MLIB_MulNeg16SS Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
11 cycles
Max
11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-163
MLIB_Mul32SS
3.38
MLIB_Mul32SS
This function returns the 32-bit fractional product of two 16-bit fractional inputs.
3.38.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Mul32SS(Frac16 f16In1, Frac16 f16In2)
3.38.2
Prototype
inline Frac32 MLIB_Mul32SSFAsmi(register Frac16 f16In1, register Frac16
f16In2)
3.38.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-76. Function Arguments
Name
3.38.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.38.5
Dependencies
The dependent files are:
• MLIB_Mul16Asm.h
• MLIB_types.h
3.38.6
Description
The MLIB_Mul32SS function returns the fractional product of two fractional
inputs. The function does not saturate the output if the saturation mode is turned
off.
MLIB_Mul32SS a b = a b
Eqn. 3-34
Math Library, Rev. 0
Freescale Semiconductor
3-165
MLIB_Mul32SS
where:
• result - Frac32
• a - Frac16
• b - Frac16
3.38.7
Returns
The function returns the 32-bit fractional product of two 16-bit fractional inputs
f16In1 and f16In2.
3.38.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.38.9
Special Issues
If the both inputs are 0x8000, the output is 0x8000 0000 if the saturation mode is
turned off.
In case of desired saturation, the MLIB_Mul32SS function requires the
saturation mode to be turned on or the MLIB_Mul32SSSat has to be used
instead.
3.38.10 Implementation
The MLIB_Mul32SS function is implemented as an inline function.
Example 3-38. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf16In1*mf16In2 */
mf32Out = MLIB_Mul32SS(mf16In1, mf16In2);
}
Math Library, Rev. 0
3-166
Freescale Semiconductor
MLIB_Mul32SS
3.38.11 See Also
See MLIB_Mul16SS, MLIB_Mul16SSSat, MLIB_MulNeg16SS,
MLIB_Mul32SSSat and MLIB_MulNeg32SS for more information.
3.38.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-77. Performance of the MLIB_Mul32SS Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
10 cycles
Max
10 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-167
MLIB_Mul32SSSat
3.39
MLIB_Mul32SSSat
This function returns the 32-bit fractional product of two 16-bit fractional inputs
with saturation.
3.39.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Mul32SSSat(Frac16 f16In1, Frac16 f16In2)
3.39.2
Prototype
inline Frac32 MLIB_Mul32SSSatFAsmi(register Frac16 f16In1, register
Frac16 f16In2)
3.39.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-78. Function Arguments
Name
3.39.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.39.5
Dependencies
The dependent files are:
• MLIB_Mul16Asm.h
• MLIB_types.h
3.39.6
Description
The MLIB_Mul32SSSat function returns the fractional product of two fractional
inputs. The function saturates the output if necessary.
MLIB_Mul32SSSat a b = a b
Eqn. 3-35
Math Library, Rev. 0
Freescale Semiconductor
3-169
MLIB_Mul32SSSat
where:
• result - Frac32
• a - Frac16
• b - Frac16
3.39.7
Returns
The function returns the 32-bit fractional product of two 16-bit fractional inputs
f16In1 and f16In2.
3.39.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.39.9
Special Issues
If the both inputs are 0x8000, the output is 0x7FFF FFFF.
The function MLIB_Mul32SSSat does not require the saturation mode to be
turned on.
3.39.10 Implementation
The MLIB_Mul32SSSat function is implemented as an inline function.
Example 3-39. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf16In1*mf16In2 */
mf32Out = MLIB_Mul32SSSat(mf16In1, mf16In2);
}
3.39.11 See Also
See MLIB_Mul16SS, MLIB_Mul16SSSat, MLIB_MulNeg16SS,
MLIB_Mul32SS and MLIB_MulNeg32SS for more information.
Math Library, Rev. 0
3-170
Freescale Semiconductor
MLIB_Mul32SSSat
3.39.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-79. Performance of the MLIB_Mul32SSSat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
11 cycles
Max
11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-171
MLIB_MulNeg32SS
3.40
MLIB_MulNeg32SS
This function returns the 32-bit fractional negative product of 16-bit fractional
inputs.
3.40.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_MulNeg32SS(Frac16 f16In1, Frac16 f16In2)
3.40.2
Prototype
inline Frac32 MLIB_MulNeg32SSFAsmi(register Frac16 f16In1, register
Frac16 f16In2)
3.40.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-80. Function Arguments
Name
3.40.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.40.5
Dependencies
The dependent files are:
• MLIB_Mul16Asm.h
• MLIB_types.h
3.40.6
Description
The MLIB_MulNeg32SS function returns the fractional negative product of two
fractional inputs.
Math Library, Rev. 0
Freescale Semiconductor
3-173
MLIB_MulNeg32SS
3.40.7
Returns
The function returns the 32-bit negative value of multiple of two 16-bit fractional
inputs f16In1 and f16In2.
MLIB_MulNeg32SS a b = – a b
Eqn. 3-36
where:
• result - Frac32
• a - Frac16
• b - Frac16
3.40.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.40.9
Special Issues
The function MLIB_MulNeg32SS is saturation mode independent.
3.40.10 Implementation
The MLIB_MulNeg32SS function is implemented as an inline function.
Example 3-40. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = -mf16In1*mf16In2 */
mf32Out = MLIB_MulNeg32SS(mf16In1, mf16In2);
}
3.40.11 See Also
See MLIB_Mul16SS, MLIB_Mul16SSSat, MLIB_MulNeg16SS,
MLIB_Mul32SS and MLIB_MulNeg32SS for more information.
Math Library, Rev. 0
3-174
Freescale Semiconductor
MLIB_MulNeg32SS
3.40.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-81. Performance of the MLIB_MulNeg32SS Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
9 cycles
Max
9 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-175
MLIB_MulRnd16SS
3.41
MLIB_MulRnd16SS
This function returns the rounded 16-bit fractional product of two 16-bit
fractional inputs.
3.41.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_MulRnd16SS(Frac16 f16In1, Frac16 f16In2)
3.41.2
Prototype
inline Frac16 MLIB_MulRnd16SSFAsmi(register Frac16 f16In1, register
Frac16 f16In2)
3.41.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-82. Function Arguments
Name
3.41.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.41.5
Dependencies
The dependent files are:
• MLIB_MulRnd16Asm.h
• MLIB_types.h
3.41.6
Description
The MLIB_MulRnd16SS function returns the rounded fractional product of two
fractional inputs. The result is rounded to the nearest. The function does not
saturate the output if the saturation mode is turned off.
Math Library, Rev. 0
Freescale Semiconductor
3-177
MLIB_MulRnd16SS
ab
MLIB_MulRnd16SS a b = round ---------------
65536
Eqn. 3-37
where:
• result - Frac16
• a - Frac16
• b - Frac16
3.41.7
Returns
The function returns the rounded 16-bit fractional product of two 16-bit fractional
inputs f16In1 and f16In2.
3.41.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.41.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the product before rounding is 0x7FFF 8000 to 0x7FFF FFF the output is
0x8000 if the saturation mode is turned off.
If the both inputs are 0x8000, the output is 0x8000 if the saturation mode is turned
off.
In case of desired saturation, the MLIB_MulRnd16SS function requires the
saturation mode to be turned on or the MLIB_MulRnd16SSSat has to be used
instead.
3.41.10 Implementation
The MLIB_MulRnd16SS function is implemented as an inline function.
Example 3-41. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac16 mf16Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
Math Library, Rev. 0
3-178
Freescale Semiconductor
MLIB_MulRnd16SS
/* mf16Out = mf16In1*mf16In2, result is rounded */
mf16Out = MLIB_MulRnd16SS(mf16In1, mf16In2);
}
3.41.11 See Also
See MLIB_MulRnd16SSSat, MLIB_MulNegRnd16SS,
MLIB_MulNegRnd16SSSat, MLIB_MulRnd32SS,
MLIB_MulRnd32SSSat, MLIB_MulNegRnd32SS and
MLIB_MulNegRnd32SSSat for more information.
3.41.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-83. Performance of the MLIB_MulRnd16SS Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
11 cycles
Max
11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-179
MLIB_MulRnd16SSSat
3.42
MLIB_MulRnd16SSSat
This function returns the rounded 16-bit fractional product of two 16-bit
fractional inputs with saturation.
3.42.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_MulRnd16SSSat(Frac16 f16In1, Frac16 f16In2)
3.42.2
Prototype
inline Frac16 MLIB_MulRnd16SSSatFAsmi(register Frac16 f16In1, register
Frac16 f16In2)
3.42.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-84. Function Arguments
Name
3.42.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.42.5
Dependencies
The dependent files are:
• MLIB_MulRnd16Asm.h
• MLIB_types.h
3.42.6
Description
The MLIB_MulRnd16SSSat returns the rounded fractional product of two
fractional inputs. The result is rounded to the nearest. The function saturates the
output if necessary.
Math Library, Rev. 0
Freescale Semiconductor
3-181
MLIB_MulRnd16SSSat
ab
MLIB_MulRnd16SSSat a b = round ---------------
65536
Eqn. 3-38
where:
• result - Frac16
• a - Frac16
• b - Frac16
3.42.7
Returns
The function returns the 16-bit rounded product of two 16-bit fractional inputs
f16In1 and f16In2.
3.42.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.42.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the product before rounding is 0x7FFF 8000 to 0x7FFF FFF the output is
0x7FFF.
If the both inputs are 0x8000, the output is 0x7FFF.
The function MLIB_MulRnd16SSSat does not require the saturation mode to be
turned on.
3.42.10 Implementation
The MLIB_MulRnd16SSSat function is implemented as an inline function.
Example 3-42. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac16 mf16Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16In1*mf16In2, result is rounded */
Math Library, Rev. 0
3-182
Freescale Semiconductor
MLIB_MulRnd16SSSat
mf16Out = MLIB_MulRnd16SSSat(mf16In1, mf16In2);
}
3.42.11 See Also
See MLIB_MulRnd16SS, MLIB_MulNegRnd16SS,
MLIB_MulNegRnd16SSSat, MLIB_MulRnd32SS,
MLIB_MulRnd32SSSat, MLIB_MulNegRnd32SS and
MLIB_MulNegRnd32SSSat for more information.
3.42.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-85. Performance of the MLIB_MulRnd16SSSat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
11 cycles
Max
11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-183
MLIB_MulNegRnd16SS
3.43
MLIB_MulNegRnd16SS
This function returns the rounded 16-bit fractional negative product of two 16-bit
fractional inputs.
3.43.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_MulNegRnd16SS(Frac16 f16In1, Frac16 f16In2)
3.43.2
Prototype
inline Frac16 MLIB_MulNegRnd16SSFAsmi(register Frac16 f16In1, register
Frac16 f16In2)
3.43.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-86. Function Arguments
Name
3.43.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.43.5
Dependencies
The dependent files are:
• MLIB_MulRnd16Asm.h
• MLIB_types.h
3.43.6
Description
The MLIB_MulNegRnd16SS function returns the rounded fractional negative
product of two fractional inputs. The result is rounded to the nearest. The function
does not saturate the output if the saturation mode is turned off.
Math Library, Rev. 0
Freescale Semiconductor
3-185
MLIB_MulNegRnd16SS
–a b
MLIB_MulRnd16SS a b = round ---------------
65536
Eqn. 3-39
where:
• result - Frac16
• a - Frac16
• b - Frac16
3.43.7
Returns
The function returns the rounded 16-bit fractional negative product of two 16-bit
fractional inputs f16In1 and f16In2.
3.43.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.43.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the product before rounding is 0x7FFF 8000 to 0x7FFF FFF the output is
0x8000 if the saturation mode is turned off.
In case of desired saturation, the MLIB_MulNegRnd16SS function requires the
saturation mode to be turned on or the MLIB_MulNegRnd16SSSat has to be
used instead.
3.43.10 Implementation
The MLIB_MulNegRnd16SS function is implemented as an inline function.
Example 3-43. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac16 mf16Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = -mf16In1*mf16In2, result is rounded */
mf16Out = MLIB_MulNegRnd16SS(mf16In1, mf16In2);
Math Library, Rev. 0
3-186
Freescale Semiconductor
MLIB_MulNegRnd16SS
}
3.43.11 See Also
See MLIB_MulRnd16SS, MLIB_MulRnd16SSSat,
MLIB_MulNegRnd16SSSat, MLIB_MulRnd32SS,
MLIB_MulRnd32SSSat, MLIB_MulNegRnd32SS and
MLIB_MulNegRnd32SSSat for more information.
3.43.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-87. Performance of the MLIB_MulNegRnd16SS Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
11 cycles
Max
11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-187
MLIB_MulNegRnd16SSSat
3.44
MLIB_MulNegRnd16SSSat
This function returns the rounded 16-bit fractional negative product of two 16-bit
fractional inputs with saturation.
3.44.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_MulNegRnd16SSSat(Frac16 f16In1, Frac16 f16In2)
3.44.2
Prototype
inline Frac16 MLIB_MulNegRnd16SSSatFAsmi(register Frac16 f16In1,
register Frac16 f16In2)
3.44.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-88. Function Arguments
Name
3.44.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.44.5
Dependencies
The dependent files are:
• MLIB_MulRnd16Asm.h
• MLIB_types.h
3.44.6
Description
The MLIB_MulNegRnd16SSSat function returns the rounded fractional
negative product of two fractional inputs. The result is rounded to the nearest. The
function saturates the output if necessary.
Math Library, Rev. 0
Freescale Semiconductor
3-189
MLIB_MulNegRnd16SSSat
–a b
MLIB_MulRnd16SSSat a b = round ---------------
65536
Eqn. 3-40
where:
• result - Frac16
• a - Frac16
• b - Frac16
3.44.7
Returns
The function returns the rounded 16-bit fractional negative product of two 16-bit
fractional inputs f16In1 and f16In2. The function saturates the output if
necessary.
3.44.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.44.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the product before rounding is 0x7FFF 8000 to 0x7FFF FFF the output is
0x7FFF.
The function MLIB_MulNegRnd16SSSat does not require the saturation mode
to be turned on.
3.44.10 Implementation
The MLIB_MulNegRnd16SSSat function is implemented as an inline function.
Example 3-44. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac16 mf16Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = -mf16In1*mf16In2, result is rounded */
mf16Out = MLIB_MulNegRnd16SSSat(mf16In1, mf16In2);
Math Library, Rev. 0
3-190
Freescale Semiconductor
MLIB_MulNegRnd16SSSat
}
3.44.11 See Also
See MLIB_MulRnd16SS, MLIB_MulRnd16SSSat,
MLIB_MulNegRnd16SS, MLIB_MulRnd32SS, MLIB_MulRnd32SSSat,
MLIB_MulNegRnd32SS and MLIB_MulNegRnd32SSSat for more
information.
3.44.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-89. Performance of the MLIB_MulNegRnd16SSSat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
11 cycles
Max
11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-191
MLIB_MulRnd32SS
3.45
MLIB_MulRnd32SS
This function returns the 32-bit product of two 16-bit fractional inputs rounded to
the upper 16 bits.
3.45.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_MulRnd32SS(Frac16 f16In1, Frac16 f16In2)
3.45.2
Prototype
inline Frac32 MLIB_MulRnd32SSFAsmi(register Frac16 f16In1, register
Frac16 f16In2)
3.45.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-90. Function Arguments
Name
3.45.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.45.5
Dependencies
The dependent files are:
• MLIB_MulRnd16Asm.h
• MLIB_types.h
3.45.6
Description
The MLIB_MulRnd32SS function returns the product of two fractional inputs
rounded to the upper 16 bits. The result is rounded to the nearest. The function
does not saturate the output if the saturation mode is turned off.
Math Library, Rev. 0
Freescale Semiconductor
3-193
MLIB_MulRnd32SS
ab
MLIB_MulRnd32SS a b = round --------------- « 16
65536
Eqn. 3-41
where:
• result - Frac32
• a - Frac16
• b - Frac16
3.45.7
Returns
The function returns the 32-bit fractional product of two 16-bit fractional inputs
f16In1 and f16In2 rounded to the upper 16 bits.
3.45.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.45.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the product before rounding is 0x7FFF 8000 to 0x7FFF FFF the output is
0x8000 0000 if the saturation mode is turned off.
If the both inputs are 0x8000, the output is 0x8000 0000 if the saturation mode is
turned off.
In case of desired saturation, the MLIB_MulRnd32SS function requires the
saturation mode to be turned on or the MLIB_MulRnd32SSSat has to be used
instead.
3.45.10 Implementation
The MLIB_MulRnd32SS function is implemented as an inline function.
Example 3-45. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
Math Library, Rev. 0
3-194
Freescale Semiconductor
MLIB_MulRnd32SS
/* mf32Out = mf16In1*mf16In2, result is rounded */
mf32Out = MLIB_MulRnd32SS(mf16In1, mf16In2);
}
3.45.11 See Also
See MLIB_MulRnd16SS, MLIB_MulRnd16SSSat,
MLIB_MulNegRnd16SS, MLIB_MulNegRnd16SSSat,
MLIB_MulRnd32SSSat, MLIB_MulNegRnd32SS and
MLIB_MulNegRnd32SSSat for more information.
3.45.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-91. Performance of the MLIB_MulRnd32SS Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
10 cycles
Max
10 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-195
MLIB_MulRnd32SSSat
3.46
MLIB_MulRnd32SSSat
This function returns the 32-bit product of two 16-bit fractional inputs rounded to
the upper 16 bits with saturation.
3.46.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_MulRnd32SSSat(Frac16 f16In1, Frac16 f16In2)
3.46.2
Prototype
inline Frac32 MLIB_MulRnd32SSSatFAsmi(register Frac16 f16In1, register
Frac16 f16In2)
3.46.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-92. Function Arguments
Name
3.46.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.46.5
Dependencies
The dependent files are:
• MLIB_MulRnd16Asm.h
• MLIB_types.h
3.46.6
Description
The MLIB_MulRnd32SSSat function returns the product of two fractional
inputs rounded to the upper 16 bits. The result is rounded to the nearest. The
function saturates the output if necessary.
Math Library, Rev. 0
Freescale Semiconductor
3-197
MLIB_MulRnd32SSSat
ab
MLIB_MulRnd32SSSat a b = round --------------- « 16
65536
Eqn. 3-42
where:
• result - Frac32
• a - Frac16
• b - Frac16
3.46.7
Returns
The function returns the 32-bit fractional product of two 16-bit fractional inputs
f16In1 and f16In2 rounded to the upper 16 bits.
3.46.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.46.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the product before rounding is 0x7FFF 8000 to 0x7FFF FFF the output is
0x7FFF FFFF.
If the both inputs are 0x8000, the output is 0x7FFF FFFF.
The function MLIB_MulRnd32SSSat does not require the saturation mode to be
turned on.
3.46.10 Implementation
The MLIB_MulRnd32SSSat function is implemented as an inline function.
Example 3-46. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf16In1*mf16In2, result is rounded */
Math Library, Rev. 0
3-198
Freescale Semiconductor
MLIB_MulRnd32SSSat
mf32Out = MLIB_MulRnd32SSSat(mf16In1, mf16In2);
}
3.46.11 See Also
See MLIB_MulRnd16SS, MLIB_MulRnd16SSSat,
MLIB_MulNegRnd16SS, MLIB_MulNegRnd16SSSat,
MLIB_MulRnd32SS, MLIB_MulNegRnd32SS and
MLIB_MulNegRnd32SSSat for more information.
3.46.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-93. Performance of the MLIB_MulRnd32SSSat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
11 cycles
Max
11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-199
MLIB_MulNegRnd32SS
3.47
MLIB_MulNegRnd32SS
This function returns the 32-bit negative product of two 16-bit fractional inputs
rounded to the upper 16 bits.
3.47.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_MulNegRnd32SS(Frac16 f16In1, Frac16 f16In2)
3.47.2
Prototype
inline Frac32 MLIB_MulNegRnd32SSFAsmi(register Frac16 f16In1, register
Frac16 f16In2)
3.47.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-94. Function Arguments
Name
3.47.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.47.5
Dependencies
The dependent files are:
• MLIB_MulRnd16Asm.h
• MLIB_types.h
3.47.6
Description
The MLIB_MulNegRnd32SS function returns the negative product of two
fractional inputs rounded to the upper 16 bits. The result is rounded to the nearest.
The function does not saturate the output if the saturation mode is turned off.
Math Library, Rev. 0
Freescale Semiconductor
3-201
MLIB_MulNegRnd32SS
–a b
MLIB_MulNegRnd32SS a b = round --------------- « 16
65536
Eqn. 3-43
where:
• result - Frac32
• a - Frac16
• b - Frac16
3.47.7
Returns
The function returns the 32-bit fractional negative product of two 16-bit
fractional inputs f16In1 and f16In2 rounded to the upper 16 bits.
3.47.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.47.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the product before rounding is 0x7FFF 8000 to 0x7FFF FFF the output is
0x8000 0000 if the saturation mode is turned off.
In case of desired saturation, the MLIB_MulNegRnd32SS function requires the
saturation mode to be turned on or the MLIB_MulNegRnd32SSSat has to be
used instead.
3.47.10 Implementation
The MLIB_MulNegRnd32SS function is implemented as an inline function.
Example 3-47. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = -mf16In1*mf16In2, result is rounded */
mf32Out = MLIB_MulNegRnd32SS(mf16In1, mf16In2);
Math Library, Rev. 0
3-202
Freescale Semiconductor
MLIB_MulNegRnd32SS
}
3.47.11 See Also
See MLIB_MulRnd16SS, MLIB_MulRnd16SSSat,
MLIB_MulNegRnd16SS, MLIB_MulNegRnd16SSSat,
MLIB_MulRnd32SS, MLIB_MulRnd32SSSat and
MLIB_MulNegRnd32SSSat for more information.
3.47.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-95. Performance of the MLIB_MulNegRnd32SS Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
10 cycles
Max
10 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-203
MLIB_MulNegRnd32SSSat
3.48
MLIB_MulNegRnd32SSSat
This function returns the 32-bit negative product of two 16-bit fractional inputs
rounded to the upper 16 bits with saturation.
3.48.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_MulNegRnd32SSSat(Frac16 f16In1, Frac16 f16In2)
3.48.2
Prototype
inline Frac16 MLIB_MulNegRnd32SSSatFAsmi(register Frac16 f16In1,
register Frac16 f16In2)
3.48.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-96. Function Arguments
Name
3.48.4
In/Out
Format
Range
Description
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.48.5
Dependencies
The dependent files are:
• MLIB_MulRnd16Asm.h
• MLIB_types.h
3.48.6
Description
The MLIB_MulNegRnd32SSSat function returns the negative product of two
fractional inputs rounded to the upper 16 bits. The result is rounded to the nearest.
The function saturates the output if necessary.
Math Library, Rev. 0
Freescale Semiconductor
3-205
MLIB_MulNegRnd32SSSat
–a b
MLIB_MulNegRnd32SSSat a b = round --------------- « 16
65536
Eqn. 3-44
where:
• result - Frac32
• a - Frac16
• b - Frac16
3.48.7
Returns
The function returns the 32-bit fractional negative product of two 16-bit
fractional inputs f16In1 and f16In2 rounded to the upper 16 bits. The function
saturates the output if necessary.
3.48.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.48.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the product before rounding is 0x7FFF 8000 to 0x7FFF FFF the output is
0x7FFF FFFF.
The function MLIB_MulNegRnd32SSSat does not require the saturation mode
to be turned on.
3.48.10 Implementation
The MLIB_MulNegRnd32SSSat function is implemented as an inline function.
Example 3-48. Implementation Code
#include "mlib.h"
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = -mf16In1*mf16In2, result is rounded */
mf32Out = MLIB_MulNegRnd32SSSat(mf16In1, mf16In2);
Math Library, Rev. 0
3-206
Freescale Semiconductor
MLIB_MulNegRnd32SSSat
}
3.48.11 See Also
See MLIB_MulRnd16SS, MLIB_MulRnd16SSSat,
MLIB_MulNegRnd16SS, MLIB_MulNegRnd16SSSat,
MLIB_MulRnd32SS, MLIB_MulRnd32SSSat and MLIB_MulNegRnd32SS
for more information.
3.48.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-97. Performance of the MLIB_MulNegRnd32SSSat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
11 cycles
Max
11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-207
MLIB_Mul32LS
3.49
MLIB_Mul32LS
This function returns the 32-bit product of a 32-bit and a 16-bit fractional input.
3.49.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Mul32LS(Frac32 f32In1, Frac16 f16In2)
3.49.2
Prototype
inline Frac32 MLIB_Mul32LSFAsmi(register Frac32 f32In1, register Frac16
f16In2)
V3 core version:
inline Frac32 MLIB_V3Mul32LSFAsmi(register Frac32 f32In1, register
Frac16 f16In2)
3.49.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-98. Function Arguments
Name
3.49.4
In/Out
Format
Range
Description
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.49.5
Dependencies
The dependent files are:
• MLIB_Mul24Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-209
MLIB_Mul32LS
3.49.6
Description
The MLIB_Mul32LS function returns the product of two fractional inputs. The
function does not saturate the output if the saturation mode is turned off.
MLIB_Mul32LS a b = a b » 16
Eqn. 3-45
where:
• result - Frac32
• a - Frac32
• b - Frac16
3.49.7
Returns
The function returns the upper 32 bits of the fractional product of a 32-bit
(f32In1) and a 16-bit (f16In2) fractional input.
3.49.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.49.9
Special Issues
If the inputs are 0x8000 0000 and 0x8000, the output is 0x8000 0000 if the
saturation mode is turned off.
In case of desired saturation, the MLIB_Mul32LS function requires the
saturation mode to be turned on or the MLIB_Mul32LSSat has to be used
instead.
3.49.10 Implementation
The MLIB_Mul32LS function is implemented as an inline function.
Example 3-49. Implementation Code
#include "mlib.h"
static Frac32 mf32In1;
static Frac16 mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf32In1 = FRAC32(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32In1*mf16In2 */
mf32Out = MLIB_Mul32LS(mf32In1, mf16In2);
Math Library, Rev. 0
3-210
Freescale Semiconductor
MLIB_Mul32LS
}
3.49.11 See Also
See MLIB_Mul32LSSat, MLIB_MulNeg32LS, MLIB_Mul32LL,
MLIB_Mul32LLSat and MLIB_MulNeg32LL for more information.
3.49.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-99. Performance of the MLIB_Mul32LS Function
Code Size (words)
V2: 5, V3: 3
Data Size (words)
0
Execution Clock
Min
V2: 13, V3: 11 cycles
Max
V2: 13, V3: 11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-211
MLIB_Mul32LSSat
3.50
MLIB_Mul32LSSat
This function returns the 32-bit product of a 32-bit and a 16-bit fractional input
with saturation.
3.50.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Mul32LSSat(Frac32 f32In1, Frac16 f16In2)
3.50.2
Prototype
inline Frac32 MLIB_Mul32LSSatFAsmi(register Frac32 f32In1, register
Frac16 f16In2)
V3 core version:
inline Frac32 MLIB_V3Mul32LSSatFAsmi(register Frac32 f32In1, register
Frac16 f16In2)
3.50.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-100. Function Arguments
Name
3.50.4
In/Out
Format
Range
Description
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.50.5
Dependencies
The dependent files are:
• MLIB_Mul24Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-213
MLIB_Mul32LSSat
3.50.6
Description
The MLIB_Mul32LSSat function returns the product of two fractional inputs.
The function saturates the output if necessary.
MLIB_Mul32LSSat a b = a b » 16
Eqn. 3-46
where:
• result - Frac32
• a - Frac32
• b - Frac16
3.50.7
Returns
The function returns the upper 32 bits of the fractional product of a 32-bit
(f32In1) and a 16-bit (f16In2) fractional input.
3.50.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.50.9
Special Issues
If the inputs are 0x8000 0000 and 0x8000, the output is 0x7FFF FFFF.
The function MLIB_Mul32LSSat does not require the saturation mode to be
turned on.
3.50.10 Implementation
The MLIB_Mul32LSSat function is implemented as an inline function.
Example 3-50. Implementation Code
#include "mlib.h"
static Frac32 mf32In1;
static Frac16 mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf32In1 = FRAC32(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32In1*mf16In2 */
mf32Out = MLIB_Mul32LSSat(mf32In1, mf16In2);
}
Math Library, Rev. 0
3-214
Freescale Semiconductor
MLIB_Mul32LSSat
3.50.11 See Also
See MLIB_Mul32LS, MLIB_MulNeg32LS, MLIB_Mul32LL,
MLIB_Mul32LLSat and MLIB_MulNeg32LL for more information.
3.50.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-101. Performance of the MLIB_Mul32LSSat Function
Code Size (words)
V2: 6, V3: 4
Data Size (words)
0
Execution Clock
Min
V2: 14, V3: 11 cycles
Max
V2: 14, V3: 11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-215
MLIB_MulNeg32LS
3.51
MLIB_MulNeg32LS
This function returns the 32-bit fractional negative product of a 32-bit and a
16-bit fractional input.
3.51.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_MulNeg32LS(Frac32 f32In1, Frac16 f16In2)
3.51.2
Prototype
inline Frac32 MLIB_MulNeg32LSFAsmi(register Frac32 f32In1, register
Frac16 f16In2)
inline Frac32 MLIB_V3MulNeg32LSFAsmi(register Frac32 f32In1, register
Frac16 f16In2)
3.51.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-102. Function Arguments
Name
3.51.4
In/Out
Format
Range
Description
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.51.5
Dependencies
The dependent files are:
• MLIB_Mul24Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-217
MLIB_MulNeg32LS
3.51.6
Description
The MLIB_MulNeg32LS function returns the negative product of two fractional
inputs.
MLIB_MulNeg32LS a b = – a b » 16
Eqn. 3-47
where:
• result - Frac32
• a - Frac32
• b - Frac16
3.51.7
Returns
The function returns the upper 32 bits of the fractional product of a 32-bit
(f32In1) and a 16-bit (f16In2) fractional input.
3.51.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.51.9
Special Issues
The MLIB_MulNeg32LS function is saturation mode independent.
3.51.10 Implementation
The MLIB_MulNeg32LS function is implemented as an inline function.
Example 3-51. Implementation Code
#include "mlib.h"
static Frac32 mf32In1;
static Frac16 mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf32In1 = FRAC32(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = -mf32In1*mf16In2 */
mf32Out = MLIB_MulNeg32LS(mf32In1, mf16In2);
}
Math Library, Rev. 0
3-218
Freescale Semiconductor
MLIB_MulNeg32LS
3.51.11 See Also
See MLIB_Mul32LS, MLIB_Mul32LSSat, MLIB_Mul32LL,
MLIB_Mul32LLSat and MLIB_MulNeg32LL for more information.
3.51.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-103. Performance of the MLIB_MulNeg32LS Function
Code Size (words)
V2: 6, V3: 4
Data Size (words)
0
Execution Clock
Min
V2: 14, V3: 11 cycles
Max
V2: 14, V3: 11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-219
MLIB_Mul32LL
3.52
MLIB_Mul32LL
This function returns the 32-bit fractional product of two 32-bit fractional inputs.
3.52.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Mul32LL(Frac32 f32In1, Frac32 f32In2)
3.52.2
Prototype
inline Frac32 MLIB_Mul32LLFAsmi(register Frac32 f32In1, register Frac32
f32In2)
V3 core version:
inline Frac32 MLIB_V3Mul32LLFAsmi(register Frac32 f32In1, register
Frac32 f32In2)
3.52.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-104. Function Arguments
Name
3.52.4
In/Out
Format
Range
Description
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f32In2
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.52.5
Dependencies
The dependent files are:
• MLIB_Mul32Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-221
MLIB_Mul32LL
3.52.6
Description
The MLIB_Mul32LL function returns the product of two fractional inputs. The
function does not saturate the output if the saturation mode is turned off.
MLIB_Mul32LL a b = a b » 32
Eqn. 3-48
where:
• result - Frac32
• a - Frac32
• b - Frac32
3.52.7
Returns
The function returns the upper 32 bits of the fractional product of two 32-bit
fractional inputs f32In1 and f32In2.
3.52.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.52.9
Special Issues
If the both inputs are 0x8000 0000, the output is 0x8000 0000 if the saturation
mode is turned off.
In case of desired saturation, the MLIB_Mul32LL function requires the
saturation mode to be turned on or the MLIB_Mul32LLSat has to be used
instead.
3.52.10 Implementation
The MLIB_Mul32LL function is implemented as an inline function.
Example 3-52. Implementation Code
#include "mlib.h"
static Frac32 mf32In1;
static Frac32 mf32In2;
static Frac32 mf32Out;
void main(void)
{
mf32In1 = FRAC32(0.1);
mf32In2 = FRAC32(-0.2);
/* mf32Out = mf32In1*mf32In2 */
mf32Out = MLIB_Mul32LL(mf32In1, mf32In2);
Math Library, Rev. 0
3-222
Freescale Semiconductor
MLIB_Mul32LL
}
3.52.11 See Also
See MLIB_Mul32LS, MLIB_Mul32LSSat, MLIB_MulNeg32LS,
MLIB_Mul32LLSat and MLIB_MulNeg32LL for more information.
3.52.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-105. Performance of the MLIB_Mul32LL Function
Code Size (words)
V2: 9, V3: 2
Data Size (words)
0
Execution Clock
Min
V2: 17, V3: 10 cycles
Max
V2: 17, V3: 10 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-223
MLIB_Mul32LLSat
3.53
MLIB_Mul32LLSat
This function returns the 32-bit fractional product of two 32-bit fractional inputs
with saturation.
3.53.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Mul32LLSat(Frac32 f32In1, Frac32 f32In2)
3.53.2
Prototype
inline Frac32 MLIB_Mul32LLSatFAsmi(register Frac32 f32In1, register
Frac32 f32In2)
V3 core version:
inline Frac32 MLIB_V3Mul32LLSatFAsmi(register Frac32 f32In1, register
Frac32 f32In2)
3.53.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-106. Function Arguments
Name
3.53.4
In/Out
Format
Range
Description
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f32In2
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.53.5
Dependencies
The dependent files are:
• MLIB_Mul32Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-225
MLIB_Mul32LLSat
3.53.6
Description
The MLIB_Mul32LLSat function returns the product of two fractional inputs.
The function saturates the output if necessary.
MLIB_Mul32LLSat a b = a b » 32
Eqn. 3-49
where:
• result - Frac32
• a - Frac32
• b - Frac32
3.53.7
Returns
The function returns the upper 32 bits of the fractional product of two 32-bit
fractional inputs f32In1 and f32In2.
3.53.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.53.9
Special Issues
If the both inputs are 0x8000 0000, the output is 0x7FFF FFFF.
The function MLIB_Mul32LLSat does not require the saturation mode to be
turned on.
3.53.10 Implementation
The MLIB_Mul32LLSat function is implemented as an inline function.
Example 3-53. Implementation Code
#include "mlib.h"
static Frac32 mf32In1;
static Frac32 mf32In2;
static Frac32 mf32Out;
void main(void)
{
mf32In1 = FRAC32(0.1);
mf32In2 = FRAC32(-0.2);
/* mf32Out = mf32In1*mf32In2 */
mf32Out = MLIB_Mul32LLSat(mf32In1, mf32In2);
}
Math Library, Rev. 0
3-226
Freescale Semiconductor
MLIB_Mul32LLSat
3.53.11 See Also
See MLIB_Mul32LS, MLIB_Mul32LSSat, MLIB_MulNeg32LS,
MLIB_Mul32LL and MLIB_MulNeg32LL for more information.
3.53.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-107. Performance of the MLIB_Mul32LLSat Function
Code Size (words)
V2: 10, V3:
Data Size (words)
0
Execution Clock
Min
V2: 18, V3: 11 cycles
Max
V2: 18, V3: 11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-227
MLIB_MulNeg32LL
3.54
MLIB_MulNeg32LL
This function returns the 32-bit fractional negative product of two 32-bit
fractional inputs.
3.54.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_MulNeg32LL(Frac32 f32In1, Frac32 f32In2)
3.54.2
Prototype
inline Frac32 MLIB_MulNeg32LLFAsmi(register Frac32 f32In1, register
Frac32 f32In2)
V3 core version:
inline Frac32 MLIB_V3MulNeg32LLFAsmi(register Frac32 f32In1, register
Frac32 f32In2)
3.54.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-108. Function Arguments
Name
3.54.4
In/Out
Format
Range
Description
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f32In2
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.54.5
Dependencies
The dependent files are:
• MLIB_Mul32Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-229
MLIB_MulNeg32LL
3.54.6
Description
The MLIB_MulNeg32LL function returns the negative product of two
fractional inputs.
MLIB_Mul32NegLL a b = – a b » 32
Eqn. 3-50
where:
• result - Frac32
• a - Frac32
• b - Frac32
3.54.7
Returns
The function returns the upper 32 bits of the fractional negative product of two
32-bit fractional inputs f32In1 and f32In2.
3.54.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.54.9
Special Issues
The MLIB_MulNeg32LL function is saturation mode independent.
3.54.10 Implementation
The MLIB_MulNeg32LL function is implemented as an inline function.
Example 3-54. Implementation Code
#include "mlib.h"
static Frac32 mf32In1;
static Frac32 mf32In2;
static Frac32 mf32Out;
void main(void)
{
mf32In1 = FRAC32(0.1);
mf32In2 = FRAC32(-0.2);
/* mf32Out = -mf32In1*mf32In2 */
mf32Out = MLIB_MulNeg32LL(mf32In1, mf32In2);
}
Math Library, Rev. 0
3-230
Freescale Semiconductor
MLIB_MulNeg32LL
3.54.11 See Also
See MLIB_Mul32LS, MLIB_Mul32LSSat, MLIB_MulNeg32LS,
MLIB_Mul32LL and MLIB_Mul16SSSat for more information.
3.54.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-109. Performance of the MLIB_MulNeg32LL Function
Code Size (words)
V2: 10, V3: 3
Data Size (words)
0
Execution Clock
Min
V2: 18, V3: 11 cycles
Max
V2: 18, V3: 11 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-231
MLIB_Mac16SSS
3.55
MLIB_Mac16SSS
This function returns the 16-bit sum of the 16-bit fractional accumulator and the
product of two 16-bit fractional inputs.
3.55.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Mac16SSS(Frac16 f16Acc, Frac16 f16In1, Frac16 f16In2)
3.55.2
Prototype
inline Frac16 MLIB_Mac16SSSFAsmi(register Frac16 f16Acc, register Frac16
f16In1, register Frac16 f16In2)
3.55.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-110. Function Arguments
Name
3.55.4
In/Out
Format
Range
Description
f16Acc
In
SF16
0x8000...
0x7FFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.55.5
Dependencies
The dependent files are:
• MLIB_Mac16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-233
MLIB_Mac16SSS
3.55.6
Description
The MLIB_Mac16SSS returns the sum of the accumulator and the product of
two fractional inputs. The function does not saturate the output if the saturation
mode is turned off.
MLIB_Mac16SSS a b c = a + b c » 16
Eqn. 3-51
where:
• result - Frac16
• a - Frac16
• b - Frac16
• c - Frac16
3.55.7
Returns
The function returns the 16-bit sum of a 16-bit fractional accumulator (f16Acc)
and the product of two 16-bit fractional inputs (f16In1 and f16In2).
3.55.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.55.9
Special Issues
If the result is greater than 0x7FFF or smaller than 0x8000, the output overflows
if the saturation mode is turned off.
In case of desired saturation, the MLIB_Mac16SSS function requires the
saturation mode to be turned on or the MLIB_Mac16SSSSat has to be used
instead.
3.55.10 Implementation
The MLIB_Mac16SSS function is implemented as an inline function.
Example 3-55. Implementation Code
#include "mlib.h"
static
static
static
static
Frac16
Frac16
Frac16
Frac16
mf16Acc;
mf16In1;
mf16In2;
mf16Out;
void main(void)
{
mf16Acc = FRAC16(0.3);
Math Library, Rev. 0
3-234
Freescale Semiconductor
MLIB_Mac16SSS
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16Acc + mf16In1*mf16In2 */
mf16Out = MLIB_Mac16SSS(mf16Acc, mf16In1, mf16In2);
}
3.55.11 See Also
See MLIB_Mac16SSSSat, MLIB_Msu16SSS, MLIB_Msu16SSSSat,
MLIB_Mac32LSS, MLIB_Mac32LSSSat, MLIB_Msu32LSS and
MLIB_Msu32LSSSat for more information.
3.55.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-111. Performance of the MLIB_Mac16SSS Function
Code Size (words)
4
Data Size (words)
0
Execution Clock
Min
14 cycles
Max
14 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-235
MLIB_Mac16SSSSat
3.56
MLIB_Mac16SSSSat
This function returns the 16-bit sum of the 16-bit fractional accumulator and the
product of two 16-bit fractional inputs with saturation.
3.56.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Mac16SSSSat(Frac16 f16Acc, Frac16 f16In1, Frac16 f16In2)
3.56.2
Prototype
inline Frac16 MLIB_Mac16SSSSatFAsmi(register Frac16 f16Acc, register
Frac16 f16In1, register Frac16 f16In2)
3.56.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-112. Function Arguments
Name
3.56.4
In/Out
Format
Range
Description
f16Acc
In
SF16
0x8000...
0x7FFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.56.5
Dependencies
The dependent files are:
• MLIB_Mac16Asm.h
• MLIB_types.h
3.56.6
Description
The MLIB_Mac16SSSSat returns the sum of the accumulator and the product of
two fractional inputs. The function saturates the output if necessary.
Math Library, Rev. 0
Freescale Semiconductor
3-237
MLIB_Mac16SSSSat
MLIB_Mac16SSSSat a b c = a + b c » 16
Eqn. 3-52
where:
• result - Frac16
• a - Frac16
• b - Frac16
• c - Frac16
3.56.7
Returns
The function returns the 16-bit sum of a 16-bit fractional accumulator (f16Acc)
and the product of two 16-bit fractional inputs f16In1 and f16In2.
3.56.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.56.9
Special Issues
If the result is greater than 0x7FFF the output is 0x7FFF. If the result is smaller
than 0x8000 the output is 0x8000.
The function MLIB_Mac16SSSSat does not require the saturation mode to be
turned on.
3.56.10 Implementation
The MLIB_Mac16SSSSat function is implemented as an inline function.
Example 3-56. Implementation Code
#include "mlib.h"
static
static
static
static
Frac16
Frac16
Frac16
Frac16
mf16Acc;
mf16In1;
mf16In2;
mf16Out;
void main(void)
{
mf16Acc = FRAC16(0.3);
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16Acc + mf16In1*mf16In2 */
mf16Out = MLIB_Mac16SSSSat(mf16Acc, mf16In1, mf16In2);
}
Math Library, Rev. 0
3-238
Freescale Semiconductor
MLIB_Mac16SSSSat
3.56.11 See Also
See MLIB_Mac16SSS, MLIB_Msu16SSS, MLIB_Msu16SSSSat,
MLIB_Mac32LSS, MLIB_Mac32LSSSat, MLIB_Msu32LSS and
MLIB_Msu32LSSSat for more information.
3.56.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-113. Performance of the MLIB_Mac16SSSSat Function
Code Size (words)
4
Data Size (words)
0
Execution Clock
Min
14 cycles
Max
14 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-239
MLIB_Msu16SSS
3.57
MLIB_Msu16SSS
This function returns the 16-bit value of the product of two 16-bit fractional
inputs subtracted from the 16-bit fractional accumulator.
3.57.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Msu16SSS(Frac16 f16Acc, Frac16 f16In1, Frac16 f16In2)
3.57.2
Prototype
inline Frac16 MLIB_Msu16SSSFAsmi(register Frac16 f16Acc, register Frac16
f16In1, register Frac16 f16In2)
3.57.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-114. Function Arguments
Name
3.57.4
In/Out
Format
Range
Description
f16Acc
In
SF16
0x8000...
0x7FFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.57.5
Dependencies
The dependent files are:
• MLIB_Mac16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-241
MLIB_Msu16SSS
3.57.6
Description
The MLIB_Msu16SSS returns the product of two fractional inputs subtracted
from the fractional accumulator. The function does not saturate the output if the
saturation mode is turned off.
MLIB_Msu16SSS a b c = a – b c » 16
Eqn. 3-53
where:
• result - Frac16
• a - Frac16
• b - Frac16
• c - Frac16
3.57.7
Returns
This function returns the 16-bit product of two 16-bit fractional inputs (f16In1
and f16In2) sustracted from the 16-bit fractional accumulator (f16Acc).
3.57.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.57.9
Special Issues
If the result is greater than 0x7FFF or smaller than 0x8000, the output overflows
if the saturation mode is turned off.
In case of desired saturation, the MLIB_Msu16SSS function requires the
saturation mode to be turned on or the MLIB_Msu16SSSSat has to be used
instead.
3.57.10 Implementation
The MLIB_Msu16SSS function is implemented as an inline function.
Example 3-57. Implementation Code
#include "mlib.h"
static Frac16 mf16Acc, mf16In1, mf16In2;
static Frac16 mf16Out;
void main(void)
{
mf16Acc = FRAC16(0.3);
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
Math Library, Rev. 0
3-242
Freescale Semiconductor
MLIB_Msu16SSS
/* mf16Out = mf16Acc - mf16In1*mf16In2 */
mf16Out = MLIB_Msu16SSS(mf16Acc, mf16In1, mf16In2);
}
3.57.11 See Also
See MLIB_Mac16SSS, MLIB_Mac16SSSSat, MLIB_Msu16SSSSat,
MLIB_Mac32LSS, MLIB_Mac32LSSSat, MLIB_Msu32LSS and
MLIB_Msu32LSSSat for more information.
3.57.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-115. Performance of the MLIB_Msu16SSS Function
Code Size (words)
4
Data Size (words)
0
Execution Clock
Min
14 cycles
Max
14 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-243
MLIB_Msu16SSSSat
3.58
MLIB_Msu16SSSSat
This function returns the 16-bit value of the product of two 16-bit fractional
inputs subtracted from the 16-bit fractional accumulator with saturation.
3.58.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Msu16SSSSat(Frac16 f16Acc, Frac16 f16In1, Frac16 f16In2)
3.58.2
Prototype
inline Frac16 MLIB_Msu16SSSSatFAsmi(register Frac16 f16Acc, register
Frac16 f16In1, register Frac16 f16In2)
3.58.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-116. Function Arguments
Name
3.58.4
In/Out
Format
Range
Description
f16Acc
In
SF16
0x8000...
0x7FFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.58.5
Dependencies
The dependent files are:
• MLIB_Mac16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-245
MLIB_Msu16SSSSat
3.58.6
Description
The MLIB_Msu16SSSSat returns the product of two fractional inputs
subtracted from the fractional accumulator. The function saturates the output if
necessary.
MLIB_Msu16SSSSat a b c = a – b c » 16
Eqn. 3-54
where:
• result - Frac16
• a - Frac16
• b - Frac16
• c - Frac16
3.58.7
Returns
This function returns the 16-bit value of multiple of two 16-bit fractional inputs
f16In1 and f16In2 subtracted from 16-bit fractional input f16Acc.
3.58.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.58.9
Special Issues
If the result is greater than 0x7FFF the output is 0x7FFF. If the result is smaller
than 0x8000 the output is 0x8000.
The function MLIB_Msu16SSSSat does not require the saturation mode to be
turned on.
3.58.10 Implementation
The MLIB_Msu16SSSSat function is implemented as an inline function.
Example 3-58. Implementation Code
#include "mlib.h"
static
static
static
static
Frac16
Frac16
Frac16
Frac16
mf16Acc;
mf16In1;
mf16In2;
mf16Out;
void main(void)
{
mf16Acc = 0;
mf16In1 = FRAC16(0.1);
Math Library, Rev. 0
3-246
Freescale Semiconductor
MLIB_Msu16SSSSat
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16Acc - mf16In1*mf16In2 */
mf16Out = MLIB_Msu16SSSSat(mf16Acc, mf16In1, mf16In2);
}
3.58.11 See Also
See MLIB_Mac16SSS, MLIB_Mac16SSSSat, MLIB_Msu16SSS,
MLIB_Mac32LSS, MLIB_Mac32LSSSat, MLIB_Msu32LSS and
MLIB_Msu32LSSSat for more information.
3.58.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-117. Performance of the MLIB_Msu16SSSSat Function
Code Size (words)
4
Data Size (words)
0
Execution Clock
Min
14 cycles
Max
14 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-247
MLIB_Mac32LSS
3.59
MLIB_Mac32LSS
This function returns the 32-bit sum of the 32-bit fractional accumulator and the
product of two 16-bit fractional inputs.
3.59.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Mac32LSS(Frac32 f32Acc, Frac16 f16In1, Frac16 f16In2)
3.59.2
Prototype
inline Frac32 MLIB_Mac32LSSFAsmi(register Frac32 f32Acc, register Frac16
f16In1, register Frac16 f16In2)
3.59.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-118. Function Arguments
Name
3.59.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.59.5
Dependencies
The dependent files are:
• MLIB_Mac16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-249
MLIB_Mac32LSS
3.59.6
Description
The MLIB_Mac32LSS returns the sum of the accumulator and the product of
two fractional inputs. The function does not saturate the output if the saturation
mode is turned off.
MLIB_Mac32LSS a b c = a + b c
Eqn. 3-55
where:
• result - Frac32
• a - Frac32
• b - Frac16
• c - Frac16
3.59.7
Returns
The function returns the 32-bit sum of 32-bit fractional accumulator (f32Acc) and
the product of two 16-bit fractional inputs f16In1 and f16In2.
3.59.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.59.9
Special Issues
If the result is greater than 0x7FFF FFFF or smaller than 0x8000 0000, the output
overflows if the saturation mode is turned off.
In case of desired saturation, the MLIB_Mac32LSS function requires the
saturation mode to be turned on or the MLIB_Mac32LSSSat has to be used
instead.
3.59.10 Implementation
The MLIB_Mac32LSS function is implemented as an inline function.
Example 3-59. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac16
Frac16
Frac32
mf32Acc;
mf16In1;
mf16In2;
mf32Out;
void main(void)
{
mf32Acc = FRAC32(0.3);
Math Library, Rev. 0
3-250
Freescale Semiconductor
MLIB_Mac32LSS
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32Acc + mf16In1*mf16In2 */
mf32Out = MLIB_Mac32LSS(mf32Acc, mf16In1, mf16In2);
}
3.59.11 See Also
See MLIB_Mac16SSS, MLIB_Mac16SSSSat, MLIB_Msu16SSS,
MLIB_Msu16SSSSat, MLIB_Mac32LSSSat, MLIB_Msu32LSS and
MLIB_Msu32LSSSat for more information.
3.59.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-119. Performance of the MLIB_Mac32LSS Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
12 cycles
Max
12 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-251
MLIB_Mac32LSSSat
3.60
MLIB_Mac32LSSSat
This function returns the 32-bit sum of the 32-bit fractional accumulator and the
product of two 16-bit fractional inputs.
3.60.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Mac32LSSSat(Frac32 f32Acc, Frac16 f16In1, Frac16 f16In2)
3.60.2
Prototype
inline Frac32 MLIB_Mac32LSSSatFAsmi(register Frac32 f32Acc, register
Frac16 f16In1, register Frac16 f16In2)
3.60.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-120. Function Arguments
Name
3.60.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.60.5
Dependencies
The dependent files are:
• MLIB_Mac16Asm.h
• MLIB_types.h
3.60.6
Description
The MLIB_Mac32LSSSat eturns the sum of the accumulator and the product of
two fractional inputs. The function saturates the output if necessary.
Math Library, Rev. 0
Freescale Semiconductor
3-253
MLIB_Mac32LSSSat
MLIB_Mac32LSSSat a b c = a + b c
Eqn. 3-56
where:
• result - Frac32
• a - Frac32
• b - Frac16
• c - Frac16
3.60.7
Returns
The function returns the 32-bit sum of 32-bit fractional accumulator (f32Acc) and
the product of two 16-bit fractional inputs f16In1 and f16In2.
3.60.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.60.9
Special Issues
If the result is greater than 0x7FFF FFFF the output is 0x7FFF FFFF. If the result
is smaller than 0x8000 0000 the output is 0x8000 0000.
The function MLIB_Mac32LSSSat does not require the saturation mode to be
turned on.
3.60.10 Implementation
The MLIB_Mac32LSSSat function is implemented as an inline function.
Example 3-60. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac16
Frac16
Frac32
mf32Acc;
mf16In1;
mf16In2;
mf32Out;
void main(void)
{
mf32Acc = FRAC32(0.3);
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32Acc + mf16In1*mf16In2 */
mf32Out = MLIB_Mac32LSSSat(mf32Acc, mf16In1, mf16In2);
}
Math Library, Rev. 0
3-254
Freescale Semiconductor
MLIB_Mac32LSSSat
3.60.11 See Also
See MLIB_Mac16SSS, MLIB_Mac16SSSSat, MLIB_Msu16SSS,
MLIB_Msu16SSSSat, MLIB_Mac32LSS, MLIB_Msu32LSS and
MLIB_Msu32LSSSat for more information.
3.60.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-121. Performance of the MLIB_Mac32LSSSat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
13 cycles
Max
13 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-255
MLIB_Msu32LSS
3.61
MLIB_Msu32LSS
This function returns the 32-bit product of two 16-bit fractional inputs subtracted
from the 32-bit fractional accumulator.
3.61.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Msu32LSS(Frac32 f32Acc, Frac16 f16In1, Frac16 f16In2)
3.61.2
Prototype
inline Frac32 MLIB_Msu32LSSFAsmi(register Frac32 f32Acc, register Frac16
f16In1, register Frac16 f16In2)
3.61.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-122. Function Arguments
Name
3.61.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.61.5
Dependencies
The dependent files are:
• MLIB_Mac16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-257
MLIB_Msu32LSS
3.61.6
Description
The MLIB_Msu32LSS returns the product of two fractional inputs subtracted
from the fractional accumulator. The function does not saturate the output if the
saturation mode is turned off.
MLIB_Msu32LSS a b c = a – b c
Eqn. 3-57
where:
• result - Frac16
• a - Frac16
• b - Frac16
• c - Frac16
3.61.7
Returns
This function returns the 32-bit product of two 16-bit fractional inputs (f16In1
and f16In2) subtracted from the 32-bit fractional accumulator (f32Acc).
3.61.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.61.9
Special Issues
If the result is greater than 0x7FFF FFFF or smaller than 0x8000 0000, the output
overflows if the saturation mode is turned off.
In case of desired saturation, the MLIB_Msu32LSS function requires the
saturation mode to be turned on or the MLIB_Msu32LSSSat has to be used
instead.
3.61.10 Implementation
The MLIB_Msu32LSS function is implemented as an inline function.
Example 3-61. Implementation Code
#include "mlib.h"
static Frac32 mf32Acc;
static Frac16 mf16In1, mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf32Acc = FRAC32(0.4);
mf16In1 = FRAC16(0.1);
Math Library, Rev. 0
3-258
Freescale Semiconductor
MLIB_Msu32LSS
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32Acc - mf16In1*mf16In2 */
mf32Out = MLIB_Msu32LSS(mf32Acc, mf16In1, mf16In2);
}
3.61.11 See Also
See MLIB_Mac16SSS, MLIB_Mac16SSSSat, MLIB_Msu16SSS,
MLIB_Msu16SSSSat, MLIB_Mac32LSS, MLIB_Mac32LSSSat and
MLIB_Msu32LSSSat for more information.
3.61.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-123. Performance of the MLIB_Msu32LSS Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
12 cycles
Max
12 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-259
MLIB_Msu32LSSSat
3.62
MLIB_Msu32LSSSat
This function returns the 32-bit product of two 16-bit fractional inputs subtracted
from the 32-bit fractional accumulator with saturation.
3.62.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Msu32LSSSat(Frac32 f32Acc, Frac16 f16In1, Frac16 f16In2)
3.62.2
Prototype
inline Frac32 MLIB_Msu32LSSSatFAsmi(register Frac32 f32Acc, register
Frac16 f16In1, register Frac16 f16In2)
3.62.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-124. Function Arguments
Name
3.62.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.62.5
Dependencies
The dependent files are:
• MLIB_Mac16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-261
MLIB_Msu32LSSSat
3.62.6
Description
The MLIB_Msu32LSSSat returns the product of two fractional inputs
subtracted from the fractional accumulator. The function saturates the output if
necessary.
MLIB_Msu32LSSSat a b c = a – b c
Eqn. 3-58
where:
• result - Frac16
• a - Frac16
• b - Frac16
• c - Frac16
3.62.7
Returns
This function returns the 32-bit product of two 16-bit fractional inputs (f16In1
and f16In2) subtracted from the 32-bit fractional accumulator (f32Acc).
3.62.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.62.9
Special Issues
If the result is greater than 0x7FFF FFFF the output is 0x7FFF FFFF. If the result
is smaller than 0x8000 0000 the output is 0x8000 0000.
The function MLIB_Msu32LSSSat does not require the saturation mode to be
turned on.
3.62.10 Implementation
The MLIB_Msu32LSSSat function is implemented as an inline function.
Example 3-62. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac16
Frac16
Frac32
mf32Acc;
mf16In1;
mf16In2;
mf32Out;
void main(void)
{
mf32Acc = FRAC32(0.4);
mf16In1 = FRAC16(0.1);
Math Library, Rev. 0
3-262
Freescale Semiconductor
MLIB_Msu32LSSSat
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32Acc - mf16In1*mf16In2 */
mf32Out = MLIB_Msu32LSSSat(mf32Acc, mf16In1, mf16In2);
}
3.62.11 See Also
See MLIB_Mac16SSS, MLIB_Mac16SSSSat, MLIB_Msu16SSS,
MLIB_Msu16SSSSat, MLIB_Mac32LSS, MLIB_Mac32LSSSat and
MLIB_Msu32LSS for more information.
3.62.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-125. Performance of the MLIB_Msu32LSSSat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
13 cycles
Max
13 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-263
MLIB_MacRnd16SSS
3.63
MLIB_MacRnd16SSS
This function returns the rounded 16-bit sum of the 16-bit fractional accumulator
and the product of two 16-bit fractional inputs.
3.63.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_MacRnd16SSS(Frac16 f16Acc, Frac16 f16In1, Frac16 f16In2)
3.63.2
Prototype
inline Frac16 MLIB_MacRnd16SSSFAsmi(register Frac16 f16Acc, register
Frac16 f16In1, register Frac16 f16In2)
3.63.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-126. Function Arguments
Name
3.63.4
In/Out
Format
Range
Description
f16Acc
In
SF16
0x8000...
0x7FFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.63.5
Dependencies
The dependent files are:
• MLIB_MacRnd16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-265
MLIB_MacRnd16SSS
3.63.6
Description
The MLIB_MacRnd16SSS returns the rounded sum of the accumulator and the
product of two fractional inputs. The result is rounded to the nearest. The function
does not saturate the output if the saturation mode is turned off.
MLIB_MacRnd16SSS a b c = round a + b c » 16
Eqn. 3-59
where:
• result - Frac16
• a - Frac16
• b - Frac16
• c - Frac16
3.63.7
Returns
The function returns the 16-bit rounded sum of a 16-bit fractional accumulator
(f16Acc) and the product of two 16-bit fractional inputs (f16In1 and f16In2).
3.63.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.63.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the result before rounding is 0x7FFF 8000 to 0x7FFF FFF the output is 0x8000
if the saturation mode is turned off.
If the result is greater than 0x7FFF or smaller than 0x8000, the output overflows
if the saturation mode is turned off.
In case of desired saturation, the MLIB_MacRnd16SSS function requires the
saturation mode to be turned on or the MLIB_MacRnd16SSSSat has to be used
instead.
3.63.10 Implementation
The MLIB_MacRnd16SSS function is implemented as an inline function.
Example 3-63. Implementation Code
#include "mlib.h"
static Frac16 mf16Acc;
Math Library, Rev. 0
3-266
Freescale Semiconductor
MLIB_MacRnd16SSS
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac16 mf16Out;
void main(void)
{
mf16Acc = FRAC16(0.5);
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16Acc + mf16In1*mf16In2, result is rounded*/
mf16Out = MLIB_MacRnd16SSS(mf16Acc, mf16In1, mf16In2);
}
3.63.11 See Also
See MLIB_MacRnd16SSSSat, MLIB_MsuRnd16SSS,
MLIB_MsuRnd16SSSSat, MLIB_MacRnd32LSS,
MLIB_MacRnd32LSSSat, MLIB_MsuRnd32LSS and
MLIB_MsuRnd32LSSSat for more information.
3.63.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-127. Performance of the MLIB_MacRnd16SSS Function
Code Size (words)
4
Data Size (words)
0
Execution Clock
Min
14 cycles
Max
14 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-267
MLIB_MacRnd16SSSSat
3.64
MLIB_MacRnd16SSSSat
This function returns the rounded 16-bit sum of the 16-bit fractional accumulator
and the product of two 16-bit fractional inputs with saturation.
3.64.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_MacRnd16SSSSat(Frac16 f16Acc, Frac16 f16In1, Frac16 f16In2)
3.64.2
Prototype
inline Frac16 MLIB_MacRnd16SSSSatFAsmi(register Frac16 f16Acc, register
Frac16 f16In1, register Frac16 f16In2)
3.64.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-128. Function Arguments
Name
3.64.4
In/Out
Format
Range
Description
f16Acc
In
SF16
0x8000...
0x7FFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.64.5
Dependencies
The dependent files are:
• MLIB_MacRnd16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-269
MLIB_MacRnd16SSSSat
3.64.6
Description
The MLIB_MacRnd16SSSSat function returns the rounded fractional product
of two fractional inputs. The result is rounded to the nearest. The function
saturates the output if necessary.
MLIB_MacRnd16SSSSat a b c = round a + b c » 16
Eqn. 3-60
where:
• result - Frac16
• a - Frac16
• b - Frac16
• c - Frac16
3.64.7
Returns
The function returns the 16-bit rounded sum of a 16-bit fractional accumulator
(f16Acc) and the product of two 16-bit fractional inputs (f16In1 and f16In2).
3.64.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.64.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the result is greater than 0x7FFF the output is 0x7FFF. If the result is smaller
than 0x8000 the output is 0x8000.
The function MLIB_MacRnd16SSSSat does not require the saturation mode to
be turned on.
3.64.10 Implementation
The MLIB_MacRnd16SSSSat function is implemented as an inline function.
Example 3-64. Implementation Code
#include "mlib.h"
static
static
static
static
Frac16
Frac16
Frac16
Frac16
mf16Acc;
mf16In1;
mf16In2;
mf16Out;
Math Library, Rev. 0
3-270
Freescale Semiconductor
MLIB_MacRnd16SSSSat
void main(void)
{
mf16Acc = FRAC16(0.4);
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16Acc + mf16In1*mf16In2, result is rounded*/
mf16Out = MLIB_MacRnd16SSSSat(mf16Acc, mf16In1, mf16In2);
}
3.64.11 See Also
See MLIB_MacRnd16SSS, MLIB_MsuRnd16SSS,
MLIB_MsuRnd16SSSSat, MLIB_MacRnd32LSS,
MLIB_MacRnd32LSSSat, MLIB_MsuRnd32LSS and
MLIB_MsuRnd32LSSSat for more information.
3.64.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-129. Performance of the MLIB_MacRnd16SSSSat Function
Code Size (words)
4
Data Size (words)
0
Execution Clock
Min
14 cycles
Max
14 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-271
MLIB_MsuRnd16SSS
3.65
MLIB_MsuRnd16SSS
This function returns the rounded 16-bit fractional product of two 16-bit
fractional inputs subtracted from the 16-bit fractional accumulator.
3.65.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_MsuRnd16SSS(Frac16 f16Acc, Frac16 f16In1, Frac16 f16In2)
3.65.2
Prototype
inline Frac16 MLIB_MsuRnd16SSSFAsmi(register Frac16 f16Acc, register
Frac16 f16In1, register Frac16 f16In2)
3.65.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-130. Function Arguments
Name
3.65.4
In/Out
Format
Range
Description
f16Acc
In
SF16
0x8000...
0x7FFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.65.5
Dependencies
The dependent files are:
• MLIB_MacRnd16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-273
MLIB_MsuRnd16SSS
3.65.6
Description
The MLIB_MsuRnd16SSS returns the rounded product of two fractional inputs
subtracted from the fractional accumulator. The result is rounded to the nearest.
The function does not saturate the output if the saturation mode is turned off.
MLIB_MsuRnd16SSS a b c = round a – b c » 16
Eqn. 3-61
where:
• result - Frac16
• a - Frac16
• b - Frac16
• c - Frac16
3.65.7
Returns
This function returns the rounded 16-bit product 16-bit fractional inputs (f16In1
and f16In2) subtracted from the 16-bit fractional accumulator (f16Acc).
3.65.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.65.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the result before rounding is 0x7FFF 8000 to 0x7FFF FFF the output is 0x8000
if the saturation mode is turned off.
If the result is greater than 0x7FFF or smaller than 0x8000, the output overflows
if the saturation mode is turned off.
In case of desired saturation, the MLIB_MsuRnd16SSS function requires the
saturation mode to be turned on or the MLIB_MsuRnd16SSSSat has to be used
instead.
3.65.10 Implementation
The MLIB_MsuRnd16SSS function is implemented as an inline function.
Example 3-65. Implementation Code
#include "mlib.h"
static Frac32 mf32Acc;
Math Library, Rev. 0
3-274
Freescale Semiconductor
MLIB_MsuRnd16SSS
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf16Acc = FRAC16(0.2);
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16Acc - mf16In1*mf16In2, result is rounded */
mf16Out = MLIB_MsuRnd16SSS(mf16Acc, mf16In1, mf16In2);
}
3.65.11 See Also
See MLIB_MacRnd16SSS, MLIB_MacRnd16SSSSat,
MLIB_MsuRnd16SSSSat, MLIB_MacRnd32LSS,
MLIB_MacRnd32LSSSat, MLIB_MsuRnd32LSS and
MLIB_MsuRnd32LSSSat for more information.
3.65.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-131. Performance of the MLIB_MsuRnd16SSS Function
Code Size (words)
4
Data Size (words)
0
Execution Clock
Min
14 cycles
Max
14 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-275
MLIB_MsuRnd16SSSSat
3.66
MLIB_MsuRnd16SSSSat
This function returns the rounded 16-bit fractional product of two 16-bit
fractional inputs subtracted from the 16-bit fractional accumulator with
saturation.
3.66.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_MsuRnd16SSSSat(Frac16 f16Acc, Frac16 f16In1, Frac16 f16In2)
3.66.2
Prototype
inline Frac16 MLIB_MsuRnd16SSSSatFAsmi(register Frac16 f16Acc, register
Frac16 f16In1, register Frac16 f16In2)
3.66.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-132. Function Arguments
Name
3.66.4
In/Out
Format
Range
Description
f16Acc
In
SF16
0x8000...
0x7FFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.66.5
Dependencies
The dependent files are:
• MLIB_MacRnd16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-277
MLIB_MsuRnd16SSSSat
3.66.6
Description
The MLIB_MsuRnd16SSSSat returns the rounded product of two fractional
inputs subtracted from the fractional accumulator. The result is rounded to the
nearest. The function saturates the output if necessary.
MLIB_MsuRnd16SSS a b c = round a – b c » 16
Eqn. 3-62
where:
• result - Frac16
• a - Frac16
• b - Frac16
• c - Frac16
3.66.7
Returns
This function returns the rounded 16-bit product 16-bit fractional inputs (f16In1
and f16In2) subtracted from the 16-bit fractional accumulator (f16Acc).
3.66.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.66.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the result is greater than 0x7FFF the output is 0x7FFF. If the result is smaller
than 0x8000 the output is 0x8000.
The function MLIB_MsuRnd16SSSSat does not require the saturation mode to
be turned on.
3.66.10 Implementation
The MLIB_MsuRnd16SSSSat function is implemented as an inline function.
Example 3-66. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac16
Frac16
Frac32
mf32Acc;
mf16In1;
mf16In2;
mf32Out;
Math Library, Rev. 0
3-278
Freescale Semiconductor
MLIB_MsuRnd16SSSSat
void main(void)
{
mf16Acc = FRAC16(0.1);
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf16Out = mf16Acc - mf16In1*mf16In2, result is rounded */
mf16Out = MLIB_MsuRnd16SSSSat(mf16Acc, mf16In1, mf16In2);
}
3.66.11 See Also
See MLIB_MacRnd16SSS, MLIB_MacRnd16SSSSat,
MLIB_MsuRnd16SSS, MLIB_MacRnd32LSS, MLIB_MacRnd32LSSSat,
MLIB_MsuRnd32LSS and MLIB_MsuRnd32LSSSat for more information.
3.66.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-133. Performance of the MLIB_MsuRnd16SSSSat Function
Code Size (words)
4
Data Size (words)
0
Execution Clock
Min
14 cycles
Max
14 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-279
MLIB_MacRnd32LSS
3.67
MLIB_MacRnd32LSS
This function returns the 32-bit sum rounded to the upper 16 bits of the 32-bit
fractional accumulator and the product of two 16-bit fractional inputs.
3.67.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_MacRnd32LSS(Frac32 f32Acc, Frac16 f16In1, Frac16 f16In2)
3.67.2
Prototype
inline Frac32 MLIB_MacRnd32LSSFAsmi(register Frac32 f32Acc, register
Frac16 f16In1, register Frac16 f16In2)
3.67.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-134. Function Arguments
Name
3.67.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.67.5
Dependencies
The dependent files are:
• MLIB_MacRnd16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-281
MLIB_MacRnd32LSS
3.67.6
Description
The MLIB_MacRnd32LSS returns the rounded sum of the accumulator and the
product of two fractional inputs. The result is rounded to the nearest. The function
does not saturate the output if the saturation mode is turned off.
a+bc
MLIB_MacRnd32LSS a b c = round ------------------- « 16
65536
Eqn. 3-63
where:
• result - Frac32
• a - Frac32
• b - Frac16
• c - Frac16
3.67.7
Returns
The function returns the 32-bit sum rounded to the upper 16 bits of the 32-bit
fractional accumulator (f32Acc) and the product of two 16-bit fractional inputs
(f16In1 and f16In2). The function does not saturate the output if the saturation
mode is turned off.
3.67.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.67.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the result before rounding is 0x7FFF 8000 to 0x7FFF FFF the output is 0x8000
0000 if the saturation mode is turned off.
If the result is greater than 0x7FFF FFFF or smaller than 0x8000 0000, the output
overflows if the saturation mode is turned off.
In case of desired saturation, the MLIB_MacRnd32LSS function requires the
saturation mode to be turned on or the MLIB_MacRnd32LSSSat has to be used
instead.
3.67.10 Implementation
The MLIB_MacRnd32LSS function is implemented as an inline function.
Math Library, Rev. 0
3-282
Freescale Semiconductor
MLIB_MacRnd32LSS
Example 3-67. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac16
Frac16
Frac32
mf32Acc;
mf16In1;
mf16In2;
mf32Out;
void main(void)
{
mf32Acc = FRAC32(0.6);
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32Acc + mf16In1*mf16In2, result is rounded*/
mf32Out = MLIB_MacRnd32LSS(mf32Acc, mf16In1, mf16In2);
}
3.67.11 See Also
See MLIB_MacRnd16SSS, MLIB_MacRnd16SSSSat,
MLIB_MsuRnd16SSS, MLIB_MsuRnd16SSSSat,
MLIB_MacRnd32LSSSat, MLIB_MsuRnd32LSS and
MLIB_MsuRnd32LSSSat for more information.
3.67.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-135. Performance of the MLIB_MacRnd32LSS Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
12 cycles
Max
12 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-283
MLIB_MacRnd32LSSSat
3.68
MLIB_MacRnd32LSSSat
This function returns the 32-bit sum rounded to the upper 16 bits of the 32-bit
fractional accumulator and the product of two 16-bit fractional inputs with
saturation.
3.68.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_MacRnd32LSSSat(Frac32 f32Acc, Frac16 f16In1, Frac16 f16In2)
3.68.2
Prototype
inline Frac32 MLIB_MacRnd32LSSSatFAsmi(register Frac32 f32Acc, register
Frac16 f16In1, register Frac16 f16In2)
3.68.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-136. Function Arguments
Name
3.68.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.68.5
Dependencies
The dependent files are:
• MLIB_MacRnd16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-285
MLIB_MacRnd32LSSSat
3.68.6
Description
The MLIB_MacRnd32LSSSat returns the rounded sum of the accumulator and
the product of two fractional inputs. The result is rounded to the nearest. The
function saturates the output if necessary.
a+bc
MLIB_MacRnd32LSSSat a b c = round ------------------- « 16
65536
Eqn. 3-64
where:
• result - Frac32
• a - Frac32
• b - Frac16
• c - Frac16
3.68.7
Returns
The function returns the 32-bit sum rounded to the upper 16 bits of the 32-bit
fractional accumulator (f32Acc) and the product of two 16-bit fractional inputs
(f16In1 and f16In2).
3.68.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.68.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the result before roudning is greater than 0x7FFF 8000 the output is 0x7FFF
FFFF. If the result is smaller than 0x8000 0000 the output is 0x8000 0000.
The function MLIB_MacRnd32LSSSat does not require the saturation mode to
be turned on.
3.68.10 Implementation
The MLIB_MacRnd32LSSSat function is implemented as an inline function.
Example 3-68. Implementation Code
#include "mlib.h"
static Frac32 mf32Acc;
static Frac16 mf16In1;
Math Library, Rev. 0
3-286
Freescale Semiconductor
MLIB_MacRnd32LSSSat
static Frac16 mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf32Acc = FRAC32(0.6);
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32Acc + mf16In1*mf16In2, result is rounded*/
mf32Out = MLIB_MacRnd32LSSSat(mf32Acc, mf16In1, mf16In2);
}
3.68.11 See Also
See MLIB_MacRnd16SSS, MLIB_MacRnd16SSSSat,
MLIB_MsuRnd16SSS, MLIB_MsuRnd16SSSSat, MLIB_MacRnd32LSS,
MLIB_MsuRnd32LSS and MLIB_MsuRnd32LSSSat for more information.
3.68.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-137. Performance of the MLIB_MacRnd32LSSSat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
13 cycles
Max
13 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-287
MLIB_MsuRnd32LSS
3.69
MLIB_MsuRnd32LSS
This function returns the rounded 32-bit value rounded to the upper 16 bits of the
product of two 16-bit fractional inputs subtracted from the 32-bit fractional
accumulator.
3.69.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_MsuRnd32LSS(Frac32 f32Acc, Frac16 f16In1, Frac16 f16In2)
3.69.2
Prototype
inline Frac32 MLIB_MsuRnd32LSSFAsmi(register Frac32 f32Acc, register
Frac16 f16In1, register Frac16 f16In2)
3.69.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-138. Function Arguments
Name
3.69.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.69.5
Dependencies
The dependent files are:
• MLIB_MacRnd16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-289
MLIB_MsuRnd32LSS
3.69.6
Description
The MLIB_MsuRnd32LSS returns the rounded product of two fractional inputs
subtracted from the fractional accumulator. The result is rounded to the nearest.
The function does not saturate the output if the saturation mode is turned off.
a–bc
MLIB_MsuRnd32LSS a b c = round ------------------- « 16
65536
Eqn. 3-65
where:
• result - Frac32
• a - Frac32
• b - Frac16
• c - Frac16
3.69.7
Returns
This function returns the 32-bit value rounded to the upper 16 bits of the product
of two 16-bit fractional inputs (f16In1 and f16In2) subtracted from the 32-bit
fractional accumulator (f32Acc).
3.69.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.69.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the result before rounding is 0x7FFF 8000 to 0x7FFF FFF the output is 0x8000
0000 if the saturation mode is turned off.
If the result is greater than 0x7FFF FFFF or smaller than 0x8000 0000, the output
overflows if the saturation mode is turned off.
In case of desired saturation, the MLIB_MsuRnd32LSS function requires the
saturation mode to be turned on or the MLIB_MsuRnd32LSSSat has to be used
instead.
3.69.10 Implementation
The MLIB_MsuRnd32LSS function is implemented as an inline function.
Math Library, Rev. 0
3-290
Freescale Semiconductor
MLIB_MsuRnd32LSS
Example 3-69. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac16
Frac16
Frac32
mf32Acc;
mf16In1;
mf16In2;
mf32Out;
void main(void)
{
mf32Acc = FRAC16(0.7);
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32Acc - mf16In1*mf16In2, result is rounded */
mf32Out = MLIB_MsuRnd32LSS(mf32Acc, mf16In1, mf16In2);
}
3.69.11 See Also
See MLIB_MacRnd16SSS, MLIB_MacRnd16SSSSat,
MLIB_MsuRnd16SSS, MLIB_MsuRnd16SSSSat, MLIB_MacRnd32LSS,
MLIB_MacRnd32LSSSat and MLIB_MsuRnd32LSSSat for more
information.
3.69.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-139. Performance of the MLIB_MsuRnd32LSS Function
Code Size (words)
2
Data Size (words)
0
Execution Clock
Min
12 cycles
Max
12 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-291
MLIB_MsuRnd32LSSSat
3.70
MLIB_MsuRnd32LSSSat
This function returns the rounded 32-bit value rounded to the upper 16 bits of the
product of two 16-bit fractional inputs subtracted from the 32-bit fractional
accumulator with saturation.
3.70.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_MsuRnd32LSSSat(Frac32 f32Acc, Frac16 f16In1, Frac16 f16In2)
3.70.2
Prototype
inline Frac32 MLIB_MsuRnd32LSSSatFAsmi(register Frac32 f32Acc, register
Frac16 f16In1, register Frac16 f16In2)
3.70.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-140. Function Arguments
Name
3.70.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f16In1
In
SF16
0x8000...
0x7FFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.70.5
Dependencies
The dependent files are:
• MLIB_MacRnd16Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-293
MLIB_MsuRnd32LSSSat
3.70.6
Description
The MLIB_MsuRnd32LSSSat returns the rounded value of multiple of two
fractional inputs substracted from the fractional value of accumulator. The result
is rounded to the nearest. The function saturates the output if necessary.
a–bc
MLIB_MsuRnd32LSSSat a b c = round ------------------- « 16
65536
Eqn. 3-66
where:
• result - Frac32
• a - Frac32
• b - Frac16
• c - Frac16
3.70.7
Returns
This function returns the 32-bit value rounded to the upper 16 bits of the product
of two 16-bit fractional inputs (f16In1 and f16In2) substracted from the 32-bit
fractional accumulator (f32Acc).
3.70.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.70.9
Special Issues
The rounding of the result where the lower 16 bits are equalt to exactly 0.5
(0x8000) depends on the rounding mode of the core (convergent or two’s
compelment.)
If the result before roudning is greater than 0x7FFF 8000 the output is 0x7FFF
FFFF. If the result is smaller than 0x8000 0000 the output is 0x8000 0000.
The function MLIB_MsuRnd32LSSSat does not require the saturation mode to
be turned on.
3.70.10 Implementation
The MLIB_MsuRnd32LSSSat function is implemented as an inline function.
Example 3-70. Implementation Code
#include "mlib.h"
static Frac32 mf32Acc;
Math Library, Rev. 0
3-294
Freescale Semiconductor
MLIB_MsuRnd32LSSSat
static Frac16 mf16In1;
static Frac16 mf16In2;
static Frac32 mf32Out;
void main(void)
{
mf32Acc = FRAC16(0.3);
mf16In1 = FRAC16(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32Acc - mf16In1*mf16In2, result is rounded */
mf32Out = MLIB_MsuRnd32LSSSat(mf32Acc, mf16In1, mf16In2);
}
3.70.11 See Also
See MLIB_MacRnd16SSS, MLIB_MacRnd16SSSSat,
MLIB_MsuRnd16SSS, MLIB_MsuRnd16SSSSat, MLIB_MacRnd32LSS,
MLIB_MacRnd32LSSSat and MLIB_MsuRnd32LSS for more information.
3.70.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-141. Performance of the MLIB_MsuRnd32LSSSat Function
Code Size (words)
3
Data Size (words)
0
Execution Clock
Min
13 cycles
Max
13 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-295
MLIB_Mac32LLS
3.71
MLIB_Mac32LLS
This function returns the 32-bit sum of the 32-bit fractional accumulator and the
upper 32 bits of the product of a 32-bit and a 16-bit fractional inputs.
3.71.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Mac32LLS(Frac32 f32Acc, Frac32 f32In1, Frac16 f16In2)
3.71.2
Prototype
inline Frac32 MLIB_Mac32LLSFAsmi(register Frac32 f32Acc, register Frac32
f32In1, register Frac16 f16In2)
V3 core version:
inline Frac32 MLIB_V3Mac32LLSFAsmi(register Frac32 f32Acc, register
Frac32 f32In1, register Frac16 f16In2)
3.71.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-142. Function Arguments
Name
3.71.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.71.5
Dependencies
The dependent files are:
• MLIB_Mac24Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-297
MLIB_Mac32LLS
3.71.6
Description
The MLIB_Mac32LLS returns the sum of the accumulator and the product of
two fractional inputs. The function does not saturate the output if the saturation
mode is turned off.
MLIB_Mac32LLS a b c = a + b c » 16
Eqn. 3-67
where:
• result - Frac32
• a - Frac32
• b - Frac32
• c - Frac16
3.71.7
Returns
The function returns the 32-bit sum of the 32-bit fractional accumulator (f32Acc)
and the upper 32 bits of the product of a 32-bit fractional input (f32In1) and 16-bit
fractional input (f16In2).
3.71.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.71.9
Special Issues
If the result is greater than 0x7FFF FFFF or smaller than 0x8000 0000, the output
overflows if the saturation mode is turned off.
In case of desired saturation, the MLIB_Mac32LLS function requires the
saturation mode to be turned on or the MLIB_Mac32LLSSat has to be used
instead.
3.71.10 Implementation
The MLIB_Mac32LLS function is implemented as an inline function.
Example 3-71. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac32
Frac16
Frac32
mf32Acc;
mf32In1;
mf16In2;
mf32Out;
void main(void)
{
Math Library, Rev. 0
3-298
Freescale Semiconductor
MLIB_Mac32LLS
mf32Acc = FRAC32(0.7);
mf32In1 = FRAC32(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32Acc + mf32In1*mf16In2 */
mf32Out = MLIB_Mac32LLS(mf32Acc, mf32In1, mf16In2);
}
3.71.11 See Also
See MLIB_Mac32LLSSat, MLIB_Msu32LLS, MLIB_Msu32LLSSat,
MLIB_Mac32LLL, MLIB_Mac32LLLSat, MLIB_Msu32LLL and
MLIB_Msu32LLLSat for more information.
3.71.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-143. Performance of the MLIB_Mac32LLS Function
Code Size (words)
V2: 6, V3: 3
Data Size (words)
0
Execution Clock
Min
V2: 16, V3: 12 cycles
Max
V2: 16, V3: 12 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-299
MLIB_Mac32LLSSat
3.72
MLIB_Mac32LLSSat
This function returns the 32-bit sum of the 32-bit fractional accumulator and the
upper 32 bits of the product of a 32-bit and a 16-bit fractional inputs with
saturation.
3.72.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Mac32LLSSat(Frac32 f32Acc, Frac32 f32In1, Frac16 f16In2)
3.72.2
Prototype
inline Frac32 MLIB_Mac32LLSSatFAsmi(register Frac32 f32Acc, register
Frac32 f32In1, register Frac16 f16In2)
V3 core version:
inline Frac32 MLIB_V3Mac32LLSSatFAsmi(register Frac32 f32Acc, register
Frac32 f32In1, register Frac16 f16In2)
3.72.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-144. Function Arguments
Name
3.72.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.72.5
Dependencies
The dependent files are:
• MLIB_Mac24Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-301
MLIB_Mac32LLSSat
3.72.6
Description
The MLIB_Mac32LLSSat returns the sum of the accumulator and the product
of two fractional inputs. The function saturates the output if necessary.
MLIB_Mac32LLSSat a b c = a + b c » 16
Eqn. 3-68
where:
• result - Frac32
• a - Frac32
• b - Frac32
• c - Frac16
3.72.7
Returns
The function returns the 32-bit sum of the 32-bit fractional accumulator (f32Acc)
and the upper 32 bits of the product of a 32-bit fractional input (f32In1) and 16-bit
fractional input (f16In2).
3.72.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.72.9
Special Issues
If the result is greater than 0x7FFF FFFF the output is 0x7FFF FFFF. If the result
is smaller than 0x8000 0000 the output is 0x8000 0000.
The function MLIB_Mac32LLSSat does not require the saturation mode to be
turned on.
3.72.10 Implementation
The MLIB_Mac32LLSSat function is implemented as an inline function.
Example 3-72. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac32
Frac16
Frac32
mf32Acc;
mf32In1;
mf16In2;
mf32Out;
void main(void)
{
mf32Acc = FRAC32(0.7);
mf32In1 = FRAC32(0.1);
Math Library, Rev. 0
3-302
Freescale Semiconductor
MLIB_Mac32LLSSat
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32Acc + mf32In1*mf16In2 */
mf32Out = MLIB_Mac32LLSSat(mf32Acc, mf32In1, mf16In2);
}
3.72.11 See Also
See MLIB_Mac32LLS, MLIB_Msu32LLS, MLIB_Msu32LLSSat,
MLIB_Mac32LLL, MLIB_Mac32LLLSat, MLIB_Msu32LLL and
MLIB_Msu32LLLSat for more information.
3.72.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-145. Performance of the MLIB_Mac32LLSSat Function
Code Size (words)
V2: 7, V3: 4
Data Size (words)
0
Execution Clock
Min
V2: 17, V3: 14 cycles
Max
V2: 17, V3: 14 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-303
MLIB_Msu32LLS
3.73
MLIB_Msu32LLS
This function returns the 32-bit value of the upper 32 bits of the product of a
32-bit and 16-bit fractional input subtracted from the 32-bit fractional
accumulator.
3.73.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Msu32LLS(Frac32 f32Acc, Frac32 f32In1, Frac16 f16In2)
3.73.2
Prototype
inline Frac32 MLIB_Msu32LLSFAsmi(register Frac32 f32Acc, register Frac32
f32In1, register Frac16 f16In2)
V3 core version:
inline Frac32 MLIB_V3Mac32LLSSatFAsmi(register Frac32 f32Acc, register
Frac32 f32In1, register Frac16 f16In2)
3.73.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-146. Function Arguments
Name
3.73.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.73.5
Dependencies
The dependent files are:
• MLIB_Mac24Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-305
MLIB_Msu32LLS
3.73.6
Description
The MLIB_Msu32LLS returns the product of two fractional inputs subtracted
from the fractional accumulator. The function does not saturate the output if the
saturation mode is turned off.
MLIB_Msu32LLS a b c = a – b c » 16
Eqn. 3-69
where:
• result - Frac32
• a - Frac32
• b - Frac32
• c - Frac16
3.73.7
Returns
This function returns the 32-bit value of the upper 32 bits of the product of a
32-bit fractional input (f32In1) and a 16-bit fractional input (f16In2) subtracted
from the 32-bit fractional accumulator (f32Acc).
3.73.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.73.9
Special Issues
If the result is greater than 0x7FFF FFFF or smaller than 0x8000 0000, the output
overflows if the saturation mode is turned off.
In case of desired saturation, the MLIB_Msu32LLS function requires the
saturation mode to be turned on or the MLIB_Msu32LLSSat has to be used
instead.
3.73.10 Implementation
The MLIB_Msu32LLS function is implemented as an inline function.
Example 3-73. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac32
Frac16
Frac32
mf32Acc;
mf32In1;
mf16In2;
mf32Out;
void main(void)
{
Math Library, Rev. 0
3-306
Freescale Semiconductor
MLIB_Msu32LLS
mf32Acc = FRAC32(0.8);
mf32In1 = FRAC32(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32Acc - mf32In1*mf16In2 */
mf32Out = MLIB_Msu32LLS(mf32Acc, mf32In1, mf16In2);
}
3.73.11 See Also
See MLIB_Mac32LLS, MLIB_Mac32LLSSat, MLIB_Msu32LLSSat,
MLIB_Mac32LLL, MLIB_Mac32LLLSat, MLIB_Msu32LLL and
MLIB_Msu32LLLSat for more information.
3.73.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-147. Performance of the MLIB_Msu32LLS Function
Code Size (words)
V2: 6, V3: 3
Data Size (words)
0
Execution Clock
Min
V2: 16, V3: 12 cycles
Max
V2: 16, V3: 12 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-307
MLIB_Msu32LLSSat
3.74
MLIB_Msu32LLSSat
This function returns the 32-bit value of the upper 32 bits of the product of a
32-bit and 16-bit fractional input subtracted from the 32-bit fractional
accumulator with saturation
3.74.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Msu32LLSSat(Frac32 f32Acc, Frac32 f32In1, Frac16 f16In2)
3.74.2
Prototype
inline Frac32 MLIB_Msu32LLSSatFAsmi(register Frac32 f32Acc, register
Frac32 f32In1, register Frac16 f16In2)
V3 core version:
inline Frac32 MLIB_V3Msu32LLSSatFAsmi(register Frac32 f32Acc, register
Frac32 f32In1, register Frac16 f16In2)
3.74.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-148. Function Arguments
Name
3.74.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f16In2
In
SF16
0x8000...
0x7FFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.74.5
Dependencies
The dependent files are:
• MLIB_Mac24Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-309
MLIB_Msu32LLSSat
3.74.6
Description
The MLIB_Msu32LLSSat returns the product of two fractional inputs
subtracted from the fractional accumulator. The function saturates the output if
necessary.
MLIB_Msu32LLSSat a b c = a – b c » 16
Eqn. 3-70
where:
• result - Frac32
• a - Frac32
• b - Frac32
• c - Frac16
3.74.7
Returns
This function returns the 32-bit value of the upper 32 bits of the product of a
32-bit fractional input (f32In1) and a 16-bit fractional input (f16In2) subtracted
from the 32-bit fractional accumulator (f32Acc).
3.74.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.74.9
Special Issues
If the result is greater than 0x7FFF FFFF the output is 0x7FFF FFFF. If the result
is smaller than 0x8000 0000 the output is 0x8000 0000.
The function MLIB_Msu32LLSSat does not require the saturation mode to be
turned on.
3.74.10 Implementation
The MLIB_Msu32LLSSat function is implemented as an inline function.
Example 3-74. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac32
Frac16
Frac32
mf32Acc;
mf32In1;
mf16In2;
mf32Out;
void main(void)
{
mf32Acc = FRAC32(0.6);
Math Library, Rev. 0
3-310
Freescale Semiconductor
MLIB_Msu32LLSSat
mf32In1 = FRAC32(0.1);
mf16In2 = FRAC16(-0.2);
/* mf32Out = mf32Acc - mf32In1*mf16In2 */
mf32Out = MLIB_Msu32LLSSat(mf32Acc, mf32In1, mf16In2);
}
3.74.11 See Also
See MLIB_Mac32LLS, MLIB_Mac32LLSSat, MLIB_Msu32LLS,
MLIB_Mac32LLL, MLIB_Mac32LLLSat, MLIB_Msu32LLL and
MLIB_Msu32LLLSat for more information.
3.74.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-149. Performance of the MLIB_Msu32LLSSat Function
Code Size (words)
V2: 7, V3: 4
Data Size (words)
0
Execution Clock
Min
V2: 17, V3: 13 cycles
Max
V2: 17, V3: 13 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-311
MLIB_Mac32LLL
3.75
MLIB_Mac32LLL
This function returns the 32-bit sum of the 32-bit fractional accumulator and the
upper 32 bits of the product of two 32-bit fractional inputs.
3.75.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Mac32LLL(Frac32 f32Acc, Frac32 f32In1, Frac32 f32In2)
3.75.2
Prototype
inline Frac32 MLIB_Mac32LLLFAsmi(register Frac32 f32Acc, register Frac32
f32In1, register Frac32 f32In2)
V3 core version:
inline Frac32 MLIB_V3Mac32LLLFAsmi(register Frac32 f32Acc, register
Frac32 f32In1, register Frac32 f32In2)
3.75.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-150. Function Arguments
Name
3.75.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f32In2
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.75.5
Dependencies
The dependent files are:
• MLIB_Mac32Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-313
MLIB_Mac32LLL
3.75.6
Description
The MLIB_Mac32LLL returns the sum of the accumulator and the product of
two fractional inputs. The function does not saturate the output if the saturation
mode is turned off.
MLIB_Mac32LLL a b c = a + b c » 32
Eqn. 3-71
where:
• result - Frac32
• a - Frac32
• b - Frac32
• c - Frac32
3.75.7
Returns
The function returns the 32-bit sum of the 32-bit fractional accumulator (f32Acc)
and the upper 32 bits of the product of two 32-bit fractional inputs (f32In1 and
f32In2).
3.75.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.75.9
Special Issues
If the result is greater than 0x7FFF FFFF or smaller than 0x8000 0000, the output
overflows if the saturation mode is turned off.
In case of desired saturation, the MLIB_Mac32LLL function requires the
saturation mode to be turned on or the MLIB_Mac32LLLSat has to be used
instead.
3.75.10 Implementation
The MLIB_Mac32LLL function is implemented as an inline function.
Example 3-75. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac32
Frac32
Frac32
mf32Acc;
mf32In1;
mf32In2;
mf32Out;
void main(void)
{
Math Library, Rev. 0
3-314
Freescale Semiconductor
MLIB_Mac32LLL
mf32Acc = FRAC32(0.5);
mf32In1 = FRAC32(0.1);
mf32In2 = FRAC32(-0.2);
/* mf32Out = mf32Acc + mf32In1*mf32In2 */
mf32Out = MLIB_Mac32LLL(mf32Acc, mf32In1, mf32In2);
}
3.75.11 See Also
See MLIB_Mac32LLS, MLIB_Mac32LLSSat, MLIB_Msu32LLS,
MLIB_Msu32LLSSat, MLIB_Mac32LLLSat, MLIB_Msu32LLL and
MLIB_Msu32LLLSat for more information.
3.75.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-151. Performance of the MLIB_Mac32LLL Function
Code Size (words)
V2: 10, V3: 2
Data Size (words)
0
Execution Clock
Min
V2: 20, V3: 12 cycles
Max
V2: 20, V3: 12 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-315
MLIB_Mac32LLLSat
3.76
MLIB_Mac32LLLSat
This function returns the 32-bit sum of the 32-bit fractional accumulator and the
upper 32 bits of the product of two 32-bit fractional inputs with saturation.
3.76.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Mac32LLLSat(Frac32 f32Acc, Frac32 f32In1, Frac32 f32In2)
3.76.2
Prototype
inline Frac32 MLIB_Mac32LLLSatFAsmi(register Frac32 f32Acc, register
Frac32 f32In1, register Frac32 f32In2)
V3 core version:
inline Frac32 MLIB_V3Mac32LLLSatFAsmi(register Frac32 f32Acc, register
Frac32 f32In1, register Frac32 f32In2)
3.76.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-152. Function Arguments
Name
3.76.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f32In2
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.76.5
Dependencies
The dependent files are:
• MLIB_Mac32Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-317
MLIB_Mac32LLLSat
3.76.6
Description
The MLIB_Mac32LLLSat returns the sum of the accumulator and the product
of two fractional inputs. The function saturates the output if necessary.
MLIB_Mac32LLLSat a b c = a + b c » 32
Eqn. 3-72
where:
• result - Frac32
• a - Frac32
• b - Frac32
• c - Frac32
3.76.7
Returns
The function returns the 32-bit sum of the 32-bit fractional accumulator (f32Acc)
and the upper 32 bits of the product of two 32-bit fractional inputs (f32In1 and
f32In2).
3.76.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.76.9
Special Issues
If the result is greater than 0x7FFF FFFF the output is 0x7FFF FFFF. If the result
is smaller than 0x8000 0000 the output is 0x8000 0000.
The function MLIB_Mac32LLLSat does not require the saturation mode to be
turned on.
3.76.10 Implementation
The MLIB_Mac32LLLSat function is implemented as an inline function.
Example 3-76. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac32
Frac32
Frac32
mf32Acc;
mf32In1;
mf32In2;
mf32Out;
void main(void)
{
mf32Acc = FRAC32(0.6);
mf32In1 = FRAC32(0.1);
Math Library, Rev. 0
3-318
Freescale Semiconductor
MLIB_Mac32LLLSat
mf32In2 = FRAC32(-0.2);
/* mf32Out = mf32Acc + mf32In1*mf32In2 */
mf32Out = MLIB_Mac32LLLSat(mf32Acc, mf32In1, mf32In2);
}
3.76.11 See Also
See MLIB_Mac32LLS, MLIB_Mac32LLSSat, MLIB_Msu32LLS,
MLIB_Msu32LLSSat, MLIB_Mac32LLL, MLIB_Msu32LLL and
MLIB_Msu32LLLSat for more information.
3.76.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-153. Performance of the MLIB_Mac32LLLSat Function
Code Size (words)
V2: 11, V3: 3
Data Size (words)
0
Execution Clock
Min
V2: 21, V3: 13 cycles
Max
V2: 21, V3: 13 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-319
MLIB_Msu32LLL
3.77
MLIB_Msu32LLL
This function returns the 32-bit value of the upper 32 bits of the product of two
32-bit fractional inputs subtracted from the 32-bit fractional accumulator.
3.77.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Msu32LLL(Frac32 f32Acc, Frac32 f32In1, Frac32 f32In2)
3.77.2
Prototype
inline Frac32 MLIB_Msu32LLLFAsmi(register Frac32 f32Acc, register Frac32
f32In1, register Frac32 f32In2)
V3 core version:
inline Frac32 MLIB_V3Msu32LLLFAsmi(register Frac32 f32Acc, register
Frac32 f32In1, register Frac32 f32In2)
3.77.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-154. Function Arguments
Name
3.77.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f32In2
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.77.5
Dependencies
The dependent files are:
• MLIB_Mac32Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-321
MLIB_Msu32LLL
3.77.6
Description
The MLIB_Msu32LLL returns the product of two fractional inputs subtracted
from the fractional accumulator. The function does not saturate the output if the
saturation mode is turned off.
MLIB_Msu32LLL a b c = a – b c » 32
Eqn. 3-73
where:
• result - Frac32
• a - Frac32
• b - Frac32
• c - Frac32
3.77.7
Returns
This function returns the 32-bit value of the upper 32 bits of the product of two
32-bit fractional inputs (f32In1 and f32In2) subtracted from the 32-bit fractional
accumulator (f32Acc).
3.77.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.77.9
Special Issues
If the result is greater than 0x7FFF FFFF or smaller than 0x8000 0000, the output
overflows if the saturation mode is turned off.
In case of desired saturation, the MLIB_Msu32LLL function requires the
saturation mode to be turned on or the MLIB_Msu32LLLSat has to be used
instead.
3.77.10 Implementation
The MLIB_Msu32LLL function is implemented as an inline function.
Example 3-77. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac32
Frac32
Frac32
mf32Acc;
mf32In1;
mf32In2;
mf32Out;
void main(void)
{
Math Library, Rev. 0
3-322
Freescale Semiconductor
MLIB_Msu32LLL
mf32Acc = FRAC32(0.6);
mf32In1 = FRAC32(0.1);
mf32In2 = FRAC32(-0.2);
/* mf32Out = mf32Acc - mf32In1*mf32In2 */
mf32Out = MLIB_Msu32LLL(mf32Acc, mf32In1, mf32In2);
}
3.77.11 See Also
See MLIB_Mac32LLS, MLIB_Mac32LLSSat, MLIB_Msu32LLS,
MLIB_Msu32LLSSat, MLIB_Mac32LLL, MLIB_Mac32LLLSat and
MLIB_Msu32LLLSat for more information.
3.77.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-155. Performance of the MLIB_Msu32LLL Function
Code Size (words)
V2: 10, V3: 2
Data Size (words)
0
Execution Clock
Min
V2: 20, V3: 12 cycles
Max
V2: 20, V3: 12 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-323
MLIB_Msu32LLLSat
3.78
MLIB_Msu32LLLSat
This function returns the 32-bit value of the upper 32 bits of the product of two
32-bit fractional inputs substracted from the 32-bit fractional accumulator with
saturation.
3.78.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Msu32LLLSat(Frac32 f32Acc, Frac32 f32In1, Frac32 f32In2)
3.78.2
Prototype
inline Frac32 MLIB_Msu32LLLSatFAsmi(register Frac32 f32Acc, register
Frac32 f32In1, register Frac32 f32In2)
V3 core version:
inline Frac32 MLIB_V3Msu32LLLSatFAsmi(register Frac32 f32Acc, register
Frac32 f32In1, register Frac32 f32In2)
3.78.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-156. Function Arguments
Name
3.78.4
In/Out
Format
Range
Description
f32Acc
In
SF32
0x8000 0000...
0x7FFF FFFF
accumulator value
f32In1
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
f32In2
In
SF32
0x8000 0000...
0x7FFF FFFF
input value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.78.5
Dependencies
The dependent files are:
• MLIB_Mac32Asm.h
• MLIB_types.h
Math Library, Rev. 0
Freescale Semiconductor
3-325
MLIB_Msu32LLLSat
3.78.6
Description
The MLIB_Msu32LLLSat returns the product of two fractional inputs
substracted from the fractional accumulator. The function saturates the output if
necessary.
MLIB_Msu32LLLSat a b c = a – b c » 32
Eqn. 3-74
where:
• result - Frac32
• a - Frac32
• b - Frac32
• c - Frac32
3.78.7
Returns
This function returns the 32-bit value of the upper 32 bits of the product of two
32-bit fractional inputs (f32In1 and f32In2) substracted from the 32-bit fractional
accumulator (f32Acc). The function saturates the output if necessary.
3.78.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-1, 1).
3.78.9
Special Issues
If the result is greater than 0x7FFF FFFF the output is 0x7FFF FFFF. If the result
is smaller than 0x8000 0000 the output is 0x8000 0000.
The function MLIB_Msu32LLLSat does not require the saturation mode to be
turned on.
3.78.10 Implementation
The MLIB_Msu32LLLSat function is implemented as an inline function.
Example 3-78. Implementation Code
#include "mlib.h"
static
static
static
static
Frac32
Frac32
Frac32
Frac32
mf32Acc;
mf32In1;
mf32In2;
mf32Out;
void main(void)
{
mf32Acc = FRAC32(-0.3);
Math Library, Rev. 0
3-326
Freescale Semiconductor
MLIB_Msu32LLLSat
mf32In1 = FRAC32(0.1);
mf32In2 = FRAC32(-0.2);
/* mf32Out = mf32Acc - mf32In1*mf32In2 */
mf32Out = MLIB_Msu32LLLSat(mf32Acc, mf32In1, mf32In2);
}
3.78.11 See Also
See MLIB_Mac32LLS, MLIB_Mac32LLSSat, MLIB_Msu32LLS,
MLIB_Msu32LLSSat, MLIB_Mac32LLL, MLIB_Mac32LLLSat and
MLIB_Msu32LLL for more information.
3.78.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-157. Performance of the MLIB_Msu32LLLSat Function
Code Size (words)
V2: 11, V3: 3
Data Size (words)
0
Execution Clock
Min
V2: 21, V3: 13 cycles
Max
V2: 21, V3: 13 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-327
MLIB_Div1Q16SS
3.79
MLIB_Div1Q16SS
This function performs the single-quadrant division of two 16-bit non-negative
fractional inputs with the 16-bit result.
3.79.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Div1Q16SS(Frac16 f16Num, Frac16 f16Denom)
3.79.2
Prototype
inline Frac16 MLIB_Div1Q16SSFAsmi(register Frac16 f16Num, register
Frac16 f16Denom)
3.79.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-158. Function Arguments
Name
3.79.4
In/Out
Format
Range
Description
f16Num
In
SF16
0x8000...
0x7FFF
numerator value
f16Denom
In
SF16
0x8000...
0x7FFF
denominator value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.79.5
Dependencies
The dependent files are:
• MLIB_DivAsm.h
• MLIB_types.h
3.79.6
Description
The MLIB_Div1Q16SS function returns the single quadrant division of two
non-negative fractional inputs. The function normalizes the inputs to get higher
precision of division.
Math Library, Rev. 0
Freescale Semiconductor
3-329
MLIB_Div1Q16SS
a « 16
MLIB_Div1Q16SS a b = --------------b
Eqn. 3-75
where:
• result - Frac16
• a - Frac16
• b - Frac16
3.79.7
Returns
The function divides a 16-bit non-negative fractional numerator (f16Num) by a
16-bit non-negative fractional denominator (f16Denom) with the 16-bit
fractional result.
3.79.8
Range Issues
The input data value is in the range of <0,1). The output data value is in the range
<0, 1).
3.79.9
Special Issues
If the denominator is equal to 0, the result is 0x7fff.
The function MLIB_Div1Q16SS does not require the saturation mode to be
turned on.
3.79.10 Implementation
The MLIB_Div1Q16SS function is implemented as an inline function.
Example 3-79. Implementation Code
#include "mlib.h"
static Frac16 mf16Num, mf16Denom, mf16Out;
void main(void)
{
mf16Num = FRAC16(0.1);
mf16Denom = FRAC16(0.2);
/* mf16Out = mf16Num/mf16Denom */
mf16Out = MLIB_Div1Q16SS(mf16Num, mf16Denom);
}
Math Library, Rev. 0
3-330
Freescale Semiconductor
MLIB_Div1Q16SS
3.79.11 See Also
See MLIB_Div4Q16SS, MLIB_Div1Q16LS, MLIB_Div4Q16LS,
MLIB_Div1Q32LS and MLIB_Div4Q32LS for more information.
3.79.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-159. Performance of the MLIB_Div1Q16SS Function
Code Size (words)
20
Data Size (words)
0
Execution Clock
Min
49 cycles
Max
49 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-331
MLIB_Div4Q16SS
3.80
MLIB_Div4Q16SS
This function performs the four-quadrant division of two 16-bit non-negative
fractional inputs with the 16-bit result.
3.80.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Div4Q16SS(Frac16 f16Num, Frac16 f16Denom)
3.80.2
Prototype
inline Frac16 MLIB_Div4Q16SSFAsmi(register Frac16 f16Num, register
Frac16 f16Denom)
3.80.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-160. Function Arguments
Name
3.80.4
In/Out
Format
Range
Description
f16Num
In
SF16
0x8000...
0x7FFF
numerator value
f16Denom
In
SF16
0x8000...
0x7FFF
denominator value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.80.5
Dependencies
The dependent files are:
• MLIB_DivAsm.h
• MLIB_types.h
3.80.6
Description
The MLIB_Div4Q16SS function returns the division of two fractional inputs.
The function normalizes the inputs to get higher precision of division.
Math Library, Rev. 0
Freescale Semiconductor
3-333
MLIB_Div4Q16SS
a « 16
MLIB_Div4Q16SS a b = --------------b
Eqn. 3-76
where:
• result - Frac16
• a - Frac16
• b - Frac16
3.80.7
Returns
The function divides a 16-bit fractional numerator (f16Num) by a 16-bit
fractional denominator (f16Denom) with the 16-bit fractional result.
3.80.8
Range Issues
The input data value is in the range of <1,1). The output data value is in the range
<1, 1).
3.80.9
Special Issues
If the denominator is equal to 0, the result is 0x7fff if the numerator is greater than
0, otherwise 0x8000.
The function MLIB_Div4Q16SS does not require the saturation mode to be
turned on.
3.80.10 Implementation
The MLIB_Div4Q16SS function is implemented as an inline function.
Example 3-80. Implementation Code
#include "mlib.h"
static Frac16 mf16Num;
static Frac16 mf16Denom;
static Frac16 mf16Out;
void main(void)
{
mf16Num = FRAC16(-0.1);
mf16Denom = FRAC16(0.2);
/* mf16Out = mf16Num/mf16Denom */
mf16Out = MLIB_Div4Q16SS(mf16Num, mf16Denom);
}
Math Library, Rev. 0
3-334
Freescale Semiconductor
MLIB_Div4Q16SS
3.80.11 See Also
See MLIB_Div1Q16SS, MLIB_Div1Q16LS, MLIB_Div4Q16LS,
MLIB_Div1Q32LS and MLIB_Div4Q32LS for more information.
3.80.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-161. Performance of the MLIB_Div4Q16SS Function
Code Size (words)
26
Data Size (words)
0
Execution Clock
Min
56 cycles
Max
56 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-335
MLIB_Div1Q16LS
3.81
MLIB_Div1Q16LS
This function performs the single-quadrant division of a 32-bit non-negative
fractional numerator and a 16-bit non-negative fractional denominator with the
16-bit fractional result.
3.81.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Div1Q16LS(Frac32 f32Num, Frac16 f16Denom)
3.81.2
Prototype
inline Frac16 MLIB_Div1Q16LSFAsmi(register Frac32 f32Num, register
Frac16 f16Denom)
3.81.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-162. Function Arguments
Name
3.81.4
In/Out
Format
Range
Description
f32Num
In
SF32
0x8000 0000...
0x7FFF FFFF
numerator value
f16Denom
In
SF16
0x8000...
0x7FFF
denominator value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.81.5
Dependencies
The dependent files are:
• MLIB_DivAsm.h
• MLIB_types.h
3.81.6
Description
The MLIB_Div1Q16LS function returns the single-quadrant division of two
non-negative fractional inputs. The function normalizes the inputs to get higher
precision of division.
Math Library, Rev. 0
Freescale Semiconductor
3-337
MLIB_Div1Q16LS
a
MLIB_Div1Q16LS a b = --b
Eqn. 3-77
where:
• result - Frac16
• a - Frac32
• b - Frac16
3.81.7
Returns
The function divides a 32-bit non-negative fractional numerator (f32Num) by a
16-bit non-negative fractional denominator (f16Denom) and returns the 16-bit
fractional result.
3.81.8
Range Issues
The input data value is in the range of <0,1). The output data value is in the range
<0, 1).
3.81.9
Special Issues
If the denominator is equal to 0, the result is 0x7fff.
The function MLIB_Div1Q16LS does not require the saturation mode to be
turned on.
3.81.10 Implementation
The MLIB_Div1Q16LS function is implemented as an inline function.
Example 3-81. Implementation Code
#include "mlib.h"
static Frac32 mf32Num;
static Frac16 mf16Denom, mf16Out;
void main(void)
{
mf32Num = FRAC32(0.1);
mf16Denom = FRAC16(0.2);
/* mf16Out = mf32Num/mf16Denom */
mf16Out = MLIB_Div1Q16LS(mf32Num, mf16Denom);
}
Math Library, Rev. 0
3-338
Freescale Semiconductor
MLIB_Div1Q16LS
3.81.11 See Also
See MLIB_Div1Q16SS, MLIB_Div4Q16SS, MLIB_Div4Q16LS,
MLIB_Div1Q32LS and MLIB_Div4Q32LS for more information.
3.81.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-163. Performance of the MLIB_Div1Q16LS Function
Code Size (words)
22
Data Size (words)
0
Execution Clock
Min
50 cycles
Max
50 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-339
MLIB_Div4Q16LS
3.82
MLIB_Div4Q16LS
This function performs the four-quadrant division of a 32-bit fractional
numerator and a 16-bit fractional denominator with the 16-bit fractional result.
3.82.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Div4Q16LS(Frac32 f32Num, Frac16 f16Denom)
3.82.2
Prototype
inline Frac16 MLIB_Div1Q16LSFAsmi(register Frac32 f32Num, register
Frac16 f16Denom)
3.82.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-164. Function Arguments
Name
3.82.4
In/Out
Format
Range
Description
f32Num
In
SF32
0x8000 0000...
0x7FFF FFFF
numerator value
f16Denom
In
SF16
0x8000...
0x7FFF
denominator value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.82.5
Dependencies
The dependent files are:
• MLIB_DivAsm.h
• MLIB_types.h
3.82.6
Description
The MLIB_Div4Q16LS function returns the four-quadrant division of two
non-negative fractional inputs. The function normalizes the inputs to get higher
precision of division.
Math Library, Rev. 0
Freescale Semiconductor
3-341
MLIB_Div4Q16LS
a
MLIB_Div4Q16LS a b = --b
Eqn. 3-78
where:
• result - Frac16
• a - Frac32
• b - Frac16
3.82.7
Returns
The function divides a 32-bit fractional numerator (f32Num) by a 16-bit
fractional denominator (f16Denom) and returns the 16-bit fractional result.
3.82.8
Range Issues
The input data value is in the range of <1,1). The output data value is in the range
<1, 1).
3.82.9
Special Issues
If the denominator is equal to 0, the result is 0x7fff if the numerator is greater than
0, otherwise 0x8000.
The function MLIB_Div4Q16LS does not require the saturation mode to be
turned on.
3.82.10 Implementation
The MLIB_Div4Q16LS function is implemented as an inline function.
Example 3-82. Implementation Code
#include "mlib.h"
static Frac32 mf32Num;
static Frac16 mf16Denom, mf16Out;
void main(void)
{
mf32Num = FRAC32(0.1);
mf16Denom = FRAC16(0.2);
/* mf16Out = mf32Num/mf16Denom */
mf16Out = MLIB_Div4Q16LS(mf32Num, mf16Denom);
}
Math Library, Rev. 0
3-342
Freescale Semiconductor
MLIB_Div4Q16LS
3.82.11 See Also
See MLIB_Div1Q16SS, MLIB_Div4Q16SS, MLIB_Div1Q16LS,
MLIB_Div1Q32LS and MLIB_Div4Q32LS for more information.
3.82.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-165. Performance of the MLIB_Div4Q16LS Function
Code Size (words)
28
Data Size (words)
0
Execution Clock
Min
56 cycles
Max
56 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-343
MLIB_Div1Q32LS
3.83
MLIB_Div1Q32LS
This function performs the single-quadrant division of 32-bit non-negative
fractional numerator and a 16-bit non-negative fractional denominator with the
32-bit fractional result.
3.83.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Div1Q32LS(Frac32 f32Num, Frac16 f16Denom)
3.83.2
Prototype
inline Frac16 MLIB_Div1Q16LSFAsmi(register Frac32 f32Num, register
Frac16 f16Denom)
3.83.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-166. Function Arguments
Name
3.83.4
In/Out
Format
Range
Description
f32Num
In
SF32
0x8000 0000...
0x7FFF FFFF
numerator value
f16Denom
In
SF16
0x8000...
0x7FFF
denominator value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.83.5
Dependencies
The dependent files are:
• MLIB_DivAsm.h
• MLIB_types.h
3.83.6
Description
The MLIB_Div1Q32LS function returns the single quadrant division of two
non-negative fractional inputs. The function normalizes the inputs to get higher
precision of division.
Math Library, Rev. 0
Freescale Semiconductor
3-345
MLIB_Div1Q32LS
a
MLIB_Div1Q32LS a b = --- « 16
b
Eqn. 3-79
where:
• result - Frac32
• a - Frac32
• b - Frac16
3.83.7
Returns
The function divides a 32-bit non-negative fractional numerator (f32Num) by a
16-bit non-negative fractional denominator (f16Denom) and returns the 32-bit
fractional result.
3.83.8
Range Issues
The input data value is in the range of <0,1). The output data value is in the range
<0, 1).
3.83.9
Special Issues
If the denominator is equal to 0, the result is 0x7fff ffff.
The function MLIB_Div1Q32LS does not require the saturation mode to be
turned on.
3.83.10 Implementation
The MLIB_Div1Q32LS function is implemented as an inline function.
Example 3-83. Implementation Code
#include "mlib.h"
static Frac32 mf32Num, mf32Out;
static Frac16 mf16Denom;
void main(void)
{
mf32Num = FRAC32(0.1);
mf16Denom = FRAC16(0.2);
/* mf32Out = mf32Num/mf16Denom */
mf32Out = MLIB_Div1Q32LS(mf32Num, mf16Denom);
}
Math Library, Rev. 0
3-346
Freescale Semiconductor
MLIB_Div1Q32LS
3.83.11 See Also
See MLIB_Div1Q16SS, MLIB_Div4Q16SS, MLIB_Div1Q16LS,
MLIB_Div1Q16LS and MLIB_Div4Q32LS for more information.
3.83.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-167. Performance of the MLIB_Div1Q32LS Function
Code Size (words)
26
Data Size (words)
0
Execution Clock
Min
70 cycles
Max
70 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-347
MLIB_Div4Q32LS
3.84
MLIB_Div4Q32LS
This function performs the four-quadrant division of 32-bit fractional numerator
and a 16-bit fractional denominator with the 32-bit fractional result.
3.84.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Div4Q32LS(Frac32 f32Num, Frac16 f16Denom)
3.84.2
Prototype
inline Frac16 MLIB_Div1Q16LSFAsmi(register Frac32 f32Num, register
Frac16 f16Denom)
3.84.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-168. Function Arguments
Name
3.84.4
In/Out
Format
Range
Description
f32Num
In
SF32
0x8000 0000...
0x7FFF FFFF
numerator value
f16Denom
In
SF16
0x8000...
0x7FFF
denominator value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.84.5
Dependencies
The dependent files are:
• MLIB_DivAsm.h
• MLIB_types.h
3.84.6
Description
The MLIB_Div4Q32LS function returns the four-quadrant division of two
fractional inputs. The function normalizes the inputs to get higher precision of
division.
Math Library, Rev. 0
Freescale Semiconductor
3-349
MLIB_Div4Q32LS
a
MLIB_Div4Q32LS a b = --- « 16
b
Eqn. 3-80
where:
• result - Frac32
• a - Frac32
• b - Frac16
3.84.7
Returns
The function divides a 32-bit fractional numerator (f32Num) by a 16-bit
non-negative fractional denominator (f16Denom) and returns the 32-bit
fractional result.
3.84.8
Range Issues
The input data value is in the range of <1,1). The output data value is in the range
<1, 1).
3.84.9
Special Issues
If the denominator is equal to 0, the result is 0x7fff ffff if the numerator is greater
than 0, otherwise 0x8000 0000.
The function MLIB_Div4Q32LS does not require the saturation mode to be
turned on.
3.84.10 Implementation
The MLIB_Div4Q32LS function is implemented as an inline function.
Example 3-84. Implementation Code
#include "mlib.h"
static Frac32 mf32Num, mf32Out;
static Frac16 mf16Denom;
void main(void)
{
mf32Num = FRAC32(0.1);
mf16Denom = FRAC16(0.2);
/* mf32Out = mf32Num/mf16Denom */
mf32Out = MLIB_Div4Q32LS(mf32Num, mf16Denom);
}
Math Library, Rev. 0
3-350
Freescale Semiconductor
MLIB_Div4Q32LS
3.84.11 See Also
See MLIB_Div1Q16SS, MLIB_Div4Q16SS, MLIB_Div1Q16LS,
MLIB_Div1Q16LS and MLIB_Div1Q32LS for more information.
3.84.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-169. Performance of the MLIB_Div4Q32LS Function
Code Size (words)
32
Data Size (words)
0
Execution Clock
Min
77 cycles
Max
77 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-351
MLIB_Rcp161Q
3.85
MLIB_Rcp161Q
This function calculates the the single-quadrant 16-bit precision reciprocal
function of the 16-bit fractional input with the 32-bit result.
3.85.1
Synopsis
#include ”mlib.h”
Frac16 MLIB_Rcp161Q(Frac16 f16Denom)
3.85.2
Prototype
inline Frac16 MLIB_Rcp161QFAsmi(register Frac16 f16Denom)
3.85.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-170. Function Arguments
Name
f16Denom
3.85.4
In/Out
In
Format
SF16
Range
0x0...
0x7FFF
Description
denominator value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.85.5
Dependencies
The dependent files are:
• MLIB_DivAsm.h
• MLIB_types.h
3.85.6
Description
The MLIB_Rcp161Q function returns the single-quadrant division of fractional
1 by a non-negative fractional input. The function normalizes the inputs to get
higher precision of division. The 16-bit precision division is performed.
1
MLIB_Rcp161Q a = --a
Eqn. 3-81
Math Library, Rev. 0
Freescale Semiconductor
3-353
MLIB_Rcp161Q
where:
• result - Frac32
— upper 16 bits: Word16 (scale: 0x4000 to 32768)
— lower 16 bits: Frac16 (scale: 0x8000 corresponds to 1)
• a - Frac16
3.85.7
Returns
The function divides the fractional 1 by a 16-bit non-negative fractional
denominator (f16Denom) with the 32-bit result.
3.85.8
Range Issues
The input data value is in the range of <0,1). The output data value is in the range
<0, 32768).
3.85.9
Special Issues
If the denominator is equal to 0, the result is 0x7fff ffff.
The result precision is below 3 LSB if the input is 4096 and greater. If higher
precision for lower inputs is required, the MLIB_Rcp321Q function has to be
used.
The function MLIB_Rcp161Q does not require the saturation mode to be turned
on.
3.85.10 Implementation
The MLIB_Rcp161Q function is implemented as an inline function.
Example 3-85. Implementation Code
#include "mlib.h"
static Frac16 mf16Denom;
static Frac32 mf132Out;
void main(void)
{
mf16Denom = FRAC16(0.2);
/* mf32Out = 1 / mf16Denom */
mf32Out = MLIB_Rcp161Q(mf16Denom);
}
Math Library, Rev. 0
3-354
Freescale Semiconductor
MLIB_Rcp161Q
3.85.11 See Also
See MLIB_Rcp164Q, MLIB_Rcp321Q and MLIB_Rcp324Q for more
information.
3.85.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-171. Performance of the MLIB_Rcp161Q Function
Code Size (words)
18
Data Size (words)
0
Execution Clock
Min
44 cycles
Max
44 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-355
MLIB_Rcp164Q
3.86
MLIB_Rcp164Q
This function calculates the the four-quadrant 16-bit precision reciprocal function
of the 16-bit fractional input with the 32-bit result.
3.86.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Rcp164Q(Frac16 f16Denom)
3.86.2
Prototype
inline Frac32 MLIB_Rcp164QFAsmi(register Frac16 f16Denom)
3.86.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-172. Function Arguments
Name
f16Denom
3.86.4
In/Out
In
Format
SF16
Range
0x8000...
0x7FFF
Description
denominator value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.86.5
Dependencies
The dependent files are:
• MLIB_DivAsm.h
• MLIB_types.h
3.86.6
Description
The MLIB_Rcp164Q function returns the four-quadrant division of fractional 1
by a fractional input. The function normalizes the inputs to get higher precision
of division. The 16-bit precision division is performed.
1
MLIB_Rcp321Q a = --a
Eqn. 3-82
Math Library, Rev. 0
Freescale Semiconductor
3-357
MLIB_Rcp164Q
where:
• result - Frac32
— upper 16 bits: Word16 (scale: 0x4000 to 32768)
— lower 16 bits: Frac16 (scale: 0x8000 corresponds to 1)
• a - Frac16
3.86.7
Returns
The function divides the fractional 1 by a 16-bit non-negative fractional
denominator (f16Denom) with the 32-bit result.
3.86.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-32768, 32768).
3.86.9
Special Issues
If the denominator is equal to 0, the result is 0x7fff ffff.
The result precision is below 3 LSB if the input is 4096 and greater. If higher
precision for lower inputs is required, the MLIB_Rcp324Q function has to be
used.
The function MLIB_Rcp164Q does not require the saturation mode to be turned
on.
3.86.10 Implementation
The MLIB_Rcp164Q function is implemented as an inline function.
Example 3-86. Implementation Code
#include "mlib.h"
static Frac16 mf16Denom;
static Frac32 mf132Out;
void main(void)
{
mf16Denom = FRAC16(0.2);
/* mf32Out = 1 / mf16Denom */
mf32Out = MLIB_Rcp164Q(mf16Denom);
}
Math Library, Rev. 0
3-358
Freescale Semiconductor
MLIB_Rcp164Q
3.86.11 See Also
See MLIB_Rcp161Q, MLIB_Rcp321Q and MLIB_Rcp324Q for more
information.
3.86.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-173. Performance of the MLIB_Rcp164Q Function
Code Size (words)
22
Data Size (words)
0
Execution Clock
Min
45 cycles
Max
45 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-359
MLIB_Rcp321Q
3.87
MLIB_Rcp321Q
This function calculates the the single-quadrant 32-bit precision reciprocal
function of the 16-bit fractional input with the 32-bit result.
3.87.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Rcp321Q(Frac16 f16Denom)
3.87.2
Prototype
inline Frac32 MLIB_Rcp321QFAsmi(register Frac16 f16Denom)
3.87.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-174. Function Arguments
Name
f16Denom
3.87.4
In/Out
In
Format
SF16
Range
0x0...
0x7FFF
Description
denominator value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.87.5
Dependencies
The dependent files are:
• MLIB_DivAsm.h
• MLIB_types.h
3.87.6
Description
The MLIB_Rcp321Q function returns the single-quadrant division of fractional
1 by a non-negative fractional input. The function normalizes the inputs to get
higher precision of division. The 32-bit precision division is performed.
1
MLIB_Rcp161Q a = --a
Eqn. 3-83
Math Library, Rev. 0
Freescale Semiconductor
3-361
MLIB_Rcp321Q
where:
• result - Frac32
— upper 16 bits: Word16 (scale: 0x4000 to 32768)
— lower 16 bits: Frac16 (scale: 0x8000 corresponds to 1)
• a - Frac16
3.87.7
Returns
The function divides the fractional 1 by a 16-bit non-negative fractional
denominator (f16Denom) with the 32-bit result.
3.87.8
Range Issues
The input data value is in the range of <0,1). The output data value is in the range
<0, 32768).
3.87.9
Special Issues
If the denominator is equal to 0, the result is 0x7fff ffff.
The function MLIB_Rcp321Q does not require the saturation mode to be turned
on.
3.87.10 Implementation
The MLIB_Rcp321Q function is implemented as an inline function.
Example 3-87. Implementation Code
#include "mlib.h"
static Frac16 mf16Denom;
static Frac32 mf132Out;
void main(void)
{
mf16Denom = FRAC16(0.2);
/* mf32Out = 1 / mf16Denom */
mf32Out = MLIB_Rcp321Q(mf16Denom);
}
3.87.11 See Also
See MLIB_Rcp161Q, MLIB_Rcp164Q and MLIB_Rcp324Q for more
information.
Math Library, Rev. 0
3-362
Freescale Semiconductor
MLIB_Rcp321Q
3.87.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-175. Performance of the MLIB_Rcp321Q Function
Code Size (words)
23
Data Size (words)
0
Execution Clock
Min
62 cycles
Max
62 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-363
MLIB_Rcp324Q
3.88
MLIB_Rcp324Q
This function calculates the the four-quadrant 32-bit precision reciprocal function
of the 16-bit fractional input with the 32-bit result.
3.88.1
Synopsis
#include ”mlib.h”
Frac32 MLIB_Rcp324Q(Frac16 f16Denom)
3.88.2
Prototype
inline Frac32 MLIB_Rcp324QFAsmi(register Frac16 f16Denom)
3.88.3
Arguments
This subsection describes the input/output arguments to a function or a macro. It
explains the algorithms being used by the functions or macro.
Table 3-176. Function Arguments
Name
f16Denom
3.88.4
In/Out
In
Format
SF16
Range
0x8000...
0x7FFF
Description
denominator value
Availability
This library module is available in the C-callable interface assembly version
format.
This library module is targeted for the 56800E and 56800Ex platforms.
3.88.5
Dependencies
The dependent files are:
• MLIB_DivAsm.h
• MLIB_types.h
3.88.6
Description
The MLIB_Rcp324Q function returns the four-quadrant division of fractional 1
by a fractional input. The function normalizes the inputs to get higher precision
of division. The 32-bit precision division is performed.
1
MLIB_Rcp324Q a = --a
Eqn. 3-84
Math Library, Rev. 0
Freescale Semiconductor
3-365
MLIB_Rcp324Q
where:
• result - Frac32
— upper 16 bits: Word16 (scale: 0x4000 to 32768)
— lower 16 bits: Frac16 (scale: 0x8000 corresponds to 1)
• a - Frac16
3.88.7
Returns
The function divides the fractional 1 by a 16-bit fractional denominator
(f16Denom) with the 32-bit result.
3.88.8
Range Issues
The input data value is in the range of <-1,1). The output data value is in the range
<-32768, 32768).
3.88.9
Special Issues
If the denominator is equal to 0, the result is 0x7fff ffff.
The function MLIB_Rcp324Q does not require the saturation mode to be turned
on.
3.88.10 Implementation
The MLIB_Rcp324Q function is implemented as an inline function.
Example 3-88. Implementation Code
#include "mlib.h"
static Frac16 mf16Denom;
static Frac32 mf132Out;
void main(void)
{
mf16Denom = FRAC16(0.2);
/* mf32Out = 1 / mf16Denom */
mf32Out = MLIB_Rcp324Q(mf16Denom);
}
3.88.11 See Also
See MLIB_Rcp161Q, MLIB_Rcp164Q and MLIB_Rcp321Q for more
information.
Math Library, Rev. 0
3-366
Freescale Semiconductor
MLIB_Rcp324Q
3.88.12 Performance
This section specifies actual requirements of the function or macro in terms of
required code memory, data memory, and number of clock cycles to execute.
Table 3-177. Performance of the MLIB_Rcp324Q Function
Code Size (words)
27
Data Size (words)
0
Execution Clock
Min
66 cycles
Max
66 cycles
Math Library, Rev. 0
Freescale Semiconductor
3-367
Appendix A Revision History
Table 0-1. Revision history
Revision number
Date
0
02/2014
Subsequent changes
Initial release
Advanced Control Library, Rev. 0
Freescale Semiconductor
A-1
How to Reach Us:
Information in this document is provided solely to enable system and software
Home Page:
freescale.com
implementers to use Freescale products. There are no express or implied copyright
Web Support:
freescale.com/support
information in this document.
licenses granted hereunder to design or fabricate any integrated circuits based on the
Freescale reserves the right to make changes without further notice to any products
herein. Freescale makes no warranty, representation, or guarantee regarding the
suitability of its products for any particular purpose, nor does Freescale assume any
liability arising out of the application or use of any product or circuit, and specifically
disclaims any and all liability, including without limitation consequential or incidental
damages. “Typical” parameters that may be provided in Freescale data sheets and/or
specifications can and do vary in different applications, and actual performance may
vary over time. All operating parameters, including “typicals,” must be validated for
each customer application by customer’s technical experts. Freescale does not convey
any license under its patent rights nor the rights of others. Freescale sells products
pursuant to standard terms and conditions of sale, which can be found at the following
address: freescale.com/SalesTermsandConditions.
Freescale, the Freescale logo, and CodeWarrior are trademarks of Freescale
Semiconductor, Inc., Reg. U.S. Pat. & Tm. Off. All other product or service names are
the property of their respective owners.
© 2014 Freescale Semiconductor, Inc.
56800Ex_MLIB
Rev. 0, 02/2014