Cx51 Compiler

Optimizing C Compiler and Library Reference

for Classic and Extended 8051 Microcontrollers

User’s Guide 11.2000

2

Information in this document is subject to change without notice and does not
represent a commitment on the part of the manufacturer. The software described
in this document is furnished under license agreement or nondisclosure
agreement and may be used or copied only in accordance with the terms of the
agreement. It is against the law to copy the software on any medium except as
specifically allowed in the license or nondisclosure agreement. The purchaser
may make one copy of the software for backup purposes. No part of this manual
may be reproduced or transmitted in any form or by any means, electronic or
mechanical, including photocopying, recording, or through information storage
and retrieval systems, for any purpose other than for the purchaser’s personal
use, without written permission.

© Copyright 1988-2000 Keil Elektronik GmbH. and Keil Software, Inc.

All rights reserved.

Keil C51™, Keil CX51™, and µVision2 are a trademarks of Keil Elektronik
GmbH.
Microsoft® and Windows™ are trademarks or registered trademarks of
Microsoft Corporation.
IBM®, PC®, and PS/2® are registered trademarks of International Business
Machines Corporation.
Intel®, MCS® 51, MCS® 251, ASM-51®, and PL/M-51® are registered
trademarks of Intel Corporation.

Every effort was made to ensure accuracy in this manual and to give appropriate
credit to persons, companies, and trademarks referenced herein.
Keil Software — Cx51 Compiler User’s Guide 3

Preface
This manual describes how to use the Cx51 Optimizing C Compilers to compile
C programs for your target 8051 environment. The Cx51 Compiler package can
be used on all 8051 family processors and is executable under the Windows 32-
Bit command line prompt. This manual assumes that you are familiar with the
Windows operating system, know how to program 8051 processors, and have a
working knowledge of the C language.

NOTE
This manual uses the term Windows to refer to the 32-bit Windows Versions
Windows 95, Windows 98, Windows NT, and Windows 2000.

If you have questions about programming in C, or if you would like more

information about the C programming language, refer to “Books About the C
Language” on page 18.

Many of the examples and descriptions in this manual discuss invoking the
compiler from the Windows command prompt. While this may not be applicable
to you if you are running Cx51 within an integrated development environment
like µVision2, examples in this manual are universal in that they apply to all
programming environments.
4 Contents

Manual Organization
This user’s guide is divided into eight chapters and six appendices:
“Chapter 1. Introduction,” describes the Cx51 compiler.

“Chapter 2. Compiling with Cx51,” explains how to compile a source file using
the Cx51 cross compiler. This chapter describes the command-line directives
that control file processing, compiling, and output.

“Chapter 3. Language Extensions,” describes the C language extensions

required to support the 8051 system architecture. This chapter provides a
detailed list of commands, functions, and controls not found in ANSI C
compilers.

“Chapter 4. Preprocessor,” describes the components of the Cx51 preprocessor

and includes examples.

“Chapter 5. 8051 Derivatives,” describes the 8051 family derivatives supported

by the Cx51 compiler. This chapter also includes tips for improving target
performance.

“Chapter 6. Advanced Programming Techniques,” lists important information

for the experienced developer. This chapter includes customization file
descriptions, as well as optimizer and segment names. This chapter also
discusses how to interface Cx51 with other 8051 programming languages.

“Chapter 7. Error Messages,” lists fatal errors, syntax errors, and warnings you
may encounter while using Cx51.

“Chapter 8. Library Reference,” provides you with an extensive Cx51 library

routine reference material. The library routines are listed by category and
include file. An alphabetical reference section, which includes example code for
each of the library routines, concludes this chapter.

The Appendix includes information on the differences between compiler

versions, writing code, and other items of interest.
Keil Software — Cx51 Compiler User’s Guide 5

Document Conventions
This document uses the following conventions:

Examples Description
README.TXT Bold capitalized text is used for the names of executable programs, data files,
source files, environment variables, and commands you enter at the Windows
command prompt. This text usually represents commands that you must type
in literally.
Example:: CLS DIR BL51.EXE

Note that you are not required to enter these commands using all capital
letters.
Language Elements of the C language are presented in bold type. This includes
Elements keywords, operators and library functions.
Example: if != long
isdigit main >>
Courier Text in this typeface is used to represent information that displays on screen or
prints at the printer.
This typeface is also used within the text when discussing or describing
command line items.
Variables Text in italics represents information that you must provide. For example,
projectfile in a syntax string means that you are required to supply the actual
project filename.
Occasionally, italics are also used to emphasize words in the text.
Elements that Ellipses (…) are used in examples to indicate an item that may be repeated.
repeat…
Omitted code Vertical ellipses are used in source code examples to indicate a fragment of the
. program is omitted.
.
Example:
.
void main (void) {
.
.
.
while (1);

Optional Items Optional arguments in command-line and option fields are indicated by double
brackets.

Example: C51 TEST.C PRINT
(filename)


{ opt1 | opt2 } Text contained within braces, separated by a vertical bar, represents a group of
items from which one must be chosen. One item in the list must be selected.

braces enclose all of the choices

vertical bars separate the choices
Keys Text in the sans serif typeface represents keys on the keyboard. For example,
“Press Enter to continue.”
Keil Software — Cx51 Compiler User’s Guide 7

Contents
Chapter 1. Introduction....................................................................................17
Support for all 8051 Variants...................................................................................... 17
Books About the C Language ..................................................................................... 18
Chapter 2. Compiling with Cx51.....................................................................19
Environment Variables ............................................................................................... 19
Running Cx51 from the Command Prompt................................................................. 20
ERRORLEVEL..................................................................................................... 21
Cx51 Output Files ................................................................................................. 21
Control Directives....................................................................................................... 22
Directive Categories.............................................................................................. 22
Reference .................................................................................................................... 25
AREGS / NOAREGS............................................................................................ 26
ASM / ENDASM .................................................................................................. 28
BROWSE.............................................................................................................. 30
CODE.................................................................................................................... 31
COMPACT ........................................................................................................... 32
COND / NOCOND ............................................................................................... 33
DEBUG................................................................................................................. 35
DEFINE ................................................................................................................ 36
DISABLE.............................................................................................................. 37
EJECT................................................................................................................... 39
FLOATFUZZY ..................................................................................................... 40
INCDIR................................................................................................................. 41
INTERVAL........................................................................................................... 42
INTPROMOTE / NOINTPROMOTE .................................................................. 43
INTVECTOR / NOINTVECTOR ........................................................................ 46
LARGE ................................................................................................................. 48
LISTINCLUDE..................................................................................................... 49
MAXARGS........................................................................................................... 50
MOD517 / NOMOD517 ....................................................................................... 51
MODA2 / NOMODA2 ......................................................................................... 53
MODDP2 / NOMODDP2..................................................................................... 54
MODP2 / NOMODP2........................................................................................... 55
NOAMAKE .......................................................................................................... 56
NOEXTEND......................................................................................................... 57
OBJECT / NOOBJECT ........................................................................................ 58
OBJECTEXTEND................................................................................................ 59
ONEREGBANK ................................................................................................... 60
OMF2.................................................................................................................... 61
OPTIMIZE............................................................................................................ 62
ORDER ................................................................................................................. 64
PAGELENGTH .................................................................................................... 65
PAGEWIDTH....................................................................................................... 66
8 Contents

PREPRINT............................................................................................................67
PRINT / NOPRINT...............................................................................................68
REGFILE ..............................................................................................................69
REGISTERBANK ................................................................................................70
REGPARMS / NOREGPARMS ...........................................................................71
RET_PSTK, RET_XSTK ....................................................................................73
ROM......................................................................................................................75
SAVE / RESTORE................................................................................................76
SMALL .................................................................................................................77
SRC .......................................................................................................................78
STRING ................................................................................................................79
SYMBOLS ............................................................................................................80
VARBANKING ....................................................................................................81
WARNINGLEVEL ...............................................................................................82
Chapter 3. Language Extensions .....................................................................85
Keywords ....................................................................................................................85
8051 Memory Areas....................................................................................................86
Program Memory ..................................................................................................86
Internal Data Memory ...........................................................................................87
External Data Memory ..........................................................................................88
Special Function Register Memory .......................................................................89
Memory Models..........................................................................................................89
Small Model ..........................................................................................................89
Compact Model .....................................................................................................90
Large Model ..........................................................................................................90
Memory Types ............................................................................................................90
Explicitly Declared Memory Types.......................................................................91
Implicit Memory Types .........................................................................................92
Data Types ..................................................................................................................92
Bit Types.....................................................................................................................93
Bit-addressable Objects...............................................................................................94
Special Function Registers ..........................................................................................96
sfr ..........................................................................................................................96
sfr16 ......................................................................................................................97
sbit.........................................................................................................................97
Absolute Variable Location ........................................................................................99
Pointers .....................................................................................................................101
Generic Pointers ..................................................................................................101
Memory-specific Pointers....................................................................................104
Pointer Conversions ............................................................................................106
Abstract Pointers .................................................................................................109
Function Declarations ...............................................................................................113
Function Parameters and the Stack......................................................................114
Passing Parameters in Registers ..........................................................................115
Function Return Values .......................................................................................115
Specifying the Memory Model for a Function.....................................................116
Specifying the Register Bank for a Function .......................................................117
Keil Software — Cx51 Compiler User’s Guide 9

Register Bank Access.......................................................................................... 119

Interrupt Functions .............................................................................................. 120
Reentrant Functions ............................................................................................ 124
Alien Function (PL/M-51 Interface) ................................................................... 127
-
Real time Function Tasks................................................................................... 128

Chapter 4. Preprocessor .................................................................................129

Directives.................................................................................................................. 129
Stringize Operator..................................................................................................... 130
Token-pasting operator ............................................................................................. 131
Predefined Macro Constants ..................................................................................... 132
Chapter 5. 8051 Derivatives ...........................................................................133
Atmel 89x8252 and variants ..................................................................................... 134
Dallas 80C320, 420, 520, and 530............................................................................ 134
Infineon C517, C517A, C509 and variants 80C537 ................................................. 135
Data Pointers....................................................................................................... 135
High-speed Arithmetic ........................................................................................ 136
Library Routines.................................................................................................. 138
Philips 8xC750, 8xC751, and 8xC752...................................................................... 139
Philips and Atmel WM Dual DPTR ......................................................................... 140
Chapter 6. Advanced Programming Techniques.........................................143
Customization Files .................................................................................................. 143
STARTUP.A51................................................................................................... 144
START751.A51 .................................................................................................. 148
INIT.A51............................................................................................................. 150
INIT751.A51....................................................................................................... 151
PUTCHAR.C ...................................................................................................... 153
GETKEY.C......................................................................................................... 153
CALLOC.C ......................................................................................................... 153
FREE.C ............................................................................................................... 153
INIT_MEM.C ..................................................................................................... 153
MALLOC.C ........................................................................................................ 154
REALLOC.C....................................................................................................... 154
Optimizer .................................................................................................................. 155
General Optimizations ........................................................................................ 156
-
8051 Specific Optimizations.............................................................................. 156
Options for Code Generation .............................................................................. 157
Segment Naming Conventions.................................................................................. 158
Data Objects........................................................................................................ 158
Program Objects.................................................................................................. 159
Interfacing C Programs to Assembler ....................................................................... 161
Function Parameters............................................................................................ 161
Parameter Passing in Registers............................................................................ 162
Parameter Passing in Fixed Memory Locations .................................................. 163
Function Return Values....................................................................................... 163
10 Contents

Using the SRC Directive .....................................................................................164

Register Usage.....................................................................................................166
Overlaying Segments...........................................................................................166
Example Routines................................................................................................166
Small Model Example .........................................................................................167
Compact Model Example ....................................................................................169
Large Model Example .........................................................................................171
Interfacing C Programs to PL/M-51..........................................................................173
Data Storage Formats................................................................................................174
Bit Variables........................................................................................................174
Signed and Unsigned Characters, Pointers to data, idata, and pdata ..................175
Signed and Unsigned Integers, Enumerations, Pointers to xdata and
code .....................................................................................................................175
Signed and Unsigned Long Integers....................................................................175
Generic and Far Pointers .....................................................................................176
-
Floating point Numbers......................................................................................177
Floating-point Errors ...........................................................................................179
Accessing Absolute Memory Locations....................................................................180
Absolute Memory Access Macros.......................................................................180
Linker Location Controls ....................................................................................181
The _at_ Keyword ...............................................................................................182
Debugging.................................................................................................................183
Chapter 7. Error Messages ............................................................................185
Fatal Errors ...............................................................................................................185
Actions ................................................................................................................186
Errors...................................................................................................................187
Syntax and Semantic Errors ......................................................................................189
Warnings ...................................................................................................................201
Chapter 8. Library Reference........................................................................205
Intrinsic Routines ......................................................................................................205
Library Files..............................................................................................................206
Standard Types..........................................................................................................207
jmp_buf ...............................................................................................................207
va_list ..................................................................................................................207
Absolute Memory Access Macros.............................................................................208
CBYTE................................................................................................................208
CWORD..............................................................................................................208
DBYTE ...............................................................................................................209
DWORD..............................................................................................................209
PBYTE ................................................................................................................210
PWORD ..............................................................................................................210
XBYTE ...............................................................................................................211
XWORD..............................................................................................................211
Routines by Category ................................................................................................212
Buffer Manipulation ............................................................................................212
Keil Software — Cx51 Compiler User’s Guide 11

Character Conversion and Classification ............................................................ 213

Data Conversion.................................................................................................. 214
Math .................................................................................................................... 214
Memory Allocation ............................................................................................. 216
Stream Input and Output ..................................................................................... 217
String Manipulation ............................................................................................ 219
Variable-length Argument Lists .......................................................................... 220
Miscellaneous...................................................................................................... 220
Include Files.............................................................................................................. 221
8051 Special Function Register Include Files ..................................................... 221
80C517.H............................................................................................................ 221
ABSACC.H......................................................................................................... 222
ASSERT.H.......................................................................................................... 223
CTYPE.H............................................................................................................ 223
INTRINS.H......................................................................................................... 223
MATH.H............................................................................................................. 223
SETJMP.H .......................................................................................................... 225
STDARG.H......................................................................................................... 225
STDDEF.H ......................................................................................................... 225
STDIO.H............................................................................................................. 225
STDLIB.H........................................................................................................... 226
STRING.H .......................................................................................................... 226
Reference .................................................................................................................. 227
abs ....................................................................................................................... 228
acos / acos517 ..................................................................................................... 229
asin / asin517....................................................................................................... 230
assert ................................................................................................................... 231
atan / atan517 ...................................................................................................... 232
atan2.................................................................................................................... 233
atof / atof517 ....................................................................................................... 234
atoi ...................................................................................................................... 235
atol ...................................................................................................................... 236
cabs ..................................................................................................................... 237
calloc................................................................................................................... 238
ceil....................................................................................................................... 239
_chkfloat_ ........................................................................................................... 240
cos / cos517......................................................................................................... 241
cosh ..................................................................................................................... 242
_crol_ .................................................................................................................. 243
_cror_ .................................................................................................................. 244
exp / exp517........................................................................................................ 245
fabs...................................................................................................................... 246
floor..................................................................................................................... 247
fmod .................................................................................................................... 248
free ...................................................................................................................... 249
getchar................................................................................................................. 250
_getkey................................................................................................................ 251
12 Contents

gets ......................................................................................................................252
init_mempool.......................................................................................................253
_irol_ ...................................................................................................................254
_iror_...................................................................................................................255
isalnum ................................................................................................................256
isalpha .................................................................................................................257
iscntrl...................................................................................................................258
isdigit...................................................................................................................259
isgraph .................................................................................................................260
islower .................................................................................................................261
isprint...................................................................................................................262
ispunct .................................................................................................................263
isspace .................................................................................................................264
isupper .................................................................................................................265
isxdigit.................................................................................................................266
labs ......................................................................................................................267
log / log517 .........................................................................................................268
log10 / log10517 .................................................................................................269
longjmp ...............................................................................................................270
_lrol_ ...................................................................................................................272
_lror_...................................................................................................................273
malloc..................................................................................................................274
memccpy .............................................................................................................275
memchr................................................................................................................276
memcmp ..............................................................................................................277
memcpy ...............................................................................................................278
memmove ............................................................................................................279
memset ................................................................................................................280
modf ....................................................................................................................281
_nop_...................................................................................................................282
offsetof ................................................................................................................283
pow......................................................................................................................284
printf / printf517 ..................................................................................................285
putchar.................................................................................................................291
puts ......................................................................................................................292
rand......................................................................................................................293
realloc..................................................................................................................294
scanf ....................................................................................................................295
setjmp ..................................................................................................................299
sin / sin517 ..........................................................................................................300
sinh ......................................................................................................................301
sprintf / sprintf517 ...............................................................................................302
sqrt / sqrt517........................................................................................................304
srand ....................................................................................................................305
sscanf / sscanf517 ................................................................................................306
strcat ....................................................................................................................308
strchr....................................................................................................................309
Keil Software — Cx51 Compiler User’s Guide 13

strcmp.................................................................................................................. 310
strcpy................................................................................................................... 311
strcspn ................................................................................................................. 312
strlen.................................................................................................................... 313
strncat.................................................................................................................. 314
strncmp................................................................................................................ 315
strncpy................................................................................................................. 316
strpbrk ................................................................................................................. 317
strpos................................................................................................................... 318
strrchr .................................................................................................................. 319
strrpbrk................................................................................................................ 320
strrpos.................................................................................................................. 321
strspn................................................................................................................... 322
strtod / strtod517 ............................................................................................... 323
strtol .................................................................................................................... 325
strtoul .................................................................................................................. 327
tan / tan517.......................................................................................................... 329
tanh...................................................................................................................... 330
_testbit_............................................................................................................... 331
toascii .................................................................................................................. 332
toint ..................................................................................................................... 333
tolower ................................................................................................................ 334
_tolower .............................................................................................................. 335
toupper ................................................................................................................ 336
_toupper .............................................................................................................. 337
ungetchar............................................................................................................. 338
va_arg.................................................................................................................. 339
va_end ................................................................................................................. 341
va_start................................................................................................................ 342
vprintf.................................................................................................................. 343
vsprintf ................................................................................................................ 345
Appendix A. Differences from ANSI C.........................................................347
Compiler-related Differences.................................................................................... 347
Library-related Differences....................................................................................... 347
Appendix B. Version Differences...................................................................351
Version 6.0 Differences ............................................................................................ 351
Version 5 Differences ............................................................................................... 351
Version 4 Differences ............................................................................................... 352
Version 3.4 Differences ............................................................................................ 355
Version 3.2 Differences ............................................................................................ 356
Version 3.0 Differences ............................................................................................ 357
Version 2 Differences ............................................................................................... 358
Using C51 Version 5 with Previous Versions........................................................... 359
Appendix C. Writing Optimum Code ...........................................................361
Memory Model ......................................................................................................... 361
14 Contents

Variable Location......................................................................................................363
Variable Size.............................................................................................................363
Unsigned Types.........................................................................................................364
Local Variables .........................................................................................................364
Other Sources............................................................................................................364
Appendix D. Compiler Limits........................................................................365
Limitations of the Cx51 Compiler Implementation ...................................................365
Limitations of the Intel Object Module Format.........................................................366
Appendix E. Byte Ordering............................................................................367
Appendix F. Hints, Tips, and Techniques.....................................................369
Recursive Code Reference Error...............................................................................369
Problems Using the printf Routines ..........................................................................370
Uncalled Functions....................................................................................................371
Trouble with the bdata Memory Type.......................................................................372
Using Monitor-51......................................................................................................373
Function Pointers ......................................................................................................374
Glossary.............................................................................................................375
Index ..................................................................................................................382
Keil Software — Cx51 Compiler User’s Guide 15
Keil Software — Cx51 Compiler User’s Guide 17

Chapter 1. Introduction
The C programming language is a general-purpose, programming language that
provides code efficiency, elements of structured programming, and a rich set of 1
operators. C is not a big language and is not designed for any one particular area
of application. Its generality combined with its absence of restrictions, makes C
a convenient and effective programming solution for a wide variety of software
tasks. Many applications can be solved more easily and efficiently with C than
with other more specialized languages.

The Cx51 Optimizing C Compiler is a complete implementation of the American

National Standards Institute (ANSI) standard for the C language. Cx51 is not a
universal C compiler adapted for the 8051 target. It is a ground-up
implementation dedicated to generating extremely fast and compact code for the
8051 microprocessor. Cx51 provides you the flexibility of programming in C
and the code efficiency and speed of assembly language.

The C language on its own is not capable of performing operations (such as input
and output) that would normally require intervention from the operating system.
Instead, these capabilities are provided as part of the standard library. Because
these functions are separate from the language itself, C is especially suited for
producing code that is portable across a wide number of platforms.

Since Cx51 is a cross compiler, some aspects of the C programming language

and standard libraries are altered or enhanced to address the peculiarities of an
embedded target processor. Refer to “Chapter 3. Language Extensions” on page
85 for more detailed information.

Support for all 8051 Variants

The 8051 Family is one of the fastest growing Microcontroller Architectures.
More than 400 device variants from various silicon vendors are today available.
New extended 8051 Devices, like the Philips 80C51MX architecture are
deticated for large application with several Mbytes code and data space.

For optimum support of these different 8051 variants, Keil provides the several
development tools that are listed in the table below. A new output file format
(OMF2) allows direct support of up to 16MB code and data space. The CX51
compiler is a variant of the C51 compiler that is design for the new Philips
80C51MX architecture.
 18 Chapter 1. Introduction

Development Tools Support Microcontrollers, Description

C51 Compiler Development Tools for classic 8051.

1
A51 Macro Assembler Includes support for 32 x 64KB code banks.
BL51 Linker/Locater
C51 Compiler (with OMF2 Output) Development Tools for classic 8051 and extended
AX51 Macro Assembler 8051 variants (like Dallas 390). Includes support for
LX51 Linker/Locater code banking and up to 16MB code and xdata memory.
CX51 Compiler Development Tools for Philips 80C51MX
AX51 Macro Assembler Supports up to 16MB code and xdata memory.
LX51 Extended Linker/Locater
The Cx51 compiler is available in different packages. The table above
refers to the entire line of the 8051 development tools.

NOTE
The term Cx51 is used to refer to both compiler variants: the C51 compiler and
the CX51 compiler.

Books About the C Language

There are a number of books that provide an introduction to the C programming
language. There are even more books that detail specific tasks using C. The
following list is by no means a complete list of books on the subject. The list is
provided only as a reference for those who wish more information.

The C Programming Language, Second Edition

Kernighan & Ritchie
Prentice-Hall, Inc.
ISBN 0-13-110370-9

C: A Reference Manual, Second Edition

Harbison & Steel
Prentice-Hall Software Series
ISBN 0-13-109810-1

C and the 8051: Programming and Multitasking

Schultz
P T R Prentice-Hall, Inc.
ISBN 0-13-753815-4
Keil Software — Cx51 Compiler User’s Guide 19

Chapter 2. Compiling with Cx51

This chapter explains how to use Cx51 to compile C source files and discusses
the control directives you may specify. These directives allow you to perform
serveral functions. For example:

Direct Cx51 to generate a listing file

Control the information included in the object file 2

Specify code optimization and memory models

NOTE
Typically you will use the Cx51 compiler within the µVision2 IDE. For more
information on using the µVision2 IDE refer to the µVision2 Getting Started for
8051 User’s Guide.

Environment Variables
If you run the Cx51 compiler within the µVision2 IDE, you need no additional
settings on your computer. If you want to run the Cx51 compiler and utilities
from the command prompt, you must manually create the following environment
variables.

Variable Path Environment Variable specifies …

PATH \C51\BIN path of the C51 and CX51 executable programs.
TMP path to use for temporary files generated by the compiler. If the
specified path does not exist, the compiler generates an error and
aborts compilation.
C51INC \C51\INC the location of the standard Cx51 include files.
C51LIB \C51\LIB the location of the standard Cx51 library files.

For Windows NT and Windows 2000 these environment variables are entered
under Control Panel – System – Advanced – Environment Variables.

For Windows 95, Windows 98 and Windows ME the settings are placed in
AUTOEXEC.BAT using the following commands:

PATH = C:\C51\BIN;%PATH%
SET TMP = D:\
SET C51INC = C:\C51\INC
SET C51LIB = C:\C51\LIB
 20 Chapter 2. Compiling with Cx51

Running Cx51 from the Command Prompt

To invoke the C51 or CX51 compiler, enter C51 or CX51 at the command
prompt. On this command line, you must include the name of the C source file
to be compiled, as well as any other necessary control directives required to
compile your source file. The format for the Cx51 command line is shown
below:

2 C51 sourcefile
directives…



CX51 sourcefile
directives…



or:
C51 @commandfile
CX51 @commandfile

where:
sourcefile is the name of the source program you want to compile.
directives are the directives you want to use to control the function of the
compiler. Refer to “Control Directives” on page 22 for a
detailed list of the available directives.
commandfile is the name of a command input file that may contain sourcefile
and directives. A commandfile is used, when the Cx51
invocation line gets complex and exceeds the limits of the
Windows command prompt.

The following command line example invokes C51, specifies the source file
SAMPLE.C, and uses the controls DEBUG, CODE, and PREPRINT.
C51 SAMPLE.C DEBUG CODE PREPRINT

The Cx51 compiler displays the following information upon successful

invocation and compilation.
C51 COMPILER V6.10

C51 COMPILATION COMPLETE. 0 WARNING(S), 0 ERROR(S)

Keil Software — Cx51 Compiler User’s Guide 21

ERRORLEVEL
After compilation, the number of errors and warnings detected is output to the
screen. Cx51 then sets the ERRORLEVEL to indicate the status of the
compilation. Values are listed in the following table:

ERRORLEVEL Meaning
0
1
No errors or warnings
Warnings only
2
2 Errors and possibly warnings
3 Fatal errors

You can access the ERRORLEVEL variable in batch files. Refer to the Windows
command index or to batch commands in the Windows on-line help for more
information on ERRORLEVEL or batch files.

Cx51 Output Files

Cx51 generates a number of output files during compilation. By default, each of
these output files shares the same filename as the source file. However, each
has a different file extension. The following table lists the files and gives a brief
description of each.

File Extension Description

filename.LST Files with this extension are listing files that contain the formatted source
text along with any errors detected by the compiler. Listing files may
optionally contain the used symbols and the generated assembly code.
See the PRINT directive in the following sections for more information.
filename.OBJ Files with this extension are object modules that contain relocatable object
code. Object modules may be linked to an absolute object module by the
Lx51 Linker/Locator.
filename.I Files with this extension contain the source text as expanded by the
preprocessor. All macros are expanded and all comments are deleted in
this listing. See the PREPRINT directive in the following sections for more
information.
filename.SRC Files with this extension are assembly source files generated from your C
source code. These files can be assembled with the A51 assembler. See
the SRC directive in the following sections for more information.
 22 Chapter 2. Compiling with Cx51

Control Directives
Cx51 offers a number of control directives that you can use to control the
operation of the compiler. Directives are composed of one or more letters or
digits and, unless otherwise specified, can be specified after the filename on the
command line or within a source file using the #pragma directive.

2 Example
C51 testfile.c SYMBOLS CODE DEBUG

#pragma SYMBOLS CODE DEBUG

In the above examples, SYMBOLS, CODE, and DEBUG are all control directives.
testfile.c is the source file to be compiled.

NOTE
The syntax is the same for the command line and the #pragma directive.
Multiple options, however, may be specified on the #pragma line.

Typically, each control directive may be specified only once at the beginning of
a source file. If a directive is specified more than once, the compiler generates a
fatal error and aborts compilation. Directives that may be specified more than
once are so noted in the following sections.

Directive Categories
Control directives can be divided into three groups: source controls, object
controls, and listing controls.

Source controls define macros on the command line and determine the name
of the file to be compiled.

Object controls affect the form and content of the generated object module
(*.OBJ). These directives allow you to specify the optimizing level or
include debugging information in the object file.

Listing controls govern various aspects of the listing file (*.LST), in
particular its format and specific content.
Keil Software — Cx51 Compiler User’s Guide 23

The following table is an alphabetical list of the control directives. Unterlined

characters denote the abbreviation of the directive.

Directive Class Description

AREGS, Object Enable or disable absolute register (ARn) addressing.
NOAREGS
ASM, ENDASM Object Marks the beginning and the end of an inline assembly block.
BROWSE †
CODE †
COMPACT †
Object
Listing
Object
Enable generation of browser information.
Add an assembly listing to the listing file.
Select COMPACT memory model.
2
COND, Listing Include or exclude source lines skipped from the preprocessor.
NOCOND †
DEBUG † Object Include debugging information in the object file.
DEFINE Source Define preprocessor names in the Cx51 invocation line.
DISABLE Object Disables interrupts for the duration of a function.
EJECT Listing Inserts a form feed character into the listing file.
FLOATFUZZY Object Specify number of bits rounded during floating compare.
INCDIR, ID † Source Specify additional path names for include files.
INTERVAL † Object Specify the interval for interrupt vectors for SIECO chips.
INTPROMOTE, Object Enable or disable ANSI integer promotion.
NOINTPROMOTE †
INTVECTOR, Object Specify base address for interrupt vectors or disable vectors.
NOINTVECTOR †
LARGE † Object Select LARGE memory model.
LISTINCLUDE Listing Display contents of include files in the listing file.
MAXARGS † Object Specify size of variable argument lists.
MOD517, Object Enable or disable code to support the additional hardware
NOMOD517 features of the 80C517 and derivatives.
MODA2, Object Enable or disable dual DPTR register support for Atmel 82x8252
NOMODA2 and variants.
MODDP2, Object Enable or disable dual DPTR register support for Dallas
NOMODDP2 Semiconductor 320, 520, 530, 550 and variants.
MODP2, Object Enable or disable dual DPTR register support for Philips and
NOMODP2 Temic derivatives.
NOAMAKE † Object Disable information records for AutoMAKE.
NOEXTEND † Source Disable Cx51 extensions to ANSI C.
OBJECT, Object Enable object file and optionally specify name or suppress the
NOOBJECT † object file.
OBJECTEXTEND † Object Include additional variable type information in the object file.
ONEREGBANK Object Assume that only registerbank 0 is used in interrupt code.
OMF2 † Object Generate OMF2 output file format.
OPTIMIZE Object Specify the level of optimization performed by the compiler.
 24 Chapter 2. Compiling with Cx51

Directive Class Description

ORDER † Object Variables are allocated in the order in which they appear in the
source file.
PAGELENGTH † Listing Specify number of rows on the page.
PAGEWIDTH † Listing Specify number of columns on the page.
PREPRINT † Listing Produce a preprocessor listing file where all macros are
expanded.
PRINT, NOPRINT † Listing Specify a name for the listing file or disable the listing file.

2 REGFILE †
REGISTERBANK
Object
Object
Specify a register definition file for global register optimization.
Select the register bank that is used for absolute register
accesses.
REGPARMS, Object Enable or disable register parameter passing.
NOREGPARMS
RET_PSTK †, Object Use reentrant stack for saving return addresses.
RET_XSTK †
ROM † Object Control generation of AJMP/ACALL instructions.
SAVE, Object Saves and restores settings for AREGS, REGPARMS and
RESTORE OPTIMIZE directives.
SMALL † Object Select SMALL memory model. (Default.)
SRC † Object Create an assembler source file instead of an object module.
STRING † Object Locate implicit string constants to xdata or far memory.
SYMBOLS † Listing Include a list of all symbols used within the module in the listing
file.
VARBANKING † Object Use library set for variable banking support.
WARNINGLEVEL † Listing Selects the level of Warning detection.
† These directives may be specified only once on the command line or at the beginning of a source
file using in the #pragma statement. They may not be used more than once in a source file.

Control directives and their arguments, with the exception of arguments

specified with the DEFINE directive, are case insensitive.
Keil Software — Cx51 Compiler User’s Guide 25

Reference
The remainder of this chapter is devoted to describing each of the available
Cx51 compiler control directives. The directives are listed in alphabetical order,
and each is divided into the following sections:

Abbreviation: Gives any abbreviations that may be substituted for the

directive name.
2
Arguments: Describes and lists optional and required directive
arguments.

Default: Shows the directive’s default setting.

µVision2 Control: Gives the dialog box in the µVision2 IDE which allows you
to specify the directive.

Description: Provides a detailed description of the directive and how to

use it.

See Also: Names related directives.

Example: Shows you an example of how to use and, sometimes, the

effects of the directive.
 26 Chapter 2. Compiling with Cx51

AREGS / NOAREGS
Abbreviation: None.

Arguments: None.

Default: AREGS

2 µVision2 Control: Options – C51 - Don’t use absolute register accesses.

Description: The AREGS control causes the compiler to use absolute

register addressing for registers R0 through R7. Absolute
addressing improves the efficiency of the generated code.
For example, PUSH and POP instructions function only
with direct or absolute addresses. Using the AREGS
directive, allows you to directly push and pop registers.

You may use the REGISTERBANK directive to define

which register bank to use.

The NOAREGS directive disables absolute register

addressing for registers R0 through R7. Functions which are
compiled with NOAREGS are not dependent on the register
bank and may use all 8051 register banks. This directive
may be used for functions that are called from other
functions using different register banks.

NOTE
Though it may be defined several times in a program, the
AREGS / NOAREGS option is valid only when defined
outside of a function declaration.
Keil Software — Cx51 Compiler User’s Guide 27

Example: The following is a source and code listing which uses both
NOAREGS and AREGS.
stmt level source
1 extern char func ();
2 char k;
3
4 #pragma NOAREGS
5 noaregfunc () {
6 1 k = func () + func ();
7
8
9
1 }

#pragma AREGS
2
10 aregfunc () {
11 1 k = func () + func ();
12 1 }

; FUNCTION noaregfunc (BEGIN)

; SOURCE LINE # 6
0000 120000 E LCALL func
0003 EF MOV A,R7
0004 C0E0 PUSH ACC
0006 120000 E LCALL func
0009 D0E0 POP ACC
000B 2F ADD A,R7
000C F500 R MOV k,A
; SOURCE LINE # 7
000E 22 RET
; FUNCTION noaregfunc (END)

; FUNCTION aregfunc (BEGIN)

; SOURCE LINE # 11
0000 120000 E LCALL func
0003 C007 PUSH AR7
0005 120000 E LCALL func
0008 D0E0 POP ACC
000A 2F ADD A,R7
000B F500 R MOV k,A
; SOURCE LINE # 12
000D 22 RET
; FUNCTION aregfunc (END)

Note the different methods of saving R7 on the stack. The

code generated for the function noaregfunc is:
MOV A,R7
PUSH ACC

while the code for the aregfunc function is:

PUSH AR7
 28 Chapter 2. Compiling with Cx51

ASM / ENDASM
Abbreviation: None.

Arguments: None.

Default: None.

2 µVision2 Control: This directive cannot be specified on the command line.

Description: The ASM directive signals the beginning merge of a block

of source text into the .SRC file generated using the SRC
directive.

This source text can be thought of as in-line assembly.

However, it is output to the source file generated only when
using the SRC directive. The source text is not assembled
and output to the object file.

In µVision2 you may set a file specific option for C source

files that contain ASM/ENDASM sections as follows:
• right click on the file in the Project Window – Files tab
• choose Options for… to open Options – Properties page
• enable Generate Assembler SRC file
• enable Assemble SRC file.
With this settings, µVision2 will generate an assembler
source file (.SRC) and translates this file with the Assembler
to an Object file (.OBJ).

The ENDASM directive is used to signal the end of the

source text block.

NOTE
The ASM / ENDASM directive can occur only in the source
file, as part of a #pragma directive.
Keil Software — Cx51 Compiler User’s Guide 29

Example: #pragma asm / #pragma endasm

The following C source file:

.
.
.
stmt level source
1 extern void test ();

2
2
3 main () {
4 1 test ();
5 1
6 1 #pragma asm
7 1 JMP $ ; endless loop
8 1 #pragma endasm
9 1 }
.
.
.

generates the following .SRC file.

; ASM.SRC generated from: ASM.C
NAME ASM
?PR?main?ASM SEGMENT CODE
EXTRN CODE (test)
EXTRN CODE (?C_STARTUP)
PUBLIC main
; extern void test ();
;
; main () {
RSEG ?PR?main?ASM
USING 0
main:
; SOURCE LINE # 3
; test ();
; SOURCE LINE # 4
LCALL test
;
; #pragma asm
JMP $ ; endless loop
; #pragma endasm
; }
; SOURCE LINE # 9
RET ; END OF main
END
 30 Chapter 2. Compiling with Cx51

BROWSE
Abbreviation: BR

Arguments: None.

Default: No browse information is created

2 µVision2 Control: Options – Output – Browse Information

Description: With BROWSE, the compiler creates browse information.

The browse information covers identifiers (including
preprocessor symbols), their memory space, type, definition-
and reference lists. This information can be displayed
within µVision2. Select View - Source Browser to open
the µVision2 Source Browser. Refer to the µVision2
Getting Started User’s Guide, Chapter 4, µVision2 Utilities,
Source Browser for more information.

Example: C51 SAMPLE.C BROWSE

#pragma browse
Keil Software — Cx51 Compiler User’s Guide 31

CODE
Abbreviation: CD

Arguments: None.

Default: No assembly code listing is generated.

µVision2 Control: Options – Listing – C Compiler Listing - Assembly Code. 2

Description: The CODE directive appends an assembly mnemonics list to
the listing file. The assembler code is represented for each
function contained in the source program. By default, no
assembly code listing is included in the listing file.

Example: C51 SAMPLE.C CD

#pragma code

The following example shows the C source followed by the

resulting object code and its representative mnemonics. The
line number of each statement that produced the code is
displayed between the assembly lines. The characters R and
E stand for Relocatable and External, respectively.

stmt level source

1 extern unsigned char a, b;
2 unsigned char c;
3
4 main()
5 {
6 1 c = 14 + 15 * ((b / c) + 252);
7 1 }
.
.
.
ASSEMBLY LISTING OF GENERATED OBJECT CODE

; FUNCTION main (BEGIN)

; SOURCE LINE # 5
; SOURCE LINE # 6
0000 E500 E MOV A,b
0002 8500F0 R MOV B,c
0005 84 DIV AB
0006 75F00F MOV B,#0FH
0009 A4 MUL AB
000A 24D2 ADD A,#0D2H
000C F500 R MOV c,A
; SOURCE LINE # 7
000E 22 RET
; FUNCTION main (END)
 32 Chapter 2. Compiling with Cx51

COMPACT
Abbreviation: CP

Arguments: None.

Default: SMALL

2 µVision2 Control: Options – Target – Memory Model

Description: This directive implements the COMPACT memory model.

In the COMPACT memory model, all function and

procedure variables and local data segments reside in the
external data memory of the 8051 system. This external
data memory may be up to 256 bytes (one page) long. With
this model, the short form of addressing the external data
memory through @R0/R1 is used.

Regardless of memory model type, you may declare

variables in any of the 8051 memory ranges. However,
placing frequently used variables (such as loop counters and
array indices) in internal data memory significantly
improves system performance.

NOTE
The stack required for function calls is always placed in
IDATA memory.

See Also: SMALL, LARGE, ROM

Example: C51 SAMPLE.C COMPACT

#pragma compact
Keil Software — Cx51 Compiler User’s Guide 33

COND / NOCOND
Abbreviation: CO

Arguments: None.

Default: COND

µVision2 Control: Options – Listing – C Compiler Listing - Conditional. 2

Description: This directive determines whether or not those portions of
the source file affected by conditional compilation are
displayed in the listing file.

The COND directive forces lines omitted from compilation

to appear in the listing file. Line numbers and nesting levels
are not output. This allows for easier visual identification.

The effect of this directive takes place one line after it is

detected by the preprocessor.

The NOCOND directive determines whether or not those

portions of the source file affected by conditional
compilation are displayed in the listing file.

This directive also prevents lines omitted from compilation

from appearing in the listing file.
 34 Chapter 2. Compiling with Cx51

Example: The following example shows the listing file for a source
file compiled with the COND directive.
.
.
.
stmt level source
1 extern unsigned char a, b;
2 unsigned char c;
3

2 4
5
6 1
main()
{
#if defined (VAX)
c = 13;
#elif defined (_ _TIME_ _)
9 1 b = 14;
10 1 a = 15;
11 1 #endif
12 1 }
.
.
.

The following example shows the listing file for a source

file compiled with the NOCOND directive.
.
.
.
stmt level source
1 extern unsigned char a, b;
2 unsigned char c;
3
4 main()
5 {
6 1 #if defined (VAX)
9 1 b = 14;
10 1 a = 15;
11 1 #endif
12 1 }
.
.
.
Keil Software — Cx51 Compiler User’s Guide 35

DEBUG
Abbreviation: DB

Arguments: None.

Default: No Debug information is generated.

µVision2 Control: Options – Output – Debug Information 2

Description: The DEBUG directive instructs the compiler to include
debugging information in the object file. By default,
debugging information is excluded from the generated
object file.

Debug information is necessary for the symbolic testing of

programs. This information contains both global and local
variable definitions and their addresses, as well as function
names and their line numbers. Debug information contained
in each object module remains valid through the
Link/Locate procedure. This information can be used by the
µVision2 debugger or by any of the Intel-compatible
emulators.

NOTE
The OBJECTEXTEND directive can be used to instruct the
compiler to include additional variable type definition
information in the object file.

See Also: OBJECTEXTEND

Example: C51 SAMPLE.C DEBUG

#pragma db
 36 Chapter 2. Compiling with Cx51

DEFINE
Abbreviation: DF

Arguments: One or more names separated by commas, in accordance

with the naming conventions of the C language. An optional
argument can be specified for each name given in the define
directive.
2 Default: None.

µVision2 Control: Options – Cx51 – Define.

Description: The DEFINE directive defines names on the invocation line

which may be queried by the preprocessor with #if, #ifdef
and #ifndef. The defined names are copied exactly as they
are entered. This command is case-sensitive. As an option
each name can be followed by an argument string.

NOTE
The DEFINE directive can be specified only on the
command line. Use the C preprocessor #define directive for
use inside a C source.

Example: C51 SAMPLE.C DEFINE (check, NoExtRam)

C51 MYPROG.C DF (X1="1+5",iofunc="getkey ()")

Keil Software — Cx51 Compiler User’s Guide 37

DISABLE
Abbreviation: None.

Arguments: None.

Default: None.

µVision2 Control: This directive cannot be specified on the command line. 2

Description: The DISABLE directive instructs the compiler to generate
code that disables all interrupts for the duration of a
function. DISABLE must be specified with a #pragma
directive immediately before a function declaration. The
DISABLE control applies to one function only and must be
re-specified for each new function.

NOTE
DISABLE may be specified using the #pragma directive
only, and may not be specified at the command line.

DISABLE can be specified more than once in a source file

and must be specified once for each function that is to
execute with interrupts disabled.

A function with disabled interrupts cannot return a bit value

to the caller.
 38 Chapter 2. Compiling with Cx51

Example: The following example is a source and code listing of a

function using the DISABLE directive. Note that the EA
special function register is cleared at the beginning of the
function (JBC EA,?C0002) and restored at the end
(MOV EA,C).
.
.

2
.
stmt level source
1 typedef unsigned char uchar;
2
3 #pragma disable /* Disable Interrupts */
4 uchar dfunc (uchar p1, uchar p2) {
5 1 return (p1 * p2 + p2 * p1);
6 1 }

; FUNCTION _dfunc (BEGIN)

0000 D3 SETB C
0001 10AF01 JBC EA,?C0002
0004 C3 CLR C
0005 ?C0002:
0005 C0D0 PUSH PSW
;---- Variable 'p1' assigned to register 'R7' ----
;---- Variable 'p2' assigned to register 'R5' ----
; SOURCE LINE # 4
; SOURCE LINE # 5
0007 ED MOV A,R5
0008 8FF0 MOV B,R7
000A A4 MUL AB
000B 25E0 ADD A,ACC
000D FF MOV R7,A
; SOURCE LINE # 6
000E ?C0001:
000E D0D0 POP PSW
0010 92AF MOV EA,C
0012 22 RET
; FUNCTION _dfunc (END)
.
.
.
Keil Software — Cx51 Compiler User’s Guide 39

EJECT
Abbreviation: EJ

Arguments: None.

Default: None.

µVision2 Control: This directive cannot be specified on the command line. 2

Description: The EJECT directive causes a form feed character to be
inserted into the listing file.

NOTE
The EJECT directive occurs only in the source file, and
must be part of a #pragma directive.

Example: #pragma eject

 40 Chapter 2. Compiling with Cx51

FLOATFUZZY
Abbreviation: FF

Arguments: A number between 0 and 7.

Default: FLOATFUZZY (3)

2 µVision2 Control: Options – Cx51 – Bits to round for float compare

Description: The FLOATFUZZY directive determines the number of

bits rounded before a floating-point compare is executed.
The default value of 3 specifies that the three least
significant bits of a float value are rounded before the
floating-point compare is executed.

Example: C51 MYFILE.C FLOATFUZZY (2)

#pragma FF (0)
Keil Software — Cx51 Compiler User’s Guide 41

INCDIR
Abbreviation: ID

Arguments: Path specifications for include files enclosed in parentheses.

Default: None.

µVision2 Control: Options – Cx51 – Include Paths. 2

Description: The INCDIR directive specifies the location of Cx51
include files. The compiler accepts a maximum of 50 path
declarations. If more then one path declaration is required,
the path names must be separated by semicolons. If you
specify #include "filename.h", the Cx51 Compiler searches
first the current directory and then directory of the source
file. When this fails or when #include <filename.h> is used,
the paths specified by the INCDIR directive are searched.
When this still fails, the paths specified by the C51INC
environment variable are used.

Example: C51 SAMPLE.C INCDIR(C:\C51\MYINC;C:\CHIP_DIR)

 42 Chapter 2. Compiling with Cx51

INTERVAL
Abbreviation: None

Arguments: An optional interval, in parentheses, for the interrupt vector

table.

Default: INTERVAL (8)

2 µVision2 Control: Options – Cx51 – Misc controls: enter the directive.

Description: The INTERVAL directive specifies an interval for interrupt

vectors. The interval specification is required for SIECO-51
derivatives which define interrupt vectors in 3-byte
intervals. Using this directive, the compiler locates interrupt
vectors at the absolute address calculated by:

(interval × n) + offset + 3,

where:
interval is the argument of the INTERVAL
directive (default 8).
n is the interrupt number.
offset is the argument of the INTVECTOR
directive (default 0).

See Also: INTVECTOR / NOINTVECTOR

Example: C51 SAMPLE.C INTERVAL(3)

#pragma interval(3)
Keil Software — Cx51 Compiler User’s Guide 43

INTPROMOTE / NOINTPROMOTE
Abbreviation: IP / NOIP

Arguments: None.

Default: INTPROMOTE

µVision2 Control: Options – Cx51 – Enable ANSI integer promotion rules. 2

Description: The INTPROMOTE directive enables ANSI integer
promotion rules. Expressions used in if statements are
promoted from smaller types to integer expressions before
comparison. This allows Microsoft C and Borland C
programs to be ported to Cx51 with fewer modifications.

Because the 8051 is an 8-bit processor, use of the

INTPROMOTE directive generates inefficient code in
some applications.

The NOINTPROMOTE directive disables automatic

integer promotions. Integer promotions are normally
enabled to provide the greatest compatibility between Cx51
and other ANSI compilers. However, integer promotions
can yield inefficient code on the 8051.

Example: C51 SAMPLE.C INTPROMOTE

#pragma intpromote

C51 SAMPLE.C NOINTPROMOTE

 44 Chapter 2. Compiling with Cx51

The following example demonstrates code generated using

the INTPROMOTE and NOINTPROMOTE control
directive.
stmt lvl source

1 char c;
2 unsigned char c1,c2;
3 int i;

2 4
5
6
7
1
1
main () {
if (c == 0xff) c = 0;
if (c == -1) c = 1;
/* never true! */
/* works */
8 1 i = c + 5;
9 1 if (c1 < c2 +4) c1 = 0;
10 1 }
Keil Software — Cx51 Compiler User’s Guide 45

Code generated with INTPROMOTE Code generated with NOINTPROMOTE

; FUNCTION main (BEGIN) ; FUNCTION main (BEGIN)
; SOURCE LINE # 6 ; SOURCE LINE # 6
0000 AF00 MOV R7,c 0000 AF00 MOV R7,c
0002 EF MOV A,R7 0002 EF MOV A,R7
0003 33 RLC A 0003 33 RLC A
0004 95E0 SUBB A,ACC 0004 95E0 SUBB A,ACC
0006 FE MOV R6,A 0006 FE MOV R6,A
0007 EF MOV A,R7 0007 EF MOV A,R7
0008 F4 CPL A 0008 F4 CPL A
0009 4E ORL A,R6 0009 4E ORL A,R6
000A
000C
000E
7002
F500
JNZ ?C0001
MOV c,A
?C0001:
; SOURCE LINE # 7
000A
000C
000E
7002
F500
JNZ ?C0001
MOV c,A
?C0001:
; SOURCE LINE # 7
2
000E E500 MOV A,c 000E E500 MOV A,c
0010 B4FF03 CJNE A,#0FFH,?C0002 0010 B4FF03 CJNE A,#0FFH,?C0002
0013 750001 MOV c,#01H 0013 750001 MOV c,#01H
0016 ?C0002: 0016
; SOURCE LINE # 8 ; SOURCE LINE # 8
0016 AF00 MOV R7,c 0016 E500 MOV A,c
0018 EF MOV A,R7 0018 2405 ADD A,#05H
0019 33 RLC A 001A FF MOV R7,A
001A 95E0 SUBB A,ACC 001B 33 RLC A
001C FE MOV R6,A 001C 95E0 SUBB A,ACC
001D EF MOV A,R7 001E F500 MOV i,A
001E 2405 ADD A,#05H 0020 8F00 MOV i+01H,R7
0020 F500 MOV i+01H,A
0022 E4 CLR A
0023 3E ADDC A,R6
0024 F500 MOV i,A
; SOURCE LINE # 9 ; SOURCE LINE # 9
0026 E500 MOV A,c2 0022 E500 MOV A,c2
0028 2404 ADD A,#04H 0024 2404 ADD A,#04H
002A FF MOV R7,A 0026 FF MOV R7,A
002B E4 CLR A 0027 E500 MOV A,c1
002C 33 RLC A 0029 C3 CLR C
002D FE MOV R6,A 002A 9F SUBB A,R7
002E C3 CLR C 002B 5003 JNC ?C0004
002F E500 MOV A,c1 002D E4 CLR A
0031 9F SUBB A,R7 002E F500 MOV c1,A
0032 EE MOV A,R6
0033 6480 XRL A,#080H
0035 F8 MOV R0,A
0036 7480 MOV A,#080H
0038 98 SUBB A,R0
0039 5003 JNC ?C0004
003B E4 CLR A
003C F500 MOV c1,A
; SOURCE LINE # 10 ; SOURCE LINE # 10
003E ?C0004: 0030 ?C0004:
003E 22 RET 0030 22 RET
; FUNCTION main (END) ; FUNCTION main (END)

CODE SIZE = 63 Bytes CODE SIZE = 49 Bytes

 46 Chapter 2. Compiling with Cx51

INTVECTOR / NOINTVECTOR
Abbreviation: IV / NOIV

Arguments: An optional offset, in parentheses, for the interrupt vector

table.

Default: INTVECTOR (0)

2 µVision2 Control: Options – Cx51 – Misc controls: enter the directive.

Description: The INTVECTOR directive instructs the compiler to

generate interrupt vectors for functions which require them.
An offset may be entered if the vector table starts at an
address other than 0.

Using this directive, the compiler generates an interrupt

vector entry using either an AJMP or LJMP instruction
depending upon the size of the program memory specified
with the ROM directive.

The NOINTVECTOR directive prevents the generation of

an interrupt vector table. This flexibility allows the user to
provide interrupt vectors with other programming tools.

The compiler normally generates an interrupt vector entry

using a 3-byte jump instruction (LJMP). Vectors are
located starting at absolute address:

(interval × n) + offset + 3,

where:
n is the interrupt number.
interval is the argument of the INTERVAL
directive (default 8).
offset is the argument of the INTVECTOR
directive (default 0).
See Also: INTERVAL
Keil Software — Cx51 Compiler User’s Guide 47

Example: C51 SAMPLE.C INTVECTOR(0x8000)

#pragma iv(0x8000)

C51 SAMPLE.C NOINTVECTOR

#pragma noiv

2
 48 Chapter 2. Compiling with Cx51

LARGE
Abbreviation: LA

Arguments: None.

Default: SMALL

2 µVision2 Control: Options – Target – Memory Model

Description: This directive implements the LARGE memory model. In

the LARGE memory model, all variables and local data
segments of functions and procedures reside (as defined) in
the external data memory of the 8051 system. Up to
64 KBytes of external data memory may be accessed. This,
however, requires the long and therefore inefficient form of
data access through the data pointer (DPTR).

Regardless of memory model type, you may declare

variables in any of the 8051 memory ranges. However,
placing frequently used variables (such as loop counters and
array indices) in internal data memory significantly
improves system performance.

NOTE
The stack required for function calls is always placed in
IDATA memory.

See Also: SMALL, COMPACT, ROM

Example: C51 SAMPLE.C LARGE

#pragma large
Keil Software — Cx51 Compiler User’s Guide 49

LISTINCLUDE
Abbreviation: LC

Arguments: None.

Default: NOLISTINCLUDE

µVision2 Control: Options – Listing – C Compiler Listing – #include files 2

Description: The LISTINCLUDE directive displays the contents of the
include files in the listing file. By default, include files are
not listed in the listing file.

Example: C51 SAMPLE.C LISTINCLUDE

#pragma listinclude
 50 Chapter 2. Compiling with Cx51

MAXARGS
Abbreviation: None.

Arguments: Number of bytes compiler reserves for variable-length

argument lists.

µVision2 Control: Options – Cx51 – Misc controls: enter the directive.

2 Default: MAXARGS(15) for small and compact models.

MAXARGS(40) for large model.

Description: With the MAXARGS directive, you specify the buffer size
for parameters passed in variable-length argument lists.
MAXARGS defines the maximum number of parameters.
The MAXARGS directive must be applied before the C
function. This directive has no impact on the maximum
number of arguments that may be passed to reentrant
functions.

Example: C51 SAMPLE.C MAXARGS(20)

#pragma maxaregs (4) /* allow 4 bytes for parameters */

#include <stdarg.h>

void func (char typ, ...) {

va_list ptr;
char c;
int i;

va_start (ptr, typ);

switch *typ) {
case 0: /* a CHAR is passed */
c = va_arg (ptr, char); break;

case 1: /* an INT is passed */

i = va_arg (ptr, int); break;
}
}

void testfunc (void) {

func (0, 'c'); /* pass a char variable */
func (1, 0x1234); /* pass an int variable */
}
Keil Software — Cx51 Compiler User’s Guide 51

MOD517 / NOMOD517
Abbreviation: None.

Arguments: Optional parameters, enclosed in parentheses, to control

support for individual components of the 80C517.

Default: NOMOD517

µVision2 Control: Options – Target – Use On-Chip Arithmetic Unit

2
Options – Target – Use multiple DPTR registers

Description: The MOD517 directive instructs the Cx51 compiler to

produce code for the additional hardware components (the
arithmetic processor and the additional data pointers) of the
Infineon C517 or variants. This feature dramatically
impacts the execution of integer, long, and floating-point
math operations, as well as functions that make use of the
additional data pointers.

The following library functions take advantage of the extra

data pointers: memcpy, memmove, memcmp, strcpy, and
strcmp.

Library functions which take advantage of the arithmetic

processor are so indicated by a 517 suffix. (Refer to
“Chapter 8. Library Reference” on page 205 for details on
these functions.)

Additional parameters may be specified with MOD517 to

control Cx51 support of the individual components of the
Infineon device. When specified, the parameters must
appear within parentheses immediately following the
MOD517 directive. Parentheses are not required if none of
these additional parameters is specified.

Directive Description
NOAU When specified, C51 uses only the additional data
pointers of the Infineon device. The arithmetic processor
is not used. The NOAU parameter is useful for functions
that are called by an interrupt while the arithmetic
processor is already being used.
 52 Chapter 2. Compiling with Cx51

Directive Description
NODP8 When specified, C51 uses only the arithmetic processor.
The additional data pointers are not used. The NODP8
parameter is useful for interrupt functions declared without
the using function attribute. In this case, the extra data
pointers are not used and, therefore, do not need to be
saved on the stack during the interrupt.

Specifying both of these additional parameters with

2 MOD517 has the same effect as using the NOMOD517
directive.

The NOMOD517 directive disables generation of code that

utilizes the additional hardware components of the C517 or
variants.

NOTE
Though it may be defined several times in a program, the
MOD517 directive is valid only when defined outside of a
function declaration.

See Also: MODA2, MODDP2, MODP2

Example: C51 SAMPL517.C MOD517

#pragma MOD517 (NOAU)

#pragma MOD517 (NODP8)

#pragma MOD517 (NODP8, NOAU)

C51 SAMPL517.C NOMOD517

#pragma NOMOD517
Keil Software — Cx51 Compiler User’s Guide 53

MODA2 / NOMODA2
Abbreviation: None.

Arguments: MODA2

Default: NOMODA2

µVision2 Control: Options – Target – Use multiple DPTR registers 2

Description: The MODA2 directive instructs the C51 compiler to
produce code for the additional hardware components
(specifically, the additional CPU data pointers) available in
the Atmel 80x8252 or variants and compatible derivatives.
Using additional data pointers can improve the performance
of the following library functions: memcpy, memmove,
memcmp, strcpy, and strcmp.

The NOMODA2 directive disables generation of code that

utilizes the additional CPU data pointers.

See Also: MOD517, MODDP2 MODP2

Example: C51 SAMPLE.C MODA2

#pragma moda2

C51 SAMPLE.C NOMODA2

#pragma nomoda2
 54 Chapter 2. Compiling with Cx51

MODDP2 / NOMODDP2
Abbreviation: None.

Arguments: MODDP2

Default: NOMODDP2

2 µVision2 Control: Options – Target – Use multiple DPTR registers

Description: The MODDP2 directive instructs the C51 compiler to

produce code for the additional hardware components
(specifically, the additional CPU data pointers) available in
the Dallas 80C320, C520, C530, C550, or variants and
compatible derivatives. Using additional data pointers can
improve the performance of the following library functions:
memcpy, memmove, memcmp, strcpy, and strcmp.

The NOMODDP2 directive disables generation of code that

utilizes the additional CPU data pointers.

See Also: MOD517, MODA2, MODP2

Example: C51 SAMPL320.C MODDP2

#pragma moddp2

C51 SAMPL320.C NOMODDP2

#pragma nomoddp2
Keil Software — Cx51 Compiler User’s Guide 55

MODP2 / NOMODP2
Abbreviation: None.

Arguments: MODP2

Default: NOMODP2

µVision2 Control: Options – Target – Use multiple DPTR registers 2

Description: The MODP2 directive instructs the C51 compiler to use the
additional DPTR registers (dual data pointers) that are
available in some 8051 variants from Philips or Temic.
Using additional data pointers can improve the performance
of the following library functions: memcpy, memmove,
memcmp, strcpy, and strcmp.

The NOMODP2 directive disables generation of code that

utilizes the dual DPTR registers.

See Also: MOD517, MODA2, MODDP2

Example: C51 SAMPLE.C MODP2

#pragma modp2

C51 SAMPLE.C NOMODP2

#pragma nomodp2
 56 Chapter 2. Compiling with Cx51

NOAMAKE
Abbreviation: NOAM

Arguments: None.

Default: AutoMAKE information is generated.

2 µVision2 Control: no possible. If AutoMAKE information is disabled the

µVision2 build process may not work.

Description: NOAMAKE disables the project information records

produced by the Cx51 compiler for use with AutoMAKE.
This option also disables the register optimization
information. Use NOAMAKE to generate object files that
can be used with older versions of the 8051 development
tool chain.

Example: C51 SAMPLE.C NOAMAKE

#pragma NOAM
Keil Software — Cx51 Compiler User’s Guide 57

NOEXTEND
Abbreviation: None.

Arguments: None.

Default: All language extensions are enabled.

µVision2 Control: Options – C51 – Misc Controls: enter the directive 2

Description: The NOEXTEND control instructs the compiler to process
only ANSI C language constructs. The Cx51 language
extensions are disabled. Reserved keywords such as bit,
reentrant and using are not recognized and generate
compilation errors or warnings.

Example: C51 SAMPLE.C NOEXTEND

#pragma NOEXTEND
 58 Chapter 2. Compiling with Cx51

OBJECT / NOOBJECT
Abbreviation: OJ / NOOJ

Arguments: An optional filename enclosed in parentheses.

Default: OBJECT (filename.OBJ)

2 µVision2 Control: Options – Output – Select Folder for Objects

Description: The OBJECT(filename) directive changes the name of the

object file to the name provided. By default, the name and
path of the source file with the extension .OBJ is used.

The NOOBJECT control disables the generation of an

object file.

Example: C51 SAMPLE.C OBJECT(sample1.obj)

#pragma oj(sample_1.obj)

C51 SAMPLE.C NOOBJECT

#pragma nooj
Keil Software — Cx51 Compiler User’s Guide 59

OBJECTEXTEND
Abbreviation: OE

Arguments: None.

Default: None.

µVision2 Control: Options – Output – Debug Information 2

Description: The OBJECTEXTEND directive instructs the compiler to
include additional variable-type, definition information in
the generated object file. This additional information is
used to identify objects within different scopes that have the
same names so that they may be correctly differentiated by
various emulators and simulators.

NOTE
Object files generated using this directive contain a superset
of the OMF-51 specification for relocatable object formats.
Emulators or simulators must provide enhanced object
loaders to use this feature. If in doubt, do not use
OBJECTEXTEND.

See Also: DEBUG, OMF2

Example: C51 SAMPLE.C OBJECTEXTEND DEBUG

#pragma oe db
 60 Chapter 2. Compiling with Cx51

ONEREGBANK
Abbreviation: OB

Arguments: None

Default: None

2 µVision2 Control: Options – C51 – Misc controls: enter the directive.

Description: Cx51 selects registerbank 0 on entry to interrupts that do not

specify the using attribute. This is done at the beginning of
the interrupt service routine with the MOV PSW,#0
instruction. This ensures that high-priority interrupts that do
not use the using attribute can interrupt lower priority
interrupts that use a different registerbank.

If your application uses only one registerbank for interrupts,

you may use the ONEREGBANK directive. This elimitates
the MOV PSW,#0 instruction.

Example: C51 SAMPLE.C ONEREGBANK

#pragma OB
Keil Software — Cx51 Compiler User’s Guide 61

OMF2
Abbreviation: O2

Arguments: None

Default: The C51 compiler generates by default the Intel OMF51 file
format. The OMF2 file format is default for the CX51
compiler. 2
µVision2 Control: Project – Select Device – Use LX51 instead of BL51.

Description: The OMF2 directive enables the OMF2 file format. The
OMF2 file format provides detailed symbol type checking
across moduls and eliminates historic limitations of the Intel
OMF51 file format. Constants can be located also into an
xdata ROM to free code space for program code. The
OMF2 file format is required when you want to use one of
the following features of the C51 compiler:

VARBANKING directive for using the far memory type.

const xdata memory type to use a ROM in xdata space.

STRING directive to locate string constants into xdata or
far space.

ROM (D512K) or ROM (D16M) for using the
contigouos CPU mode on Dallas 390 and variants.
The OMF2 file format requires the extended LX51
linker/locater and cannot be used with the BL51
linker/locater.

See Also: OBJECTEXTEND

Example: C51 SAMPLE.C OMF2

#pragma O2
 62 Chapter 2. Compiling with Cx51

OPTIMIZE
Abbreviation: OT

Arguments: A decimal number between 0 and 9 enclosed in parentheses.

In addition, OPTIMIZE (SIZE) or OPTIMIZE (SPEED)
may be used to select whether the optimization emphasis
should be placed on code size or on execution speed.
2 Default: OPTIMIZE (8, SPEED)

µVision2 Control: Options – C51 – Code Optimization

Description: The OPTIMIZE directive sets the optimization level and

emphasis.

NOTE
Each higher optimization level contains all of the
characteristics of the preceding lower optimization level.

Level Description
0 Constant Folding: The compiler performs calculations that
reduce expressions to numeric constants, where possible.
This includes calculations of run-time addresses.

Simple Access Optimizing: The compiler optimizes access

of internal data and bit addresses in the 8051 system.

Jump Optimizing: The compiler always extends jumps to the

final target. Jumps to jumps are deleted.
1 Dead Code Elimination: Unused code fragments and
artifacts are eliminated.

Jump Negation: Conditional jumps are closely examined to

see if they can be streamlined or eliminated by the inversion of
the test logic.
2 Data Overlaying: Data and bit segments suitable for static
overlay are identified and internally marked. The BL51
Linker/Locator has the capability, through global data flow
analysis, of selecting segments which can then be overlaid.
3 Peephole Optimizing: Redundant MOV instructions are
removed. This includes unnecessary loading of objects from
the memory as well as load operations with constants.
Complex operations are replaced by simple operations when
memory space or execution time can be saved.
Keil Software — Cx51 Compiler User’s Guide 63

Level Description
4 Register Variables: Automatic variables and function
arguments are located in registers when possible.
Reservation of data memory for these variables is omitted.

Extended Access Optimizing: Variables from the IDATA,

XDATA, PDATA and CODE areas are directly included in
operations. The use of intermediate registers is not necessary
most of the time.

Local Common Subexpression Elimination: If the same

calculations are performed repetitively in an expression, the
result of the first calculation is saved and used further
2
whenever possible. Superfluous calculations are eliminated
from the code.

Case/Switch Optimizing: Code involving switch and case

statements is optimized as jump tables or jump strings.
5 Global Common Subexpression Elimination: Identical sub
expressions within a function are calculated only once when
possible. The intermediate result is stored in a register and
used instead of a new calculation.

Simple Loop Optimizing: Program loops that fill a memory

range with a constant are converted and optimized.
6 Loop Rotation: Program loops are rotated if the resulting
program code is faster and more efficient.
7 Extended Index Access Optimizing: Uses the DPTR for
register variables where appropriate. Pointer and array
access are optimized for both execution speed and code size.
8 Reuse of Common Entry Code: When there are multiple
calls to a single function, some of the setup code can be
reused, thereby reducing program size.
9 Common Block Subroutines: Detects recurring instruction
sequences and converts them into subroutines. Cx51 even
rearranges code to obtain larger recurring sequences.

OPTIMIZE level 9 includes all optimizations of levels

0 to 8.

Example: C51 SAMPLE.C OPTIMIZE (9)

C51 SAMPLE.C OPTIMIZE (0)

#pragma ot(6, SIZE)

#pragma ot(size)
 64 Chapter 2. Compiling with Cx51

ORDER
Abbreviation: OR

Arguments: None.

Default: The variables are not ordered.

2 µVision2 Control: Options – C51 – Keep Variables in Order

Description: The ORDER directive instructs Cx51 to order all variables

in memory according to their order of definition in the C
source file. ORDER disables the hash algorithm used by
the C compiler. The Cx51 compiler runs a little slower.

Example: C51 SAMPLE.C ORDER

#pragma OR
Keil Software — Cx51 Compiler User’s Guide 65

PAGELENGTH
Abbreviation: PL

Arguments: A decimal number up to 65535 enclosed in parentheses.

Default: PAGELENGTH (60)

µVision2 Control: Options – Listing – Page Length 2

Description: The PAGELENGTH directive specifies the number of lines
printed per page in the listing file. The default is 60 lines
per page. This includes headers and empty lines.

See Also: PAGEWIDTH

Example: C51 SAMPLE.C PAGELENGTH (70)

#pragma pl (70)
 66 Chapter 2. Compiling with Cx51

PAGEWIDTH
Abbreviation: PW

Arguments: A decimal number in range 78 to 132 enclosed in

parentheses.

Default: PAGEWIDTH (132)

2 µVision2 Control: Options – Listing – Page Width

Description: The PAGEWIDTH directive specifies the number of

characters per line that can be printed to the listing file.
Lines containing more than the specified number of
characters are broken into two or more lines.

See Also: PAGELENGTH

Example: C51 SAMPLE.C PAGEWIDTH(79)

#pragma pw(79)
Keil Software — Cx51 Compiler User’s Guide 67

PREPRINT
Abbreviation: PP

Arguments: An optional filename enclosed in parentheses.

Default: No preprocessor listing is generated.

µVision2 Control: Options – C51 – C Preprocessor Listing 2

Description: The PREPRINT directive instructs the compiler to produce
a preprocessor listing. Macro calls are expanded and
comments are deleted. If PREPRINT is used without an
argument, the source filename with the extension .I is
defined as the list filename. If this is not desired, you must
specify a filename. By default, Cx51 does not generate a
preprocessor output file.

NOTE
The PREPRINT directive may be specified only on the
command line. It may not be specified in the C source file
by means of the #pragma directive.

Example: C51 SAMPLE.C PREPRINT

C51 SAMPLE.C PP (PREPRO.LSI)

 68 Chapter 2. Compiling with Cx51

PRINT / NOPRINT
Abbreviation: PR / NOPR

Arguments: An optional filename enclosed in parentheses.

Default: PRINT (filename.LST)

2 µVision2 Control: Options – Listing – Select Folder for List Files

Description: The compiler produces a listing of each compiled program

using the extension .LST. Using the PRINT directive, you
may redefine the name of the listing file.

The NOPRINT directive prevents the compiler from

generating a listing file.

Example: C51 SAMPLE.C PRINT(CON:)

#pragma pr (\usr\list\sample.lst)

C51 SAMPLE.C NOPRINT

#pragma nopr
Keil Software — Cx51 Compiler User’s Guide 69

REGFILE
Abbreviation: RF

Arguments: A file name enclosed in parentheses.

Default: None.

µVision2 Control: Options – C51 – Global Register Coloring 2

Description: With REGFILE, the Cx51 compiler reads a register
definition file for global register optimization. The register
definition file specifies the register usage of external
functions. With this information the Cx51 compiler knows
about the register utilization of external functions. This
enables global program-wide register optimization.

Example: C51 SAMPLE.C REGFILE(sample.reg)

#pragma REGFILE(sample.reg)
 70 Chapter 2. Compiling with Cx51

REGISTERBANK
Abbreviation: RB

Arguments: A number between 0 and 3 enclosed in parentheses.

Default: REGISTERBANK (0)

2 µVision2 Control: Options – C51 – Misc controls: enter the directive.

Description: The REGISTERBANK directive selects which register

bank to use for subsequent functions declared in the source
file. Resulting code may use the absolute form of register
access when the absolute register number can be computed.
The using function attribute supersedes the effects of the
REGISTERBANK directive.

NOTE
Unlike the using function attribute, the REGISTERBANK
control does not switch the register bank.

Functions that return a value to the caller, must always use

the same register bank as the caller. If the register banks
are not the same, return values may be returned in registers
of the wrong register bank.

The REGISTERBANK directive may appear more than

once in a source program; however, the directive is ignored
if used within a function declaration.

Example: C51 SAMPLE.C REGISTERBANK(1)

#pragma rb(3)
Keil Software — Cx51 Compiler User’s Guide 71

REGPARMS / NOREGPARMS
Abbreviation: None.

Arguments: None.

Default: REGPARMS

µVision2 Control: Options – C51 – Misc controls: enter the directive. 2

Description: The REGPARMS directive enables the compiler to
generate code that passes up to three function arguments in
registers. This type of parameter passing is similar to what
you would use when writing in assembly and is significantly
faster than storing function arguments in memory.
Parameters that cannot be located in registers are passed
using fixed memory areas.

The NOREGPARMS directive forces all function

arguments to be passed in fixed memory areas. This
directive generates parameter passing code which is
compatible with C51, Version 2 and Version 1.

NOTE
You may specify both the REGPARMS and
NOREGPARMS directives several times within a source
program. This allows you to create some program sections
with register parameters and other sections using the old
style of parameter passing. Use NOREGPARMS to access
existing older assembler functions or library files without
having to reassemble or recompile them. This is illustrated
in the following example program.

#pragma NOREGPARMS /* Parm passing-old method */

extern int old_func (int, char);

#pragma REGPARMS /* Parm passing-new method */

extern int new_func (int, char);

main () {
char a;
int x1, x2;
x1 = old_func (x2, a);
x1 = new_func (x2, a);
}
 72 Chapter 2. Compiling with Cx51

Example: C51 SAMPLE.C NOREGPARMS

2
Keil Software — Cx51 Compiler User’s Guide 73

RET_PSTK, RET_XSTK
Abbreviation: RP, RX

Arguments: None.

Default: None.

µVision2 Control: Options – C51 – Misc controls: enter the directive. 2

Description: The RET_PSTK, and RET_XSTK directives cause the
pdata or xdata reentrant stacks to be used for return
address. Normally, return addresses are stored on the
8051’s hardware stack. These directives instruct the
compiler to generate code that pops the return address from
the hardware stack and store it on the reentrant stack
specified.

RET_PSTK Uses the compact model reentrant stack.

RET_XSTK Uses the large model reentrant stack.

NOTES
You may use the RET_xSTK directives to unload return
addresses from the on-chip or hardware stack. These
directives may be selectively used on the modules that
contain the deepest stack nesting.
If you use one of these directives you must initilize the
reentrant stack pointer defined in the startup code. Refer to
“STARTUP.A51” on page 144 for more information on how
to initilize the reentrant stacks.

1 #pragma RET_XSTK
2 extern void func2 (void);
3
4 void func (void) {
5 1 func2 ();
6 1 }

ASSEMBLY LISTING OF GENERATED OBJECT CODE

; FUNCTION func (BEGIN)
0000 120000 E LCALL ?C?CALL_XBP
; SOURCE LINE # 5
0003 120000 E LCALL func2
; SOURCE LINE # 6
0006 020000 E LJMP ?C?RET_XBP
; FUNCTION func (END)
 74 Chapter 2. Compiling with Cx51

Example: C51 SAMPLE.C RET_XSTK

2
Keil Software — Cx51 Compiler User’s Guide 75

ROM
Abbreviation: None.

Arguments: (SMALL), (COMPACT), (LARGE), (D512K) or (D16M)

Default: ROM (LARGE)

µVision2 Control: Options – Target – Code Rom Size 2

Description: You use the ROM directive to specify the size of the
program memory. This directive affects the coding of the
JMP and CALL instructions.O

Option Description
SMALL CALL and JMP instructions are coded as ACALL and
AJMP. The maximum program size may be 2 KBytes. The
entire program must be allocated within the 2 KByte
program memory space.
COMPACT CALL instructions are coded as LCALL. JMP instructions
are coded as AJMP within a function. The size of a
function must not exceed 2 KBytes. The entire program
may, however, comprise a maximum of 64 KBytes. The
type of application determines whether or not
ROM (COMPACT) is more advantageous than
ROM (LARGE). Any code space saving advantages in
using ROM (COMPACT) must be empirically determined.
LARGE CALL and JMP instructions are coded as LCALL and
LJMP. This allows you to use the entire address space
without any restrictions. Program size is limited to
64 KBytes. Function size is also limited to 64 KBytes.
D512K C51 generates ACALL and AJMP instructions with 19-bit
(Dallas 390 address for using the contiguous mode of Dallas 390 and
& variants) variants. The maximum program size may be 512 KBytes.
D16M C51 generates LCALL with 24-bit address and AJMP
(Dallas 390 instructions with 19-bit address for using the contiguous
& variants) mode of Dallas 390 and variants. The maximum program
size may be 16MBytes.
The option D512K and D16M require the OMF251 directive.

See Also: SMALL, COMPACT, LARGE

Example: C51 SAMPLE.C ROM (SMALL)

#pragma ROM (SMALL)

 76 Chapter 2. Compiling with Cx51

SAVE / RESTORE
Abbreviation: None.

Arguments: None.

Default: None.

2 µVision2 Control: This directive cannot be specified on the command line.

Description: The SAVE directive stores the current settings of AREGS,

REGPARMS and the current OPTIMIZE level and
emphasis. These settings are saved, for example, before an
#include directive and restored afterwards using
RESTORE.

The RESTORE directive fetches the values of the last

SAVE directive from the save stack.

The maximum nesting depth for SAVE directives is eight

levels.

NOTE
SAVE and RESTORE may be specified only as an argument
of a #pragma statement. You may not specify this control
option in the command line.

Example: #pragma save

#pragma noregparms

extern void test1 (char c, int i);

extern char test2 (long l, float f);

#pragma restore

In the above example, parameter passing in registers is

disabled for the two external functions, test1 and test2.
The settings at the time of the SAVE directive are restored
by the RESTORE directive.
Keil Software — Cx51 Compiler User’s Guide 77

SMALL
Abbreviation: SM

Arguments: None.

Default: SMALL

µVision2 Control: Options – Target – Memory Model 2

Description: This directive implements the SMALL memory model. The
SMALL memory model places all function variables and
local data segments in the internal data memory of the 8051
system. This allows for very efficient access to data objects.
The address space of the SMALL memory model, however,
is limited.

Regardless of memory model type, you may declare

variables in any of the 8051 memory ranges. However,
placing frequently used directives (such as loop counters and
array indices) in internal data memory significantly
improves system performance.

NOTE
The stack required for function calls is always placed in
IDATA memory.

Always start by using the SMALL memory model. Then, as

your application grows, you can place large variables and
data in other memory areas by explicitly declaring the
memory area with the variable declaration.

See Also: COMPACT, LARGE, ROM

Example: C51 SAMPLE.C SMALL

#pragma small
 78 Chapter 2. Compiling with Cx51

SRC
Abbreviation: None.

Arguments: An optional filename in parentheses.

Default: None.

2 µVision2 Control: Can be set under µVision2 as follows:

• right click on the file in the Project Window – Files tab
• choose Options for… to open Options – Properties page
• enable Generate Assembler SRC file

Description: Use the SRC directive to create an assembler source file

instead of an object file. This source file may be assembled
with the A51 assembler. If a filename is not specified in
parentheses, the base name and path of the C source file are
used with the .SRC extension.

NOTE
The compiler cannot simultaneously produce a source file
and an object file.

See Also: ASM, ENDASM

Example: C51 SAMPLE.C SRC

C51 SAMPLE.C SRC(SML.A51)

Keil Software — Cx51 Compiler User’s Guide 79

STRING
Abbreviation: ST

Arguments: (CODE), (XDATA), or (FAR)

Default: STRING (CODE)

µVision2 Control: Options – C51 – Misc controls: enter the directive. 2

Description: The STRING directive allows you to specify the memory
type used to locate implicit strings. By default implicit
strings are located in the code memory. In the following
example the string “hello world” is located in the code
memory.
void main (void) {
printf ("hello world\n");
}

By using the STRING directive you can change the location

of such strings. This option must be used carefully, since
existing programs might use memory typed pointers to
access strings. By allocating strings into the xdata or far
memory space, you may avoid the use of code banking in
your application. Especially for extended 8051 devices like
the Philips 80C51MX this option is useful.

Option Description
CODE Implicit strings are located in code space. This is the
default setting of the Cx51 compiler.
XDATA Implicit strings are located in const xdata space.
FAR Implicit strings are located in const far space.
The option XDATA and FAR require the OMF2 directive.

Example: C51 SAMPLE.C STRING (XDATA)

#pragma STRING (FAR)

 80 Chapter 2. Compiling with Cx51

SYMBOLS
Abbreviation: SB

Arguments: None.

Default: No list of symbols is generated.

2 µVision2 Control: Options – Listing – C Compiler Listing - Symbols

Description: The SYMBOLS directive instructs the compiler to generate

a list of all symbols used in and by the program module
being compiled. This list is included in the listing file. The
memory category, memory type, offset, and size are listed
for each symbolic object.

Example: C51 SAMPLE.C SYMBOLS

#pragma SYMBOLS

The following excerpt from a listing file shows the symbol

listing:
NAME CLASS MSPACE TYPE OFFSET SIZE
==== ===== ====== ==== ====== ====
EA . . . . . . . . ABSBIT ----- BIT 00AFH 1
update . . . . . . PUBLIC CODE PROC ----- -----
dtime. . . . . . PARAM DATA PTR 0000H 3
setime . . . . . . PUBLIC CODE PROC ----- -----
mode . . . . . . PARAM DATA PTR 0000H 3
dtime. . . . . . PARAM DATA PTR 0003H 3
setuptime. . . . AUTO DATA STRUCT 0006H 3
time . . . . . . . * TAG * ----- STRUCT ----- 3
hour . . . . . . MEMBER DATA U_CHAR 0000H 1
min. . . . . . . MEMBER DATA U_CHAR 0001H 1
sec. . . . . . . MEMBER DATA U_CHAR 0002H 1
SBUF . . . . . . . SFR DATA U_CHAR 0099H 1
ring . . . . . . . PUBLIC DATA BIT 0001H 1
SCON . . . . . . . SFR DATA U_CHAR 0098H 1
TMOD . . . . . . . SFR DATA U_CHAR 0089H 1
TCON . . . . . . . SFR DATA U_CHAR 0088H 1
mnu. . . . . . . . PUBLIC CODE ARRAY 00FDH 119
Keil Software — Cx51 Compiler User’s Guide 81

VARBANKING
Abbreviation: VB

Arguments: None.

Default: The standard C51 library set is used.

µVision2 Control: Options – C51 – Misc controls: enter the directive. 2

Description: The VARBANKING directive selects a different set of
library functions that is required to support variable banking
in the xdata space.

NOTE
For extended 8051 devices and variable banking with
classic 8051 devices several Keil Application Notes will
become available on www.keil.com or the Keil development
tools CD-ROM that explain the banking and memory
configuration for these devices.

Example: C51 SAMPLE.C VARBANKING

#pragma VARBANKING
 82 Chapter 2. Compiling with Cx51

WARNINGLEVEL
Abbreviation: WL

Arguments: A number from 0-2.

Default: WARNINGLEVEL (2)

2 µVision2 Control: Options – C51 – Warnings

Description: The WARNINGLEVEL directive allows you to suppress

compiler warnings. Refer to “Chapter 7. Error Messages”
on page 185 for a full list of the compiler warnings.

Warning Level Description

0 Disables almost all compiler warnings.
1 Lists only those warnings which may generate
incorrect code.
2 (Default) Lists all WARNING messages including
warnings about unused variables, expressions,
or labels.

Example: C51 SAMPLE.C WL (1)

#pragma WARNINGLEVEL (0)

Keil Software — Cx51 Compiler User’s Guide 83

2
Keil Software — Cx51 Compiler User’s Guide 85

Chapter 3. Language Extensions

Cx51 provides a number of extensions for ANSI Standard C. Most of these
provide direct support for elements of the 8051 architecture. Cx51 includes
extensions for:

Memory Types and Areas on the 8051

Memory Models

Memory Type Specifiers

Variable Data Type Specifiers

Bit variables and bit-addressable data
Special Function Registers
3

Pointers

Function Attributes

The following sections describe each of these in detail.

Keywords
To facilitate many of the features of the 8051, Cx51 adds a number of new
keywords to the scope of the C language. The following is a list of the keywords
available in Cx51:

_at_ idata sfr

alien interrupt sfr16
bdata large small
bit pdata _task_
code _priority_ using
compact reentrant xdata
data sbit

You can disable these extensions using the NOEXTEND control directive.
Refer to “Chapter 2. Compiling with Cx51” on page 19 for more information.
 86 Chapter 3. Language Extensions

8051 Memory Areas

The 8051 architecture supports a number of physically separate memory areas or
memory spaces for program and data. Each memory area offers certain
advantages and disadvantages. There are memory spaces that can be read from
but not written to, memory spaces that can be read from or written to, and
memory spaces that can be read from or written to more quickly than other
memory spaces. This wide variety of memory space is quite different from most
mainframe, minicomputer, and microcomputer architectures where the program,
data, and constants are all loaded into the same physical memory space within
the computer. Refer to the Intel 8-Bit Embedded Controllers handbook or other
8051 data books for more information about the 8051 memory architecture.
3
Program Memory
Program (CODE) memory is read only; it cannot be written to. Program memory
may reside within the 8051 CPU, it may be external, or it may be both,
depending upon the 8051 derivative and the hardware design. There may be up
to 64 KBytes of program memory. Program code including all functions and
library routines are stored in program memory. Constant variables may be stored
in program memory, as well. The 8051 executes programs stored in program
memory only.

Program memory can be accessed by using the code memory type specifier in
Cx51.
Keil Software — Cx51 Compiler User’s Guide 87

Internal Data Memory

Internal data memory resides within the 8051 CPU and can be read from and
written to. Up to 256 bytes of internal data memory are available depending
upon the 8051 derivative. The first 128 bytes of internal data memory are both
directly addressable and indirectly addressable. The upper 128 bytes of data
memory (from 0x80 to 0xFF) can be addressed only indirectly. There is also a
16 byte area starting at 20h that is bit-addressable.

Access to internal data memory is very fast because it can be accessed using an
8-bit address. However, internal data memory is limited to a maximum of 256
bytes.

Internal data can be broken down into three distinct data types when using Cx51:
data, idata, and bdata.
3
The data memory specifier always refers to the first 128 bytes of internal data
memory. Variables stored here are accessed using direct addressing.

The idata memory specifier refers to all 256 bytes of internal data memory;
however, this memory type specifier code is generated by indirect addressing
which is slower than direct addressing.

The bdata memory specifier refers to the 16 bytes of bit-addressable memory in

the internal data area (20h to 2Fh). This memory type specifier allows you to
declare data types that can also be accessed at the bit level.
 88 Chapter 3. Language Extensions

External Data Memory

External data memory can be read from and written to and is physically located
externally from the 8051 CPU. Access to external data is very slow when
compared to access to internal data. This is because external data memory is
accessed indirectly through the data pointer (DPTR) register which must be
loaded with a 16-bit address before accessing the external memory.

There may be up to 64 KBytes of external data memory; though, this address

space does not necessarily have to be used as memory. Your hardware design
may map peripheral devices into the memory space. If this is the case, your
program would access external data memory to program and control the

3 peripheral. This technique is referred to as memory-mapped I/O.

There are two different data types in Cx51 with which you may access external
data: xdata and pdata.

The xdata memory specifier refers to any location in the 64 KByte address space
of external data memory.

The pdata memory type specifier refers to only 1 page or 256 bytes of external
data memory. See “Compact Model” on page 90 for more information on pdata.
Keil Software — Cx51 Compiler User’s Guide 89

Special Function Register Memory

The 8051 also provides 128 bytes of memory for Special Function Registers
(SFRs). SFRs are bit, byte, or word-sized registers that are used to control
timers, counters, serial I/O, port I/O, and peripherals. Refer to “Special Function
Registers” on page 96 for more information on SFRs.

Memory Models
The memory model determines which default memory type to use for function
arguments, automatic variables, and declarations with no explicit memory type
specifier. You specify the memory model on the Cx51 command line using the
SMALL, COMPACT and LARGE control directives. Refer to “Control
3
Directives” on page 22 for more information about these directives.

NOTE
Except in very special selected applications, always use the default SMALL
memory model. It generates the fastest, most efficient code.

By explicitly declaring a variable with a memory type specifier, you may

override the default memory type imposed by the memory model. Refer to
“Memory Types” on page 90 for more information.

Small Model
In this model, all variables, by default, reside in the internal data memory of the
8051 system. (This is the same as if they were declared explicitly using the data
memory type specifier.) In this memory model, variable access is very efficient.
However, all objects, as well as the stack must fit into the internal RAM. Stack
size is critical because the real stack size depends upon the nesting depth of the
various functions. Typically, if the linker/locator is configured to overlay
variables in the internal data memory, the small model is the best model to use.
 90 Chapter 3. Language Extensions

Compact Model
Using the compact model, all variables, by default, reside in one page of external
data memory. (This is as if they were explicitly declared using the pdata
memory type specifier.) This memory model can accommodate a maximum of
256 bytes of variables. The limitation is due to the addressing scheme used,
which is indirect through registers R0 and R1 (@R0, @R1). This memory
model is not as efficient as the small model, therefore, variable access is not as
fast. However, the compact model is faster than the large model.

When using the compact model, Cx51 accesses external memory with
instructions that utilize the @R0 and @R1 operands. R0 and R1 are byte

3 registers and provide only the low-order byte of the address. If the compact
model is used with more than 256 bytes of external memory, the high-order
address byte (or page) is provided by Port 2 on the 8051. In this case, you must
initialize Port 2 with the proper external memory page to use. This can be done
in the startup code. You must also specify the starting address for PDATA to
the linker. Refer to “STARTUP.A51” on page 144 for more information on
using the compact model.

Large Model
In the large model, all variables, by default, reside in external data memory (up
to 64 KBytes). (This is the same as if they were explicitly declared using the
xdata memory type specifier.) The data pointer (DPTR) is used for addressing.
Memory access through this data pointer is inefficient, especially on variables
with a length of two or more bytes. This type of data access mechanism
generates more code than the small or compact models.

Memory Types
The Cx51 compiler explicitly supports the architecture of the 8051 and its
derivatives and provides access to all memory areas of the 8051. Each variable
may be explicitly assigned to a specific memory space.

Accessing the internal data memory is considerably faster than accessing the
external data memory. For this reason, place frequently used variables in
internal data memory. Place larger, less frequently used variables in external
data memory.
Keil Software — Cx51 Compiler User’s Guide 91

Explicitly Declared Memory Types

By including a memory type specifier in the variable declaration, you may
specify where variables are stored.

The following table summarizes the available memory type specifiers.

Memory Type Description

code Program memory (64 KBytes); accessed by opcode MOVC @A+DPTR.
data Directly addressable internal data memory; fastest access to variables
(128 bytes).
idata Indirectly addressable internal data memory; accessed across the full internal

bdata
address space (256 bytes).
Bit-addressable internal data memory; allows mixed bit and byte access
(16 bytes).
3
xdata External data memory (64 KBytes); accessed by opcode MOVX @DPTR.
pdata Paged (256 bytes) external data memory; accessed by opcode MOVX @Rn.

As with the signed and unsigned attributes, you may include memory type
specifiers in the variable declaration.

Example:
char data var1;
char code text[] = "ENTER PARAMETER:";
unsigned long xdata array[100];
float idata x,y,z;
unsigned int pdata dimension;
unsigned char xdata vector[10][4][4];
char bdata flags;

NOTE
For compatibility with previous versions of the Cx51 compiler, you may specify
the memory area before the data type. For example, the following declaration

data char x;

is equivalent to

char data x;

Nonetheless, this feature should not be used in new programs because it may not
be supported in future versions of the Cx51 compiler.
 92 Chapter 3. Language Extensions

Implicit Memory Types

If the memory type specifier is omitted in a variable declaration, the default or
implicit memory type is automatically selected. Function arguments and
automatic variables which cannot be located in registers are also stored in the
default memory area.

The default memory type is determined by the SMALL, COMPACT and

LARGE compiler control directives. Refer to “Memory Models” on page 89 for
more information.

3 Data Types
Cx51 provides you with a number of basic data types to use in your C programs.
Cx51 offers you the standard C data types and also supports several data types
that are unique to the 8051 platform. The following table lists the available
Cx51 data types.

Data Type Bits Bytes Value Range

bit † 1 0 to 1
signed char 8 1 -128 to +127
unsigned char 8 1 0 to 255
enum 8 / 16 1 or 2 -128 to +127 or -32768 to +32767
signed short 16 2 -32768 to +32767
unsigned short 16 2 0 to 65535
signed int 16 2 -32768 to +32767
unsigned int 16 2 0 to 65535
signed long 32 4 -2147483648 to 2147483647
unsigned long 32 4 0 to 4294967295
float 32 4 ±1.175494E-38 to ±3.402823E+38
sbit † 1 0 to 1
sfr † 8 1 0 to 255
sfr16 † 16 2 0 to 65535
† The bit, sbit, sfr, and sfr16 data types are not provided in ANSI C and are unique to Cx51.
These data types are described in detail in the following sections.
Keil Software — Cx51 Compiler User’s Guide 93

Bit Types
Cx51 provides you with a bit data type that may be used for variable
declarations, argument lists, and function return values. A bit variable is
declared the same as other C data types.

Example:
static bit done_flag = 0; /* bit variable */

bit testfunc ( /* bit function */

bit flag1, /* bit arguments */
bit flag2)
{

3
.
.
.
return (0); /* bit return value */
}

All bit variables are stored in a bit segment located in the internal memory area
of the 8051. Because this area is only 16 bytes long, a maximum of 128 bit
variables may be declared within any one scope.

Memory types may be included in the declaration of a bit variable. However,

because bit variables are stored in the internal data area of the 8051, the data
and idata memory types only may be included in the declaration. Any other
memory types are invalid.

The following restrictions apply to bit variables and bit declarations:

Functions which use disabled interrupts (#pragma disable), and functions

that are declared using an explicit register bank (using n) cannot return a bit
value. The Cx51 compiler generates an error message for functions of this
type that attempt to return a bit type.

A bit cannot be declared as a pointer. For example:
bit *ptr; /* invalid */

An array of type bit is invalid. For example:

bit ware [5]; /* invalid */
 94 Chapter 3. Language Extensions

Bit-addressable Objects
Bit-addressable objects are objects which can be addressed as bytes or as bits.
Only data objects that occupy the bit-addressable area of the 8051 internal
memory fall into this category. The Cx51 compiler places variables declared
with the bdata memory type into this bit-addressable area. You may declare
these variables as shown below:
int bdata ibase; /* Bit-addressable int */

char bdata bary [4]; /* Bit-addressable array */

The variables ibase and bary are bit-addressable. Therefore, the individual
3 bits of these variables may be directly accessed and modified. Use the sbit
keyword to declare new variables that access the bits of variables declared using
bdata. For example:
sbit mybit0 = ibase ^ 0; /* bit 0 of ibase */
sbit mybit15 = ibase ^ 15; /* bit 15 of ibase */

sbit Ary07 = bary[0] ^ 7; /* bit 7 of bary[0] */

sbit Ary37 = bary[3] ^ 7; /* bit 7 of bary[3] */

The above example represents declarations, not assignments to the bits of the
ibase and bary variables declared above. The expression following the carat
symbol (‘^’) in the example, specifies the position of the bit to access with this
declaration. This expression must be a constant value. The range depends on
the type of the base variable included in the declaration. The range is 0 to 7 for
char and unsigned char, 0 to 15 for int, unsigned int, short, and unsigned
short, and 0 to 31 for long and unsigned long.

You may provide external variable declarations for the sbit type to access these
types in other modules. For example:
extern bit mybit0; /* bit 0 of ibase */
extern bit mybit15; /* bit 15 of ibase */

extern bit Ary07; /* bit 7 of bary[0] */

extern bit Ary37; /* bit 7 of bary[3] */

Declarations involving the sbit type require that the base object be declared with
the memory type bdata. The only exceptions are the variants for special
function bits. Refer to “Special Function Registers” on page 96 for more
information.
Keil Software — Cx51 Compiler User’s Guide 95

The following example shows how to change the ibase and bary bits using the
above declarations.
Ary37 = 0; /* clear bit 7 in bary[3] */
bary[3] = 'a'; /* Byte addressing */
ibase = -1; /* Word addressing */
mybit15 = 1; /* set bit 15 in ibase */

The bdata memory type is handled like the data memory type except that
variables declared with bdata reside in the bit-addressable portion of the
internal data memory. Note that the total size of this area of memory may not
exceed 16 bytes.

In addition to declaring sbit variables for scalar types, you may also declare sbit
variables for structures and unions. For example:
union lft
3
{
float mf;
long ml;
};

bdata struct bad

{
char m1;
union lft u;
} tcp;

sbit tcpf31 = tcp.u.ml ^ 31; /* bit 31 of float */

sbit tcpm10 = tcp.m1 ^ 0;
sbit tcpm17 = tcp.m1 ^ 7;

NOTE
You may not specify bit variables for the bit positions of a float. However, you
may include the float and a long in a union. Then, you may declare bit
variables to access the bits in the long type.

The sbit data type uses the specified variable as a base address and adds the bit
position to obtain a physical bit address. Physical bit addresses are not
equivalent to logical bit positions for certain data types. Physical bit position 0
refers to bit position 0 of the first byte. Physical bit position 8 refers to bit
position 0 of the second byte. Because int variables are stored high-byte first,
bit 0 of the integer is located in bit position 0 of the second byte. This is physical
bit position 8 when accessed using an sbit data type.
 96 Chapter 3. Language Extensions

Special Function Registers

The 8051 family of microprocessors provides you with a distinct memory area
for accessing Special Function Registers (SFRs). SFRs are used in your program
to control timers, counters, serial I/Os, port I/Os, and peripherals. SFRs reside
from address 0x80 to 0xFF and can be accessed as bits, bytes, and words. For
more information about special function registers, refer to the Intel 8-Bit
Embedded Controllers handbook or other 8051 data books.

Within the 8051 family, the number and type of SFRs vary. Note that no SFR
names are predefined by the Cx51 compiler. However, declarations for SFRs
are provided in include files.
3 Cx51 provides you with a number of include files for various 8051 derivatives.
Each file contains declarations for the SFRs available on that derivative. See
“8051 Special Function Register Include Files” on page 221 for more
information about include files.

Cx51 provides access to SFRs with the sfr, sfr16, and sbit data types. The
following sections describe each of these data types.

sfr
SFRs are declared in the same fashion as other C variables. The only difference
is that the data type specified is sfr rather than char or int. For example:
sfr P0 = 0x80; /* Port-0, address 80h */
sfr P1 = 0x90; /* Port-1, address 90h */
sfr P2 = 0xA0; /* Port-2, address 0A0h */
sfr P3 = 0xB0; /* Port-3, address 0B0h */

P0, P1, P2, and P3 are the SFR name declarations. Names for sfr variables are
defined just like other C variable declarations. Any symbolic name may be used
in an sfr declaration.

The address specification after the equal sign (=) must be a numeric constant.
(Expressions with operators are not allowed.) This constant expression must lie
in the SFR address range (0x80 to 0xFF).
Keil Software — Cx51 Compiler User’s Guide 97

sfr16
Many of the newer 8051 derivatives use two SFRs with consecutive addresses to
specify 16-bit values. For example, the 8052 uses addresses 0xCC and 0xCD for
the low and high bytes of timer/counter 2. Cx51 provides the sfr16 data type to
access 2 SFRs as a 16-bit SFR.

Access to 16-bit SFRs is possible only when the low byte immediately precedes
the high byte. The low byte is used as the address in the sfr16 declaration. For
example:
sfr16 T2 = 0xCC; /* Timer 2: T2L 0CCh, T2H 0CDh */
sfr16 RCAP2 = 0xCA; /* RCAP2L 0CAh, RCAP2H 0CBh */

In this example, T2 and RCAP2 are declared as 16-bit special function registers. 3
The sfr16 declarations follow the same rules as outlined for sfr declarations.
Any symbolic name can be used in an sfr16 declaration. The address
specification after the equal sign (‘=’) must be a numeric constant. Expressions
with operators are not allowed. The address must be the low byte of the SFR
low-byte, high-byte pair.

sbit
With typical 8051 applications, it is often necessary to access individual bits
within an SFR. The Cx51 compiler makes this possible with the sbit data type.
The sbit data type allows you to access bit-addressable SFRs. For example:
sbit EA = 0xAF;

This declaration defines EA to be the SFR bit at address 0xAF. On the 8051,
this is the enable all bit in the interrupt enable register.

NOTE
Not all SFRs are bit-addressable. Only those SFRs whose address is evenly
divisible by 8 are bit-addressable. The lower nibble of the SFR’s address must
be 0 or 8. For example, SFRs at 0xA8 and 0xD0 are bit-addressable, whereas
SFRs at 0xC7 and 0xEB are not. To calculate an SFR bit address, add the bit
position to the SFR byte address. So, to access bit 6 in the SFR at 0xC8, the SFR
bit address would be 0xCE (0xC8 + 6).
 98 Chapter 3. Language Extensions

Any symbolic name can be used in an sbit declaration. The expression to the
right of the equal sign (=) specifies an absolute bit address for the symbolic
name. There are three variants for specifying the address:

Variant 1: sfr_name ^ int_constant

This variant uses a previously declared sfr (sfr_name) as the
base address for the sbit. The address of the existing SFR
must be evenly divisible by 8. The expression following the
carat symbol (^) specifies the position of the bit to access
with this declaration. The bit position must be a number in
the 0 to 7 range. For example:
sfr PSW = 0xD0;

3 sfr
sbit
sbit
IE = 0xA8;
OV = PSW ^ 2;
CY = PSW ^ 7;
sbit EA = IE ^ 7;

Variant 2: int_constant ^ int_constant

This variant uses an integer constant as the base address for
the sbit. The base address value must be evenly divisible by
8. The expression following the carat symbol (‘^’) specifies
the position of the bit to access with this declaration. The
bit position must be a number in the 0 to 7 range. For
example:
sbit OV = 0xD0 ^ 2;
sbit CY = 0xD0 ^ 7;
sbit EA = 0xA8 ^ 7;

Variant 3: int_constant
This variant uses an absolute bit address for the sbit. For
example:
sbit OV = 0xD2;
sbit CY = 0xD7;
sbit EA = 0xAF;

NOTE
Special function bits represent an independent declaration class that may not be
interchangeable with other bit declarations or bit fields.

The sbit data type declaration may be used to access individual bits of variables
declared with the bdata memory type specifier. Refer to “Bit-addressable
Objects” on page 94 for more information.
Keil Software — Cx51 Compiler User’s Guide 99

Absolute Variable Location

Variables may be located at absolute memory locations in your C program
source modules using the _at_ keyword. The usage for this feature is:

memory_space type variable_name _at_ constant;

where:
memory_space is the memory space for the variable. If missing from the
declaration, the default memory space is used. Refer to
“Memory Models” on page 89 for more information about

type
the default memory space.
is the variable type.
3
variable_name is the variable name.
constant is the address at which to locate the variable.

The absolute address following _at_ must conform to the physical boundaries of
the memory space for the variable. Cx51 checks for invalid address
specifications.

The following restrictions apply to absolute variable location:

1. Absolute variables cannot be initialized.

2. Functions and variables of type bit cannot be located at an absolute address.

 100 Chapter 3. Language Extensions

The following example demonstrates how to locate several different variable

types using the _at_ keyword.
struct link
{
struct link idata *next;
char code *test;
};

idata struct link list _at_ 0x40; /* list at idata 0x40 */

xdata char text[256] _at_ 0xE000; /* array at xdata 0xE000 */
xdata int i1 _at_ 0x8000; /* int at xdata 0x8000 */

void main ( void ) {

link.next = (void *) 0;
i1 = 0x1234;
text [0] = 'a';

3 }

Often, you may wish to declare your variables in one source module and access
them in another. Use the following external declarations to access the _at_
variables defined above in another source file.
struct link
{
struct link idata *next;
char code *test;
};

extern idata struct link list; /* list at idata 0x40 */

extern xdata char text[256]; /* array at xdata 0xE000 */
extern xdata int i1; /* int at xdata 0x8000 */
Keil Software — Cx51 Compiler User’s Guide 101

Pointers
Cx51 supports the declaration of variable pointers using the * character. Cx51
pointers can be used to perform all operations available in standard C. However,
because of the unique architecture of the 8051 and its derivatives, Cx51 provides
two different types of pointers: memory-specific pointers and generic pointers.
Each of these pointer types, as well as conversion methods are discussed in the
following sections.

Generic Pointers
Generic pointers are declared in the same fashion as standard C pointers. For
example:
3
char *s; /* string ptr */
int *numptr; /* int ptr */
long *state; /* Texas */

Generic pointers are always stored using three bytes. The first byte is the
memory type, the second is the high-order byte of the offset, and the third is the
low-order byte of the offset. Generic pointers may be used to access any variable
regardless of its location in 8051 memory space. Many of the Cx51 library
routines use these pointer types for this reason. By using these generic pointers,
a function can access data regardless of the memory in which it is stored.

NOTE
The code generated for a generic pointer executes more slowly than the
equivalent code generated for a memory-specific pointer. This is because the
memory area is not known until run-time. The compiler cannot optimize memory
accesses and must generate generic code that can access any memory area. If
execution speed is a priority, you should use memory-specific pointers instead of
generic pointers wherever possible.

The following code and assembly listing shows the values assigned to generic
pointers for variables in different memory areas. Note that the first value is the
memory space followed by the high-order byte and low-order byte of the
address.
stmt level source
1 char *c_ptr; /* char ptr */
2 int *i_ptr; /* int ptr */
3 long *l_ptr; /* long ptr */
 102 Chapter 3. Language Extensions

4
5 void main (void)
6 {
7 1 char data dj; /* data vars */
8 1 int data dk;
9 1 long data dl;
10 1
11 1 char xdata xj; /* xdata vars */
12 1 int xdata xk;
13 1 long xdata xl;
14 1
15 1 char code cj = 9; /* code vars */
16 1 int code ck = 357;
17 1 long code cl = 123456789;
18 1
19 1
20 1 c_ptr = &dj; /* data ptrs */
21 1 i_ptr = &dk;

3 22
23
24
1
1
1
l_ptr = &dl;

c_ptr = &xj; /* xdata ptrs */

25 1 i_ptr = &xk;
26 1 l_ptr = &xl;
27 1
28 1 c_ptr = &cj; /* code ptrs */
29 1 i_ptr = &ck;
30 1 l_ptr = &cl;
31 1 }

ASSEMBLY LISTING OF GENERATED OBJECT CODE

; FUNCTION main (BEGIN)

; SOURCE LINE # 5
; SOURCE LINE # 6
; SOURCE LINE # 20
0000 750000 R MOV c_ptr,#00H
0003 750000 R MOV c_ptr+01H,#HIGH dj
0006 750000 R MOV c_ptr+02H,#LOW dj
; SOURCE LINE # 21
0009 750000 R MOV i_ptr,#00H
000C 750000 R MOV i_ptr+01H,#HIGH dk
000F 750000 R MOV i_ptr+02H,#LOW dk
; SOURCE LINE # 22
0012 750000 R MOV l_ptr,#00H
0015 750000 R MOV l_ptr+01H,#HIGH dl
0018 750000 R MOV l_ptr+02H,#LOW dl
; SOURCE LINE # 24
001B 750001 R MOV c_ptr,#01H
001E 750000 R MOV c_ptr+01H,#HIGH xj
0021 750000 R MOV c_ptr+02H,#LOW xj
; SOURCE LINE # 25
0024 750001 R MOV i_ptr,#01H
0027 750000 R MOV i_ptr+01H,#HIGH xk
002A 750000 R MOV i_ptr+02H,#LOW xk
; SOURCE LINE # 26
002D 750001 R MOV l_ptr,#01H
0030 750000 R MOV l_ptr+01H,#HIGH xl
0033 750000 R MOV l_ptr+02H,#LOW xl
; SOURCE LINE # 28
0036 7500FF R MOV c_ptr,#0FFH
Keil Software — Cx51 Compiler User’s Guide 103

0039 750000 R MOV c_ptr+01H,#HIGH cj

003C 750000 R MOV c_ptr+02H,#LOW cj
; SOURCE LINE # 29
003F 7500FF R MOV i_ptr,#0FFH
0042 750000 R MOV i_ptr+01H,#HIGH ck
0045 750000 R MOV i_ptr+02H,#LOW ck
; SOURCE LINE # 30
0048 7500FF R MOV l_ptr,#0FFH
004B 750000 R MOV l_ptr+01H,#HIGH cl
004E 750000 R MOV l_ptr+02H,#LOW cl
; SOURCE LINE # 31
0051 22 RET
; FUNCTION main (END)

In the above example listing, the generic pointers c_ptr, i_ptr, and l_ptr
are all stored in the internal data memory of the 8051. However, you may
specify the memory area in which a generic pointer is stored by using a memory
type specifier. For example: 3
char * xdata strptr; /* generic ptr stored in xdata */
int * data numptr; /* generic ptr stored in data */
long * idata varptr; /* generic ptr stored in idata */

These examples are pointers to variables that may be stored in any memory area.
The pointers, however, are stored in xdata, data, and idata respectively.
 104 Chapter 3. Language Extensions

Memory-specific Pointers
Memory-specific pointers always include a memory type specification in the
pointer declaration and always refer to a specific memory area. For example:
char data *str; /* ptr to string in data */
int xdata *numtab; /* ptr to int(s) in xdata */
long code *powtab; /* ptr to long(s) in code */

Because the memory type is specified at compile-time, the memory type byte
required by generic pointers is not needed by memory-specific pointers.
Memory-specific pointers can be stored using only one byte (idata, data, bdata,
and pdata pointers) or two bytes (code and xdata pointers).

3 NOTE
The code generated for a memory-specific pointer executes more quickly than
the equivalent code generated for a generic pointer. This is because the memory
area is known at compile-time rather than at run-time. The compiler can use
this information to optimize memory accesses. If execution speed is a priority,
you should use memory-specific pointers instead of generic pointers wherever
possible.

Like generic pointers, you may specify the memory area in which a
memory-specific pointer is stored. To do so, prefix the pointer declaration with
a memory type specifier. For example:
char data * xdata str; /* ptr in xdata to data char */
int xdata * data numtab; /* ptr in data to xdata int */
long code * idata powtab; /* ptr in idata to code long */

Memory-specific pointers may be used to access variables in the declared 8051

memory area only. Memory-specific pointers provide the most efficient method
of accessing data objects, but at the cost of reduced flexibility.
Keil Software — Cx51 Compiler User’s Guide 105

The following code and assembly listing shows how pointer values are assigned
to memory-specific pointers. Note that the code generated for these pointers is
much less involved than the code generated in the generic pointers example
listing in the previous section.
stmt level source

1 char data *c_ptr; /* memory-specific char ptr */

2 int xdata *i_ptr; /* memory-specific int ptr */
3 long code *l_ptr; /* memory-specific long ptr */
4
5 long code powers_of_ten [] =
6 {
7 1L,
8 10L,
9 100L,
10
11
12
1000L,
10000L,
100000L,
3
13 1000000L,
14 10000000L,
15 100000000L
16 };
17
18 void main (void)
19 {
20 1 char data strbuf [10];
21 1 int xdata ringbuf [1000];
22 1
23 1 c_ptr = &strbuf [0];
24 1 i_ptr = &ringbuf [0];
25 1 l_ptr = &powers_of_ten [0];
26 1 }

ASSEMBLY LISTING OF GENERATED OBJECT CODE

; FUNCTION main (BEGIN)

; SOURCE LINE # 18
; SOURCE LINE # 19
; SOURCE LINE # 23
0000 750000 R MOV c_ptr,#LOW strbuf
; SOURCE LINE # 24
0003 750000 R MOV i_ptr,#HIGH ringbuf
0006 750000 R MOV i_ptr+01H,#LOW ringbuf
; SOURCE LINE # 25
0009 750000 R MOV l_ptr,#HIGH powers_of_ten
000C 750000 R MOV l_ptr+01H,#LOW powers_of_ten
; SOURCE LINE # 26
000F 22 RET
; FUNCTION main (END)
 106 Chapter 3. Language Extensions

Pointer Conversions
Cx51 can convert between memory-specific pointers and generic pointers.
Pointer conversions can be forced by explicit program code using type casts or
can be coerced by the compiler.

The Cx51 compiler coverts a memory-specific pointer into a generic pointer

when the memory-specific pointer is passed as an argument to a function which
requires a generic pointer. This is the case for functions such as printf, sprintf,
and gets which use generic pointers as arguments. For example:
extern int printf (void *format, ...);

3 extern int myfunc (void code *p, int xdata *pq);

int xdata *px;

char code *fmt = "value = %d | %04XH\n";

void debug_print (void) {

printf (fmt, *px, *px); /* fmt is converted */
myfunc (fmt, px); /* no conversions */
}

In the call to printf, the argument fmt which represents a 2-byte code pointer is
automatically converted or coerced into a 3-byte generic pointer. This is done
because the prototype for printf requires a generic pointer as the first argument.

NOTE
A memory-specific pointer used as an argument to a function is always
converted into a generic pointer if no function prototype is present. This can
cause errors if the called function actually expects a shorter pointer as an
argument. In order to avoid these kinds of errors in programs, use #include
files, and prototype all external functions. This guarantees conversion of the
necessary types by the compiler and increases the likelihood that the compiler
detects type conversion errors.
Keil Software — Cx51 Compiler User’s Guide 107

The following table details the process involved in converting generic pointers
(generic *) to memory-specific pointers (code *, xdata *, idata *, data *,
pdata *).

Conversion Type Description

generic * to code * The offset section (2 bytes) of the generic pointer is used.
generic * to xdata * The offset section (2 bytes) of the generic pointer is used.
generic * to data * The low-order byte of the generic pointer offset is used.
The high-order byte is discarded.
generic * to idata * The low-order byte of the generic pointer offset is used.
The high-order byte is discarded.
generic * to pdata * The low-order byte of the generic pointer offset is used.
The high-order byte is discarded.

The following table describes the process involved in converting

3
memory-specific pointers (code *, xdata *, idata *, data *, pdata *) to generic
pointers (generic *).

Conversion Type Description

xdata * to generic * The memory type of the generic pointer is set to 0x01 for xdata.
The 2-byte offset of the xdata * is used.
code * to generic * The memory type of the generic pointer is set to 0xFF for code.
The 2-byte offset of the code * is used.
idata * to generic * The memory type of the generic pointer is set to 0x00 for idata / data.
data * to generic * The 1-byte offset of the idata * / data * is converted to an unsigned
int and used as the offset.
pdata * to generic * The memory type of the generic pointer is set to 0xFE for pdata.
The 1-byte offset of the pdata * is converted to an unsigned int and
used as the offset.
 108 Chapter 3. Language Extensions

The following listing illustrates a few pointer conversions and the resulting code:
stmt level source
1 int *p1; /* generic ptr (3 bytes) */
2 int xdata *p2; /* xdata ptr (2 bytes) */
3 int idata *p3; /* idata ptr (1 byte) */
4 int code *p4; /* code ptr (2 bytes */
5
6 void pconvert (void) {
7 1 p1 = p2; /* xdata* to generic* */
8 1 p1 = p3; /* idata* to generic* */
9 1 p1 = p4; /* code* to generic* */
10 1
11 1 p4 = p1; /* generic* to code* */
12 1 p3 = p1; /* generic* to idata* */
13 1 p2 = p1; /* generic* to xdata* */
14 1

3 15

16
1
*** WARNING
1
*** WARNING
p2 = p3;

p3 = p4;
/* idata* to
259 IN LINE 15 OF P.C: pointer:
/* code* to
259 IN LINE 16 OF P.C: pointer:
xdata* (WARN) */
different mspace
idata* (WARN) */
different mspace
17 1 }

ASSEMBLY LISTING OF GENERATED OBJECT CODE

; FUNCTION pconvert (BEGIN)
; SOURCE LINE # 7
0000 750001 R MOV p1,#01H
0003 850000 R MOV p1+01H,p2
0006 850000 R MOV p1+02H,p2+01H
; SOURCE LINE # 8
0009 750000 R MOV p1,#00H
000C 750000 R MOV p1+01H,#00H
000F 850000 R MOV p1+02H,p3
; SOURCE LINE # 9
0012 7B05 MOV R3,#0FFH
0014 AA00 R MOV R2,p4
0016 A900 R MOV R1,p4+01H
0018 8B00 R MOV p1,R3
001A 8A00 R MOV p1+01H,R2
001C 8900 R MOV p1+02H,R1
; SOURCE LINE # 11
001E AE02 MOV R6,AR2
0020 AF01 MOV R7,AR1
0022 8E00 R MOV p4,R6
0024 8F00 R MOV p4+01H,R7
; SOURCE LINE # 12
0026 AF01 MOV R7,AR1
0028 8F00 R MOV p3,R7
; SOURCE LINE # 13
002A AE02 MOV R6,AR2
002C 8E00 R MOV p2,R6
002E 8F00 R MOV p2+01H,R7
; SOURCE LINE # 15
0030 750000 R MOV p2,#00H
0033 8F00 R MOV p2+01H,R7
; SOURCE LINE # 16
0035 850000 R MOV p3,p4+01H
; SOURCE LINE # 17
0038 22 RET
; FUNCTION pconvert (END)
Keil Software — Cx51 Compiler User’s Guide 109

Abstract Pointers
Abstract pointer types let you access fixed memory locations in any memory
area. You may also use abstract pointers to call functions located at absolute or
fixed addresses.

Abstract pointer types are described here through code examples which use the
following variables.
char xdata *px; /* ptr to xdata */
char idata *pi; /* ptr to idata */
char code *pc; /* ptr to code */

char c; /* char variable in data space */

int i; /* int variable in data space */
3
The following example assigns the address of the main C function to a pointer
(stored in data memory) to a char stored in code memory.
Source pc = (void *) main;
Object 0000 750000 R MOV pc,#HIGH main
0003 750000 R MOV pc+01H,#LOW main

The following example casts the address of the variable i (which is an

int data *) to a pointer to a char in idata. Since i is stored in data and since
indirectly accessed data is idata, this pointer conversion is valid.
Source pi = (char idata *) &i;
Object 0000 750000 R MOV pi,#LOW i

The following example casts a pointer to a char in xdata to a pointer to a char

in idata. Since xdata pointers occupy 2 bytes and idata pointers occupy 1 byte,
this pointer conversion may not yield the desired results since the upper byte of
the xdata pointer is ignored. Refer to “Pointer Conversions” on page 106 for
more information about converting between different pointer types.
Source pi = (char idata *) px;
Object 0000 850000 R MOV pi,px+01H

The following example casts 0x1234 as a pointer to a char in code memory.

Source pc = (char code *) 0x1234;
Object 0000 750012 R MOV pc,#012H
0003 750034 R MOV pc+01H,#034H
 110 Chapter 3. Language Extensions

The following example casts 0xFF00 as a function pointer that takes no

arguments and returns an int, invokes the function, and assigns the return value
to the variable i. The portion of this example that performs the function pointer
type cast is: ((int (code *)(void)) 0xFF00). By adding the argument list to the
end of the function pointer, the compiler can correctly invoke the function.
Source i = ((int (code *)(void)) 0xFF00) ();
Object 0000 12FF00 LCALL 0FF00H
0003 8E00 R MOV i,R6
0005 8F00 R MOV i+01H,R7

The following example casts 0x8000 as a pointer to a char in code memory,

extracts the char pointed to, and assigns it to the variable c.

3 Source
Object
c = *((char code *) 0x8000);
0000 908000 MOV DPTR,#08000H
0003 E4 CLR A
0004 93 MOVC A,@A+DPTR
0005 F500 R MOV c,A

The following example casts 0xFF00 as a pointer to a char in xdata memory,

extracts the char pointed to, and adds it to the variable c.
Source c += *((char xdata *) 0xFF00);
Object 0000 90FF00 MOV DPTR,#0FF00H
0003 E0 MOVX A,@DPTR
0004 2500 R ADD A,c
0006 F500 R MOV c,A

The following example casts 0xF0 as a pointer to a char in idata memory,

extracts the char pointed to, and adds it to the variable c.
Source c += *((char idata *) 0xF0);
Object 0000 78F0 MOV R0,#0F0H
0002 E6 MOV A,@R0
0003 2500 R ADD A,c
0005 F500 R MOV c,A
Keil Software — Cx51 Compiler User’s Guide 111

The following example casts 0xE8 as a pointer to a char in pdata memory,

extracts the char pointed to, and adds it to the variable c.
Source c += *((char pdata *) 0xE8);
Object 0000 78E8 MOV R0,#0E8H
0002 E2 MOVX A,@R0
0003 2500 R ADD A,c
0005 F500 R MOV c,A

The following example casts 0x2100 as a pointer to an int in code memory,

extracts the int pointed to, and assigns it to the variable i.
Source i = *((int code *) 0x2100);
Object 0000 902100 MOV DPTR,#02100H
0003
0004
0005
E4
93
FE
CLR
MOVC
MOV
A
A,@A+DPTR
R6,A
3
0006 7401 MOV A,#01H
0008 93 MOVC A,@A+DPTR
0009 8E00 R MOV i,R6
000B F500 R MOV i+01H,A

The following example casts 0x4000 as a pointer to a pointer in xdata that

points to a char in xdata. The assignment extracts the pointer stored in xdata
that points to the char which is also stored in xdata.
Source px = *((char xdata * xdata *) 0x4000);
Object 0000 904000 MOV DPTR,#04000H
0003 E0 MOVX A,@DPTR
0004 FE MOV R6,A
0005 A3 INC DPTR
0006 E0 MOVX A,@DPTR
0007 8E00 R MOV px,R6
0009 F500 R MOV px+01H,A
 112 Chapter 3. Language Extensions

Like the previous example, this example casts 0x4000 as a pointer to a pointer in
xdata that points to a char in xdata. However, the pointer is accessed as an
array of pointers in xdata. The assignment accesses array element 0 (which is
stored at 0x4000 in xdata) and extracts the pointer there that points to the char
stored in xdata.
Source px = ((char xdata * xdata *) 0x4000) [0];
Object 0000 904000 MOV DPTR,#04000H
0003 E0 MOVX A,@DPTR
0004 FE MOV R6,A
0005 A3 INC DPTR
0006 E0 MOVX A,@DPTR
0007 8E00 R MOV px,R6
0009 F500 R MOV px+01H,A

3 The following example is identical to the previous one except that the
assignment accesses element 1 from the array. Since the object pointed to is a
pointer in xdata (to a char), the size of each element in the array is 2 bytes. The
assignment accesses array element 1 (which is stored at 0x4002 in xdata) and
extracts the pointer there that points to the char stored in xdata.
Source px = ((char xdata * xdata *) 0x4000) [1];
Object 0000 904002 MOV DPTR,#04002H
0003 E0 MOVX A,@DPTR
0004 FE MOV R6,A
0005 A3 INC DPTR
0006 E0 MOVX A,@DPTR
0007 8E00 R MOV px,R6
0009 F500 R MOV px+01H,A
Keil Software — Cx51 Compiler User’s Guide 113

Function Declarations
Cx51 provides you with a number of extensions for standard C function
declarations. These extensions allow you to:

Specify a function as an interrupt procedure

Choose the register bank used

Select the memory model

Specify reentrancy

Specify alien (PL/M-51) functions

You include these extensions or attributes (many of which may be combined) in

3
the function declaration. Use the following standard format for your Cx51
function declarations.

return_type funcname (
args)
{small | compact | large}

reentrant
interrupt n
using n

where:
return_type is the type of the value returned from the function.
If no type is specified, int is assumed.
funcname is the name of the function.
args is the argument list for the function.
small, compact, or large is the explicit memory model for the function.
reentrant indicates that the function is recursive or reentrant.
interrupt indicates that the function is an interrupt function.
using specifies which register bank the function uses.

Descriptions of these attributes and other features are described in detail in the
following sections.
 114 Chapter 3. Language Extensions

Function Parameters and the Stack

The stack pointer on the classic 8051 accesses internal data memory only. Cx51
locates the stack area immediately following all variables in the internal data
memory. The stack pointer accesses internal memory indirectly and can use all
of the internal data memory up to the 0xFF limit.

The total stack space of the classic 8051 is limited: only 256 bytes maximum.
Rather than consume stack space with function parameters or arguments, Cx51
assigns a fixed memory location for each function parameter. When a function
is called, the caller must copy the arguments into the assigned memory locations
before transferring control to the desired function. The function then extracts its

3 parameters, as needed, from these fixed memory locations. Only the return
address is stored on the stack during this process. Interrupt functions require
more stack space because they must switch register banks and save the values of
a few registers on the stack.

NOTE
Cx51 allows you to use also extended stack areas that are available in some
enhanced 8051 variants. In this way the stack space can be increase to several
Kbytes.

By default, the Cx51 compiler passes up to three function arguments in registers.

This enhances speed performance. For more information, refer to “Passing
Parameters in Registers” on page 115.

NOTE
Some 8051 derivatives provide only 64 bytes of on-chip data memory; most
devices have just 256 bytes. Take this into consideration when determining
which memory model to use, because the amount of on-chip data and idata
memory used directly affects the amount of stack space.
Keil Software — Cx51 Compiler User’s Guide 115

Passing Parameters in Registers

The Cx51 compiler allows up to three function arguments to be passed in CPU
registers. This mechanism significantly improves system performance as
arguments do not have to be written to and read from memory. Argument or
parameter passing can be controlled by the REGPARMS and NOREGPARMS
control directives defined in the previous chapter.

The following table details the registers used for different argument positions
and data types.

Argument Number char, 1-byte ptr int, 2-byte ptr long, float generic ptr
1
2
R7
R5
R6 & R7
R4 & R5
R4—R7
R4—R7
R1—R3
R1—R3
3
3 R3 R2 & R3 R1—R3

If no registers are available for argument passing, fixed memory locations are
used for function parameters.

Function Return Values

CPU registers are always used for function return values. The following table
lists the return types and the registers used for each.

Return Type Register Description

bit Carry Flag
char, unsigned char, R7
1-byte ptr
int, unsigned int, R6 & R7 MSB in R6, LSB in R7
2-byte ptr
long, unsigned long R4-R7 MSB in R4, LSB in R7
float R4-R7 32-Bit IEEE format
generic ptr R1-R3 Memory type in R3, MSB R2, LSB R1

NOTE
If the first parameter of a function is of type bit, other parameters are not passed
in registers. This is because the parameters that can be passed in registers are
out of sequence with the numbering scheme shown above. For this reason, bit
parameters should be declared at the end of the argument list.
 116 Chapter 3. Language Extensions

Specifying the Memory Model for a Function

Cx51 functions normally use the default memory model to determine which
memory space to use for function arguments and local variables. Refer to
“Memory Models” on page 89 for more information.

You may, however, specify which memory model to use for a single function by
including the small, compact, or large function attribute in the function
declaration. For example:
#pragma small /* Default to small model */

extern int calc (char i, int b) large reentrant;

3 extern int func (int i, float f) large;

extern void *tcp (char xdata *xp, int ndx) small;

int mtest (int i, int y) /* Small model */

{
return (i * y + y * i + func(-1, 4.75));
}

int large_func (int i, int k) large /* Large model */

{
return (mtest (i, k) + 2);
}

The advantage of functions using the SMALL memory model is that the local
data and function argument parameters are stored in the internal 8051 RAM.
Therefore, data access is very efficient. The internal memory is limited,
however. Occasionally, the limited amount of internal data memory available
when using the small model cannot satisfy the requirements of a very large
program, and other memory models must be used. In this situation, you may
declare that a function uses a different memory model, as shown above.

By specifying the function model attribute in the function declaration, you can
select which of the three possible reentrant stacks and frame pointers are used.
Stack access in the SMALL model is more efficient than in the LARGE model.
Keil Software — Cx51 Compiler User’s Guide 117

Specifying the Register Bank for a Function

The lowest 32 bytes of all members of the 8051 family are grouped into 4 banks
of 8 registers each. Programs can access these registers as R0 through R7. The
register bank is selected by two bits of the program status word (PSW). Register
banks are useful when processing interrupts or when using a real-time operating
system. Rather than saving the 8 registers, the CPU can switch to a different
register bank for the duration of the interrupt service routine.

The using function attribute is used to specify which register bank a function
uses. For example:

3
void rb_function (void) using 3
{
.
.
.
}

The using attribute takes as an argument an integer constant in the 0 to 3 range

value. Expressions with operators are not allowed. The using attribute is not
allowed in function prototypes. The using attribute affects the object code of the
function as follows:

The currently selected register bank is saved on the stack at function entry.

The specified register bank is set.

The former register bank is restored before the function is exited.
 118 Chapter 3. Language Extensions

The following example shows how to specify the using function attribute and
what the generated assembly code for the function entry and exit looks like.
stmt level source

1
2 extern bit alarm;
3 int alarm_count;
4 extern void alfunc (bit b0);
5
6 void falarm (void) using 3 {
7 1 alarm_count++;
8 1 alfunc (alarm = 1);
9 1 }

ASSEMBLY LISTING OF GENERATED OBJECT CODE

3 ; FUNCTION falarm (BEGIN)

0000 C0D0 PUSH PSW
0002 75D018 MOV PSW,#018H
; SOURCE LINE # 6
; SOURCE LINE # 7
0005 0500 R INC alarm_count+01H
0007 E500 R MOV A,alarm_count+01H
0009 7002 JNZ ?C0002
000B 0500 R INC alarm_count
000D ?C0002:
; SOURCE LINE # 8
000D D3 SETB C
000E 9200 E MOV alarm,C
0010 9200 E MOV ?alfunc?BIT,C
0012 120000 E LCALL alfunc
; SOURCE LINE # 9
0015 D0D0 POP PSW
0017 22 RET
; FUNCTION falarm (END)

In the previous example, the code starting at offset 0000h saves the initial PSW
on the stack and sets the new register bank. The code starting at offset 0015h
restores the original register bank by popping the original PSW from the stack.

The using attribute may not be used in functions that return a value in registers.
You must exercise extreme care to ensure that register bank switches are
performed only in carefully controlled areas. Failure to do so may yield
incorrect function results. Even when you use the same register bank, functions
declared with the using attribute cannot return a bit value.

Typically, the using attribute is most useful in functions that also specify the
interrupt attribute. It is most common to specify a different register bank for
each interrupt priority level. Therefore, you could use one register bank for all
non-interrupt code, one for the high-level interrupt, and one for the low-level
interrupt.
Keil Software — Cx51 Compiler User’s Guide 119

Register Bank Access

The Cx51 compiler allows you to define the default register bank in a function.
The REGISTERBANK control directive allows you to specify which default
register bank to use for all functions in a source file. This directive, however,
does not generate code to switch the register bank.

Upon reset, the 8051 loads the PSW with 00h which selects register bank 0. By
default, all non-interrupt functions use register bank 0. To change this, you
must:

Modify the startup code to select a different register bank

Specify the REGISTERBANK control directive along with the new register
bank number
3
By default, the Cx51 compiler generates code that accesses the registers R0—R7
using absolute addresses. This is done for maximum performance. Absolute
register accesses are controlled by the AREGS and NOAREGS control
directives. Functions which employ absolute register accesses must not be called
from another function that uses a different register bank. Doing so causes
unpredictable results because the called function assumes that a different register
bank is selected. To make a function insensitive to the current register bank, the
function must be compiled using the NOAREGS control directive. This would
be useful for a function that was called from the main program and also from an
interrupt function that uses a different register bank.

NOTE
The Cx51 compiler does not and cannot detect a register bank mismatch
between functions. Therefore, make sure that functions using alternate register
banks call only other functions that do not assume a default register bank.

Refer to “Chapter 2. Compiling with Cx51” on page 19 for more information

regarding the REGISTERBANK, AREGS, and NOARGES directives.
 120 Chapter 3. Language Extensions

Interrupt Functions
The 8051 and its derivatives provide a number of hardware interrupts that may
be used for counting, timing, detecting external events, and sending and
receiving data using the serial interface. The standard interrupts found on an
8051 are listed in the following table:

Interrupt Number Interrupt Description Address

0 EXTERNAL INT 0 0003h
1 TIMER/COUNTER 0 000Bh
2 EXTERNAL INT 1 0013h
3 TIMER/COUNTER 1 001Bh

3 4 SERIAL PORT 0023h

As 8051 vendors created new parts, more interrupts were added. The Cx51
compiler supports interrupt functions for 32 interrupts (0-31). Use the interrupt
vector address in the following table to determine the interrupt number.

Interrupt Number Address Interrupt Number Address

0 0003h 16 0083h
1 000Bh 17 008Bh
2 0013h 18 0093h
3 001Bh 19 009Bh
4 0023h 20 00A3h
5 002Bh 21 00ABh
6 0033h 22 00B3h
7 003Bh 23 00BBh
8 0043h 24 00C3h
9 004Bh 25 00CBh
10 0053h 26 00D3h
11 005Bh 27 00DBh
12 0063h 28 00E3h
13 006Bh 29 00EBh
14 0073h 30 00F3h
15 007Bh 31 00FBh
Keil Software — Cx51 Compiler User’s Guide 121

The Cx51 compiler provides you with a method of calling a C function when an
interrupt occurs. This support lets you create interrupt service routines in C.
You need only be concerned with the interrupt number and register bank
selection. The compiler automatically generates the interrupt vector and entry
and exit code for the interrupt routine. The interrupt function attribute, when
included in a declaration, specifies that the associated function is an interrupt
function. For example:
unsigned int interruptcnt;
unsigned char second;

void timer0 (void) interrupt 1 using 2 {

if (++interruptcnt == 4000) { /* count to 4000 */
second++; /* second counter */
interruptcnt = 0; /* clear int counter */

}
}
3
The interrupt attribute takes as an argument an integer constant in the 0 to 31
value range. Expressions with operators and the interrupt attribute are not
allowed in function prototypes. The interrupt attribute affects the object code
of the function as follows:

The contents of the SFR ACC, B, DPH, DPL, and PSW, when required, are
saved on the stack at the function invocation time.

All working registers that are used in the interrupt function are stored on the
stack if a register bank is not specified with the using attribute.

The working registers and special registers that were saved on the stack are
restored before exiting the function.

The function is terminated by the 8051 RETI instruction.
 122 Chapter 3. Language Extensions

The following sample program shows you how to use the interrupt attribute.
The program also shows you what the code generated to enter and exit the
interrupt function looks like. The using function attribute is also used in the
example to select a register bank different from that of the non-interrupt program
code. However, because no working registers are needed in this function, the
code generated to switch the register bank is eliminated by the optimizer.
stmt level source

1 extern bit alarm;

2 int alarm_count;
3
4
5 void falarm (void) interrupt 1 using 3 {
6 1 alarm_count *= 2;

3 7
8
1
1
alarm = 1;
}

ASSEMBLY LISTING OF GENERATED OBJECT CODE

; FUNCTION falarm (BEGIN)

0000 C0E0 PUSH ACC
0002 C0D0 PUSH PSW
; SOURCE LINE # 5
; SOURCE LINE # 6
0004 E500 R MOV A,alarm_count+01H
0006 25E0 ADD A,ACC
0008 F500 R MOV alarm_count+01H,A
000A E500 R MOV A,alarm_count
000C 33 RLC A
000D F500 R MOV alarm_count,A
; SOURCE LINE # 7
000F D200 E SETB alarm
; SOURCE LINE # 8
0011 D0D0 POP PSW
0013 D0E0 POP ACC
0015 32 RETI
; FUNCTION falarm (END)

In the example above, note that the ACC and PSW registers are saved at offset
0000h and restored at offset 0011h. Note also the RETI instruction generated
to exit the interrupt.
Keil Software — Cx51 Compiler User’s Guide 123

The following rules apply to interrupt functions.

No function arguments may be specified for an interrupt function. The

compiler emits an error message if an interrupt function is declared with any
arguments.

Interrupt function declarations may not include a return value. They must be
declared as void (see the above examples). The compiler emits an error
message if any attempt is made to define a return value for the interrupt
function. The implicit int return value, however, is ignored by the compiler.

The compiler recognizes direct invocations of interrupt functions and
summarily rejects them. It is pointless to invoke interrupt procedures
directly, because exiting the procedure causes execution of the RETI
instruction which affects the hardware interrupt system of the 8051 chip.
Because no interrupt request on the part of the hardware existed, the effect of
3
this instruction is indeterminate and usually fatal. Do not call an interrupt
function indirectly through a function pointer.

The compiler generates an interrupt vector for each interrupt function. The
code generated for the vector is a jump to the beginning of the interrupt
function. Generation of interrupt vectors can be suppressed by including the
NOINTVECTOR control directive in the Cx51 command line. In this case,
you must provide interrupt vectors from separate assembly modules. Refer to
the INTVECTOR and INTERVAL control directives for more information
about the interrupt vector table.

The Cx51 compiler allows interrupt numbers within the 0 to 31 range.
Refer to your 8051 derivative document to determine which interrupts are
available.

Functions that are invoked from an interrupt procedure must function with the
same register bank as the interrupt procedure. When the NOAREGS
directive is not explicitly specified, the compiler may generate absolute
register accesses using the register bank selected (by the using attribute or by
the REGISTERBANK control) for that function. Unpredictable results may
occur when a function assumes a register bank other than the one currently
selected. Refer to “Register Bank Access” on page 119 for more information.
 124 Chapter 3. Language Extensions

Reentrant Functions
A reentrant function can be shared by several processes at the same time. When
a reentrant function is executing, another process can interrupt the execution and
then begin to execute that same reentrant function. Normally, functions in Cx51
cannot be called recursively or in a fashion which causes reentrancy. The reason
for this limitation is that function arguments and local variables are stored in
fixed memory locations. The reentrant function attribute allows you to declare
functions that may be reentrant and, therefore, may be called recursively. For
example:
int calc (char i, int b) reentrant {
int x;

3 }
x = table [i];
return (x * b);

Reentrant functions can be called recursively and can be called simultaneously

by two or more processes. Reentrant functions are often required in real-time
applications or in situations where interrupt code and non-interrupt code must
share a function.

As in the above example, you may selectively define (using the reentrant
attribute) functions as being reentrant. For each reentrant function, a reentrant
stack area is simulated in internal or external memory depending upon the
memory model used, as follows:

Small model reentrant functions simulate the reentrant stack in idata

memory.

Compact model reentrant functions simulate the reentrant stack in pdata
memory.

Large model reentrant functions simulate the reentrant stack in xdata
memory.
Keil Software — Cx51 Compiler User’s Guide 125

Reentrant functions use the default memory model to determine which memory
space to use for the reentrant stack. You may specify (with the small, compact,
and large function attributes) which memory model to use for a function. Refer
to “Specifying the Memory Model for a Function” on page 116 for more
information about memory models and function declarations.

The following rules apply to functions declared with the reentrant attribute.

bit type function arguments may not be used. Local bit scalars are also not
available. The reentrant capability does not support bit-addressable variables.

Reentrant functions must not be called from alien functions.

Reentrant function cannot use the alien attribute specifier to enable PL/M-51
argument passing conventions. 3

A reentrant function may simultaneously have other attributes like using and
interrupt and may include an explicit memory model attribute (small,
compact, large).

Return addresses are stored in the 8051 hardware stack. Any other required
PUSH and POP operations also affect the 8051 hardware stack.

Reentrant functions using different memory models may be intermixed.
However, each reentrant function must be properly prototyped and must
include its memory model attribute in the prototype. This is necessary for
calling routines to place the function arguments in the proper reentrant stack.

Each of the three possible reentrant models contains its own reentrant stack
area and stack pointer. For example, if small and large reentrant functions
are declared in a module, both small and large reentrant stacks are created
along with two associated stack pointers (one for small and one for large).

The reentrant stack simulation architecture is inefficient, but necessary due to a

lack of suitable addressing methods available on the 8051. For this reason, use
reentrant functions sparingly.

The simulated stack used by reentrant functions has its own stack pointer which
is independent of the 8051 stack and stack pointer. The stack and stack pointer
are defined and initialized in the STARTUP.A51 file.
 126 Chapter 3. Language Extensions

The following table details the stack pointer assembler variable name, data area,
and size for each of the three memory models.

Model Stack Pointer Stack Area

SMALL ?C_IBP (1 Byte) Indirectly accessible internal memory (idata).
256 bytes maximum stack area.
COMPACT ?C_PBP (1 Byte) Page-addressable external memory (pdata).
256 bytes maximum stack area.
LARGE ?C_XBP (2 Bytes) Externally accessible memory (xdata). 64 KBytes
maximum stack area.

The simulated stack area for reentrant functions is organized from top to bottom.
The 8051 hardware stack is just the opposite and is organized bottom to top.
3 When using the SMALL memory model, both the simulated stack and the 8051
hardware stack share the same memory area but from opposite directions.

The simulated stack and stack pointers are declared and initialized in the Cx51
startup code in STARTUP.A51 which can be found in the LIB subdirectory. You
must modify the startup code to specify which simulated stack(s) to initialize in
order to use reentrant functions. You can also modify the starting address for the
top of the simulated stack(s) in the startup code. Refer to “STARTUP.A51” on
page 144 for more information on reentrant function stack areas.
Keil Software — Cx51 Compiler User’s Guide 127

Alien Function (PL/M-51 Interface)

Cx51 lets you call routines written in PL/M-51 from your C programs. You can
access PL/M-51 routines from C by declaring them external along with the alien
function type specifier. For example:
extern alien char plm_func (int, char);

char c_func (void) {

int i;
char c;

for (i = 0; i < 100; i++) {

c = plm_func (i, c); /* call PL/M func */
}

}
return (c);
3
You may also create functions in C that can be invoked by PL/M-51 routines.
To do this, use the alien function type specifier in the C function declaration.
For example:
alien char c_func (char a, int b) {
return (a * b);
}

Parameters and return values of PL/M-51 functions may be any of the following
types: bit, char, unsigned char, int, and unsigned int. Other types, including
long, float, and all types of pointers, can be declared in C functions with the
alien type specifier. However, use these types with care because PL/M-51 does
not directly support 32-bit binary integers or floating-point numbers.

Public variables declared in the PL/M-51 module are available to your C

programs by declaring them external like you would for any C variable.
 128 Chapter 3. Language Extensions

Real-time Function Tasks

The Cx51 compiler provides support for the RTX51 Full and RTX51 Tiny
real-time multitasking operating systems through use of the _task_ and
_priority_ keywords. The _task_ keyword lets you define a function as a
real-time task. The _priority_ keyword lets you specify the priority for the task.

For example:
void func (void) _task_ num _priority_ pri

where:
is a task ID number from 0 to 255 for RTX51 Full or 0 to 15
3 num
for RTX51 Tiny.
pri is the priority for the task. Refer to the RTX51 User’s Guide
or the RTX51 Tiny User’s Guide for more information.

Task functions must be declared with a void return type and a void argument list.
Keil Software — Cx51 Compiler User’s Guide 129

Chapter 4. Preprocessor
The preprocessor built into the Cx51 compiler handles directives found in the
source file. Cx51 supports all of the ANSI Standard C directives. This chapter
provides a brief overview of the directives and elements provided by the
preprocessor.

Directives
Preprocessor directives must be the first non-whitespace text specified on a line.
All directives are prefixed with the pound or number-sign character (‘#’). For
example:
#pragma
#include <stdio.h>
#define DEBUG 1

The following table lists the preprocessor directives and gives a brief description
of each.
4
Directive Description
define Defines a preprocessor macro or constant.
elif Initiates an alternative branch of the if condition, when the previous if, ifdef, ifndef,
or elif branch was not taken.
else Initiates an alternative branch when the previous if, ifdef, or ifndef branch was not
taken.
endif Ends an if, ifdef, ifndef, elif, or else block.
error Outputs an error message defined by the user. This directive instructs the
compiler to emit the specified error message.
ifdef Evaluates an expression for conditional compilation. The argument to be evaluated
is the name of a definition.
ifndef Same as ifdef but the evaluation succeeds if the definition is not defined.
if Evaluates an expression for conditional compilation.
include Reads source text from an external file. The notation sequence determines the
search sequence of the included files. Cx51 searches for include files specified
with less-than/greater-than symbols (‘<’ ‘>’) in the include file directory. Cx51
searches for include files specified with double-quotes (“ “) in the current directory.
line Specifies a line number together with an optional filename. These specifications
are used in error messages to identify the error position.
pragma Allows you to specify control directives that may be included on the C51 command
line. Pragmas may contain the same control directives that are specified on the
command line.
undef Deletes a preprocessor macro or constant definition.
 130 Chapter 4. Preprocessor

Stringize Operator
The stringize or number-sign operator (‘#’), when used within a macro
definition, converts a macro parameter into a string constant. This operator may
be used only in a macro that has a specified argument or parameter list.

When the stringize operator immediately precedes the name of one of the macro
parameters, the parameter passed to the macro is enclosed within quotation
marks and is treated as a string literal. For example:
#define stringer(x) printf (#x "\n")

stringer (text)

results in the following actual output from the preprocessor.

printf ("text\n")

4 The expansion shows that the parameter is converted literally as if it were a

string. When the preprocessor stringizes the x parameter, the resulting line is:
printf ("text" "\n")

Because strings separated by whitespace are concatenated at compile time, these

two strings are combined into "text\n".

If the string passed as a parameter contains characters that should normally be

literalized or escaped (for example, " and \), the required \ character is
automatically added.
Keil Software — Cx51 Compiler User’s Guide 131

Token-pasting operator
The token-pasting operator (##) within a macro definition combines two
arguments. It permits two separate tokens in the macro definition to be joined
into a single token.

If the name of a macro parameter used in the macro definition is immediately

preceded or followed by the token-pasting operator, the macro parameter and the
token-pasting operator are replaced by the value of the passed parameter. Text
that is adjacent to the token-pasting operator that is not the name of a macro
parameter is not affected. For example:
#define paster(n) printf ("token" #n " = %d", token##n)

paster (9);

results in the following actual output from the preprocessor.

printf ("token9 = %d", token9);

This example shows the concatenation of token##n into token9. Both the
4
stringize and the token-pasting operators are used in this example.
 132 Chapter 4. Preprocessor

Predefined Macro Constants

Cx51 provides you with predefined constants to use in preprocessor directives
and C code for more portable programs. The following table lists and describes
each one.

Constant Description
_ _C51_ _ Version number of the Cx51 compiler (for example, 610 for version 6.10).
_ _DATE_ _ Date when the compilation was started.
_ _FILE_ _ Name of the file being compiled.
_ _LINE_ _ Current line number in the file being compiled.
_ _MODEL_ _ Memory model selected (0 for small, 1 for compact, 2 for large).
_ _TIME_ _ Time when the compilation was started.
__STDC_ _ Defined to 1 to indicate full conformance with the ANSI C Standard.

4
Keil Software — Cx51 Compiler User’s Guide 133

Chapter 5. 8051 Derivatives

A number of 8051 derivatives are available that provide enhanced performance
while remaining compatible with the 8051 core. These derivatives provide
additional data pointers, very fast math operations, and reduced instruction sets.

The Cx51 compiler directly supports the enhanced features of the following
8051-based microcontrollers:

Atmel 89x8252 and variants (2 data pointers).

Dallas 80C320, 80C420, 80C520, 80C530, 80C550 an variants (2 data
pointers).

Infineon C517, C517A, C509, and variants (high-speed 32-bit and 16-bit
binary arithmetic operations, 8 data pointers).

Philips 8xC750, 8xC751, and 8xC752 (maximum code space of 2 KBytes, no
LCALL or LJMP instructions, 64 bytes internal, no external data memory).

Philips and Temic support on several device variants 2 data pointers.

The C51 compiler provides you with support for these CPUs through the use of
special libraries, library routines, and the MODxxx command-line directives.
These directives enable C51 to generate object code that takes advantage of the
enhancements mentioned above. Refer to “Chapter 2. Compiling with Cx51” on
5
page 19 for more information about these directives.
 134 Chapter 5. 8051 Derivatives

Atmel 89x8252 and variants

The Atmel 89x8252 and variants provide 2 data pointers which can be used for
memory access. Using multiple data pointers can improve the speed of library
functions like memcpy, memmove, memcmp, strcpy, and strcmp.

The MODA2 control directive instructs the C51 compiler to generate code that
uses both data pointers in your program.

The C51 compiler uses at least one data pointer in an interrupt function. If an
interrupt function is compiled using the MODA2 directive, both data pointers
are saved on the stack. This happens even if the interrupt function uses only one
data pointer.

To conserve stack space, you may compile interrupt functions with the
NOMODA2 directive. The C51 compiler does not use the second data pointer
when this directive is used.

Dallas 80C320, 420, 520, and 530

5 The Dallas Semiconductor 80C320, 80C420, 80C520, and 80C530 provides 2
data pointers which can be used for memory access. Using multiple data
pointers can improve the speed of library functions like memcpy, memmove,
memcmp, strcpy, and strcmp.

The MODDP2 control directive instructs the C51 compiler to generate code that
uses both data pointers in your program.

The C51 compiler uses at least one data pointer in an interrupt function. If an
interrupt function is compiled using the MODDP2 directive, both data pointers
are saved on the stack. This happens even if the interrupt function uses only one
data pointer.

To conserve stack space, you may compile interrupt functions with the
NOMODDP2 directive. The C51 compiler does not use the second data pointer
when this directive is used.
Keil Software — Cx51 Compiler User’s Guide 135

Infineon C517, C517A, C509 and variants

80C537
The Infineon C517, C517A, and C509 provide high-speed 32-bit and 16-bit
arithmetic operations as well as 8 data pointers which can be used for memory
access. Using the high-speed arithmetic unit improves the performance of many
int, long, and float operations. On the C515C there are also 8 data pointers
available that can be enabled with the MOD517(DP8) directive.

The MOD517 control directive instructs the C51 compiler to generate code that
utilizes the advanced features of these CPUs.

Data Pointers
The Infineon C515C, C517, C517A, and C509 provide 8 data pointers which can
be used to improve memory accesses. Using multiple data pointers can improve
the execution of library functions such as: memcpy, memmove, memcmp,
strcpy, and strcmp. The 8 data pointers of the C515C, C517, C517 and C509
can also reduce the stack load of interrupt functions.

C51 uses only 2 of the 8 data pointers at a time. In order to keep the stack load
in the interrupt routines low, C51 switches to 2 unused data pointers when
5
switching the register bank. In this case, the contents of the register DPSEL are
saved on the stack, and a new pair of data pointers is selected. Saving the data
pointers on the stack is no longer required.

If an interrupt routine does not switch to another register bank (for example, the
function is declared without the using attribute), the data pointers must be saved
on the stack (using 4 bytes of stack space). To keep the size of the stack as small
as possible, use the MOD517(NODP8) directive to compile the interrupt routine
and the functions called from within the interrupt. This generates code for the
interrupt that uses only one data pointer and, therefore, only 2 bytes of stack
space.
 136 Chapter 5. 8051 Derivatives

High-speed Arithmetic
C51 uses the 32-bit and 16-bit arithmetic operations of the C517, C517A, and
C509 to improve performance of a number of math-intensive operations. C
language programs execute considerably faster when using either of these CPUs.

The following tables show execution times for various arithmetic operations and
compare the performance of the standard 8051 to that of the 80C517 CPU.

16-bit Binary Integer Operations

Operation CPU Routine Min. Avg. Max.
Signed/unsigned multiplication 8051 IMUL 29 29 29
C517 intrinsic 17 17 17
Unsigned division 8051 UIDIV 16 128 153
C517 UIDIV517 22 22 22
Signed division 8051 SIDIV 53 141 181
C517 SIDIV517 35 52 60
Times are shown in CPU cycles.

32-bit Binary Integer Operations

Operation CPU Routine Min. Avg. Max.

5 Signed/unsigned multiplication

Unsigned division
8051
C517
8051
LMUL
LMUL517
ULDIV
106
62
227
106
62
497
106
62
650
C517 ULDIV517 36 52 101
Signed division 8051 SLDIV 267 564 709
C517 SLDIV517 49 75 141
Left shift 8051 LSHL 5 237 470
C517 LSHL517 5 28 29
Unsigned right shift 8051 ULSHR 5 237 470
C517 ULSHR517 5 29 30
Signed right shift 8051 SLSHR 5 237 470
C517 — — — —
Times are shown in CPU cycles.
Keil Software — Cx51 Compiler User’s Guide 137

Floating-point Operations
Operation CPU Routine Min. Avg. Max.
Addition 8051 FPADD 8 107 202
C517 FPADD517 8 107 202
Subtraction 8051 FPSUB 11 113 214
C517 FPSUB517 11 113 214
Multiplication 8051 FPMUL 13 114 198
C517 FPMUL517 13 86 141
Division 8051 FPDIV 48 687 999
C517 FPDIV517 48 165 209
Comparison 8051 FPCMP 42 54 59
C517 FPCMP517 42 54 59
Square root 8051 SQRT 12 1936 2360
C517 SQRT517 12 755 882
Sine 8051 SIN 1565 2928 3476
C517 SIN517 1422 2519 3048
Cosine 8051 COS 1601 2921 3665
C517 COS517 1458 2514 3180
Tangent 8051 TAN 1982 4966 5699
C517 TAN517 1839 3753 4329
Arcsine 8051 ASIN 912 6991 8554
C517 ASIN517 912 3984 4717
Arccosine 8051 ACOS 796 7578 8579
C517 ACOS517 796 4255 4871
Arctangent

Exponential
8051
C517
8051
ATAN
ATAN517
EXP
1069
1037
233
3320
2444
3314
3712
2737
5308
5
C517 EXP517 176 2879 4724
Natural Logarithm 8051 LOG 32 3432 4128
C517 LOG517 32 2405 2926
Common Logarithm 8051 LOG10 34 3607 4328
C517 LOG10517 34 2530 3069
ASCII to float conversion 8051 FPATOF 960 3006 5611
C517 FPATOF517 722 2202 4144
Times are shown in CPU cycles.
 138 Chapter 5. 8051 Derivatives

NOTES
The execution times specified in the preceding tables do not take access times for
variables or stack operations into consideration. Actual processing times may
consume up to 100 additional cycles depending on the stack load and address
space used.

When using the arithmetic features of the C517, C517A and C509, note that
operations involving the arithmetic processor are exclusive and may not be
interrupted. Do not use the arithmetic extensions in both the main program and
an interrupt service routine.

Use the following suggestions to help guarantee that only one thread of
execution uses the arithmetic processor:

Use the MOD517 directive to compile functions which are guaranteed to

execute only in the main program or functions used by one interrupt service
routine, but not both.

Compile all remaining functions with the MOD517(NOAU) directive.

Library Routines
5 The extra features of the C517, C517A and C509 are used in several library
routines to enhance performance. These routines are listed below and are
described in detail in “Chapter 8. Library Reference” on page 205.

acos517 log10517 sqrt517

asin517 log517 sscanf517
atan517 printf517 strtod517
atof517 scanf517 tan517
cos517 sin517
exp517 sprintf517
Keil Software — Cx51 Compiler User’s Guide 139

Philips 8xC750, 8xC751, and 8xC752

The Philips 8xC750, 8xC751, and 8xC752 derivatives support a maximum of
2 KBytes of internal program memory. The CPU cannot execute LCALL and
LJMP instructions. The following must be considered when using these
devices:

A special library, 80C751.LIB, which does not use these instructions is

necessary for these devices.

The C51 compiler must be set to avoid using LJMP and LCALL
instructions. This is accomplished using the ROM(SMALL) directive.

Note that the following restrictions apply when creating programs for the
8xC750, 8xC751, and 8xC752:

Stream functions such as printf and putchar may not be used. These
functions are usually not necessary for this chip because it is only equipped
with a maximum of 2 KBytes and has no serial interface.

Floating-point operations may not be used. Only operations using char,
unsigned char, int, unsigned int, long, unsigned long, and bit data types
are allowed.

The C51 compiler must be invoked with the ROM(SMALL) control 5
directive. This control statement instructs the C51 compiler to use only
AJMP and ACALL instructions.

The library file 80C751.LIB must be included in the input module list of the
linker. For example:
BL51 myprog.obj, startup751.obj, 80C751.LIB

A special startup module, START751.A51, is required. This file contains

startup code that is comparable to that found in STARTUP.A51, but contains
no LJMP or LCALL instructions. Refer to “Customization Files” on page
143 for more information.
 140 Chapter 5. 8051 Derivatives

Philips and Atmel WM Dual DPTR

Philips and Atmel WM provide on several 8051 variants 2 data pointers which
can be used for memory access. Using multiple data pointers can improve the
speed of library functions like memcpy, memmove, memcmp, strcpy, and
strcmp.

The MODP2 control directive instructs the C51 compiler to generate code that
uses both data pointers in your program.

The C51 compiler uses at least one data pointer in an interrupt function. If an
interrupt function is compiled using the MODP2 directive, both data pointers are
saved on the stack. This happens even if the interrupt function uses only one
data pointer.

To conserve stack space, you may compile interrupt functions with the
NOMODP2 directive. The C51 compiler does not use the second data pointer
when this directive is used.

5
Keil Software — Cx51 Compiler User’s Guide 141

5
Keil Software — Cx51 Compiler User’s Guide 143

Chapter 6. Advanced Programming

Techniques
This chapter describes advanced programming information that the experienced
software engineer will find invaluable. Knowledge of most of these topics is not
necessary to successfully create an embedded 8051 target program using the
Cx51 compiler. However, the following sections provide insight into how many
nonstandard procedures can be accomplished (for example, interfacing to
PL/M-51).

This chapter discusses the following topics:

Files you can alter to customize the startup procedures or run-time execution
of several library routines in your target program

The conventions Cx51 uses to name code and data segments

How to interface Cx51 functions to assembly and PL/M-51 routines

Data storage formats for the different Cx51 data types

Different optimizing features of the Cx51 optimizing compiler

Customization Files
The Cx51 compiler includes a number of source files you can modify to adapt
your target program to a specific hardware platform. These files contain: code
that is executed upon startup (STARTUP.A51), code that is used to initialize static
variables (INIT.A51), and code that is used to perform low-level stream I/O
6
(GETKEY.C and PUTCHAR.C). Source code for the memory allocation routines
is also included in the files CALLOC.C, FREE.C, INIT_MEM.C, MALLOC.C, and
REALLOC.C. All of these source files are described in detail in the sections that
follow.

The code contained in these files is already compiled or assembled and included
in the C library. When you link, the code from the library is automatically
included.

To include custom startup or initialization routines, you must include them in the
linker command line. The following example shows you how to include custom
replacement files for STARTUP.A51 and PUTCHAR.C.
BL51 MYMODUL1.OBJ, MYMODUL2.OBJ, STARTUP.OBJ, PUTCHAR.OBJ
 144 Chapter 6. Advanced Programming Techniques

STARTUP.A51
The STARTUP.A51 file contains the startup code for a Cx51 target program. This
source file is located in the LIB directory. Include a copy of this file in each
8051 project that needs custom startup code.

This code is executed immediately upon reset of the target system and optionally
performs the following operations, in order:

Clears internal data memory

Clears external data memory

Clears paged external data memory

Initializes the small model reentrant stack and pointer

Initializes the large model reentrant stack and pointer

Initializes the compact model reentrant stack and pointer

Initializes the 8051 hardware stack pointer

Transfers control to the main C function

The STARTUP.A51 file provides you with assembly constants that you may
change to control the actions taken at startup. These are defined in the following
table.

Constant Name Description

6 IDATALEN Indicates the number of bytes of idata that are to be initialized to 0.

The default is 80h because most 8051 derivatives contain at least
128 bytes of internal data memory. Use a value of 100h for the
8052 and other derivatives that have 256 bytes of internal data
memory.
XDATASTART Specifies the xdata address to start initializing to 0.
XDATALEN Indicates the number of bytes of xdata to be initialized to 0. The
default is 0.
PDATASTART Specifies the pdata address to start initializing to 0.
PDATALEN Indicates the number of bytes of pdata to be initialized to 0. The
default is 0.
IBPSTACK Indicates whether or not the small model reentrant stack pointer
(?C_IBP) should be initialized. A value of 1 causes this pointer to
be initialized. A value of 0 prevents initialization of this pointer. The
default is 0.
Keil Software — Cx51 Compiler User’s Guide 145

Constant Name Description

IBPSTACKTOP Specifies the top start address of the small model reentrant stack
area. The default is 0xFF in idata memory.
Cx51 does not check to see if the stack area available satisfies the
requirements of the applications. It is your responsibility to perform
such a test.
XBPSTACK Indicates whether or not the large model reentrant stack pointer
(?C_XBP) should be initialized. A value of 1 causes this pointer to
be initialized. A value of 0 prevents initialization of this pointer. The
default is 0.
XBPSTACKTOP Specifies the top start address of the large model reentrant stack
area. The default is 0xFFFF in xdata memory.
Cx51 does not check to see if the available stack area satisfies the
requirements of the applications. It is your responsibility to perform
such a test.
PBPSTACK Indicates whether the compact model reentrant stack pointer
(?C_PBP) should be initialized. A value of 1 causes this pointer to
be initialized. A value of 0 prevents initialization of this pointer. The
default is 0.
PBPSTACKTOP Specifies the top start address of the compact model reentrant stack
area. The default is 0xFF in pdata memory.
Cx51 does not check to see if the available stack area satisfies the
requirements of the applications. It is your responsibility to perform
such a test.
PPAGEENABLE Enables (a value of 1) or disables (a value of 0) the initialization of
port 2 of the 8051 device. The default is 0. The addressing of port 2
allows the mapping of 256 byte variable memory in any arbitrary
xdata page.
PPAGE Specifies the value to write to Port 2 of the 8051 for pdata memory
access. This value represents the xdata memory page to use for
pdata. This is the upper 8 bits of the absolute address range to use
for pdata.
For example, if the pdata area begins at address 1000h (page 10h)
in the xdata memory, PPAGEENABLE should be set to 1, and
PPAGE should be set to 10h. The BL51 Linker/Locator must
6
contain a value between 1000h and 10FFh in the PDATA control
directive. For example:
BL51 <input modules> PDATA (1050H)
Neither BL51 nor Cx51 checks to see if the PDATA control directive
and the PPAGE assembler constant are correctly specified. You
must ensure that these parameters contain suitable values.

The following is a listing of STARTUP.A51.

;-----------------------------------------------------------------
; This file is part of the C51 Compiler package
;-----------------------------------------------------------------
; STARTUP.A51: This code is executed after processor reset.
;
; To translate this file use A51 with the following invocation:
;
; A51 STARTUP.A51
 146 Chapter 6. Advanced Programming Techniques

;
; To link the modified STARTUP.OBJ file to your application use
; the following BL51 invocation:
;
; BL51 <your object file list>, STARTUP.OBJ <controls>
;
;-----------------------------------------------------------------
; User-defined Power-On Initialization of Memory
;
; With the following EQU statements the initialization of memory
; at processor reset can be defined:
;
; the absolute start-address of IDATA memory is always 0
IDATALEN EQU 80H ; the length of IDATA memory in bytes.
;
XDATASTART EQU 0H ; the absolute start-address of XDATA memory
XDATALEN EQU 0H ; the length of XDATA memory in bytes.
;
PDATASTART EQU 0H ; the absolute start-address of PDATA memory
PDATALEN EQU 0H ; the length of PDATA memory in bytes.
;
; Notes: The IDATA space overlaps physically the DATA and BIT
; areas of the 8051 CPU. At minimum the memory space occupied from
; the C-51 run-time routines must be set to zero.
;-----------------------------------------------------------------
; Reentrant Stack Initialization
;
; The following EQU statements define the stack pointer for
; reentrant functions and initialized it:
;
; Stack Space for reentrant functions in the SMALL model.
IBPSTACK EQU 0 ; set to 1 if small reentrant is used.
IBPSTACKTOP EQU 0FFH+1 ; set top of stack to highest location+1.
;
; Stack Space for reentrant functions in the LARGE model.
XBPSTACK EQU 0 ; set to 1 if large reentrant is used.
XBPSTACKTOP EQU 0FFFFH+1 ; set top of stack to highest location+1.
;
; Stack Space for reentrant functions in the COMPACT model.

6 PBPSTACK
PBPSTACKTOP
EQU 0 ; set to 1 if compact reentrant is used.
EQU 0FFFFH+1 ; set top of stack to highest location+1.
;-----------------------------------------------------------------
; Page Definition for Using the Compact Model with 64 KByte xdata
; RAM
;
; The following EQU statements define the xdata page used for pdata
; variables. The EQU PPAGE must conform with the PPAGE control used
; in the linker invocation.
;
PPAGEENABLE EQU 0 ; set to 1 if pdata object are used.
PPAGE EQU 0 ; define PPAGE number.
;-----------------------------------------------------------------
NAME ?C_STARTUP

?C_C51STARTUP SEGMENT CODE

?STACK SEGMENT IDATA

RSEG ?STACK
DS 1

EXTRN CODE (?C_START)

Keil Software — Cx51 Compiler User’s Guide 147

PUBLIC ?C_STARTUP

CSEG AT 0
?C_STARTUP: LJMP STARTUP1

RSEG ?C_C51STARTUP
STARTUP1:

IF IDATALEN <> 0
MOV R0,#IDATALEN - 1
CLR A
IDATALOOP: MOV @R0,A
DJNZ R0,IDATALOOP
ENDIF

IF XDATALEN <> 0
MOV DPTR,#XDATASTART
MOV R7,#LOW (XDATALEN)
IF (LOW (XDATALEN)) <> 0
MOV R6,#(HIGH XDATALEN) +1
ELSE
MOV R6,#HIGH (XDATALEN)
ENDIF
CLR A
XDATALOOP: MOVX @DPTR,A
INC DPTR
DJNZ R7,XDATALOOP
DJNZ R6,XDATALOOP
ENDIF

IF PPAGEENABLE <> 0
MOV P2,#PPAGE
ENDIF

IF PDATALEN <> 0
MOV R0,#PDATASTART
MOV R7,LOW (PDATALEN)
CLR A
PDATALOOP: MOVX @R0,A

ENDIF
INC
DJNZ
R0
R7,PDATALOOP 6
IF IBPSTACK <> 0
EXTRN DATA (?C_IBP)

MOV ?C_IBP,#LOW IBPSTACKTOP

ENDIF

IF XBPSTACK <> 0
EXTRN DATA (?C_XBP)

MOV ?C_XBP,#HIGH XBPSTACKTOP

MOV ?C_XBP+1,#LOW XBPSTACKTOP
ENDIF

IF PBPSTACK <> 0
EXTRN DATA (?C_PBP)
MOV ?C_PBP,#LOW PBPSTACKTOP
ENDIF
 148 Chapter 6. Advanced Programming Techniques

MOV SP,#?STACK-1
LJMP ?C_START

END

START751.A51
The START751.A51 file contains the startup code for a C51 target program that is
to run on the Signetics 8xC751 CPU. This source file is located in the LIB
directory. To use this file, follow the instructions on how to use STARTUP.A51 in
the previous section. The only difference between the two files is that
START751.A51 is specifically used for the 8xC751 which cannot access more
than 2 KBytes of code space and can access no external data memory. For these
reasons, there are no assembler constants that can affect xdata and pdata
memory.

The following is a listing of START751.A51.

;-------------------------------------------------------------------------
; This file is part of the C51 Compiler package
;
;-------------------------------------------------------------------------
; START751.A51: This code is executed after processor reset.
;
; To translate this file use A51 with the following invocation:
;
; A51 START751.A51
;
; To link the modified START751.OBJ file to your application use the
; following BL51 invocation:

6 ;
;
;
BL51 <your object file list>, START751.OBJ <controls>

;-------------------------------------------------------------------------
;
; User-defined Power-On Initialization of Memory
;
; With the following EQU statements the initialization of memory
; at processor reset can be defined:
;
; the absolute start-address of IDATA memory is always 0
IDATALEN EQU 40H ; the length of IDATA memory in bytes.
;
; Notes: The IDATA space physically overlaps the DATA and BIT areas of
; the 80751 CPU. At minimum the memory space occupied by C51
; run-time routines must be set to zero.
;-------------------------------------------------------------------------
;
; Reentrant Stack Initialization
;
; The following EQU statements define the stack pointer for reentrant
; functions and initialized it:
Keil Software — Cx51 Compiler User’s Guide 149

;
; Stack Space for reentrant functions in the SMALL model.
IBPSTACK EQU 0 ; set to 1 if small reentrant is used.
IBPSTACKTOP EQU 0FFH+1 ; set top of stack to highest location+1.
;
;-------------------------------------------------------------------------

NAME ?C_STARTUP

?C_C51STARTUP SEGMENT CODE

?STACK SEGMENT IDATA

RSEG ?STACK
DS 1

EXTRN CODE (?C_START)

PUBLIC ?C_STARTUP

CSEG AT 0
?C_STARTUP: AJMP STARTUP1

RSEG ?C_C51STARTUP

STARTUP1:

IF IDATALEN <> 0
MOV R0,#IDATALEN - 1
CLR A
IDATALOOP: MOV @R0,A
DJNZ R0,IDATALOOP
ENDIF

IF IBPSTACK <> 0
EXTRN DATA (?C_IBP)

MOV ?C_IBP,#LOW IBPSTACKTOP

ENDIF

MOV
AJMP
SP,#?STACK-1
?C_START
6
END
 150 Chapter 6. Advanced Programming Techniques

INIT.A51
The INIT.A51 file contains the initialization routine for variables that were
explicitly initialized. If your system is equipped with a watchdog timer, you can
integrate a watchdog refresh into the initialization code using the watchdog
macro. This macro need be defined only if the initialization takes longer than
the watchdog cycle time. If you are using an 80515, the macro could be defined
as follows:
WATCHDOG MACRO
SETB WDT
SETB SWDT
ENDM

The following is a partial listing of INIT.A51.

;-------------------------------------------------------------------------
; This file is part of the C51 Compiler package
;-------------------------------------------------------------------------
; INIT.A51: This code is executed, if the application program contains
; initialized variables at file level.
;
; To translate this file use A51 with the following invocation:
;
; A51 INIT.A51
;
; To link the modified INIT.OBJ file to your application use the following
; BL51 invocation:
;
; BL51 <your object file list>, INIT.OBJ <controls>
;
;-------------------------------------------------------------------------
; User-defined Watch-Dog Refresh.

6
;
; If the C application contains many initialized variables & uses a
; watchdog it might be possible that the user has to include a watchdog
; refresh into the initialization process. The watchdog refresh routine
; can be defined in the following MACRO and can alter all CPU registers
; except DPTR.

WATCHDOG MACRO
; Include any Watchdog refresh code here
ENDM
;-------------------------------------------------------------------------
?C_START:
MOV DPTR,#?C_INITSEG
LOOP:
WATCHDOG
CLR A
MOV R6,#1
MOVC A,@A+DPTR
JZ INITEND
.
.
.
Keil Software — Cx51 Compiler User’s Guide 151

INIT751.A51
The INIT751.A51 file contains the initialization routine for variables that were
explicitly initialized. Use this initialization routine for the Signetics 8xC751.
The following is a listing of the INIT751.A51 file.
;-------------------------------------------------------------------------
; This file is part of the C51 Compiler package
;
;-------------------------------------------------------------------------
; INIT751.A51: This code is executed, if the application program
; contains initialized variables at file level.
;
; To translate this file use A51 with the following invocation:
;
; A51 INIT751.A51
;
; To link the modified INIT.OBJ file to your application use the
; following BL51 invocation:
;
; BL51 <your object file list>, INIT751.OBJ <controls>
;
;-------------------------------------------------------------------------

NAME ?C_INIT

?C_C51STARTUP SEGMENT CODE

?C_INITSEG SEGMENT CODE ; Segment with Initializing Data

EXTRN CODE (?C_INITSEGSTART)

EXTRN CODE (MAIN)

PUBLIC ?C_START

INITEND:
RSEG
AJMP MAIN
?C_C51STARTUP
6
IorPData: ; If CY=1 PData Values
CLR A
MOVC A,@A+DPTR
INC DPTR
MOV R0,A ; Start Address
IorPLoop: CLR A
MOVC A,@A+DPTR
INC DPTR
MOV @R0,A
Common: INC R0
DJNZ R7,IorPLoop
SJMP Loop

Bits: CLR A
MOVC A,@A+DPTR
INC DPTR
MOV R0,A
ANL A,#007H
ADD A,#Table-LoadTab
 152 Chapter 6. Advanced Programming Techniques

XCH A,R0
CLR C
RLC A ; Bit Condition to Carry
SWAP A
ANL A,#00FH
ORL A,#20H ; Bit Address
XCH A,R0 ; convert to Byte Address
MOVC A,@A+PC
LoadTab: JC SetIt
CPL A
ANL A,@R0
SJMP BitReady
SetIt: ORL A,@R0
BitReady: MOV @R0,A
DJNZ R7,Bits
SJMP Loop

Table: DB 00000001B
DB 00000010B
DB 00000100B
DB 00001000B
DB 00010000B
DB 00100000B
DB 01000000B
DB 10000000B

?C_START: MOV DPTR,#?C_INITSEGSTART

LOOP: CLR A
MOV R6,#1
MOVC A,@A+DPTR
JZ INITEND
INC DPTR
MOV R7,A
ANL A,#3FH
JNB ACC.5,NOBIG
ANL A,#01FH
MOV R6,A
CLR A

6 MOVC
INC
JZ
A,@A+DPTR
DPTR
NOBIG
INC R6
NOBIG: XCH A,R7
ANL A,#0C0H ; Typ is in Bit 6 and Bit 7
ADD A,ACC
JZ IorPDATA
JC Bits
SJMP $

RSEG ?C_INITSEG
DB 0

END
Keil Software — Cx51 Compiler User’s Guide 153

PUTCHAR.C
This file contains the putchar function which is the low-level character output
routine for the stream I/O routines. All stream routines that output character data
do so through this routine. You may adapt this routine to your individual
hardware (for example, LCD or LED displays).

The default PUTCHAR.C file delivered with the Cx51 compiler outputs
characters via the serial interface. An XON/XOFF protocol is used for flow
control. Linefeed characters (‘\n’) are automatically converted into carriage
return/linefeed sequences (‘\r\n’).

GETKEY.C
This file contains the _getkey function which is the low-level character input
routine for the stream I/O routines. All stream routines that input character data
do so through this routine. You may adapt this routine to your individual
hardware (for example, for matrix keyboards). The default GETKEY.C file
delivered with the Cx51 compiler reads a character via the serial interface. No
data conversions are performed.

CALLOC.C
This file contains the source code for the calloc function. This routine allocates
memory for an array from the memory pool.
6
FREE.C
This file contains the source code for the free function. This routine returns a
previously allocated memory block to the memory pool.

INIT_MEM.C
This file contains the source code for the init_mempool function. This routine
allows you to specify the location and size of a memory pool from which
memory may be allocated using the malloc, calloc, and realloc functions.
 154 Chapter 6. Advanced Programming Techniques

MALLOC.C
This file contains the source code for the malloc function. This routine allocates
memory from the memory pool.

REALLOC.C
This file contains the source code for the realloc function. This routine resizes a
previously allocated memory block.

6
Keil Software — Cx51 Compiler User’s Guide 155

Optimizer
The Cx51 compiler is an optimizing compiler. This means that the compiler
takes certain steps to ensure that the code that is generated and output to the
object file is the most efficient (smaller and/or faster) code possible. The
compiler analyzes the generated code to produce more efficient instruction
sequences. This ensures that your Cx51 program runs as quickly as possible.

The Cx51 compiler provides six different levels of optimizing. Each increasing
level includes the optimizations of the levels below it.

Level Description
0 Constant Folding: The compiler performs calculations that reduce expressions to
numeric constants, where possible. This includes calculations of run-time addresses.

Simple Access Optimizing: The compiler optimizes access of internal data and bit
addresses in the 8051 system.

Jump Optimizing: The compiler always extends jumps to the final target. Jumps to
jumps are deleted.
1 Dead Code Elimination: Unused code fragments and artifacts are eliminated.

Jump Negation: Conditional jumps are closely examined to see if they can be
streamlined or eliminated by the inversion of the test logic.
2 Data Overlaying: Data and bit segments suitable for static overlay are identified and
internally marked. The BL51 Linker/Locator has the capability, through global data
flow analysis, of selecting segments which can then be overlaid.
3 Peephole Optimizing: Redundant MOV instructions are removed. This includes
unnecessary loading of objects from the memory as well as load operations with
constants. Complex operations are replaced by simple operations when memory

4
space or execution time can be saved.
Register Variables: Automatic variables and function arguments are located in
registers when possible. Reservation of data memory for these variables is omitted.
6
Extended Access Optimizing: Variables from the IDATA, XDATA, PDATA and
CODE areas are directly included in operations. The use of intermediate registers is
not necessary most of the time.

Local Common Subexpression Elimination: If the same calculations are performed

repetitively in an expression, the result of the first calculation is saved and used further
whenever possible. Superfluous calculations are eliminated from the code.

Case/Switch Optimizing: Code involving switch and case statements is optimized as

jump tables or jump strings.
5 Global Common Subexpression Elimination: Identical sub expressions within a
function are calculated only once when possible. The intermediate result is stored in a
register and used instead of a new calculation.

Simple Loop Optimizing: Program loops that fill a memory range with a constant are
converted and optimized.
 156 Chapter 6. Advanced Programming Techniques

Level Description
6 Loop Rotation: Program loops are rotated if the resulting program code is faster and
more efficient.
7 Extended Access Optimization: In addition, uses the DPTR for register variables.
Pointer and array accesses are optimized for both speed
and code size.
8

General Optimizations
Optimization Description
Constant Folding Several constant values occurring in an expression or
address calculation are combined as a constant.
Jump Optimizing Jumps are inverted or extended to the final target address
when the program efficiency is thereby increased.
Dead Code Elimination Code which cannot be reached (dead code) is removed from
the program.
Register Variables Automatic variables and function arguments are located in
registers when possible. Reservation of data memory for
these variables is omitted.
Parameter Passing Via Registers A maximum of three function arguments can be passed in
registers.
Global Common Subexpression Identical subexpressions or address calculations that occur
Elimination multiple times in a function are recognized and calculated
only once when possible.

8051-Specific Optimizations
6 Optimization Description
Peephole Optimization Complex operations are replaced by simplified operations when
memory space or execution time can be saved as a result.
Extended Access Optimizing Constants and variables are included directly in operations.
Data Overlaying Data and bit segments of functions are identified as
OVERLAYABLE and are overlaid with other data and bit
segments by the BL51 Linker/Locator.
Case/Switch Optimizing Any switch and case statements are optimized by using a jump
table or string of jumps.
Keil Software — Cx51 Compiler User’s Guide 157

Options for Code Generation

Optimization Description
OPTIMIZE(SIZE) Common C operations are replaced by subprograms. Program code
is thereby reduced.
NOAREGS Cx51 no longer uses absolute register access. Program code is
independent of the register bank.
NOREGPARMS Parameter passing is always performed in local data segments. The
program code is compatible to earlier versions of Cx51.

6
 158 Chapter 6. Advanced Programming Techniques

Segment Naming Conventions

Objects generated by the Cx51 compiler (program code, program data, and
constant data) are stored in segments which are units of code or data memory. A
segment may be relocatable or may be absolute. Each relocatable segment has a
type and a name. This section describes the conventions used by Cx51 for
naming these segments.

Segment names include a module_name. The module_name is the name of the

source file in which the object is declared and excludes the drive letter, path
specification, and file extension. In order to accommodate a wide variety of
existing software and hardware tools, all segment names are converted and
stored in uppercase.

Each segment name has a prefix that corresponds to the memory type used for
the segment. The prefix is enclosed in question marks (?). The following is a
list of the standard segment name prefixes:

Segment Prefix Memory Type Description

?PR? program Executable program code
?CO? code Constant data in program memory
?BI? bit Bit data in internal data memory
?BA? bdata Bit-addressable data in internal data memory
?DT? data Internal data memory
?FD? far far memory (RAM space)
?FC? const far far memory (constant ROM space)

6 ?ID?
?PD?
idata
pdata
Indirectly-addressable internal data memory
Paged data in external data memory
?XD? xdata xdata memory (RAM space)
?XC? const xdata xdata memory (constant ROM space)

Data Objects
Data objects are the variables and constants you declare in your C programs.
Cx51 generates a separate segment for each memory type for which a variable is
declared. The following table lists the segment names generated for different
variable data objects.

Segment Name Description

?BA?module_name Bit-addressable data objects
Keil Software — Cx51 Compiler User’s Guide 159

Segment Name Description

?BI?module_name bit objects
?CO?module_name Constants (strings and initialized variables)
?DT?module_name Objects declared in data
?XC?module_name Objects declared in const far (requires OMF251 directive)
?XD?module_name Objects declared in far (requires OMF251 directive)
?ID?module_name Objects declared in idata
?PD?module_name Objects declared in pdata
?XC?module_name Objects declared in const xdata (requires OMF251 directive)
?XD?module_name Objects declared in xdata

Program Objects
Program objects include the code generated for C program functions by the
Cx51 compiler. Each function in a source module is assigned a separate code
segment using the ?PR?function_name?module_name naming convention. For
example, the function error_check in the file SAMPLE.C would result in a
segment name of ?PR?ERROR_CHECK?SAMPLE.

Segments are also created for local variables that are declared within the body of
a function. These segment names follow the above conventions and have a
different prefix depending upon the memory area in which the local variables are
stored.

Function arguments were historically passed using fixed memory locations. This
is still true for routines written in PL/M-51. However, Cx51 can pass up to 3
function arguments in registers. Other arguments are passed using the traditional
fixed memory areas. Memory space is reserved for all function arguments
6
regardless of whether or not some of these arguments may be passed in registers.
The parameter areas must be publicly known to any calling module. So, they are
publicly defined using the following segment names:
?function_name?BYTE
?function_name?BIT

For example, if func1 is a function that accepts both bit arguments as well as
arguments of other data types, the bit arguments are passed starting at
?FUNC1?BIT, and all other parameters are passed starting at ?FUNC1?BYTE.
Refer to “Interfacing C Programs to Assembler” on page 161 for examples of the
function argument segments.
 160 Chapter 6. Advanced Programming Techniques

Functions that have parameters, local variables, or bit variables contain all
additional segments for these variables. These segments can be overlaid by the
BL51 Linker/Locator.

They are created as follows based on the memory model used.

Small model segment naming conventions

Information Segment Type Segment Name

Program code code ?PR?function_name?module_name
Local variables data ?DT?function_name?module_name
Local bit variables bit ?BI?function_name?module_name

Compact model segment naming conventions

Information Segment Type Segment Name

Program code code ?PR?function_name?module_name
Local variables pdata ?PD?function_name?module_name
Local bit variables bit ?BI?function_name?module_name

Large model segment naming conventions

Information Segment Type Segment Name

Program code code ?PR?function_name?module_name
Local variables xdata ?XD?function_name?module_name

6 Local bit variables bit ?BI?function_name?module_name

The names for functions with register parameters and reentrant attributes are
modified slightly to avoid run-time errors. The following table lists deviations
from the standard segment names.

Declaration Symbol Description

void func (void) … FUNC Names of functions that have no arguments or
whose arguments are not passed in registers are
transferred to the object file without any changes.
The function name is converted to uppercase.
void func1 (char) … _FUNC1 For functions with arguments passed in registers,
the underscore character (‘_’) is prefixed to the
function name. This identifies those functions that
transfer arguments in CPU registers.
Keil Software — Cx51 Compiler User’s Guide 161

Declaration Symbol Description

void func2 (void) reentrant _?FUNC2 For functions that are reentrant, the string “_?” is
… prefixed to the function name. This is used to
identify reentrant functions.

Interfacing C Programs to Assembler

You can easily interface Cx51 to routines written in 8051 Assembler. The A51
Assembler is an 8051 macro assembler that emits object modules in OMF-51
format. By observing a few programming rules, you can call assembly routines
from C and vice versa. Public variables declared in the assembly module are
available to your C programs.

There are several reasons why you might want to call an assembly routine from
your C program. You may have assembly code already written that you wish to
use, you may need to improve the speed of a particular function, or you may
want to manipulate SFRs or memory-mapped I/O devices directly from
assembly. This section describes how to write assembly routines that can be
directly interfaced to C programs.

For an assembly routine to be called from C, it must be aware of the parameter

passing and return value conventions used in C functions. For all practical
purposes, it must appear to be a C function.

Function Parameters
By default, C functions pass up to three parameters in registers. The remaining 6
parameters are passed in fixed memory locations. You may use the directive
NOREGPARMS to disable parameter passing in registers. Parameters are
passed in fixed memory locations if parameter passing in registers is disabled or
if there are too many parameters to fit in registers. Functions that pass
parameters in registers are flagged by Cx51 with an underscore character (‘_’)
prefixed to the function name at code generation time. Functions that pass
parameters only in fixed memory locations are not prefixed with an underscore.
Refer to “Using the SRC Directive” on page 164 for an example.
 162 Chapter 6. Advanced Programming Techniques

Parameter Passing in Registers

C functions may pass parameters in registers and fixed memory locations. A
maximum of 3 parameters may be passed in registers. All other parameters are
passed using fixed memory locations. The following tables define what registers
are used for passing parameters.

Arg Number char, 1-byte ptr int, 2-byte ptr long, float generic ptr
1 R7 R6 & R7 R4—R7 R1—R3
(MSB in R6, (Mem type in R3,
LSB in R7) MSB in R2,
LSB in R1)
2 R5 R4 & R5 R4—R7 R1—R3
(MSB in R4, (Mem type in R3,
LSB in R5) MSB in R2,
LSB in R1)
3 R3 R2 & R3 R1—R3
(MSB in R2, (Mem type in R3,
LSB in R3) MSB in R2,
LSB in R1)

The following examples clarify how registers are selected for parameter passing.

Declaration Description
func1 ( The first and only argument, a, is passed in registers R6 and R7.
int a)
func2 ( The first argument, b, is passed in registers R6 and R7. The second
int b, argument, c, is passed in registers R4 and R5. The third argument, d, is
int c, passed in registers R1, R2, and R3.

6 int
func3 (
*d)

long e,
The first argument, e, is passed in registers R4, R5, R6, and R7. The
second argument, f, cannot be located in registers since those available for
long f) a second parameter with a type of long are already used by the first argument.
This parameter is passed using fixed memory locations.
func4 ( The first argument, g, passed in registers R4, R5, R6, and R7. The second
float g, parameter, h, cannot be passed in registers and is passed in fixed memory
char h) locations.
Keil Software — Cx51 Compiler User’s Guide 163

Parameter Passing in Fixed Memory Locations

Parameters passed to assembly routines in fixed memory locations use segments
named ?function_name?BYTE and ?function_name?BIT to hold the parameter
values passed to the function function_name. Bit parameters are copied into the
?function_name?BIT segment prior to calling the function. All other parameters
are copied into the ?function_name?BYTE segment. All parameters are assigned
space in these segments even if they are passed using registers. Parameters are
stored in the order in which they are declared in each respective segment.

The fixed memory locations used for parameter passing may be in internal data
memory or external data memory depending upon the memory model used. The
small memory model is the most efficient and uses internal data memory for
parameter segments. The compact and large models use external data memory
for the parameter passing segments.

Function Return Values

Function return values are always passed using CPU registers. The following
table lists the possible return types and the registers used for each.

Return Type Register Description

bit Carry Flag Single bit returned in the carry flag
char / unsigned char, R7 Single byte typed returned in R7
1-byte pointer
int / unsigned int,
2-byte ptr
long / unsigned long
R6 & R7

R4-R7
MSB in R6, LSB in R7

MSB in R4, LSB in R7

6
float R4-R7 32-Bit IEEE format
generic pointer R1-R3 Memory type in R3, MSB R2, LSB R1
 164 Chapter 6. Advanced Programming Techniques

Using the SRC Directive

You may use the Cx51 compiler to generate the shell for an assembly routine
you want to write or to help determine the passing conventions your assembly
routine should use. The SRC command-line directive specifies that Cx51
generate an assembly file instead of an object file. For example, the following C
source file:
#pragma SRC
#pragma SMALL

unsigned int asmfunc1 (

unsigned int arg)
{
return (1 + arg);
}

generates the following assembly output file when compiled using the SRC
directive.
; ASM1.SRC generated from: ASM1.C

NAME ASM1

?PR?_asmfunc1?ASM1 SEGMENT CODE

PUBLIC _asmfunc1
; #pragma SRC
; #pragma SMALL
;
; unsigned int asmfunc1 (

RSEG ?PR?_asmfunc1?ASM1
USING 0

6
_asmfunc1:
;---- Variable 'arg?00' assigned to Register 'R6/R7' ----
; SOURCE LINE # 4
; SOURCE LINE # 6
; return (1 + arg);
; SOURCE LINE # 7
MOV A,R7
ADD A,#01H
MOV R7,A
CLR A
ADDC A,R6
MOV R6,A
; }
; SOURCE LINE # 8
?C0001:
RET
; END OF _asmfunc1

END
Keil Software — Cx51 Compiler User’s Guide 165

In this example, note that the function name, asmfunc1, is prefixed with an
underscore character signifying that arguments are passed in registers. The arg
parameter is passed using R6 and R7.

The following example shows the assembly source generated for the same
function; however, register parameter passing has been disabled using the
NOREGPARMS directive.
; ASM2.SRC generated from: ASM2.C

NAME ASM2

?PR?asmfunc1?ASM2 SEGMENT CODE

?DT?asmfunc1?ASM2 SEGMENT DATA
PUBLIC ?asmfunc1?BYTE
PUBLIC asmfunc1

RSEG ?DT?asmfunc1?ASM2
?asmfunc1?BYTE:
arg?00: DS 2
; #pragma SRC
; #pragma SMALL
; #pragma NOREGPARMS
;
; unsigned int asmfunc1 (

RSEG ?PR?asmfunc1?ASM2
USING 0
asmfunc1:
; SOURCE LINE # 5
; SOURCE LINE # 7
; return (1 + arg);
; SOURCE LINE # 8
MOV A,arg?00+01H
ADD A,#01H

6
MOV R7,A
CLR A
ADDC A,arg?00
MOV R6,A
; }
; SOURCE LINE # 9
?C0001:
RET
; END OF asmfunc1

END

Note in this example that the function name, asmfunc1, is not prefixed with an
underscore character and that the arg parameter is passed in the
?asmfunc1?BYTE segment.
 166 Chapter 6. Advanced Programming Techniques

Register Usage
Assembler functions can change all register contents in the current selected
register bank as well as the contents of the registers ACC, B, DPTR, and PSW.
When invoking a C function from assembly, assume that these registers may be
destroyed by the C function that is called.

Overlaying Segments
If the overlay process is executed during program linking and locating, it is
important that each assembler subroutine have a unique program segment. This
is necessary so that during the overlay process, the references between the
functions are calculated using the references of the individual segments. The
data areas of the assembler subprograms may be included in the overlay analysis
when the following points are observed:

All segment names must be created using the Cx51 segment naming
conventions.

Each assembler function with local variables must be assigned its own data
segment. This data segment may be accessed by other functions only for
passing parameters. Parameters must be passed in order.

Example Routines

6 The following program examples show you how to pass parameters to and from
assembly routines. The following C functions are used in all of these examples:
int function (
int v_a, /* passed in R6 & R7 */
char v_b, /* passed in R5 */
bit v_c, /* passed in fixed memory location */
long v_d, /* passed in fixed memory location */
bit v_e); /* passed in fixed memory location */
Keil Software — Cx51 Compiler User’s Guide 167

Small Model Example

In the small model, parameters passed in fixed memory locations are stored in
internal data memory. The parameter passing segment for variables is located in
the data area.

The following are two assembly code examples. The first shows how the
example function is invoked from assembly. The second example displays the
assembly code for the example function.

Function invocation from assembly.

.
.
.
EXTRN CODE (_function) ; Ext declarations for function names
EXTRN DATA (?_function?BYTE) ; Seg for local variables
EXTRN BIT (?_function?BIT) ; Seg for local bit variables
.
.
.
MOV R6,#HIGH intval ; int a
MOV R7,#LOW intval ; int a
MOV R7,#charconst ; char b
SETB ?_function?BIT+0 ; bit c
MOV ?_function?BYTE+3,longval+0 ; long d
MOV ?_function?BYTE+4,longval+1 ; long d
MOV ?_function?BYTE+5,longval+2 ; long d
MOV ?_function?BYTE+6,longval+3 ; long d
MOV C,bitvalue
MOV ?_function?BIT+1,C ; bit e
LCALL _function
MOV intresult+0,R6 ; store int
MOV intresult+1,R7 ; retval
.
.
.
6
 168 Chapter 6. Advanced Programming Techniques

Function implementation in assembly.

NAME MODULE ; Names of the program module
?PR?FUNCTION?MODULE SEGMENT CODE ; Seg for prg code in 'function'
?DT?FUNCTION?MODULE SEGMENT DATA OVERLAYABLE
; Seg for local vars in 'function'
?BI?FUNCTION?MODULE SEGMENT BIT OVERLAYABLE
; Seg for local bit vars in 'function'

PUBLIC _function, ?_function?BYTE, ?_function?BIT

; Public symbols for 'C' function call

RSEG ?PD?FUNCTION?MODULE ; Segment for local variables

?_function?BYTE: ; Start of parameter passing segment
v_a: DS 2 ; int variable: v_a
v_b: DS 1 ; char variable: v_b
v_d: DS 4 ; long variable: v_d
.
. ; Additional local variables
.

RSEG ?BI?FUNCTION?MODULE ; Segment for local bit variables

?_function?BIT: ; Start of parameter passing segment
v_c: DBIT 1 ; bit variable: v_c
v_e: DBIT 1 ; bit variable: v_e
.
. ; Additional local bit variables
.

RSEG ?PR?FUNCTION?MODULE ; Program segment

_function: MOV v_a,R6 ; A function prolog and epilog is
MOV v_a+1,R7 ; not necessary. All variables can
MOV v_b,R5 ; immediately be accessed.
.
.
.
MOV R6,#HIGH retval ; Return value
MOV R7,#LOW retval ; int constant
RET ; Return

6
Keil Software — Cx51 Compiler User’s Guide 169

Compact Model Example

In the compact model, parameters passed in fixed memory locations are stored in
external data memory. The parameter passing segment for variables is located in
the pdata area.

The following are two assembly code examples. The first shows you how the
example function is invoked from assembly. The second example displays the
assembly code for the example function.

Function invocation from assembly.

EXTRN CODE (_function) ; Ext declarations for function names
EXTRN XDATA (?_function?BYTE) ; Seg for local variables
EXTRN BIT (?_function?BIT) ; Seg for local bit variables
.
.
.
MOV R6,#HIGH intval ; int a
MOV R7,#LOW intval ; int a
MOV R5,#charconst ; char b
SETB ?_function?BIT+0 ; bit c
MOV R0,#?_function?BYTE+3 ; Addr of 'v_d' in the passing area
MOV A,longval+0 ; long d
MOVX @R0,A ; Store parameter byte
INC R0 ; Inc parameter passing address
MOV A,longval+1 ; long d
MOVX @R0,A ; Store parameter byte
INC R0 ; Inc parameter passing address
MOV A,longval+2 ; long d
MOVX @R0,A ; Store parameter byte
INC R0 ; Inc parameter passing address
MOV A,longval+3 ; long d
MOVX @R0,A ; Store parameter byte
MOV
MOV
LCALL
MOV
C,bitvalue
?_function?BIT+1,C
_function
intresult+0,R6
; bit e

; Store int
6
MOV intresult+1,R7 ; Retval
.
.
.
 170 Chapter 6. Advanced Programming Techniques

Function implementation in assembly.

NAME MODULE ; Name of the program module
?PR?FUNCTION?MODULE SEGMENT CODE ; Seg for program code in 'function';
?PD?FUNCTION?MODULE SEGMENT XDATA OVERLAYABLE IPAGE
; Seg for local vars in 'function'
?BI?FUNCTION?MODULE SEGMENT BIT OVERLAYABLE
; Seg for local bit vars in
'function'

PUBLIC _function, ?_function?BYTE, ?_function?BIT

; Public symbols for C function call

RSEG ?PD?FUNCTION?MODULE ; Segment for local variables

?_function?BYTE: ; Start of the parameter passing seg
v_a: DS 2 ; int variable: v_a
v_b: DS 1 ; char variable: v_b
v_d: DS 4 ; long variable: v_d
.
. ; Additional local variables
.

RSEG ?BI?FUNCTION?MODULE ; Segment for local bit variables

?_function?BIT: ; Start of the parameter passing seg
v_c: DBIT 1 ; bit variable: v_c
v_e: DBIT 1 ; bit variable: v_e
.
. ; Additional local bit variables
.

RSEG ?PR?FUNCTION?MODULE ; Program segment

_function: MOV R0,#?_function?BYTE+0 ; Special function prolog
MOV A,R6 ; and epilog is not
MOVX @R0,A ; necessary. All
INC R0 ; vars can immediately
MOV A,R7 ; be accessed
MOVX @R0,A
INC R0
MOV A,R5

6 .
.
MOVX @R0,A

.
MOV R6,#HIGH retval ; Return value
MOV R7,#LOW retval ; int constant
RET ; Return
Keil Software — Cx51 Compiler User’s Guide 171

Large Model Example

In the large model, parameters passed in fixed memory locations are stored in
external data memory. The parameter passing segment for variables is located in
the xdata area.

The following are two assembly code examples. The first shows you how the
example function is invoked from assembly. The second example displays the
assembly code for the example function.

Function invocation from assembly

EXTRN CODE (_function) ; Ext declarations for function names
EXTRN XDATA (?_function?BYTE) ; Start of transfer for local vars
EXTRN BIT (?_function?BIT) ; Start of transfer for local bit vars
.
.
.
MOV R6,#HIGH intval ; int a
MOV R7,#LOW intval ; int a
MOV R5,#charconst ; char b
SETB ?_function?BIT+0 ; bit c
MOV R0,#?_function?BYTE+3 ; Address of 'v_d' in the passing area
MOV A,longval+0 ; long d
MOVX @DPTR,A ; Store parameter byte
INC DPTR ; Increment parameter passing address
MOV A,longval+1 ; long d
MOVX @DPTR,A ; Store parameter byte
INC DPTR ; Increment parameter passing address
MOV A,longval+2 ; long d
MOVX @DPTR,A ; Store parameter byte
INC DPTR ; Increment parameter passing address
MOV A,longval+3 ; long d
MOVX @DPTR,A ; Store parameter byte
MOV
MOV
LCALL
MOV
C,bitvalue
?_function?BIT+1,C
_function
intresult+0,R6
; bit e

; Store int
6
MOV intresult+1,R7 ; Retval
.
.
.
 172 Chapter 6. Advanced Programming Techniques

Function implementation in assembly

NAME MODULE ; Name of the program module
?PR?FUNCTION?MODULE SEGMENT CODE ; Seg for program code in 'functions'
?XD?FUNCTION?MODULE SEGMENT XDATA OVERLAYABLE
; Seg for local vars in 'function'
?BI?FUNCTION?MODULE SEGMENT BIT OVERLAYABLE
; Seg for local bit vars in 'function'

PUBLIC _function, ?_function?BYTE, ?_function?BIT

; Public symbols for C function call

RSEG ?XD?FUNCTION?MODULE ; Segment for local variables

?_function?BYTE: ; Start of the parameter passing seg
v_a: DS 2 ; int variable: v_a
v_b: DS 1 ; char variable: v_b
v_d: DS 4 ; long variable: v_l
.
.; Additional local variables from 'function'
.

RSEG ?BI?FUNCTION?MODULE ; Segment for local bit variables

?_function?BIT: ; Start of the parameter passing seg
v_c: DBIT 1 ; bit variable: v_c
v_e: DBIT 1 ; bit variable: v_e
.
. ; Additional local bit variables
.

RSEG ?PR?FUNCTION?MODULE ; Program segment

_function: MOV DPTR,#?_function?BYTE+0 ; Special function prolog
MOV A,R6 ; and epilog is not
MOVX @DPTR,A ; necessary. All vars
INC R0 ; can immediately be
MOV A,R7 ; accessed.
MOVX @DPTR,A
INC R0
MOV A,R5
MOVX @DPTR,A

6 .
.
.
MOV R6,#HIGH retval ; Return value
MOV R7,#LOW retval ; int constant
RET ; Return
Keil Software — Cx51 Compiler User’s Guide 173

Interfacing C Programs to PL/M-51

You can easily interface Cx51 to routines written in PL/M-51. Intel’s PL/M-51
is a popular programming language that is similar to C in many ways. The
PL/M-51 compiler generates object files in the OMF-51 format. You can access
PL/M-51 functions from C by declaring them with the alien function type
specifier. Public variables declared in the PL/M-51 module are available to your
C programs.

Cx51 can optionally operate with PL/M-51 parameter passing conventions. The
alien function type specifier is used to declare public or external functions that
are compatible with PL/M-51 in any memory model. For example:
extern alien char plm_func (int, char);

alien unsigned int c_func (unsigned char x, unsigned char y) {

return (x * y);
}

Parameters and return values of PL/M-51 functions may be any of the following
types: bit, char, unsigned char, int, and unsigned int. Other types, including
long, float, and all types of pointers, can be declared in C functions with the
alien type specifier. However, use these types with care because PL/M-51 does
not directly support 32-bit binary integers or floating-point numbers.

PL/M-51 does not support variable-length argument lists. Therefore, functions

declared using the alien type specifier must have a fixed number of arguments.
The ellipsis notation used for variable-length argument lists is not allowed for
alien functions and causes Cx51 to generate an error message. For example:
extern alien unsigned int plm_i (char, int, ...);
6
*** ERROR IN LINE 1 OF A.C: 'plm_i': Var_parms on alien function
 174 Chapter 6. Advanced Programming Techniques

Data Storage Formats

This section describes the storage formats of the data types available in Cx51.
Cx51 provides you with a number of basic data types to use in your C programs.
The following table lists these data types along with their size requirements and
value ranges.

Data Type Bits Bytes Value Range

bit 1 — 0 to 1
signed char 8 1 -128 to +127
unsigned char 8 1 0 to 255
enum 8 / 16 1 or 2 -128 to +127 or -32768 to +32767
signed short 16 2 -32768 to +32767
unsigned short 16 2 0 to 65535
signed int 16 2 -32768 to +32767
unsigned int 16 2 0 to 65535
signed long 32 4 -2147483648 to 2147483647
unsigned long 32 4 0 to 4294967295
float 32 4 ±1.175494E-38 to ±3.402823E+38
data *, idata *, pdata * 8 1 0x00 to 0xFF
code*, xdata * 16 2 0x0000 to 0xFFFF
generic pointer 24 3 Memory type (1 byte); Offset (2 bytes) 0 to 0xFFFF

Other data types, like structures and unions, may contain scalars from this table.
All elements of these data types are allocated sequentially and are byte-aligned
due to the 8-bit architecture of the 8051 family.
6
Bit Variables
Scalars of type bit are stored using a single bit. Pointers to and arrays of bit are
not allowed. Bit objects are always located in the bit-addressable internal
memory space of the 8051 CPU. The BL51 Linker/Locator overlays bit objects
if possible.
Keil Software — Cx51 Compiler User’s Guide 175

Signed and Unsigned Characters,

Pointers to data, idata, and pdata
Scalars of type char are stored in a single byte (8 bits). Memory-specific
pointers that reference data, idata, and pdata are also stored using a single byte
(8 bits). If an enum can be represented with an 8 bit value, the enum is also store
in a single byte.

Signed and Unsigned Integers,

Enumerations, Pointers to xdata and code
Scalars of type int, scalars of type short, enum types, and memory-specific
pointers that reference xdata or code are all stored using 2 bytes (16 bits). The
high-order byte is stored first, followed by the low-order byte. For example, an
integer value of 0x1234 is stored in memory as follows:

Address +0 +1
Contents 0x12 0x34

Signed and Unsigned Long Integers

Scalars of type long are stored using 4 bytes (32 bits). The bytes are stored in
high to low order. For example, the long value 0x12345678 is stored in memory
as follows:
6
Address +0 +1 +2 +3
Contents 0x12 0x34 0x56 0x78
 176 Chapter 6. Advanced Programming Techniques

Generic and Far Pointers

Generic pointers have no declared explicit memory type. They may point to any
memory area on the 8051. These pointers are stored using 3 bytes (24 bits). The
first byte contains a value that indicates the memory area or memory type. The
remaining two bytes contain the address offset with the high-order byte first.
The following memory format is used:

Address +0 +1 +2
Contents Memory Type Offset; High-Order Byte Offset; Low-Order Byte

Depending on the compiler version that you are using the memory type byte has
the following values:

Memory Type idata / data / bdata xdata pdata code

C51 Compiler (8051 devices) 0x00 0x01 0xFE 0xFF
CX51 Compiler (Philips 80C51MX) 0x7F 0x00 0x00 0x80
The Philips 80C51MX architecture supports new CPU instructions that operation on
universal pointer. Universal pointers are identical with CX51 generic pointers.

The format of the generic pointers is also used for pointers with the memory type
far. Therefore, any other memory type values are may be used to address far
memory space.

The following example shows the memory storage of a generic pointer (on the
C51 compiler) that references address 0x1234 in the xdata memory area.

6 Address +0 +1 +2
Contents 0x01 0x12 0x34
Keil Software — Cx51 Compiler User’s Guide 177

Floating-point Numbers
Scalars of type float are stored using 4 bytes (32 bits). The format used
corresponds to that of the IEEE-754 standard.

There are two components of a floating-point number: the mantissa and the
exponent. The mantissa stores the actual digits of the number. The exponent
stores the power to which the mantissa must be raised. The exponent is an 8-bit
value in the 0 to 255 range and is stored relative to 127. The actual value of the
exponent is calculated by subtracting 127 from the stored value (0 to 255). The
value of the exponent can be anywhere from +128 to -127. The mantissa is a
24-bit value whose most significant bit (MSB) is always 1 and is, therefore, not
stored. There is also a sign bit which indicates if the floating-point number is
positive or negative.

Floating-point numbers are stored in 8051 memory using the following format:

Address +0 +1 +2 +3
Contents SEEE EEEE EMMM MMMM MMMM MMMM MMMM MMMM

where:
S represents the sign bit where 1 is negative and 0 is positive.
E is the two’s complement exponent with an offset of 127.
M is the 23-bit normal mantissa. The highest bit is always 1
and, therefore, is not stored
6
Using the above format, the floating-point number -12.5 would be stored as a
hexadecimal value of 0xC1480000. In memory, this appears as follows:

Address +0 +1 +2 +3
Contents 0xC1 0x48 0x00 0x00

It is fairly simple to convert floating-point numbers to and from their

hexadecimal storage equivalents. The following example demonstrates how this
is done for the value -12.5 shown above.
 178 Chapter 6. Advanced Programming Techniques

The floating-point storage representation is not an intuitive format. To convert

this to a floating-point number, the bits must be separated as specified in the
storage format table above. For example:

Address +0 +1 +2 +3
Format SEEEEEEE EMMMMMMM MMMMMMMM MMMMMMMM
Binary 11000001 01001000 00000000 00000000
Hex C1 48 00 00

From this illustration, you can determine the following information:

The sign bit is 1, indicating a negative number.

The exponent value is 10000010 binary or 130 decimal. Subtracting 127
from 130 leaves 3 which is the actual exponent.

The mantissa appears as the following binary number:

10010000000000000000000

There is an understood decimal point at the left of the mantissa that is always
preceded by a 1. This digit is not stored in the hexadecimal representation of
the floating-point number. Adding 1 and the decimal point to the beginning of
the mantissa gives the following:
1.10010000000000000000000

Now, adjust the mantissa for the exponent. A negative exponent moves the
6 decimal point to the left. A positive exponent moves the decimal point to the
right. Because the exponent is 3, the mantissa is adjusted as follows:

1100.10000000000000000000

The result is now a binary floating-point number. Binary digits left of the
decimal point represent the power of two corresponding to the position: 1100
represents (1 × 23) + (1 × 22) + (0 × 21) + (0 × 20) which equals 12.

Binary digits that are right of the decimal point also represent the power of two
corresponding to their position. However, the powers are negative: .100…
represents (1 × 2-1) + (0 × 2-2) + (0 × 2-3) + … which equals .5.

Adding these values together gives 12.5 which must be negated since the sign
bit is set. So, the floating-point hexadecimal value 0xC1480000 is -12.5.
Keil Software — Cx51 Compiler User’s Guide 179

Floating-point Errors
The 8051 does not contain an interrupt vector to trap floating-point errors;
therefore, your software must appropriately respond to these error conditions. In
addition to the normal floating-point values, a floating-point number may contain
a binary error value. These values are defined as a part of the IEEE standard and
are used whenever an error occurs during normal processing of floating-point
operations. Your code should check for possible arithmetic errors at the end of
each floating-point operation.

Name Value Meaning

NaN 0xFFFFFFF Not a number
+INF 0x7F80000 Positive infinity (positive overflow)
-INF 0xFF80000 Negative infinity (negative overflow)

NOTE
The Cx51 library function _chkfloat_ lets you quickly check floating-point
status.

You can use the following union to store floating-point values.

union f {
float f; /* Floating-point value */
unsigned long ul; /* Unsigned long value */
};

This union contains a float and an unsigned long in order to perform floating-
point math operations and to respond to the IEEE error states. For example:
#define NaN 0xFFFFFFFF /* Not a number (error) */
6
#define plusINF 0x7F800000 /* Positive overflow */
#define minusINF 0xFF800000 /* Negative overflow */

union f {
float f; /* Floating-point value */
unsigned long ul; /* Unsigned long value */
};

void main (void) {

float a, b;
union f x;

x.f = a * b;
if (x.ul == NaN || x.ul == plusINF || x.ul == minusINF) {
/* handle the error */
}
else {
/* result is correct */
 180 Chapter 6. Advanced Programming Techniques

}
}

Accessing Absolute Memory Locations

The C programming language does not support a method of explicitly specifying
the memory location of a static or global variable. There are three ways to
reference explicit memory location. You can use the:

Absolute memory access macros

Linker location controls

The _at_ keyword

Each of these three methods is described below.

Absolute Memory Access Macros

First, you may use the absolute memory access macros provided as part of the
Cx51 library. Use the following macros to directly access the memory areas of
the 8051.

CBYTE XBYTE PWORD

DBYTE CWORD XWORD
PBYTE DWORD
6 Refer to “Absolute Memory Access Macros” on page 208 for definitions of these
macros.
Keil Software — Cx51 Compiler User’s Guide 181

Linker Location Controls

The second method of referencing explicit memory location is to declare the
variables in a stand-alone C module, and use the location control directives of
the BL51 Linker/Locator to specify an absolute memory address.

In the following example, assume that we have a structure called

alarm_control that we want to reside at address 2000h in xdata. We start by
entering a source file named ALMCTRL.C that contains only the declaration for
this structure.
.
.
.
struct alarm_st {
unsigned int alarm_number;
unsigned char enable flag;
unsigned int time_delay;
unsigned char status;
};

xdata struct alarm_st alarm_control;

.
.
.

The Cx51 compiler generates an object file for ALMCTRL.C and includes a
segment for variables in the xdata memory area. Because it is the only variable
declared in this module, alarm_control is the only variable in that segment.
The name of the segment is ?XD?ALMCTRL.

The BL51 Linker/Locator allows you to specify the base address of any segment
by using the location control directives. Because the alarm_control variable
was declared to reside in xdata, the XDATA BL51 directive must be used as
6
follows:
BL51 … almctrl.obj XDATA(?XD?ALMCTRL(2000h)) …

This instructs the linker to locate the segment named ?XD?ALMCTRL at

address 2000h in the xdata memory area.

There are linker directives for locating segments in the code, xdata, pdata,
idata, and data memory areas. Refer to the 8051 Utilities User’s Guide for
more information about the Linker/Locator.
 182 Chapter 6. Advanced Programming Techniques

The _at_ Keyword

The third method of accessing absolute memory locations is to use the _at_
keyword when you declare variables in your C source files. The following
example demonstrates how to locate several different variable types using the
_at_ keyword.
struct link {
struct link idata *next;
char code *test;
};

idata struct link list _at_ 0x40; /* list at idata 0x40 */

xdata char text[256] _at_ 0xE000; /* array at xdata 0xE000 */
xdata int i1 _at_ 0x8000; /* int at xdata 0x8000 */

void main ( void ) {

link.next = (void *) 0;
i1 = 0x1234;
text [0] = 'a';
}

Refer to “Absolute Variable Location” on page 99 for more information about

the _at_ keyword.

6
Keil Software — Cx51 Compiler User’s Guide 183

Debugging
By default, the C51 compiler uses the Intel Object Format (OMF-51) for object
files and generates complete symbol information. All Intel compatible emulators
may be used for program debugging. The DEBUG control directive embeds
debugging information in the object file. In addition, the OBJECTEXTEND
control directive embeds additional variable type information in the object file
which allows type-specific display of variables and structures when using certain
emulators.

The CX51 compiler uses the OMF2 object file format. The OMF2 format is also
used by the C51 compiler when the directive OMF2 is active. The OMF2 format
requires the extended LX51 linker/locater and cannot be used with the BL51
linker/locater. The OMF2 object file format provides extensive debug
information and is supported by the µVision2 debugger and some emulators.

6
Keil Software — Cx51 Compiler User’s Guide 185

Chapter 7. Error Messages

This chapter lists Fatal Error, Syntax Error, and Warning messages that you may
encounter as you develop a program. Each section includes a brief description of
the message as well as corrective actions you can take to eliminate the error or
warning condition.

Fatal Errors
Fatal errors cause immediate termination of the compilation. These errors
normally occur as the result of invalid options specified on the command line.
Fatal errors are also generated when the compiler cannot access a specified
source include file.

Fatal error messages conform to one of the following formats:

C51 FATAL-ERROR -
ACTION: <current action>
LINE: <line in which the error is detected>
ERROR: <corresponding error message>
C51 TERMINATED.

C51 FATAL-ERROR -
ACTION: <current action>
FILE: <file in which the error is detected>
ERROR: <corresponding error message>
C51 TERMINATED.

The following are descriptions of the possible text for the Action and Error
fields in the above messages.

7
 186 Chapter 7. Error Messages

Actions
ALLOCATING MEMORY
The compiler could not allocate enough memory to compile the specified
source file.
CREATING LIST-FILE / OBJECT-FILE / WORKFILE
The compiler could not create the list file, object file, or work file. This error
may occur if the disk is full or write-protected, or if the file already exists and
is read only.
GENERATING INTERMEDIATE CODE
The source file contains a function that is too large to be translated into
pseudo-code by the compiler. Try breaking the function into smaller
functions and re-compiling.
OPENING INPUT-FILE
The compiler failed to find or open the selected source or include file.
PARSING INVOKE-/#PRAGMA-LINE
An error was detected while evaluating arguments on the command line or
while evaluating parameters in a #pragma statement.
PARSING SOURCE-FILE / ANALYZING DECLARATIONS
The source file contains too many external references. Reduce the number of
external variables and functions accessed by the source file.
WRITING TO FILE
An error was encountered while writing to the list file, object file, or work
file.

7
Keil Software — Cx51 Compiler User’s Guide 187

Errors
'(' AFTER CONTROL EXPECTED
Some control parameters need an argument enclosed in parentheses. This
message is displayed when the left parenthesis is missing.
')' AFTER PARAMETER EXPECTED
This message indicates that the right parenthesis of the enclosed argument is
missing.
BAD DIGIT IN NUMBER
The numerical argument of a control parameter contains invalid characters.
Only decimal digits are acceptable.
CAN'T CREATE FILE
The filename defined on the FILE line cannot be created.
CAN'T HAVE GENERAL CONTROL IN INVOCATION LINE
General controls (for example, EJECT) cannot be included on the command
line. Place these controls in the source file using the #pragma statement.
FILE DOES NOT EXIST
The filename defined on the FILE line, cannot be found.
FILE WRITE-ERROR
An error occurred while writing to the list, preprint, work, or object file
because of insufficient disk space.
IDENTIFIER EXPECTED
This message is generated when the DEFINE control has no arguments.
DEFINE requires an identifier as its argument. This is the same convention
as in the C language.
MEMORY SPACE EXHAUSTED
The compiler could not allocate enough memory to compile the specified
source file. If you receive this message consistently, you should break the
source file into two or more smaller files and re-compile.
MORE THAN 100 ERRORS IN SOURCE-FILE
7
During the compilation more than 100 errors were detected. This causes the
termination of the compiler.
MORE THAN 256 SEGMENTS/EXTERNALS
More than 256 total references were encountered in a source file. A single
source file cannot contain more than 256 functions or external references.
This is a historical restriction mandated by the Intel Object Module Format
(OMF-51). Functions which contain scalar and/or bit declarations produce
two and sometimes three segment definitions in the object file.
 188 Chapter 7. Error Messages

NON-NULL ARGUMENT EXPECTED

The selected control parameter needs an argument (for example, a filename
or a number) enclosed in parentheses.
OUT OF RANGE NUMBER
The numerical argument of a control parameter is out of range. For instance,
the OPTIMIZE control allows only the numbers 0 through 6. A value of 7
would generate this error message.
PARSE STACK OVERFLOW
The parse stack has overflowed. This can occur if the source program
contains extremely complex expressions or if blocks are nested more than 31
levels deep.
PREPROCESSOR: LINE TOO LONG (32K)
An intermediate expansion exceeded 32K characters in length.
PREPROCESSOR: MACROS TOO NESTED
During macro expansion the stack consumption of the preprocessor grew too
large to continue. This message usually indicates a recursive macro
definition, but can also indicate a macro with too many levels of nesting.
RESPECIFIED OR CONFLICTING CONTROL
A command-line parameter was specified twice or conflicting command-line
parameters were specified.
SOURCE MUST COME FROM A DISK-FILE
The source and include files must exist on either a hard disk or diskette. The
console CON:, :CI:, or similar devices are not allowed as input files.
UNKNOWN CONTROL
The selected control parameter is unrecognized by the compiler.

7
Keil Software — Cx51 Compiler User’s Guide 189

Syntax and Semantic Errors

Syntax and semantic errors typically occur in the source program. They identify
actual programming errors. When one of these errors is encountered, the
compiler attempts to recover from the error and continue processing the source
file. As more errors are encountered, the compiler outputs additional error
messages. However, no object file is produced.

Syntax and semantic errors produce a message in the list file. These error
messages are in the following format:

*** ERROR number IN LINE line OF file: error message

where:
number is the error number.
line corresponds to the line number in the source file or include
file.
file is the name of the source or include file in which the error
was detected.
error message is descriptive text and is dependent upon the type of error
encountered.

The following table lists syntax and semantic errors by error number. The error
message displayed is listed along with a brief description and possible cause and
correction.

Number Error Message and Description

100 Unprintable character 0x?? skipped
An illegal character was found in the source file. (Note that characters inside a
comment are not checked.)
101 Unclosed string
A string is not terminated with a quote (").
7
102 String too long
A string may not contain more than 4096 characters. Use the concatenation
symbol (‘\’) to logically continue strings longer than 4096 characters. Lines
terminated in this fashion are concatenated during lexical analysis.
103 Invalid character constant
A character constant has an invalid format. The notation ‘\c’ is valid only when c
is any printable ASCII character.
125 Declarator too complex (20)
The declaration of an object may contain a maximum of 20 type modifiers (‘[‘, ‘]’,
‘*’, ‘(‘, ‘)’). This error is almost always followed by error 126.
 190 Chapter 7. Error Messages

Number Error Message and Description

126 Type-stack underflow
The type declaration stack has underflowed. This error is usually a side-effect of
error 125.
127 Invalid storage class
An object was declared with an invalid memory space specification. This occurs if
an object is declared with storage class of auto or register outside of a function.
129 Missing ‘;’ before ‘token’
This error usually indicates that a semicolon is missing from the previous line.
When this error occurs, the compiler may generate an excess of error messages.
130 Value out of range
The numerical argument after a using or interrupt specifier is invalid. The using
specifier requires a register bank number between 0 and 3. The interrupt
specifier requires an interrupt vector number between 0 and 31.
131 Duplicate function-parameter
A formal parameter name exists more than once within a function. The formal
parameter names must be unique in function declarations.
132 Not in formal parameter list
The parameter declarations inside a function use a name not present in the
parameter name list. For example:

char function (v0, v1, v2)

char *v0, *v1, *v5;
/* 'v5' is unknown in the formal list */
{
/* ... */
}
134 xdata/idata/pdata/data on function not permitted
Functions always reside in code memory and cannot be executed out of other
memory areas. Functions are implicitly defined as memory type code.
135 Bad storage class for bit
Declarations of bit scalars may include one of the static or extern storage
classes. The register or alien classes are invalid.
136 ‘void’ on variable
The type void is allowed only as a non-existent return value or an empty argument
list for functions (void func (void)), or in combination with a pointer (void *).
138 Interrupt() may not receive or return value(s)
An interrupt function was defined with one or more formal parameters or with a
return value. Interrupt functions may not contain invocation parameters or return

7 140
values.
Bit in illegal memory-space
Definitions of bit scalars may contain the optional memory type data. If the
memory type is missing then the type data is assumed, because bits always
reside in the internal data memory. This error can occur when an attempt is made
to use another data type with a bit scalar definition.
141 Syntax error near token: expected other_token, …
The token seen by the compiler is wrong. Depending upon the context the
expected token is displayed.
142 Invalid base address
The base-address of an sfr or sbit declaration is in error. Valid bases are values
in the 0x80 to 0xFF range. If the declaration uses the notation base^pos, then the
base address must also be a multiple of eight.
Keil Software — Cx51 Compiler User’s Guide 191

Number Error Message and Description

143 Invalid absolute bit address
The absolute address in sbit declarations must be in the 0x80 to 0xFF range.
144 Base^pos: invalid bit position
The definition of the bit position within an sbit declaration must be in the 0 to 7
range.
145 Undeclared sfr
146 Invalid sfr
The declaration of an absolute bit (base^pos) contains an invalid
base-specification. The base must be the name of a previously declared sfr. Any
other names are invalid.
147 Object too large
The size of a single object may not exceed the absolute limit of 65535 (64 Kbytes -
1).
149 Function member in struct/union
A struct or union may not contain a function-type member. However, pointers to
functions are perfectly valid.
150 Bit member in struct/union
A union-aggregate may not contain members of type bit. This restriction is
imposed due to the architecture of the 8051.
151 Self relative struct/union
A structure cannot contain an instance of itself.
152 Bit-field type too small for number of bits
The number of bits specified in the bit-field declaration exceeds the number of bits
in the given base type.
153 Named bit-field cannot have zero width
The named field had a zero width. Only unnamed bit-fields are allowed to have
zero width.
154 Ptr to field
Pointers to bit-fields are not valid types.
155 char/int required for fields
The base type for bit-fields requires one of the types char or int. unsigned char
and unsigned int types are also valid.
156 Alien permitted on functions only
157 Var_parms on alien function
The storage class alien is allowed only for external PL/M-51 functions. The formal
notation (char *, …) is not legal on alien functions. PL/M-51 functions always
require a fixed number of parameters.
158 Function contains unnamed parameter
The parameter list of a function definition contains an unnamed abstract type
definition. This notation is permitted only in function prototypes.
7
159 Type follows void
Prototype declarations of functions may contain an empty parameter list (for
example, int func (void)). This notation may not contain further type definitions
after void.
160 void invalid
The void type is legal only in combination with pointers or as the non-existent
return value of a function.
161 Formal parameter ignored
A declaration of an external function inside a function used a parameter name list
without any type specification (for example, extern yylex(a,b,c);).
 192 Chapter 7. Error Messages

Number Error Message and Description

162 Duplicate function-parameter
The name of a defined object inside a function duplicates the name of a
parameter.
163 Unknown array size
In general, a formal size specifier is not required for external, single, or
multi-dimensional arrays. Typically, the compiler calculates the size at initialization
time. For external arrays, the size is of no great interest. This error is the result of
attempting to use the sizeof operator on an undimensioned array or on a
multi-dimensional array with undefined element sizes.
164 Ptr to nul
This error is usually the result of a previous error for a pointer declaration.
165 Ptr to bit
The type combination pointer to bit is not a legal type.
166 Array of functions
Arrays cannot contain functions; however, they may contain pointers to functions.
167 Array of fields
Bit-fields may not be arranged as arrays.
168 Array of bit
An array may not have type bit as its basic type. This limitation is imposed by the
architecture of the 8051.
169 Function returns function
A function cannot return a function; however, a function may return a pointer to a
function.
170 Function returns array
A function cannot return an array; however, a pointer to an array is valid.
171 Missing enclosing loop
A break or continue statement may occur only within a for, while, do, or switch
statement.
172 Missing enclosing switch
A case statement may occur only within a switch statement.
173 Missing return-expression
A function which returns a value of any type but int, must contain a return
statement including an expression. Because of compatibility to older programs, no
check is done on functions which return an int value.
174 Return-expression on void-function
A void function cannot return a value and thus may not contain a return
statement.

7 175 Duplicate case value

Each case statement must contain a constant expression as its argument. The
value must not occur more than once in the given level of the switch statement.
176 More than one ‘default’
A switch statement may not contain more than one default statement.
177 Different struct/union
Different types of structures are used in an assignment or as an argument to a
function.
178 Struct/union comparison illegal
The comparison of two structures or unions is not allowed according to ANSI.
179 Illegal type conversation from/to ‘void’
Type casts to or from void are invalid.
Keil Software — Cx51 Compiler User’s Guide 193

Number Error Message and Description

180 Can’t cast to ‘function’
Type casts to function types are invalid. Try casting to a pointer to a function.
181 Incompatible operand
At least one operand type is not valid with the given operator (for example,
~float_type).
183 Unmodifiable lvalue
The object to be changed resides in code memory or has const attribute and
therefore cannot be modified.
184 Sizeof: illegal operand
The sizeof operator cannot determine the size of a function or bit-field.
185 Different memory space
The memory space of an object declaration differs from the memory space of a
prior declaration for the same object.
186 Invalid dereference
This error message may be caused by an internal compiler problem. Please
contact technical support if this error is repeated.
187 Not an lvalue
The needed argument must be the address of an object that can be modified.
188 Unknown object size
The size of an object cannot be computed because of a missing dimension size
on an array or indirection via a void pointer.
189 ‘&’ on bit/sfr illegal
The address-of operator (‘&’) is not allowed on bit objects or special function
registers (sfr).
190 ‘&’: not an lvalue
An attempt was made to construct a pointer to an anonymous object.
193 Illegal op-type(s)
193 Illegal add/sub on ptr
193 Illegal operation on bit(s)
193 Bad operand type
This error results when an expression uses illegal operand-types with the given
operator. Examples of invalid expressions are bit * bit, ptr + ptr, or ptr * anything.
The error message includes the operator which caused the error.

The following operations may be executed with bit-type operands:

Assignment (=)

OR / Compound OR (|, |=)
AND / Compound AND (&, &=)
XOR / Compound XOR (^, ^=)
7

Compare bit with bit or constant (==, !=)

Negation (~)
bit operands may be used in expressions with other data types. In this case a
type cast is automatically performed.
194 ‘*’ indirection to object of unknown size
The indirection operator * may not be used with void pointers because the object
size, which the pointer refers to, is unknown.
195 ‘*’ illegal indirection
The * operator may not be applied on non-pointer arguments.
 194 Chapter 7. Error Messages

Number Error Message and Description

196 Mspace probably invalid
The conversion of a constant to a pointer constant yields an invalid memory space,
for example char *p = 0x91234.
198 Sizeof returns zero
The sizeof operator returns a zero value.
199 Left side of ‘->’ requires struct/union pointer
The argument on the left side of the -> operator must be a struct pointer or a
union pointer.
200 Left side of ‘.’ requires struct/union
The argument on the left side of the . operator must have type struct or union.
201 Undefined struct/union tag
The given struct or union tag name is unknown.
202 Undefined identifier
The given identifier is undefined.
203 Bad storage class (nameref)
This error indicates a problem within the compiler. Please contact technical
support if this error is repeated.
204 Undefined member
The given member name in a struct or union reference is undefined.
205 Can’t call an interrupt function
An interrupt function should not be called like a normal function. The entry and
exit code for these functions is specially coded for interrupts.
207 Declared with ‘void’ parameter list
A function declared with a void parameter list cannot receive parameters from the
caller.
208 Too many actual parameters
The function call includes more parameters than previously declared.
209 Too few actual parameters
Too few actual parameters were included in a function call.
210 Too many nested calls
Function calls can be nested at most 10 levels deep.
211 Call not to a function
The term of a function call does not evaluate to a function or pointer to function.
212 Indirect call: parameters do not fit within registers
An indirect function call through a pointer cannot contain actual parameters. An

7
exception to this rule is when all parameters can be passed in registers. This is
due to the method of parameter passing employed by Cx51. The name of the
called function must be known because parameters are written into the data
segment of the called function. For indirect calls, however, the name of the called
function is not known.
213 Left side of asn-op not an lvalue
The address of a changeable object is required at the right side of the assignment
operator.
214 Illegal pointer conversion
Objects of type bit, float or aggregates cannot be converted to pointers.
215 Illegal type conversion
Struct/union/void cannot be converted to any other types.
Keil Software — Cx51 Compiler User’s Guide 195

Number Error Message and Description

216 Subscript on non-array or too many dimensions
An array reference contained either too many dimension specifiers or the object
was not an array.
217 Non-integral index
The dimension expression of an array must be of the type char, unsigned char,
int, or unsigned int. All other types are illegal.
218 Void-type in controlling expression
The limit expression in a while, for, or do statement cannot be of type void.
219 Long constant truncated to int
The value of a constant expression must be capable of being represented by an
int type.
220 Illegal constant expression
A constant expression is expected. Object names, variables or functions, are not
allowed in constant expressions.
221 Non-constant case/dim expression
A case value or a dimension specification ([ ]) must be a constant expression.
222 Div by zero
223 Mod by zero
The compiler detected a division or a modulo by zero.
225 Expression too complex, simplify
An expression is too complex and must be broken into two or more sub
expressions.
226 Duplicate struct/union/enum tag
The name for a struct, union, or enum is already defined within current scope.
227 Not a union tag
The name for a union is already defined as a different type.
228 Not a struct tag
The name for a struct is already defined as a different type.
229 Not an enum tag
The name for an enum is already defined as a different type.
230 Unknown struct/union/enum tag
The specified struct, union, or enum name is undefined.
231 Redefinition
The specified name is already defined and cannot be redefined.
232 Duplicate label

233
The specified label is already defined.
Undefined label
This message indicates a label that was accessed but was not defined.
7
Sometimes this message appears several lines after the actual label reference.
This is caused by the method used to search for undefined labels.
234 ‘{‘, scope stack overflow(31)
The maximum of 31 nested blocks has been exceeded. Additional levels of
nested blocks are ignored.
235 Parameter <number>: different types
Parameter types in the function declaration are different from those in the function
prototype.
 196 Chapter 7. Error Messages

Number Error Message and Description

236 Different length of parameter lists
The number of parameters in the function declaration is different from the number
of parameters in the function prototype.
237 Function already has a body
An attempt was made to declare a body for a function twice.
238 Duplicate member
239 Duplicate parameter
An attempt was made to define an already defined struct member or function
parameter.
240 More than 128 local bit’s
No more than 128 bit-scalars may be defined inside a function.
241 Auto segment too large
The required space for local objects exceeds the model-dependent maximum. The
maximum segment sizes are defined as follows:

SMALL 128 bytes

COMPACT 256 bytes
LARGE 65535 bytes
242 Too many initializers
The number of initializers exceeded the number of objects to be initialized.
243 String out of bounds
The number of characters in the string exceeds the number of characters required
to initialize the array of characters.
244 Can’t initialize, bad type or class
An attempt was made to initialize a bit or an sfr.
245 Unknown pragma, line ignored
The #pragma statement is unknown so, the entire line is ignored.
246 Floating-point error
This error occurs when a floating-point argument lies outside of the valid range for
32-bit floating values. The numeric range of the 32-bit IEEE values is:
±1.175494E-38 to ±3.402823E+38.
247 Non-address/constant initializer
A valid initializer expression must evaluate to a constant value or the name of an
object plus or minus a constant.
248 Aggregate initialization needs curly braces
The braces ({ }) around the given struct or union initializer were missing.

7
249 Segment <name>: Segment too large
The compiler detected a data segment that was too large. The maximum size of a
data segment depends on memory space.
250 ‘\esc’; value exceeds 255
An escape sequence in a string constant exceeds the valid value range. The
maximum value is 255.
251 Illegal octal digit
The specified character is not a valid octal digit.
252 Misplaced primary control, line ignored
Primary controls must be specified at the start of the C module before any
#include directives or declarations.
Keil Software — Cx51 Compiler User’s Guide 197

Number Error Message and Description

253 Internal error (ASMGEN\CLASS)
This error can occur under the following circumstances:

An intrinsic function (for example, _testbit_) was activated incorrectly. This is
the case when no prototype of the function exists and the number of actual
parameters or their type is incorrect. For this reason, the appropriate
declaration files must always be used (INTRINS.H, STRING.H). See Chapter 8
for more information on intrinsic functions.

Cx51 recognized an internal consistency problem. Please contact technical
support if this error occurs.
255 Switch expression has illegal type
The expression in a switch statement has not a legal data type.
256 Conflicting memory model
A function which contains the alien attribute may contain only the model
specification small. The parameters of the function must lie in internal data
memory. This applies to all external alien declarations and alien functions. For
example:

alien plm_func (char c) large { ... }

generates error 256.

257 Alien function cannot be reentrant
A function that contains the alien attribute cannot simultaneously contain the
attribute reentrant. The parameters of the function cannot be passed via the
virtual stack. This applies to all external alien declarations and alien functions.
258 Mspace illegal on struct/union member
Mspace on parameter ignored
A member of a structure or a parameter may not contain the specification of a
memory type. The object to which the pointer refers may, however, contain a
memory type. For example:

struct vp { char code c; int xdata i; };

generates error 258.

struct v1 { char c; int xdata *i; };

is the correct declaration for the struct.

259 Pointer: different mspace

7
A spaced pointer has been assigned another spaced pointer with a different
memory space. For example:

char xdata *p1;

char idata *p2;
p1 = p2; /* different memory spaces */
260 Pointer truncation
A spaced pointer has been assigned some constant value which exceeds the
range covered by the pointers memory space. For example:

char idata *p1 = 0x1234; /* result is 0x34 */

 198 Chapter 7. Error Messages

Number Error Message and Description

261 Bit(s) in reentrant ( )
A function with the attribute reentrant cannot have bit objects declared inside the
function. For example:

|int func1 (int i1) reentrant {

bit b1, b2; /* not allowed ! */
return (i1 - 1);
}
262 ‘using/disable’: can’t return bit value
Functions declared with the using attribute and functions which rely on disabled
interrupts (#pragma disable) cannot return a bit value to the caller. For example:

bit test (void) using 3

{
bit b0;
return (b0);
}

produces error 262.

263 Save/restore: save-stack overflow/underflow
The maximum nesting depth #pragma save comprises eight levels. The pragmas
save and restore work with a stack according to the LIFO (last in, first out)
principal.
264 Intrinsic ‘<intrinsic_name>’: declaration/activation error
This error indicates that an intrinsic function was defined incorrectly (parameter
number or ellipsis notation). This error should not occur if you are using the
standard .H files. Make sure that you are using the .H files that were included with
Cx51. Do not try to define your own prototypes for intrinsic library functions.
265 Recursive call to non-reentrant function
Non reentrant functions cannot be called recursively since such calls would
overwrite the parameters and local data of the function. If you need recursive
calls, you should declare the function with the reentrant attribute.
267 Funcdef requires ANSI-style prototype
A function was invoked with parameters but the declaration specifies an empty
parameter list. The prototype should be completed with the parameter types in
order to give the compiler the opportunity to pass parameters in registers and have
the calling mechanism consistent over the application.
268 Bad taskdef (taskid/priority/using)
The task declaration is incorrect.

7 271 Misplaced ‘asm/endasm’ control

The asm and endasm statements may not be nested. Endasm requires that an
asm block be opened by a previous asm statement. For example:

#pragma asm
.
.
.
assembler instruction(s)
.
.
.
#pragma endasm
Keil Software — Cx51 Compiler User’s Guide 199

Number Error Message and Description

272 ‘asm’ requires SRC control to be active
The use of asm and endasm in a source file requires that the file be compiled
using the SRC directive. The compiler then generates an assembly source file
which may then be assembled with A51.
273 ‘asm/endasm’ not allowed in include file
The use of asm and endasm is not permitted within include files. For debug
reasons executable code should be avoided in include files anyway.
274 Absolute specifier illegal
The absolute address specification is not allowed on bit objects, functions, and
function locals. The address must conform to the memory space of the object.
For example, the following declaration is invalid because the range of the indirectly
addressable data space is 0x00 to 0xFF.

idata int _at_ 0x1000;

278 Constant too big
This error occurs when a floating-point argument lies outside of the valid range for
32-bit floating values. The numeric range of the 32-bit IEEE values is:
±1.175494E-38 to ±3.402823E+38.
279 Multiple initialization
An attempt has been made to initialize some object more than once.
280 unreferenced symbol/label/parameter
A local program symbol is not used in the function.
281 non-pointer type converted to pointer
The referenced program object cannot be converted to a pointer.
282 not a sfr reference
This function invocation requires a SFR location as parameter.
283 asmparms: parameters do not fit within registers
Parameters do not fit within the available CPU registers.
284 <name>: in overlayable space, function no longer reentrant
A reentrant function contains explicit memory type specifiers for local variables.
The function will no longer be fully reentrant.
300 Unterminated comment
This message occurs when a comment does not have a closing delimiter (*/).
301 Identifier expected
The syntax of a preprocessor directive expects an identifier.
302 Misused # operator
This message occurs if the stringize operator ‘#’ is not followed by an identifier.
303 Formal argument expected
This message occurs if the stringize operator ‘#’ is not followed by an identifier
representing a formal parameter name of the macro currently being defined.
7
304 Bad macro parameter list
The macro parameter list does not represent a brace enclosed, comma separated
list of identifiers.
305 Unterminated string/char constant
A string or character constant is invalid. Typically, this error is encountered if the
closing quote is missing.
306 Unterminated macro call
The end of the input file was reached while the preprocessor was collecting and
expanding actual parameters of a macro call.
 200 Chapter 7. Error Messages

Number Error Message and Description

307 Macro ‘name’: parameter count mismatch
The number of actual parameters in a macro call does not match the number of
parameters of the macro definition. This error indicates that too few parameters
were specified.
308 Invalid integer constant expression
The numerical expression of an if/elif directive contains a syntax error.
309 Bad or missing file name
The filename argument in an include directive is invalid or missing.
310 Conditionals too nested(20)
The source file contains too many nested directives for conditional compilation.
The maximum nesting level allowed is 20.
311 Misplaced elif/else control
312 Misplaced endif control
The directives elif, else, and endif are legal only within an if, ifdef, or ifndef
directive.
313 Can’t remove predefined macro ‘name’
An attempt was made to remove a predefined macro. Existing macros may be
deleted using the #undef directive. Predefined macros cannot be removed.
314 Bad # directive syntax
In a preprocessor directive, the character ‘#’ must be followed by either a newline
character or the name of a preprocessor command (for example, if/define/ifdef,
…).
315 Unknown # directive ‘name’
The name of the preprocessor directive is not known to the compiler.
316 Unterminated conditionals
The number of endifs does not match the number of if or ifdefs after the end of the
input file.
318 Can’t open file ‘filename’
The given file could not be opened.
319 ‘File’ is not a disk file
The given file is not a disk file. Files other than disk files are not legal for
compilation.
320 User_error_text
This error number is reserved for errors introduced with the #error directive of the
preprocessor. The #error directive causes the user error text to come up with error
320 which counts like some other error and prevents the compiler from generating
code.

7 321 Missing <character>

In the filename argument of an include directive, the closing character is missing.
For example: #include <stdio.h
325 Duplicate formal parameter ‘name’
A formal parameter of a macro may be define only once.
326 Macro body cannot start or end with ‘##’
The concat operator (‘##’) cannot be the first or last token of a macro body.
327 Macro ‘macroname’: more than 50 parameters
The number of parameters per macro is limited to 50.
Keil Software — Cx51 Compiler User’s Guide 201

Warnings
Warnings produce information about potential problems which may occur during
the execution of the resulting program. Warnings do not hinder compilation of
the source file.

Warnings produce a message in the list file. These warning messages are in the
following format:

*** WARNING number IN LINE line OF file: warning message

where:
number is the error number.
line corresponds to the line number in the source file or include
file.
file is the name of the source or include file in which the error
was detected.
warning message is descriptive text that is dependent upon the type of warning
encountered.

The following table lists warnings by number. The warning message displayed
is listed along with a brief description and possible cause and correction.

Number Warning Message and Description

173 Missing return-expression
A function which returns a value of any type but int, must contain a return
statement including an expression. Because of compatibility to older programs, no
check is done on functions which return an int value.
182 Pointer to different objects
A pointer was assigned the address of a different type.
185 Different memory space
The memory space of an object declaration differs from the memory space of a
prior declaration for the same object.
7
196 Mspace probably invalid
This warning is caused by the assignment of an invalid constant value to a pointer.
Valid pointer constants are long or unsigned long. The compiler uses 24 bits
(3 bytes) for pointer objects. The low-order 16 bits represent the offset. The
high-order 8 bits represent the memory space selector.
198 Sizeof returns zero
The calculation of the size of an object yields zero. This value may be wrong if the
object is external or if not all dimension sizes of an array are known.
 202 Chapter 7. Error Messages

Number Warning Message and Description

206 Missing function prototype
The called function is unknown because no prototype declaration exists. Calls to
unknown functions are always at risk that the number of parameters does not
correspond to the actual requirements. If this is the case, the function is called
incorrectly.

The compiler has no way to check for missing or excessive parameters and their
types. Include prototypes of the functions used in your program. Prototypes must
be specified before the functions are actually called. The definition of a function
automatically produces a prototype.
209 Too few actual parameters
Too few actual parameters were included in a function call.
219 Long constant truncated to int
The value of a constant expression must be capable of being represented by an int
type.
245 Unknown pragma, line ignored
The #pragma statement is unknown, so the entire pragma line is ignored.
258 Mspace illegal on struct/union member
Mspace on parameter ignored
A member of a structure or a parameter may not contain the specification of a
memory type. The object to which the pointer refers may, however, contain a
memory type. For example:

struct vp { char code c; int xdata i; };

generates error 258.

struct v1 { char c; int xdata *i; };

is the correct declaration for the struct.

259 Pointer: different mspace
This warning is generated when two pointers that do not refer to the same memory
type of object are compared.
260 Pointer truncation
This error or warning occurs when converting a pointer to a pointer with a smaller
offset area. The conversion takes place, but the offset of the larger pointer is
truncated to fit into the smaller pointer.
261 Bit in reentrant function
A reentrant function cannot contain bits because bit scalars cannot be stored on

7 265
the virtual stack.
‘name’: recursive call to non-reentrant function
A direct recursion to a non-reentrant function was discovered. This can be
intentional but should be functionally checked (through the generated code) for
each individual case. Indirect recursions are discovered by the linker/locator.
Keil Software — Cx51 Compiler User’s Guide 203

Number Warning Message and Description

271 Misplaced ‘asm/endasm’ control
The asm and endasm statements may not be nested. Endasm requires that an
asm block be opened by a previous asm statement. For example:

#pragma asm
.
.
.
assembler instruction(s)
.
.
.
#pragma endasm
275 Expression with possibly no effect
The compiler detected an expression which does not generate code. For example:

void test (void) {

int i1, i2, i3;
i1, i2, i3; /* dead expression */
i1 << 3; /* result is not used */
}
276 Constant in condition expression
The compiler detected a conditional expression with a constant value. In most
cases this is a typing mistake. For example:

void test (void) {

int i1, i2, i3;
if (i1 = 1) i2 = 3; /* const assigned with = */
while (i3 = 2); /* const assigned with = */
}
277 Different mspaces to pointer
A typedef declaration has a conflict of the memory spaces. For example:

typedef char xdata XCC; /* mspace xdata */

typedef XCC idata PICC; /* mspace collision */
280 Unreferenced symbol/label
This message identifies a symbol or label which has been defined but not used.
307 Macro ‘name’: parameter count mismatch
The number of actual parameters in a macro call does not match the number of
parameters of the macro definition. This warning indicates that too many

7
parameters were used. Excess parameters are ignored.
317 Macro ‘name’: invalid redefinition
A predefined macro cannot be redefined or removed. The compiler recognizes the
following predefined macros:

_ _C51_ _ _ _DATE_ _ _ _FILE_ _ _ _MODEL_ _

_ _LINE_ _ _ _STDC_ _ _ _TIME_ _
322 Unknown identifier
The identifier in an #if directive line is undefined (evaluates to FALSE).
323 Newline expected, extra characters found
A #directive line is correct but contains extra non commented characters. For
example:

#include <stdio.h> foo

 204 Chapter 7. Error Messages

Number Warning Message and Description

324 Preprocessor token expected
A preprocessor token was expected but a newline character was found in input.
For example: #line where the arguments to the #line directive are missing.

7
Keil Software — Cx51 Compiler User’s Guide 205

Chapter 8. Library Reference

The Cx51 run-time library provides you with more than 100 predefined functions
and macros to use in your 8051 C programs. This library makes embedded
software development easier by providing you with routines that perform
common programming tasks such as string and buffer manipulation, data
conversion, and floating-point math operations.

Typically, the routines in this library conform to the ANSI C Standard.

However, some functions differ slightly in order to take advantage of the
features found in the 8051 architecture. For example, the function isdigit returns
a bit value as opposed to an int. Where possible, function return types and
argument types are adjusted to use the smallest possible data type. In addition,
unsigned data types are favored over signed types. These alterations to the
standard library provide a maximum of performance while also reducing
program size.

All routines in this library are implemented to be independent of and to function

using any register bank.

Intrinsic Routines
The Cx51 compiler supports a number of intrinsic library functions.
Non-intrinsic functions generate ACALL or LCALL instructions to perform the
library routine. Intrinsic functions generate in-line code to perform the library
routine. The generated in-line code is much faster and more efficient than a
called routine would be. The following functions are available in intrinsic form:

_crol_ _iror_ _nop_

_cror_ _lrol_ _testbit_
_irol_ _lror_

These routines are described in detail in the following sections.

8
 206 Chapter 8. Library Reference

Library Files
The Cx51 library includes six different compile-time libraries which are
optimized for various functional requirements. These libraries support most of
the ANSI C function calls.

Library File Description

C51S.LIB Small model library without floating-point arithmetic
C51FPS.LIB Small model floating-point arithmetic library
C51C.LIB Compact model library without floating-point arithmetic
C51FPC.LIB Compact model floating-point arithmetic library
C51L.LIB Large model library without floating-point arithmetic
C51FPL.LIB Large model floating-point arithmetic library
80C751.LIB Library for use with the Signetics 8xC751 and derivatives.
The Philips 80C51MX, Dallas 390 contigouos mode and variable code banking
requires a different set of Cx51 run-time libraries. The LX51 linker/locater
automatically adds the correct library set to your project.

Several library modules are provided in source code form. These routines are
used to perform low-level hardware-related I/O for the stream I/O functions.
You can find the source for these routines in the LIB directory. You may
modify these source files and substitute them for the library routines. By using
these routines, you can quickly adapt the library to perform (using any hardware
I/O device available in your target) stream I/O. Refer to “Stream Input and
Output” on page 217 for more information.

8
C51 Compiler 207

Standard Types
The C51 standard library contains definitions for a number of standard types
which may be used by the library routines. These standard types are declared in
include files which you may access from your C programs.

jmp_buf
The jmp_buf type is defined in SETJMP.H and specifies the buffer used by the
setjmp and longjmp routines to save and restore the program environment. The
jmp_buf type is defined as follows:
#define _JBLEN 7
typedef char jmp_buf[_JBLEN];

va_list
The va_list array type is defined in STDARG.H. This type holds data required by
the va_arg and va_end routines. The va_list type is defined as follows:
typedef char *va_list;

8
 208 Chapter 8. Library Reference

Absolute Memory Access Macros

The C51 standard library contains definitions for a number of macros that allow
you to access explicit memory addresses. These macros are defined in
ABSACC.H. Each of these macros is defined to be used like an array.

CBYTE
The CBYTE macro allows you to access individual bytes in the program
memory of the 8051 and is defined as follows:
#define CBYTE ((unsigned char volatile code *)0)

You may use this macro in your programs as follows:

rval = CBYTE [0x0002];

to read the contents of the byte in program memory at address 0002h.

CWORD
The CWORD macro allows you to access individual words in the program
memory of the 8051 and is defined as follows:
#define CWORD ((unsigned int volatile code *) 0)

You may use this macro in your programs as follows:

rval = CWORD [0x0002];

to read the contents of the word in program memory at address 0004h

(2 × sizeof (unsigned int) = 4).

NOTE
The index used with this macro does not represent the memory address of the

8 integer value. To obtain the memory address, you must multiply the index by the
sizeof an integer (2 bytes).
C51 Compiler 209

DBYTE
The DBYTE macro allows you to access individual bytes in the internal data
memory of the 8051 and is defined as follows:
#define DBYTE ((unsigned char volatile idata *) 0)

You may use this macro in your programs as follows:

rval = DBYTE [0x0002];
DBYTE [0x0002] = 5;

to read or write the contents of the byte in internal data memory at address
0002h.

DWORD
The DWORD macro allows you to access individual words in the internal data
memory of the 8051 and is defined as follows:
#define DWORD ((unsigned int volatile idata *) 0)

You may use this macro in your programs as follows:

rval = DWORD [0x0002];
DWORD [0x0002] = 57;

to read or write the contents of the word in internal data memory at address
0004h (2 × sizeof (unsigned int) = 4).

NOTE
The index used with this macro does not represent the memory address of the
integer value. To obtain the memory address, you must multiply the index by the
sizeof an integer (2 bytes).

8
 210 Chapter 8. Library Reference

PBYTE
The PBYTE macro allows you to access individual bytes in one page of the
external data memory of the 8051 and is defined as follows:
#define PBYTE ((unsigned char volatile pdata *) 0)

You may use this macro in your programs as follows:

rval = PBYTE [0x0002];
PBYTE [0x0002] = 38;

to read or write the contents of the byte in pdata memory at address 0002h.

PWORD
The PWORD macro allows you to access individual words in one page of the
external data memory of the 8051 and is defined as follows:
#define PWORD ((unsigned int volatile pdata*) 0)

You may use this macro in your programs as follows:

rval = PWORD [0x0002];
PWORD [0x0002] = 57;

to read or write the contents of the word in pdata memory at address 0004h
(2 × sizeof (unsigned int) = 4).

NOTE
The index used with this macro does not represent the memory address of the
integer value. To obtain the memory address, you must multiply the index by the
sizeof an integer (2 bytes).

8
C51 Compiler 211

XBYTE
The XBYTE macro allows you to access individual bytes in the external data
memory of the 8051 and is defined as follows:
#define XBYTE ((unsigned char volatile xdata*) 0)

You may use this macro in your programs as follows:

rval = XBYTE [0x0002];
XBYTE [0x0002] = 57;

to read or write the contents of the byte in external data memory at address
0002h.

XWORD
The XWORD macro allows you to access individual words in the external data
memory of the 8051 and is defined as follows:
#define XWORD ((unsigned int volatile xdata*) 0)

You may use this macro in your programs as follows:

rval = XWORD [2];
XWORD [2] = 57;

to read or write the contents of the word in external data memory at address
0004h (2 × sizeof (unsigned int) = 4).

NOTE
The index used with this macro does not represent the memory address of the
integer value. To obtain the memory address, you must multiply the index by the
sizeof an integer (2 bytes).

8
 212 Chapter 8. Library Reference

Routines by Category
This sections gives a brief overview of the major categories of routines available
in the Cx51 standard library. Refer to “Reference” on page 227 for a complete
description of routine syntax and usage.

NOTE
Many of the routines in the Cx51 standard library are reentrant, intrinsic, or
both. These specifications are listed under attributes in the following tables.
Unless otherwise noted, routines are non-reentrant and non-intrinsic.

Buffer Manipulation
Routine Attributes Description
memccpy Copies data bytes from one buffer to another until a
specified character or specified number of characters has
been copied.
memchr reentrant Returns a pointer to the first occurrence of a specified
character in a buffer.
memcmp reentrant Compares a given number of characters from two different
buffers.
memcpy reentrant Copies a specified number of data bytes from one buffer
to another.
memmove reentrant Copies a specified number of data bytes from one buffer
to another.
memset reentrant Initializes a specified number of data bytes in a buffer to a
specified character value.

The buffer manipulation routines are used to work on memory buffers on a

character-by-character basis. A buffer is an array of characters like a string,
however, a buffer is usually not terminated with a null character (‘\0’). For this
reason, these routines require a buffer length or count argument.

All of these routines are implemented as functions. Function prototypes are

included in the STRING.H include file.

8
C51 Compiler 213

Character Conversion and Classification

Routine Attributes Description
isalnum reentrant Tests for an alphanumeric character.
isalpha reentrant Tests for an alphabetic character.
iscntrl reentrant Tests for a Control character.
isdigit reentrant Tests for a decimal digit.
isgraph reentrant Tests for a printable character with the exception of space.
islower reentrant Tests for a lowercase alphabetic character.
isprint reentrant Tests for a printable character.
ispunct reentrant Tests for a punctuation character.
isspace reentrant Tests for a whitespace character.
isupper reentrant Tests for an uppercase alphabetic character.
isxdigit reentrant Tests for a hexadecimal digit.
toascii reentrant Converts a character to an ASCII code.
toint reentrant Converts a hexadecimal digit to a decimal value.
tolower reentrant Tests a character and converts it to lowercase if it is
uppercase.
_tolower reentrant Unconditionally converts a character to lowercase.
toupper reentrant Tests a character and converts it to uppercase if it is
lowercase.
_toupper reentrant Unconditionally converts a character to uppercase.

The character conversion and classification routines allow you to test individual
characters for a variety of attributes and convert characters to different formats.

The _tolower, _toupper, and toascii routines are implemented as macros. All
other routines are implemented as functions. All macro definitions and function
prototypes are found in the CTYPE.H include file.

8
 214 Chapter 8. Library Reference

Data Conversion
Routine Attributes Description
abs reentrant Generates the absolute value of an integer type.
atof / atof517 Converts a string to a float.
atoi Converts a string to an int.
atol Converts a string to a long.
cabs reentrant Generates the absolute value of a character type.
labs reentrant Generates the absolute value of a long type.
strtod / Converts a string to a float.
strtod517
strtol Converts a string to a long.
strtoul Converts a string to an unsigned long.

The data conversion routines convert strings of ASCII characters to numbers.

All of these routines are implemented as functions and most are prototyped in
the include file STDLIB.H. The abs, cabs, and labs functions are prototyped in
the MATH.H include file. The atof517, and strtod517 function are prototyped
in the include file 80C517.H.

Math
Routine Attributes Description
acos / acos517 Calculates the arc cosine of a specified number.
asin / asin517 Calculates the arc sine of a specified number.
atan / atan517 Calculates the arc tangent of a specified number.
atan2 Calculates the arc tangent of a fraction.
ceil Finds the integer ceiling of a specified number.
cos / cos517 Calculates the cosine of a specified number.
cosh Calculates the hyperbolic cosine of a specified number.
exp / exp517 Calculates the exponential function of a specified number.
fabs reentrant Finds the absolute value of a specified number.
floor Finds the largest integer less than or equal to a specified
number.

8 fmod
log / log517
Calculates the floating-point remainder.
Calculates the natural logarithm of a specified number.
log10 / log10517 Calculates the common logarithm of a specified number.
modf Generates integer and fractional components of a
specified number.
pow Calculates a value raised to a power.
C51 Compiler 215

Routine Attributes Description

rand reentrant Generates a pseudo random number.
sin / sin517 Calculates the sine of a specified number.
sinh Calculates the hyperbolic sine of a specified number.
srand Initializes the pseudo random number generator.
sqrt / sqrt517 Calculates the square root of a specified number.
tan / tan517 Calculates the tangent of a specified number.
tanh Calculates the hyperbolic tangent of a specified number.
_chkfloat_ intrinsic, Checks the status of float numbers.
reentrant
_crol_ intrinsic, Rotates an unsigned char left a specified number of bits.
reentrant
_cror_ intrinsic, Rotates an unsigned char right a specified number of bits.
reentrant
_irol_ intrinsic, Rotates an unsigned int left a specified number of bits.
reentrant
_iror_ intrinsic, Rotates an unsigned int right a specified number of bits.
reentrant
_lrol_ intrinsic, Rotates an unsigned long left a specified number of bits.
reentrant
_lror_ intrinsic, Rotates an unsigned long right a specified number of bits.
reentrant

The math routines perform common mathematical calculations. Most of these

routines work with floating-point values and therefore include the floating-point
libraries and support routines.

All of these routines are implemented as functions. Most are prototyped in the
include file MATH.H. Functions which end in 517 (acos517, asin517, atan517,
cos517, exp517, log517, log10517, sin517, sqrt517, and tan517) are prototyped
in the 80C517.H include file. The rand and srand functions are prototyped in the
STDLIB.H include file.

The _chkfloat_, _crol_, _cror_, _irol_, _iror_, _lrol_, and _lror_ functions are
prototyped in the INTRINS.H include file.

8
 216 Chapter 8. Library Reference

Memory Allocation
Routine Attributes Description
calloc Allocates storage for an array from the memory pool.
free Frees a memory block that was allocated using calloc,
malloc, or realloc.
init_mempool Initializes the memory location and size of the memory
pool.
malloc Allocates a block from the memory pool.
realloc Reallocates a block from the memory pool.

The memory allocation functions provide you with a means to specify, allocate,
and free blocks of memory from a memory pool. All memory allocation
functions are implemented as functions and are prototyped in the STDLIB.H
include file.

Before using any of these functions to allocate memory, you must first specify,
using the init_mempool routine, the location and size of a memory pool from
which subsequent memory requests are satisfied.

The calloc and malloc routines allocate blocks of memory from the pool. The
calloc routine allocates an array with a specified number of elements of a given
size and initializes the array to 0. The malloc routine allocates a specified
number of bytes.

The realloc routine changes the size of an allocated block, while the free routine
returns a previously allocated memory block to the memory pool.

8
C51 Compiler 217

Stream Input and Output

Routine Attributes Description
getchar reentrant Reads and echoes a character using the _getkey and
putchar routines.
_getkey Reads a character using the 8051 serial interface.
gets Reads and echoes a character string using the getchar
routine.
printf / printf517 Writes formatted data using the putchar routine.
putchar Writes a character using the 8051 serial interface.
puts reentrant Writes a character string and newline (‘\n’) character
using the putchar routine.
scanf / scanf517 Reads formatted data using the getchar routine.
sprintf / sprintf517 Writes formatted data to a string.
sscanf / sscanf517 Reads formatted data from a string.
ungetchar Places a character back into the getchar input buffer.
vprintf Writes formatted data using the putchar function.
vsprintf Writes formatted data to a string.

The stream input and output routines allow you to read and write data to and
from the 8051 serial interface or a user-defined I/O interface. The default
_getkey and putchar functions found in the Cx51 library read and write
characters using the 8051 serial interface. You can find the source for these
functions in the LIB directory. You may modify these source files and
substitute them for the library routines. When this is done, other stream
functions then perform input and output using the new _getkey and putchar
routines.

If you want to use the existing _getkey and putchar functions, you must first
initialize the 8051 serial port. If the serial port is not properly initialized, the
default stream functions do not function. Initializing the serial port requires
manipulating special function registers SFRs of the 8051. The include file
REG51.H contains definitions for the required SFRs.

8
 218 Chapter 8. Library Reference

The following example code must be executed immediately after reset, before
any stream functions are invoked.
.
.
.
#include <reg51.h>
.
.
.
SCON = 0x50; /* Setup serial port control register */
/* Mode 1: 8-bit uart var. baud rate */
/* REN: enable receiver */

PCON &= 0x7F; /* Clear SMOD bit in power ctrl reg */

/* This bit doubles the baud rate */

TMOD &= 0xCF /* Setup timer/counter mode register */

/* Clear M1 and M0 for timer 1 */
TMOD |= 0x20; /* Set M1 for 8-bit autoreload timer */

TH1 = 0xFD; /* Set autoreload value for timer 1 */

/* 9600 baud with 11.0592 MHz xtal */

TR1 = 1; /* Start timer 1 */

TI = 1; /* Set TI to indicate ready to xmit */

.
.
.

The stream routines treat input and output as streams of individual characters.
There are routines that process characters as well as functions that process
strings. Choose the routines that best suit your requirements.

All of these routines are implemented as functions. Most are prototyped in the
STDIO.H include file. The printf517, scanf517, sprintf517, and sscanf517
functions are prototyped in the 80C517.H include file.

8
C51 Compiler 219

String Manipulation
Routine Attributes Description
strcat Concatenates two strings.
strchr reentrant Returns a pointer to the first occurrence of a specified
character in a string.
strcmp reentrant Compares two strings.
strcpy reentrant Copies one string to another.
strcspn Returns the index of the first character in a string that
matches any character in a second string.
strlen reentrant Returns the length of a string.
strncat Concatenates up to a specified number of characters from
one string to another.
strncmp Compares two strings up to a specified number of
characters.
strncpy Copies up to a specified number of characters from one
string to another.
strpbrk Returns a pointer to the first character in a string that
matches any character in a second string.
strpos reentrant Returns the index of the first occurrence of a specified
character in a string.
strrchr reentrant Returns a pointer to the last occurrence of a specified
character in a string.
strrpbrk Returns a pointer to the last character in a string that
matches any character in a second string.
strrpos reentrant Returns the index of the last occurrence of a specified
character in a string.
strspn Returns the index of the first character in a string that
does not match any character in a second string.

The string routines are implemented as functions and are prototyped in the
STRING.H include file. They perform the following operations:

Copying strings

Appending one string to the end of another

Comparing two strings

Locating one or more characters from a specified set in a string

All string functions operate on null-terminated character strings. To work on 8

non-terminated strings, use the buffer manipulation routines described earlier in
this section.
 220 Chapter 8. Library Reference

Variable-length Argument Lists

Routine Attributes Description
va_arg reentrant Retrieves an argument from an argument list.
va_end reentrant Resets an argument pointer.
va_start reentrant Sets a pointer to the beginning of an argument list.

The variable-length argument list routines are implemented as macros and are
defined in the STDARG.H include file. These routines provide you with a
portable method of accessing arguments in a function that takes a variable
number of arguments. These macros conform to the ANSI C Standard for
variable-length argument lists.

Miscellaneous
Routine Attributes Description
setjmp reentrant Saves the current stack condition and program address.
longjmp reentrant Restores the stack condition and program address.
_nop_ intrinsic, Inserts an 8051 NOP instruction.
reentrant
_testbit_ intrinsic, Tests the value of a bit and clears it to 0.
reentrant

Routines found in the miscellaneous category do not fit easily into any other
library routine category. The setjmp and longjmp routines are implemented as
functions and are prototyped in the STDJMP.H include file.

The _nop_ and _testbit_ routines are used to direct the compiler to generate a
NOP instruction and a JBC instruction respectively. These routines are
prototyped in the INTRINS.H include file.

8
C51 Compiler 221

Include Files
The include files that are provided with the C51 standard library are found in the
INC subdirectory. These files contain constant and macro definitions, type
definitions, and function prototypes. The following sections describe the use
and contents of each include file. Macros and functions included in the file are
listed as well.

8051 Special Function Register Include Files

The Cx51 compiler package provides you with a number of include files that
define manifest constants for the special function registers found on many 8051
derivatives. These files can be found in the folder KEIL\C51\INC and the sub-
folders. For example, the Special Function Registers (SFR) of the Philips
80C554 device are defined in the file KEIL\C51\INC\PHILIPS\REG554.H.

Within the µVision2 editor context menu that opens on a right mouse click in a
editor window, you can insert the SFR defintion that matches the selected
device.

SFR definition files for all 8051 variants can be downloaded from
www.keil.com. The device database available on this web page contains the
header file for the Special Function Registers file of almost all 8051 devices.

80C517.H
The 80C517.H include file contains routines that use the enhanced operational
features of the 80C517 CPU and its derivatives. These routines are:

acos517 log10517 sqrt517

asin517 log517 sscanf517
atan517 printf517 strtod517
atof517 scanf517 tan517
cos517 sin517
exp517 sprintf517 8
 222 Chapter 8. Library Reference

ABSACC.H
The ABSACC.H include file contains definitions for macros that allow you to
directly access the different memory areas of the 8051.

CBYTE DWORD XBYTE

CWORD PBYTE XWORD
DBYTE PWORD

8
Keil Software — Cx51 Compiler User’s Guide 223

ASSERT.H
The ASSERT.H include file defines the assert macro you can use to create test
conditions in your programs.

CTYPE.H
The CTYPE.H include file contains definitions and prototypes for routines
which classify ASCII characters and routines which perform character
conversions. The following is a list of these routines:

isalnum isprint toint

isalpha ispunct tolower
iscntrl isspace _tolower
isdigit isupper toupper
isgraph isxdigit _toupper
islower toascii

INTRINS.H
The INTRINS.H include file contains prototypes for routines that instruct the
compiler to generate in-line intrinsic code.

_chkfloat_ _irol_ _lror_

_crol_ _iror_ _nop_
_cror_ _lrol_ _testbit_

MATH.H
The MATH.H include file contains prototypes and definitions for all routines
that perform floating-point math calculations. Other math functions are also
included in this file. All of the math function routines are listed below:

abs cosh log

acos exp log10
asin
atan
fabs
floor
modf
pow 8
atan2 fmod sin
cabs fprestore sinh
ceil fpsave sqrt
cos labs tan
 224 Chapter 8. Library Reference

tanh

8
Keil Software — Cx51 Compiler User’s Guide 225

SETJMP.H
The SETJMP.H include file defines the jmp_buf type and prototypes the setjmp
and longjmp routines which use it.

STDARG.H
The STDARG.H include file defines macros that allow you to access arguments in
functions with variable-length argument lists. The macros include:

va_arg va_end va_start

In addition, the va_list type is defined in this file.

STDDEF.H
The STDDEF.H include file defines the offsetof macro you can use to determine
the offset of members of a structure.

STDIO.H
The STDIO.H include file contains prototypes and definitions for stream I/O
routines. They are:

getchar putchar sscanf

_getkey puts ungetchar
gets scanf vprintf
printf sprintf vsprintf

The STDIO.H include file also defines the EOF manifest constant.

8
 226 Chapter 8. Library Reference

STDLIB.H
The STDLIB.H include file contains prototypes and definitions for the type
conversion and memory allocation routines listed below:

atof init_mempool strtod

atoi malloc strtol
atol rand strtoul
calloc realloc
free srand

The STDLIB.H include file also defines the NULL manifest constant.

STRING.H
The STRING.H include file contains prototypes for the following string and
buffer manipulation routines:

memccpy strchr strncpy

memchr strcmp strpbrk
memcmp strcpy strpos
memcpy strcspn strrchr
memmove strlen strrpbrk
memset strncat strrpos
strcat strncmp strspn

The STRING.H include file also defines the NULL manifest constant.

8
Keil Software — Cx51 Compiler User’s Guide 227

Reference
The following pages constitute the C51 standard library reference. The routines
included in the standard library are described here in alphabetical order and each
is divided into several sections:

Summary: Briefly describes the routine’s effect, lists include file(s)

containing its declaration and prototype, illustrates the
syntax, and describes any arguments.

Description: Provides you with a detailed description of the routine and

how it is used.

Return Value: Describes the value returned by the routine.

See Also: Names related routines.

Example: Gives a function or program fragment demonstrating proper

use of the function.

8
 228 Chapter 8. Library Reference

abs
Summary: #include <math.h>
int abs (
int val); /* number to take absolute value
of */

Description: The abs function determines the absolute value of the

integer argument val.

Return Value: The abs function returns the absolute value of val.

See Also: cabs, fabs, labs

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_abs (void) {

int x;
int y;

x = -42;

y = abs (x);

printf ("ABS(%d) = %d\n", x, y);

}

8
Keil Software — Cx51 Compiler User’s Guide 229

acos / acos517
Summary: #include <math.h>
float acos (
float x); /* number to calculate arc
cosine of */

Description: The acos function calculates the arc cosine of the

floating-point number x. The value of x must be between -
1 and 1. The floating-point value returned by acos is a
number in the 0 to π range.

The acos517 function is identical to acos, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. For using this function include the header
file 80C517.H. Do not use this routine with a CPU that does
not support this feature.

Return Value: The acos function returns the arc cosine of x.

See Also: asin, atan, atan2

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_acos (void) {

float x;
float y;

for (x = -1.0; x <= 1.0; x += 0.1) {

y = acos (x);

printf ("ACOS(%f) = %f\n", x, y);

}
}

8
 230 Chapter 8. Library Reference

asin / asin517
Summary: #include <math.h>
float asin (
float x); /* number to calculate arc sine
of */

Description: The asin function calculates the arc sine of the

floating-point number x. The value of x must be in the
range -1 to 1. The floating-point value returned by asin is a
number in the -π/2 to π/2 range.

The asin517 function is identical to asin, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. For using this function include the header
file 80C517.H. Do not use this routine with a CPU that does
not support this feature.

Return Value: The asin function returns the arc sine of x.

See Also: acos, atan, atan2

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_asin (void) {

float x;
float y;

for (x = -1.0; x <= 1.0; x += 0.1) {

y = asin (x);

printf ("ASIN(%f) = %f\n", x, y);

}
}

8
Keil Software — Cx51 Compiler User’s Guide 231

assert
Summary: #include <assert.h>
void assert (
expression);

Description: The assert macro tests expression and prints a diagnostic

message using the printf library routine if it is false.

Return Value: None.

Example: #include <assert.h>

#include <stdio.h>

void check_parms (
char *string)
{
assert (string != NULL); /* check for NULL ptr */
printf ("String %s is OK\n", string);
}

8
 232 Chapter 8. Library Reference

atan / atan517
Summary: #include <math.h>
float atan (
float x); /* number to calculate arc
tangent of */

Description: The atan function calculates the arc tangent of the

floating-point number x. The floating-point value returned
by atan is a number in the -π/2 to π/2 range.

The atan517 function is identical to atan, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. For using this function include the header
file 80C517.H. Do not use this routine with a CPU that does
not support this feature.

Return Value: The atan function returns the arc tangent of x.

See Also: acos, asin, atan2

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_atan (void) {

float x;
float y;

for (x = -10.0; x <= 10.0; x += 0.1) {

y = atan (x);

printf ("ATAN(%f) = %f\n", x, y);

}
}

8
Keil Software — Cx51 Compiler User’s Guide 233

atan2
Summary: #include <math.h>
float atan2 (
float y, /* denominator for arc tangent */
float x); /* numerator for arc tangent */

Description: The atan2 function calculates the arc tangent of the

floating-point ratio y / x. This function uses the signs of
both x and y to determine the quadrant of the return value.
The floating-point value returned by atan2 ia a number in
the -π to π range.

Return Value: The atan2 function returns the arc tangent of y / x.

See Also: acos, asin, atan

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_atan2 () {
float x;
float y;
float z;

x = -1.0;

for (y = -10.0; y < 10.0; y += 0.1) {

z = atan2 (y,x);

printf ("ATAN2(%f/%f) = %f\n", y, x, z);

}

/* z approaches -pi as y goes from -10 to 0 */

/* z approaches +pi as y goes from +10 to 0 */

8
 234 Chapter 8. Library Reference

atof / atof517
Summary: #include <stdlib.h>
float atof (
void *string); /* string to convert */

Description: The atof function converts string into a floating-point

value. The input string is a sequence of characters that can
be interpreted as a floating-point number. This function
stops processing characters from string at the first one it
cannot recognize as part of the number.

The atof517 function is identical to atof, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. For using this function include the header
file 80C517.H. Do not use this routine with a CPU that does
not support this feature.

The atof function requires string to have the following

format:

{+ | -} digits
. digits
{e | E}
{+ | -} digits
where:

digits may be one or more decimal digits.

Return Value: The atof function returns the floating-point value that is
produced by interpreting the characters in string as a
number.

See Also: atoi, atol, strtod, strtol, strtoul

Example: #include <stdlib.h>

#include <stdio.h> /* for printf */

void tst_atof (void) {

float f;
char s [] = "1.23";

8 }
f = atof (s);
printf ("ATOF(%s) = %f\n", s, f);
Keil Software — Cx51 Compiler User’s Guide 235

atoi
Summary: #include <stdlib.h>
int atoi (
void *string); /* string to convert */

Description: The atoi function converts string into an integer value.

The input string is a sequence of characters that can be
interpreted as an integer. This function stops processing
characters from string at the first one it cannot recognize as
part of the number.
The atoi function requires string to have the following
format:

whitespace
{+ | -} digits
where:

digits may be one or more decimal digits.

Return Value: The atoi function returns the integer value that is produced
by interpreting the characters in string as a number.

See Also: atof, atol, strtod, strtol, strtoul

Example: #include <stdlib.h>

#include <stdio.h> /* for printf */

void tst_atoi (void) {

int i;
char s [] = "12345";

i = atoi (s);
printf ("ATOI(%s) = %d\n", s, i);
}

8
 236 Chapter 8. Library Reference

atol
Summary: #include <stdlib.h>
long atol (
void *string); /* string to convert */

Description: The atol function converts string into a long integer value.
The input string is a sequence of characters that can be
interpreted as a long. This function stops processing
characters from string at the first one it cannot recognize as
part of the number.
The atol function requires string to have the following
format:

whitespace
{+ | -} digits

where:

digits may be one or more decimal digits.

Return Value: The atol function returns the long integer value that is
produced by interpreting the characters in string as a
number.

See Also: atof, atoi, strtod, strtol, strtoul

Example: #include <stdlib.h>

#include <stdio.h> /* for printf */

void tst_atol (void) {

long l;
char s [] = "8003488051";

l = atol (s);
printf ("ATOL(%s) = %ld\n", s, l);

8
Keil Software — Cx51 Compiler User’s Guide 237

cabs
Summary: #include <math.h>
char cabs (
char val); /* character to take absolute value of */

Description: The cabs function determines the absolute value of the

character argument val.

Return Value: The cabs function returns the absolute value of val.

See Also: abs, fabs, labs

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_cabs (void) {

char x;
char y;

x = -23;

y = cabs (x);

printf ("CABS(%bd) = %bd\n", x, y);

}

8
 238 Chapter 8. Library Reference

calloc
Summary: #include <stdlib.h>
void *calloc (
unsigned int num, /* number of items */
unsigned int len); /* length of each item */

Description: The calloc function allocates memory for an array of num

elements. Each element in the array occupies len bytes and
is initialized to 0. The total number of bytes of memory
allocated is num × len.

NOTE
Source code is provided for this routine in the LIB
directory. You can modify the source to customize this
function for your hardware environment. Refer to
“Chapter 6. Advanced Programming Techniques” on page
143 for more information.

Return Value: The calloc function returns a pointer to the allocated

memory or a null pointer if the memory allocation request
cannot be satisfied.

See Also: free, init_mempool, malloc, realloc

Example: #include <stdlib.h>

#include <stdio.h> /* for printf */

void tst_calloc (void) {

int xdata *p; /* ptr to array of 100 ints */

p = calloc (100, sizeof (int));

if (p == NULL)
printf ("Error allocating array\n");
else
printf ("Array address is %p\n", (void *) p);

8
Keil Software — Cx51 Compiler User’s Guide 239

ceil
Summary: #include <math.h>
float ceil (
float val); /* number to calculate ceiling for */

Description: The ceil function calculates the smallest integer value that is
greater than or equal to val.

Return Value: The ceil function returns a float that contains the smallest
integer value that is not less than val.

See Also: floor

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_ceil (void) {

float x;
float y;

x = 45.998;
y = ceil (x);

printf ("CEIL(%f) = %f\n", x, y);

/* output is "CEIL(45.998) = 46" */

8
 240 Chapter 8. Library Reference

_chkfloat_
Summary: #include <intrins.h>
unsigned char _chkfloat_ (
float val); /* number for error checking */

Description: The _chkfloat_ function checks the status of a floating-point

number.

Return Value: The _chkfloat_ function returns an unsigned char that

contains the following status information:

Return Value Meaning

0 Standard floating-point numbers
1 Floating-point value 0
2 +INF (positive overflow)
3 -INF (negative overflow)
4 NaN (Not a Number) error status

Example: #include <intrins.h>

#include <stdio.h> /* for printf */

char _chkfloat_ (float);

float f1, f2, f3;

void tst_chkfloat (void) {

f1 = f2 * f3;

switch (_chkfloat_ (f1)) {

case 0:
printf ("result is a number\n"); break;
case 1:
printf ("result is zero\n"); break;
case 2:
printf ("result is +INF\n"); break;
case 3:
printf ("result is -INF\n"); break;
case 4:
printf ("result is NaN\n"); break;
}
}

8
Keil Software — Cx51 Compiler User’s Guide 241

cos / cos517
Summary: #include <math.h>
float cos (
float x); /* number to calculate cosine
for */

Description: The cos function calculates the cosine of the floating-point

value x. The value of x must be between -65535 and
65535. Values outside this range result in an NaN error.

The cos517 function is identical to cos, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. For using this function include the header
file 80C517.H. Do not use this routine with a CPU that does
not support this feature.

Return Value: The cos function returns the cosine for the value x.

See Also: sin, tan

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_cos (void) {

float x;
float y;

for (x = 0; x < (2 * 3.1415); x += 0.1) {

y = cos (x);

printf ("COS(%f) = %f\n", x, y);

}
}

8
 242 Chapter 8. Library Reference

cosh
Summary: #include <math.h>
float cosh (
float x); /* value for hyperbolic cos
function */

Description: The cosh function calculates the hyperbolic cosine of the

floating-point value x.

Return Value: The cosh function returns the hyperbolic cosine for the
value x.

See Also: sinh, tanh

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_cosh (void) {

float x;
float y;

for (x = 0; x < (2 * 3.1415); x += 0.1) {

y = cosh (x);

printf ("COSH(%f) = %f\n", x, y);

}
}

8
Keil Software — Cx51 Compiler User’s Guide 243

_crol_
Summary: #include <intrins.h>
unsigned char _crol_ (
unsigned char c, /* character to rotate left */
unsigned char b); /* bit positions to rotate */

Description: The _crol_ routine rotates the bit pattern for the character c
left b bits. This routine is implemented as an intrinsic
function. The code required is included in-line rather than
being called.

Return Value: The _crol_ routine returns the rotated value of c.

See Also: _cror_, _irol_, _iror_, _lrol_, _lror_

Example: #include <intrins.h>

void tst_crol (void) {

char a;
char b;

a = 0xA5;

b = _crol_(a,3); /* b now is 0x2D */

8
 244 Chapter 8. Library Reference

_cror_
Summary: #include <intrins.h>
unsigned char _cror_ (
unsigned char c, /* character to rotate right */
unsigned char b); /* bit positions to rotate */

Description: The _cror_ routine rotates the bit pattern for the character c
right b bits. This routine is implemented as an intrinsic
function. The code required is included in-line rather than
being called.

Return Value: The _cror_ routine returns the rotated value of c.

See Also: _crol_, _irol_, _iror_, _lrol_, _lror_

Example: #include <intrins.h>

void tst_cror (void) {

char a;
char b;

a = 0xA5;

b = _crol_(a,1); /* b now is 0xD2 */

8
Keil Software — Cx51 Compiler User’s Guide 245

exp / exp517
Summary: #include <math.h>
float exp (
float x); /* power to use for ex function
*/

Description: The exp function calculates the exponential function for the
floating-point value x.

The exp517 function is identical to exp, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. For using this function include the header
file 80C517.H. Do not use this routine with a CPU that does
not support this feature.

Return Value: The exp function returns the floating-point value ex.

See Also: log, log10

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_exp (void) {

float x;
float y;

x = 4.605170186;

y = exp (x); /* y = 100 */

printf ("EXP(%f) = %f\n", x, y);

8
 246 Chapter 8. Library Reference

fabs
Summary: #include <math.h>
float fabs (
float val); /* number to calc absolute value for */

Description: The fabs function determines the absolute value of the

floating-point number val.

Return Value: The fabs function returns the absolute value of val.

See Also: abs, cabs, labs

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_fabs (void) {

float x;
float y;

x = 10.2;
y = fabs (x);
printf ("FABS(%f) = %f\n", x, y);

x = -3.6;
y = fabs (x);
printf ("FABS(%f) = %f\n", x, y);

8
Keil Software — Cx51 Compiler User’s Guide 247

floor
Summary: #include <math.h>
float floor (
float val); /* value for floor function */

Description: The floor function calculates the largest integer value that is
less than or equal to val.

Return Value: The floor function returns a float that contains the largest
integer value that is not greater than val.

See Also: ceil

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_floor (void) {

float x;
float y;

x = 45.998;

y = floor (x);

printf ("FLOOR(%f) = %f\n", x, y); /* prints 45 */

}

8
 248 Chapter 8. Library Reference

fmod
Summary: #include <math.h>
float fmod (
float x, /* value to calculate modulo for */
float y); /* integer portion of modulo */

Description: The fmod function returns a value f such that f has the
same sign as x, the absolute value of f is less than the
absolute value of y, and there exists an integer k such that
k*y+f equals x. If the quotient x / y cannot be represented,
the result is undefined.

Return Value: The fmod function returns the floating-point remainder of

x / y.
Example: #include <math.h>
#include <stdio.h> /* for printf */

void tst_fmod (void) {

float f;

f = fmod (15.0, 4.0);

printf ("fmod (15.0, 4.0) = %f\n", f);

8
Keil Software — Cx51 Compiler User’s Guide 249

free
Summary: #include <stdlib.h>
void free (
void xdata *p); /* block to free */

Description: The free function returns a memory block to the memory

pool. The p argument points to a memory block allocated
with the calloc, malloc, or realloc functions. Once it has
been returned to the memory pool by the free function, the
block is available for subsequent allocation.

If p is a null pointer, it is ignored.

NOTE
Source code for this routine is located in the folder
\KEIL\C51\LIB. You may modify the source to customize this
function for your particular hardware environment. Refer
to “Chapter 6. Advanced Programming Techniques” on
page 143 for more information.

Return Value: None.

See Also: calloc, init_mempool, malloc, realloc

Example: #include <stdlib.h>

#include <stdio.h> /* for printf */

void tst_free (void) {

void *mbuf;

printf ("Allocating memory\n");

mbuf = malloc (1000);

if (mbuf == NULL) {
printf ("Unable to allocate memory\n");
}
else {
free (mbuf);
printf ("Memory free\n");
}

}
8
 250 Chapter 8. Library Reference

getchar
Summary: #include <stdio.h>
char getchar (void);

Description: The getchar function reads a single character from the input
stream using the _getkey function. The character read is
then passed to the putchar function to be echoed.

NOTE
This function is implementation-specific and is based on the
operation of the _getkey and/or putchar functions. These
functions, as provided in the standard library, read and
write characters using the serial port of the 8051. Custom
functions may use other I/O devices.

Return Value: The getchar function returns the character read.

See Also: _getkey, putchar, ungetchar

Example: #include <stdio.h>

void tst_getchar (void) {

char c;

while ((c = getchar ()) != 0x1B) {

printf ("character = %c %bu %bx\n", c, c, c);
}

8
Keil Software — Cx51 Compiler User’s Guide 251

_getkey
Summary: #include <stdio.h>
char _getkey (void);

Description: The _getkey function waits for a character to be received

from the serial port.

NOTE
This routine is implementation-specific, and its function may
deviate from that described above. Source is included for
the _getkey and putchar functions which may be modified to
provide character level I/O for any hardware device. Refer
to “Customization Files” on page 143 for more information.

Return Value: The _getkey routine returns the received character.

See Also: getchar, putchar, ungetchar

Example: #include <stdio.h>

void tst_getkey (void) {

char c;

while ((c = _getkey ()) != 0x1B) {

printf ("key = %c %bu %bx\n", c, c, c);
}

8
 252 Chapter 8. Library Reference

gets
Summary: #include <stdio.h>
char *gets (
char *string, /* string to read */
int len); /* maximum characters to read
*/

Description: The gets function calls the getchar function to read a line of
characters into string. The line consists of all characters up
to and including the first newline character (‘\n’). The
newline character is replaced by a null character (‘\0’) in
string.

The len argument specifies the maximum number of

characters that may be read. If len characters are read
before a newline is encountered, the gets function terminates
string with a null character and returns.

NOTE
This function is implementation-specific and is based on the
operation of the _getkey and/or putchar functions. These
functions, as provided in the standard library, read and
write characters using the serial port of the 8051. Custom
functions may use other I/O devices.

Return Value: The gets function returns string.

See Also: printf, puts, scanf

Example: #include <stdio.h>

void tst_gets (void) {

xdata char buf [100];

do {
gets (buf, sizeof (buf));
printf ("Input string \"%s\"", buf);
} while (buf [0] != '\0');

8 }
Keil Software — Cx51 Compiler User’s Guide 253

init_mempool
Summary: #include <stdlib.h>
void init_mempool (
void xdata *p, /* start of memory pool */
unsigned int size); /* length of memory pool */

Description: The init_mempool function initializes the memory

management routines and provides the starting address and
size of the memory pool. The p argument points to a
memory area in xdata which is managed using the calloc,
free, malloc, and realloc library functions. The size
argument specifies the number of bytes to use for the
memory pool.

NOTE
This function must be used to setup the memory pool before
any other memory management functions (calloc, free,
malloc, realloc) can be called. Call the init_mempool
function only once at the beginning of your program.

Source code is provided for this routine in the folder

\KEIL\C51\LIB. You can modify the source to customize this
function for your hardware environment. Refer to
“Chapter 6. Advanced Programming Techniques” on page
143 for more information.

Return Value: None.

See Also: calloc, free, malloc, realloc

Example: #include <stdlib.h>

void tst_init_mempool (void) {

xdata void *p;
int i;

init_mempool (&XBYTE [0x2000], 0x1000);

/* initialize memory pool at xdata 0x2000
for 4096 bytes */

p = malloc (100);
8
for (i = 0; i < 100; i++) ((char *) p)[i] = i;
free (p);

}
 254 Chapter 8. Library Reference

_irol_
Summary: #include <intrins.h>
unsigned int _irol_ (
unsigned int i, /* integer to rotate left */
unsigned char b); /* bit positions to rotate */

Description: The _irol_ routine rotates the bit pattern for the integer i
left b bits. This routine is implemented as an intrinsic
function. The code required is included in-line rather than
being called.

Return Value: The _irol_ routine returns the rotated value of i.

See Also: _cror_, _crol_, _iror_, _lrol_, _lror_

Example: #include <intrins.h>

void tst_irol (void) {

int a;
int b;

a = 0xA5A5;

b = _irol_(a,3); /* b now is 0x2D2D */

8
Keil Software — Cx51 Compiler User’s Guide 255

_iror_
Summary: #include <intrins.h>
unsigned int _iror_ (
unsigned int i, /* integer to rotate right */
unsigned char b); /* bit positions to rotate */

Description: The _iror_ routine rotates the bit pattern for the integer i
right b bits. This routine is implemented as an intrinsic
function. The code required is included in-line rather than
being called.

Return Value: The _iror_ routine returns the rotated value of i.

See Also: _cror_, _crol_, _irol_, _lrol_, _lror_

Example: #include <intrins.h>

void tst_iror (void) {

int a;
int b;

a = 0xA5A5;

b = _irol_(a,1); /* b now is 0xD2D2 */

8
 256 Chapter 8. Library Reference

isalnum
Summary: #include <ctype.h>
bit isalnum (
char c); /* character to test */

Description: The isalnum function tests c to determine if it is an

alphanumeric character (‘A’-‘Z’, ‘a’-‘z’, ‘0’-‘9’).

Return Value: The isalnum function returns a value of 1 if c is an

alphanumeric character or a value of 0 if it is not.

See Also: isalpha, iscntrl, isdigit, isgraph, islower, isprint, ispunct,

isspace, isupper, isxdigit

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_isalnum (void) {

unsigned char i;
char *p;

for (i = 0; i < 128; i++) {

p = (isalnum (i) ? "YES" : "NO");

printf ("isalnum (%c) %s\n", i, p);

}

8
Keil Software — Cx51 Compiler User’s Guide 257

isalpha
Summary: #include <ctype.h>
bit isalpha (
char c); /* character to test */

Description: The isalpha function tests c to determine if it is an

alphabetic character (‘A’-‘Z’ or ‘a’-‘z’).

Return Value: The isalpha function returns a value of 1 if c is an

alphabetic character and a value of 0 if it is not.

See Also: isalnum, iscntrl, isdigit, isgraph, islower, isprint, ispunct,

isspace, isupper, isxdigit

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_isalpha (void) {

unsigned char i;
char *p;

for (i = 0; i < 128; i++) {

p = (isalpha (i) ? "YES" : "NO");

printf ("isalpha (%c) %s\n", i, p);

}

8
 258 Chapter 8. Library Reference

iscntrl
Summary: #include <ctype.h>
bit iscntrl (
char c); /* character to test */

Description: The iscntrl function tests c to determine if it is a control

character (0x00-0x1F or 0x7F).

Return Value: The iscntrl function returns a value of 1 if c is a control

character and a value of 0 if it is not.

See Also: isalnum, isalpha, isdigit, isgraph, islower, isprint,

ispunct, isspace, isupper, isxdigit

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_iscntrl (void) {

unsigned char i;
char *p;

for (i = 0; i < 128; i++) {

p = (iscntrl (i) ? "YES" : "NO");

printf ("iscntrl (%c) %s\n", i, p);

}

8
Keil Software — Cx51 Compiler User’s Guide 259

isdigit
Summary: #include <ctype.h>
bit isdigit (
char c); /* character to test */

Description: The isdigit function tests c to determine if it is a decimal

digit (‘0’-‘9’).

Return Value: The isdigit function returns a value of 1 if c is a decimal

digit and a value of 0 if it is not.

See Also: isalnum, isalpha, iscntrl, isgraph, islower, isprint,

ispunct, isspace, isupper, isxdigit

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_isdigit (void) {

unsigned char i;
char *p;

for (i = 0; i < 128; i++) {

p = (isdigit (i) ? "YES" : "NO");

printf ("isdigit (%c) %s\n", i, p);

}

8
 260 Chapter 8. Library Reference

isgraph
Summary: #include <ctype.h>
bit isgraph (
char c); /* character to test */

Description: The isgraph function tests c to determine if it is a printable

character (not including space). The character values tested
for are 0x21-0x7E.

Return Value: The isgraph function returns a value of 1 if c is a printable

character and a value of 0 if it is not.

See Also: isalnum, isalpha, iscntrl, isdigit, islower, isprint, ispunct,

isspace, isupper, isxdigit

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_isgraph (void) {

unsigned char i;
char *p;

for (i = 0; i < 128; i++) {

p = (isgraph (i) ? "YES" : "NO");

printf ("isgraph (%c) %s\n", i, p);

}

8
Keil Software — Cx51 Compiler User’s Guide 261

islower
Summary: #include <ctype.h>
bit islower (
char c); /* character to test */

Description: The islower function tests c to determine if it is a

lowercase alphabetic character (‘a’-‘z’).

Return Value: The islower function returns a value of 1 if c is a lowercase

letter and a value of 0 if it is not.

See Also: isalnum, isalpha, iscntrl, isdigit, isgraph, isprint, ispunct,

isspace, isupper, isxdigit

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_islower (void) {

unsigned char i;
char *p;

for (i = 0; i < 128; i++) {

p = (islower (i) ? "YES" : "NO");

printf ("islower (%c) %s\n", i, p);

}

8
 262 Chapter 8. Library Reference

isprint
Summary: #include <ctype.h>
bit isprint (
char c); /* character to test */

Description: The isprint function tests c to determine if it is a printable

character (0x20-0x7E).

Return Value: The isprint function returns a value of 1 if c is a printable

character and a value of 0 if it is not.

See Also: isalnum, isalpha, iscntrl, isdigit, isgraph, islower,

ispunct, isspace, isupper, isxdigit

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_isprint (void) {

unsigned char i;
char *p;

for (i = 0; i < 128; i++) {

p = (isprint (i) ? "YES" : "NO");

printf ("isprint (%c) %s\n", i, p);

}

8
Keil Software — Cx51 Compiler User’s Guide 263

ispunct
Summary: #include <ctype.h>
bit ispunct (
char c); /* character to test */

Description: The ispunct function tests c to determine if it is a

punctuation character. The following symbols are
punctuation characters:

! " # $ % & ' (

) * + , - . / :
; < = > ? @ [ \
] ^ _ ` { | } ~

Return Value: The ispunct function returns a value of 1 if c is a

punctuation character and a value of 0 if it is not.

See Also: isalnum, isalpha, iscntrl, isdigit, isgraph, islower, isprint,

isspace, isupper, isxdigit

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_ispunct (void) {

unsigned char i;
char *p;

for (i = 0; i < 128; i++) {

p = (ispunct (i) ? "YES" : "NO");

printf ("ispunct (%c) %s\n", i, p);

}

8
 264 Chapter 8. Library Reference

isspace
Summary: #include <ctype.h>
bit isspace (
char c); /* character to test */

Description: The isspace function tests c to determine if it is a

whitespace character (0x09-0x0D or 0x20).

Return Value: The isspace function returns a value of 1 if c is a

whitespace character and a value of 0 if it is not.

See Also: isalnum, isalpha, iscntrl, isdigit, isgraph, islower, isprint,

ispunct, isupper, isxdigit

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_isspace (void) {

unsigned char i;
char *p;

for (i = 0; i < 128; i++) {

p = (isspace (i) ? "YES" : "NO");

printf ("isspace (%c) %s\n", i, p);

}

8
Keil Software — Cx51 Compiler User’s Guide 265

isupper
Summary: #include <ctype.h>
bit isupper (
char c); /* character to test */

Description: The isupper function tests c to determine if it is an

uppercase alphabetic character (‘A’-‘Z’).

Return Value: The isupper function returns a value of 1 if c is an

uppercase character and a value of 0 if it is not.

See Also: isalnum, isalpha, iscntrl, isdigit, isgraph, islower, isprint,

ispunct, isspace, isxdigit

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_isupper (void) {

unsigned char i;
char *p;

for (i = 0; i < 128; i++) {

p = (isupper (i) ? "YES" : "NO");

printf ("isupper (%c) %s\n", i, p);

}

8
 266 Chapter 8. Library Reference

isxdigit
Summary: #include <ctype.h>
bit isxdigit (
char c); /* character to test */

Description: The isxdigit function tests c to determine if it is a

hexadecimal digit (‘A’-‘Z’, ‘a’-‘z’, ‘0’-‘9’).

Return Value: The isxdigit function returns a value of 1 if c is a

hexadecimal digit and a value of 0 if it is not.

See Also: isalnum, isalpha, iscntrl, isdigit, isgraph, islower, isprint,

ispunct, isspace, isupper

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_isxdigit (void) {

unsigned char i;
char *p;

for (i = 0; i < 128; i++) {

p = (isxdigit (i) ? "YES" : "NO");

printf ("isxdigit (%c) %s\n", i, p);

}

8
Keil Software — Cx51 Compiler User’s Guide 267

labs
Summary: #include <math.h>
long labs (
long val); /* value to calc. abs. value for */

Description: The labs function determines the absolute value of the long
integer val.

Return Value: The labs function returns the absolute value of val.

See Also: abs, cabs, fabs

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_labs (void) {

long x;
long y;

x = -12345L;

y = labs (x);

printf ("LABS(%ld) = %ld\n", x, y);

8
 268 Chapter 8. Library Reference

log / log517
Summary: #include <math.h>
float log (
float val); /* value to take natural logarithm of */

Description: The log function calculates the natural logarithm for the
floating-point number val. The natural logarithm uses the
base e or 2.718282…

The log517 function is identical to log, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. When using this function, include the
header file 80C517.H. Do not use this routine with a CPU
that does not support this feature.

Return Value: The log function returns the floating-point natural logarithm
of val.

See Also: exp, log10

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_log (void) {

float x;
float y;

x = 2.71838;
x *= x;

y = log (x); /* y = 2 */

printf ("LOG(%f) = %f\n", x, y);

8
Keil Software — Cx51 Compiler User’s Guide 269

log10 / log10517
Summary: #include <math.h>
float log10 (
float val); /* value to take common logarithm of */

Description: The log10 function calculates the common logarithm for the
floating-point number val. The common logarithm uses
base 10.

The log10517 function is identical to log10, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. When using this function, include the
header file 80C517.H. Do not use this routine with a CPU
that does not support this feature.

Return Value: The log10 function returns the floating-point common

logarithm of val.

See Also: exp, log

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_log10 (void) {

float x;
float y;

x = 1000;

y = log10 (x); /* y = 3 */

printf ("LOG10(%f) = %f\n", x, y);

8
 270 Chapter 8. Library Reference

longjmp
Summary: #include <setjmp.h>
void longjmp (
jmp_buf env, /* environment to restore */
int retval); /* return value */

Description: The longjmp function restores the state which was

previously stored in env by the setjmp function. The
retval argument specifies the value to return from the
setjmp function call.

The longjmp and setjmp functions can be used to execute a

non-local goto and are usually utilized to pass control to an
error recovery routine.

Local variables and function arguments are restored only if

declared with the volatile attribute.

Return Value: None.

See Also: setjmp

8
Keil Software — Cx51 Compiler User’s Guide 271

Example: #include <setjmp.h>

#include <stdio.h> /* for printf */

jmp_buf env; /* jump environment (must be global) */

bit error_flag;

void trigger (void) {

.
.
.
/* put processing code here */
.
.
.
if (error_flag != 0) {
longjmp (env, 1); /* return 1 to setjmp */
}
.
.
.
}

void recover (void) {

/* put recovery code here */
}

void tst_longjmp (void) {

.
.
.
if (setjmp (env) != 0) { /* setjmp returns a 0 */
printf ("LONGJMP called\n");
recover ();
}
else {
printf ("SETJMP called\n");

error_flag = 1; /* force an error */

trigger ();
}
}

8
 272 Chapter 8. Library Reference

_lrol_
Summary: #include <intrins.h>
unsigned long _lrol_ (
unsigned long l, /* 32-bit integer to rotate left */
unsigned char b); /* bit positions to rotate */

Description: The _lrol_ routine rotates the bit pattern for the long integer
l left b bits. This routine is implemented as an intrinsic
function. The code required is included in-line rather than
being called.

Return Value: The _lrol_ routine returns the rotated value of l.

See Also: _cror_, _crol_, _irol_, _iror_, _lror_

Example: #include <intrins.h>

void tst_lrol (void) {

long a;
long b;

a = 0xA5A5A5A5;

b = _lrol_(a,3); /* b now is 0x2D2D2D2D */

8
Keil Software — Cx51 Compiler User’s Guide 273

_lror_
Summary: #include <intrins.h>
unsigned long _lror_ (
unsigned long l, /* 32-bit int to rotate right */
unsigned char b); /* bit positions to rotate */

Description: The _lror_ routine rotates the bit pattern for the long integer
l right b bits. This routine is implemented as an intrinsic
function. The code required is included in-line rather than
being called.

Return Value: The _lror_ routine returns the rotated value of l.

See Also: _cror_, _crol_, _irol_, _iror_, _lrol_

Example: #include <intrins.h>

void tst_lror (void) {

long a;
long b;

a = 0xA5A5A5A5;

b = _lrol_(a,1); /* b now is 0xD2D2D2D2 */

8
 274 Chapter 8. Library Reference

malloc
Summary: #include <stdlib.h>
void *malloc (
unsigned int size); /* block size to allocate */

Description: The malloc function allocates a memory block from the

memory pool of size bytes in length.

NOTE
Source code is provided for this routine in the LIB
directory. You may modify the source to customize this
function for your hardware environment. Refer to
“Chapter 6. Advanced Programming Techniques” on page
143 for more information.

Return Value: The malloc function returns a pointer to the allocated block
or a null pointer if there was not enough memory to satisfy
the allocation request.

See Also: calloc, free, init_mempool, realloc

Example: #include <stdlib.h>

#include <stdio.h> /* for printf */

void tst_malloc (void) {

unsigned char xdata *p;

p = malloc (1000); /* allocate 1000 bytes */

if (p == NULL)
printf ("Not enough memory space\n");
else
printf ("Memory allocated\n");

8
Keil Software — Cx51 Compiler User’s Guide 275

memccpy
Summary: #include <string.h>
void *memccpy (
void *dest, /* destination buffer */
void *src, /* source buffer */
char c, /* character which ends copy */
int len); /* maximum bytes to copy */

Description: The memccpy function copies 0 or more characters from

src to dest. Characters are copied until the character c is
copied or until len bytes have been copied, whichever
comes first.

Return Value: The memccpy function returns a pointer to the byte in dest
that follows the last character copied or a null pointer if the
last character copied was c.

See Also: memchr, memcmp, memcpy, memmove, memset

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_memccpy (void) {

static char src1 [100] = "Copy this string
to dst1";
static char dst1 [100];

void *c;

c = memccpy (dst1, src1, 'g', sizeof (dst1));

if (c == NULL)
printf ("'g' was not found in the src
buffer\n");
else
printf ("characters copied up to 'g'\n");

8
 276 Chapter 8. Library Reference

memchr
Summary: #include <string.h>
void *memchr (
void *buf, /* buffer to search */
char c, /* byte to find */
int len); /* maximum buffer length */

Description: The memchr function scans buf for the character c in the
first len bytes of the buffer.

Return Value: The memchr function returns a pointer to the character c

in buf or a null pointer if the character was not found.

See Also: memccpy, memcmp, memcpy, memmove, memset

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_memchr (void) {

static char src1 [100] =
"Search this string from the start";

void *c;

c = memchr (src1, 'g', sizeof (src1));

if (c == NULL)
printf ("'g' was not found in the buffer\n");
else
printf ("found 'g' in the buffer\n");

8
Keil Software — Cx51 Compiler User’s Guide 277

memcmp
Summary: #include <string.h>
char memcmp (
void *buf1, /* first buffer */
void *buf2, /* second buffer */
int len); /* maximum bytes to compare
*/

Description: The memcmp function compares two buffers buf1 and

buf2 for len bytes and returns a value indicating their
relationship as follows:

Value Meaning
<0 buf1 less than buf2
=0 but1 equal to buf2
>0 buf1 greater than buf2

Return Value: The memcmp function returns a positive, negative, or zero

value indicating the relationship of buf1 and buf2.

See Also: memccpy, memchr, memcpy, memmove, memset

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_memcmp (void) {

static char hexchars [] = "0123456789ABCDEF";
static char hexchars2 [] = "0123456789abcdef";

char i;

i = memcmp (hexchars, hexchars2, 16);

if (i < 0)
printf ("hexchars < hexchars2\n");

else if (i > 0)
printf ("hexchars > hexchars2\n");

else

}
printf ("hexchars == hexchars2\n");
8
 278 Chapter 8. Library Reference

memcpy
Summary: #include <string.h>
void *memcpy (
void *dest, /* destination buffer */
void *src, /* source buffer */
int len); /* maximum bytes to copy */

Description: The memcpy function copies len bytes from src to dest.
If these memory buffers overlap, the memcpy function
cannot guarantee that bytes in src are copied to dest
before being overwritten. If these buffers do overlap, use
the memmove function.

Return Value: The memcpy function returns dest.

See Also: memccpy, memchr, memcmp, memmove, memset

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_memcpy (void) {

static char src1 [100] =
"Copy this string to dst1";

static char dst1 [100];

char *p;

p = memcpy (dst1, src1, sizeof (dst1));

printf ("dst = \"%s\"\n", p);

8
Keil Software — Cx51 Compiler User’s Guide 279

memmove
Summary: #include <string.h>
void *memmove (
void *dest, /* destination buffer */
void *src, /* source buffer */
int len); /* maximum bytes to move */

Description: The memmove function copies len bytes from src to

dest. If these memory buffers overlap, the memmove
function ensures that bytes in src are copied to dest before
being overwritten.

Return Value: The memmove function returns dest.

See Also: memccpy, memchr, memcmp, memcpy, memset

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_memmove (void) {

static char buf [] = "This is line 1 "
"This is line 2 "
"This is line 3 ";

printf ("buf before = %s\n", buf);

memmove (&buf [0], &buf [16], 32);

printf ("buf after = %s\n", buf);

8
 280 Chapter 8. Library Reference

memset
Summary: #include <string.h>
void *memset (
void *buf, /* buffer to initialize */
char c, /* byte value to set */
int len); /* buffer length */

Description: The memset function sets the first len bytes in buf to c.

Return Value: The memset function returns dest.

See Also: memccpy, memchr, memcmp, memcpy, memmove

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_memset (void) {

char buf [10];

memset (buf, '\0', sizeof (buf));

/* fill buffer with null characters */

8
Keil Software — Cx51 Compiler User’s Guide 281

modf
Summary: #include <math.h>
float modf (
float val, /* value to calculate modulo for */
float *ip); /* integer portion of modulo */

Description: The modf function splits the floating-point number val into
integer and fractional components. The fractional part of
val is returned as a signed floating-point number. The
integer part is stored as a floating-point number at ip.

Return Value: The modf function returns the signed fractional part of val.

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_modf (void) {

float x;
float int_part, frc_part;

x = 123.456;

frc_part = modf (x, &int_part);

printf ("%f = %f + %f\n", x, int_part,frc_part);

8
 282 Chapter 8. Library Reference

_nop_
Summary: #include <intrins.h>
void _nop_ (void);

Description: The _nop_ routine inserts an 8051 NOP instruction into the
program. This routine can be used to pause for 1 CPU
cycle. This routine is implemented as an intrinsic function.
The code required is included in-line rather than being
called.

Return Value: None.

Example: #include <intrins.h>

#include <stdio.h> /* for printf */

void tst_nop (void) {

P1 = 0xFF;

_nop_ (); /* delay for hardware */

_nop_ ();
_nop_ ();

P1 = 0x00;

8
Keil Software — Cx51 Compiler User’s Guide 283

offsetof
Summary: #include <stddef.h>
int offsetof (
structure, /* structure to use */
member); /* member to get offset for */

Description: The offsetof macro calculates the offset of the member

structure element from the beginning of the structure. The
structure argument must specify the name of a structure.
The member argument must specify the name of a member
of the structure.

Return Value: The offsetof macro returns the offset, in bytes, of the
member element from the beginning of struct structure.

Example: #include <stddef.h>

struct index_st
{
unsigned char type;
unsigned long num;
unsigned ing len;
};

typedef struct index_st index_t;

void main (void)

{
int x, y;

x = offsetof (struct index_st, len); /* x = 5 */

y = offsetof (index_t, num); /* x = 1 */
}

8
 284 Chapter 8. Library Reference

pow
Summary: #include <math.h>
float pow (
float x, /* value to use for base */
float y); /* value to use for exponent */

Description: The pow function calculates x raised to the yth power.

Return Value: The pow function returns the value xy. If x ≠ 0 and y = 0,
pow returns a value of 1. If x = 0 and y ≤ 0, pow returns
NaN. If x < 0 and y is not an integer, pow returns NaN.

See Also: sqrt

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_pow (void) {

float base;
float power;
float y;

base = 2.0;
power = 8.0;

y = pow (base, power); /* y = 256 */

printf ("%f ^ %f = %f\n", base, power, y);

8
Keil Software — Cx51 Compiler User’s Guide 285

printf / printf517
Summary: #include <stdio.h>
int printf (
const char *fmtstr /* format string */

, arguments…); /* additional arguments */

Description: The printf function formats a series of strings and numeric

values and builds a string to write to the output stream using
the putchar function. The fmtstr argument is a format
string and may be composed of characters, escape
sequences, and format specifications.

Ordinary characters and escape sequences are copied to the

stream in the order in which they are interpreted. Format
specifications always begin with a percent sign (‘%’) and
require additional arguments to be included in the function
call.

The format string is read from left to right. The first format
specification encountered references the first argument after
fmtstr and converts and outputs it using the format
specification. The second format specification accesses the
second argument after fmtstr, and so on. If there are more
arguments than format specifications, the extra arguments
are ignored. Results are unpredictable if there are not
enough arguments for the format specifications.

Format specifications have the following format:

%
flags
width
. precision
{b | B | l | L} type

Each field in the format specification can be a single

character or a number which specifies a particular format
option.

8
 286 Chapter 8. Library Reference

The type field is a single character that specifies whether

the argument is interpreted as a character, string, number, or
pointer, as shown in the following table.

Character Argument Type Output Format

d int Signed decimal number
u unsigned int Unsigned decimal number
o unsigned int Unsigned octal number
x unsigned int Unsigned hexadecimal number
using “0123456789abcdef”
X unsigned int Unsigned hexadecimal number
using “0123456789ABCEDF”
f float Floating-point number using the
format [-]dddd.dddd
e float Floating-point number using the
format [-]d.dddde[-]dd
E float Floating-point number using the
format [-]d.ddddE[-]dd
g float Floating-point number using either e
or f format, whichever is more
compact for the specified value and
precision
G float Identical to the g format except that
(where applicable) E precedes the
exponent instead of e
c char Single character
s generic * String with a terminating null
character
p generic * Pointer using the format t:aaaa
where t is the memory type the
pointer references (c: code,
i: data/idata, x: xdata, p: pdata) and
aaaa is the hexadecimal address

The optional characters b or B and l or L may immediately

precede the type character to respectively specify char or
long versions of the integer types d, i, u, o, x, and X.

8
Keil Software — Cx51 Compiler User’s Guide 287

The flags field is a single character used to justify the

output and to print +/- signs and blanks, decimal points, and
octal and hexadecimal prefixes, as shown in the following
table.

Flag Meaning
- Left justify the output in the specified field width.
+ Prefix the output value with a + or - sign if the output is a
signed type.
blank (‘ ’) Prefix the output value with a blank if it is a signed positive
value. Otherwise, no blank is prefixed.
# Prefixes a non-zero output value with 0, 0x, or 0X when
used with o, x, and X field types, respectively.

When used with the e, E, f, g, and G field types, the # flag

forces the output value to include a decimal point.

The # flag is ignored in all other cases.

* Ignore format specifier.

The width field is a non-negative number that specifies the

minimum number of characters printed. If the number of
characters in the output value is less than width, blanks are
added on the left or right (when the - flag is specified) to pad
to the minimum width. If width is prefixed with a ‘0’,
zeros are padded instead of blanks. The width field never
truncates a field. If the length of the output value exceeds
the specified width, all characters are output.

The width field may be an asterisk (‘*’), in which case an

int argument from the argument list provides the width
value. Specifying a ‘b’ in front of the asterisk specifies that
the argument used is an unsigned char.

8
 288 Chapter 8. Library Reference

The precision field is a non-negative number that specifies

the number of characters to print, the number of significant
digits, or the number of decimal places. The precision
field can cause truncation or rounding of the output value in
the case of a floating-point number as specified in the
following table.

Type Meaning of Precision Field

d, u, o, x, X The precision field is where you specify the minimum
number of digits that are included in the output value.
Digits are not truncated if the number of digits in the
argument exceeds that defined in the precision field. If
the number of digits in the argument is less than the
precision field, the output value is padded on the left with
zeros.
f The precision field is where you specify the number of
digits to the right of the decimal point. The last digit is
rounded.
e, E The precision field is where you specify the number of
digits to the right of the decimal point. The last digit is
rounded.
g, G The precision field is where you specify the maximum
number of significant digits in the output value.
c, p The precision field has no effect on these field types.
s The precision field is where you specify the maximum
number of characters in the output value. Excess
characters are not output.

The precision field may be an asterisk (‘*’), in which case

an int argument from the argument list provides the value
for the precision. Specifying a ‘b’ in front of the asterisk
specifies that the argument used is an unsigned char.

The printf517 function is identical to printf, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. When using this function, include the
header file 80C517.H. Do not use this routine with a CPU
that does not support this feature.

8
Keil Software — Cx51 Compiler User’s Guide 289

NOTE
This function is implementation-specific and is based on the
operation of the putchar function. This function, as
provided in the standard library, writes characters using the
serial port of the 8051. Custom functions may use other I/O
devices.

You must ensure that the argument type matches that of the
format specification. You can use type casts to ensure that
the proper type is passed to printf.

The total number of bytes that may be passed to printf is

limited due to the memory restrictions imposed by the 8051.
A maximum of 15 bytes may be passed in small model or
compact model. A maximum of 40 bytes may be passed in
large model.

Return Value: The printf function returns the number of characters

actually written to the output stream.

See Also: gets, puts, scanf, sprintf, sscanf, vprintf, vsprintf

8
 290 Chapter 8. Library Reference

Example: #include <stdio.h>

void tst_printf (void) {

char a;
int b;
long c;
unsigned char x;
unsigned int y;
unsigned long z;
float f,g;
char buf [] = "Test String";
char *p = buf;

a = 1;
b = 12365;
c = 0x7FFFFFFF;
x = 'A';
y = 54321;
z = 0x4A6F6E00;
f = 10.0;
g = 22.95;

printf ("char %bd int %d long %ld\n",a,b,c);

printf ("Uchar %bu Uint %u Ulong %lu\n",x,y,z);
printf ("xchar %bx xint %x xlong %lx\n",x,y,z);
printf ("String %s is at address %p\n",buf,p);
printf ("%f != %g\n", f, g);
printf ("%*f != %*g\n", 8, f, 8, g);
}

8
Keil Software — Cx51 Compiler User’s Guide 291

putchar
Summary: #include <stdio.h>
char putchar (
char c); /* character to output */

Description: The putchar function transmits the character c using the

8051 serial port.

NOTE
This routine is implementation-specific and its function may
deviate from that described above. Source is included for
the _getkey and putchar functions which may be modified to
provide character level I/O for any hardware device. Refer
to “Customization Files” on page 143 for more information.

Return Value: The putchar routine returns the character output, c.

See Also: getchar, _getkey, ungetchar

Example: #include <stdio.h>

void tst_putchar (void) {

unsigned char i;

for (i = 0x20; i < 0x7F; i++)

putchar (i);
}

8
 292 Chapter 8. Library Reference

puts
Summary: #include <stdio.h>
int puts (
const char *string); /* string to output */

Description: The puts function writes string followed by a newline

character (‘\n’) to the output stream using the putchar
function.

NOTE
This function is implementation-specific and is based on the
operation of the putchar function. This function, as
provided in the standard library, writes characters using the
serial port of the 8051. Custom functions may use other I/O
devices.

Return Value: The puts function returns EOF if an error occurred and a
value of 0 if no errors were encountered.

See Also: gets, printf, scanf

Example: #include <stdio.h>

void tst_puts (void) {

puts ("Line #1");

puts ("Line #2");
puts ("Line #3");

8
Keil Software — Cx51 Compiler User’s Guide 293

rand
Summary: #include <stdlib.h>
int rand (void);

Description: The rand function generates a pseudo-random number

between 0 and 32767.

Return Value: The rand function returns a pseudo-random number.

See Also: srand

Example: #include <stdlib.h>

#include <stdio.h> /* for printf */

void tst_rand (void) {

int i;
int r;

for (i = 0; i < 10; i++) {

printf ("I = %d, RAND = %d\n", i, rand ());
}

8
 294 Chapter 8. Library Reference

realloc
Summary: #include <stdlib.h>
void *realloc (
void xdata *p, /* previously allocated block */
unsigned int size); /* new size for block */

Description: The realloc function changes the size of a previously

allocated memory block. The p argument points to the
allocated block and the size argument specifies the new
size for the block. The contents of the existing block are
copied to the new block. Any additional area in the new
block, due to a larger block size, is not initialized.

NOTE
Source code is provided for this routine in the folder
\KEIL\C51\LIB. You can modify the source to customize this
function for your hardware environment. Refer to
“Chapter 6. Advanced Programming Techniques” on page
143 for more information.

Return Value: The realloc function returns a pointer to the new block. If
there is not enough memory in the memory pool to satisfy
the memory request, a null pointer is returned and the
original memory block is not affected.

See Also: calloc, free, init_mempool, malloc

Example: #include <stdlib.h>

#include <stdio.h> /* for printf */

void tst_realloc (void) {

void xdata *p;
void xdata *new_p;

p = malloc (100);
if (p != NULL) {
new_p = realloc (p, 200);

if (new_p != NULL) p = new_p;

8 }
}
else printf ("Reallocation failed\n");
Keil Software — Cx51 Compiler User’s Guide 295

scanf
Summary: #include <stdio.h>
int scanf (
const char *fmtstr /* format string */

, argument…); /* additional arguments */

Description: The scanf function reads data using the getchar routine.
Data input are stored in the locations specified by argument
according to the format string fmtstr. Each argument must
be a pointer to a variable that corresponds to the type
defined in fmtstr which controls the interpretation of the
input data. The fmtstr argument is composed of one or
more whitespace characters, non-whitespace characters, and
format specifications as defined below.

The scanf517 function is identical to scanf, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. When using this function include the
header file 80C517.H. Do not use this routine with a CPU
that does not support this feature.

Whitespace characters, blank (‘ ’), tab (‘\t’), or newline

(‘\n’), causes scanf to skip whitespace characters in the
input stream. A single whitespace character in the format
string matches 0 or more whitespace characters in the
input stream.

Non-whitespace characters, with the exception of the
percent sign (‘%’), cause scanf to read but not store a
matching character from the input stream. The scanf
function terminates if the next character in the input
stream does not match the specified non-whitespace
character.

Format specifications begin with a percent sign (‘%’)
and cause scanf to read and convert characters from the
input stream to the specified type values. The converted
value is stored to an argument in the parameter list.
Characters following a percent sign that are not
8
recognized as a format specification are treated as an
ordinary character. For example, %% matches a single
percent sign in the input stream.
 296 Chapter 8. Library Reference

The format string is read from left to right. Characters that

are not part of the format specifications must match
characters in the input stream. These characters are read
from the input stream but are discarded and not stored. If a
character in the input stream conflicts with the format string,
scanf terminates. Any conflicting characters remain in the
input stream.

The first format specification encountered in the format

string references the first argument after fmtstr and
converts input characters and stores the value using the
format specification. The second format specification
accesses the second argument after fmtstr, and so on. If
there are more arguments than format specifications, the
extra arguments are ignored. Results are unpredictable if
there are not enough arguments for the format specifications.

Values in the input stream are called input fields and are
delimited by whitespace characters. When converting input
fields, scanf ends a conversion for an argument when a
whitespace character is encountered. Additionally, any
unrecognized character for the current format specification
ends a field conversion.

Format specifications have the following format:

%
*
width
{b | h | l} type

Each field in the format specification can be a single

character or a number which specifies a particular format
option.

The type field is where a single character specifies whether

input characters are interpreted as a character, string, or
number. This field can be any one of the characters in the
following table.

8 Character
d
Argument Type
int *
Input Format
Signed decimal number
i int * Signed decimal, hexadecimal, or
octal integer
u unsigned int * Unsigned decimal number
Keil Software — Cx51 Compiler User’s Guide 297

Character Argument Type Input Format

o unsigned int * Unsigned octal number
x unsigned int * Unsigned hex number
e float * Floating-point number
f float * Floating-point number
g float * Floating-point number
c char * A single character
s char * A string of characters terminated by
whitespace

An asterisk (*) as the first character of a format specification

causes the input field to be scanned but not stored. The
asterisk suppresses assignment of the format specification.

The width field is a non-negative number that specifies the

maximum number of characters read from the input stream.
No more than width characters are read from the input
stream and converted for the corresponding argument.
However, fewer than width characters may be read if a
whitespace character or an unrecognized character is
encountered first.

The optional characters b, h, and l may immediately precede

the type character to respectively specify char, short, or
long versions of the integer types d, i, u, o, and x.

NOTE
This function is implementation-specific and is based on the
operation of the _getkey and/or putchar functions. These
functions, as provided in the standard library, read and
write characters using the serial port of the 8051. Custom
functions may use other I/O devices.

The total number of bytes that may be passed to scanf is

limited due to the memory restrictions imposed by the 8051.
A maximum of 15 bytes may be passed in small model or
compact model. A maximum of 40 bytes may be passed in
large model.
8
 298 Chapter 8. Library Reference

Return Value: The scanf function returns the number of input fields that
were successfully converted. An EOF is returned if an error
is encountered.

See Also: gets, printf, puts, sprintf, sscanf, vprintf, vsprintf

Example: #include <stdio.h>

void tst_scanf (void) {

char a;
int b;
long c;

unsigned char x;
unsigned int y;
unsigned long z;

float f,g;

char d, buf [10];

int argsread;

printf ("Enter a signed byte, int, and long\n");

argsread = scanf ("%bd %d %ld", &a, &b, &c);
printf ("%d arguments read\n", argsread);

printf ("Enter an unsigned byte, int, and long\n");

argsread = scanf ("%bu %u %lu", &x, &y, &z);
printf ("%d arguments read\n", argsread);

printf ("Enter a character and a string\n");

argsread = scanf ("%c %9s", &d, buf);
printf ("%d arguments read\n", argsread);

printf ("Enter two floating-point numbers\n");

argsread = scanf ("%f %f", &f, &g);
printf ("%d arguments read\n", argsread);

8
Keil Software — Cx51 Compiler User’s Guide 299

setjmp
Summary: #include <setjmp.h>
int setjmp (
jmp_buf env); /* current environment */

Description: The setjmp function saves the current state of the CPU in
env. The state can be restored by a subsequent call to the
longjmp function. When used together, the setjmp and
longjmp functions provide you with a way to execute a
non-local goto.

A call to the setjmp function saves the current instruction

address as well as other CPU registers. A subsequent call to
the longjmp function restores the instruction pointer and
registers, and execution resumes at the point just after the
setjmp call.

Local variables and function arguments are restored only if

declared with the volatile attribute.

Return Value: The setjmp function returns a value of 0 when the current
state of the CPU has been copied to env. A non-zero value
indicates that the longjmp function was executed to return
to the setjmp function call. In such a case, the return value
is the value passed to the longjmp function.

See Also: longjmp

Example: See longjmp

8
 300 Chapter 8. Library Reference

sin / sin517
Summary: #include <math.h>
float sin (
float x); /* value to calculate
sine for */

Description: The sin function calculates the sine of the floating-point

value x. The value of x must be in the -65535 to +65535
range or an NaN error value is generated.

The sin517 function is identical to sin, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. When using this function include the
header file 80C517.H. Do not use this routine with a CPU
that does not support this feature.

Return Value: The sin function returns the sine of x.

See Also: cos, tan

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_sin (void) {

float x;
float y;

for (x = 0; x < (2 * 3.1415); x += 0.1) {

y = sin (x);

printf ("SIN(%f) = %f\n", x, y);

}

8
Keil Software — Cx51 Compiler User’s Guide 301

sinh
Summary: #include <math.h>
float sinh (
float val); /* value to calc hyperbolic sine for */

Description: The sinh function calculates the hyperbolic sine of the

floating-point value x. The value of x must be in the
-65535 to +65535 range or an NaN error value is generated.

Return Value: The sinh function returns the hyperbolic sine of x.

See Also: cosh, tanh

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_sinh (void) {

float x;
float y;

for (x = 0; x < (2 * 3.1415); x += 0.1) {

y = sinh (x);
printf ("SINH(%f) = %f\n", x, y);
}

8
 302 Chapter 8. Library Reference

sprintf / sprintf517
Summary: #include <stdio.h>
int sprintf (
char *buffer, /* storage buffer */
const char *fmtstr /* format string */

, argument…); /* additional arguments */

Description: The sprintf function formats a series of strings and numeric

values and stores the resulting string in buffer. The fmtstr
argument is a format string and has the same requirements as
specified for the printf function. Refer to “printf /
printf517” on page 285 for a description of the format string
and additional arguments.

The sprintf517 function is identical to sprintf, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. When using this function include the
header file 80C517.H. Do not use this routine with a CPU
that does not support this feature.

NOTE
The total number of bytes that may be passed to sprintf is
limited due to the memory restrictions imposed by the 8051.
A maximum of 15 bytes may be passed in small model or
compact model. A maximum of 40 bytes may be passed in
large model.

Return Value: The sprintf function returns the number of characters

actually written to buffer.

See Also: gets, printf, puts, scanf, sscanf, vprintf, vsprintf

8
Keil Software — Cx51 Compiler User’s Guide 303

Example: #include <stdio.h>

void tst_sprintf (void) {

char buf [100];
int n;

int a,b;
float pi;

a = 123;
b = 456;
pi = 3.14159;

n = sprintf (buf, "%f\n", 1.1);

n += sprintf (buf+n, "%d\n", a);
n += sprintf (buf+n, "%d %s %g", b, "---", pi);
printf (buf);
}

8
 304 Chapter 8. Library Reference

sqrt / sqrt517
Summary: #include <math.h>
float sqrt (
float x); /* value to calculate square root
of */

Description: The sqrt function calculates the square root of x.

The sqrt517 function is identical to sqrt, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. When using this function include the
header file 80C517.H. Do not use this routine with a CPU
that does not support this feature.

Return Value: The sqrt function returns the positive square root of x.

See Also: exp, log, pow

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_sqrt (void) {

float x;
float y;

x = 25.0;

y = sqrt (x); /* y = 5 */

printf ("SQRT(%f) = %f\n", x, y);

8
Keil Software — Cx51 Compiler User’s Guide 305

srand
Summary: #include <stdlib.h>
void srand (
int seed); /* random number generator seed */

Description: The srand function sets the starting value seed used by the
pseudo-random number generator in the rand function. The
random number generator produces the same sequence of
pseudo-random numbers for any given value of seed.

Return Value: None.

See Also: rand

Example: #include <stdlib.h>

#include <stdio.h> /* for printf */

void tst_srand (void) {

int i;
int r;

srand (56);

for (i = 0; i < 10; i++) {

printf ("I = %d, RAND = %d\n", i, rand ());
}

8
 306 Chapter 8. Library Reference

sscanf / sscanf517
Summary: #include <stdio.h>
int sscanf (
char *buffer, /* scanf input buffer */
const char *fmtstr /* format string */

, argument…); /* additional arguments */

Description: The sscanf function reads data from the string buffer. Data
input are stored in the locations specified by argument
according to the format string fmtstr. Each argument must
be a pointer to a variable that corresponds to the type
defined in fmtstr which controls the interpretation of the
input data. The fmtstr argument is composed of one or
more whitespace characters, non-whitespace characters, and
format specifications, as defined in the scanf function
description. Refer to “scanf” on page 295 for a complete
description of the formation string and additional arguments.

The sscanf517 function is identical to sscanf, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. When using this function include the
header file 80C517.H. Do not use this routine with a CPU
that does not support this feature.

NOTE
The total number of bytes that may be passed to sscanf is
limited due to the memory restrictions imposed by the 8051.
A maximum of 15 bytes may be passed in small model or
compact model. A maximum of 40 bytes may be passed in
large model.

Return Value: The sscanf function returns the number of input fields that
were successfully converted. An EOF is returned if an error
is encountered.

8 See Also: gets, printf, puts, scanf, sprintf, vprintf, vsprintf

Keil Software — Cx51 Compiler User’s Guide 307

Example: #include <stdio.h>

void tst_sscanf (void) {

char a;
int b;
long c;

unsigned char x;
unsigned int y;
unsigned long z;

float f,g;

char d, buf [10];

int argsread;

printf ("Reading a signed byte, int,and long\n");

argsread = sscanf ("1 -234 567890",
"%bd %d %ld", &a, &b, &c);
printf ("%d arguments read\n", argsread);

printf ("Reading an unsigned byte, int, and long\n");

argsread = sscanf ("2 44 98765432",
"%bu %u %lu", &x, &y, &z);
printf ("%d arguments read\n", argsread);

printf ("Reading a character and a string\n");

argsread = sscanf ("a abcdefg", "%c %9s", &d, buf);
printf ("%d arguments read\n", argsread);

printf ("Reading two floating-point numbers\n");

argsread = sscanf ("12.5 25.0", "%f %f", &f, &g);
printf ("%d arguments read\n", argsread);

8
 308 Chapter 8. Library Reference

strcat
Summary: #include <string.h>
char *strcat (
char *dest, /* destination string */
char *src); /* source string */

Description: The strcat function concatenates or appends src to dest

and terminates dest with a null character.

Return Value: The strcat function returns dest.

See Also: strcpy, strlen, strncat, strncpy

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strcat (void) {

char buf [21];
char s [] = "Test String";

strcpy (buf, s);

strcat (buf, " #2");

printf ("new string is %s\n", buf);

8
Keil Software — Cx51 Compiler User’s Guide 309

strchr
Summary: #include <string.h>
char *strchr (
const char *string, /* string to search */
char c); /* character to find */

Description: The strchr function searches string for the first occurrence
of c. The null character terminating string is included in
the search.

Return Value: The strchr function returns a pointer to the character c

found in string or a null pointer if no matching character
was found.

See Also: strcspn, strpbrk, strpos, strrchr, strrpbrk, strrpos,

strspn

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strchr (void) {

char *s;
char buf [] = "This is a test";

s = strchr (buf, 't');

if (s != NULL)
printf ("found a 't' at %s\n", s);
}

8
 310 Chapter 8. Library Reference

strcmp
Summary: #include <string.h>
char strcmp (
char *string1, /* first string */
char *string2); /* second string */

Description: The strcmp function compares the contents of string1 and

string2 and returns a value indicating their relationship.

Return Value: The strcmp function returns the following values to indicate
the relationship of string1 to string2:

Value Meaning
<0 string1 less than string2
=0 string1 equal to string2
>0 string1 greater than string2

See Also: memcmp, strncmp

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strcmp (void) {

char buf1 [] = "Bill Smith";
char buf2 [] = "Bill Smithy";
char i;

i = strcmp (buf1, buf2);

if (i < 0)
printf ("buf1 < buf2\n");

else if (i > 0)
printf ("buf1 > buf2\n");

else
printf ("buf1 == buf2\n");
}

8
Keil Software — Cx51 Compiler User’s Guide 311

strcpy
Summary: #include <string.h>
char *strcpy (
char *dest, /* destination string */
char *src); /* source string */

Description: The strcpy function copies src to dest and appends a null
character to the end of dest.

Return Value: The strcpy function returns dest.

See Also: strcat, strlen, strncat, strncpy

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strcpy (void) {

char buf [21];
char s [] = "Test String";

strcpy (buf, s);

strcat (buf, " #2");

printf ("new string is %s\n", buf);

}

8
 312 Chapter 8. Library Reference

strcspn
Summary: #include <string.h>
int strcspn (
char *src, /* source string */
char *set); /* characters to find */

Description: The strcspn function searches the src string for any of the
characters in the set string.

Return Value: The strcspn function returns the index of the first character
located in src that matches a character in set. If the first
character in src matches a character in set, a value of 0 is
returned. If there are no matching characters in src, the
length of the string is returned.

See Also: strchr, strpbrk, strpos, strrchr, strrpbrk, strrpos, strspn

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strcspn (void) {

char buf [] = "13254.7980";
int i;

i = strcspn (buf, ".,");

if (buf [i] != '\0')

printf ("%c was found in %s\n", (char)
buf [i], buf);
}

8
Keil Software — Cx51 Compiler User’s Guide 313

strlen
Summary: #include <string.h>
int strlen (
char *src); /* source string */

Description: The strlen function calculates the length, in bytes, of src.

This calculation does not include the null terminating
character.

Return Value: The strlen function returns the length of src.

See Also: strcat, strcpy, strncat, strncpy

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strlen (void) {

char buf [] = "Find the length of this string";
int len;

len = strlen (buf);

printf ("string length is %d\n", len);

8
 314 Chapter 8. Library Reference

strncat
Summary: #include <string.h>
char *strncat (
char *dest, /* destination string */
char *src, /* source string */
int len); /* max. chars to concatenate */

Description: The strncat function appends at most len characters from

src to dest and terminates dest with a null character. If
src is shorter than len characters, src is copied up to and
including the null terminating character.

Return Value: The strncat function returns dest.

See Also: strcat, strcpy, strlen, strncpy

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strncat (void) {

char buf [21];

strcpy (buf, "test #");

strncat (buf, "three", sizeof (buf) - strlen
(buf));
}

8
Keil Software — Cx51 Compiler User’s Guide 315

strncmp
Summary: #include <string.h>
char strncmp (
char *string1, /* first string */
char *string2, /* second string */
int len); /* max characters to
compare */

Description: The strncmp function compares the first len bytes of

string1 and string2 and returns a value indicating their
relationship.

Return Value: The strncmp function returns the following values to

indicate the relationship of the first len bytes of string1 to
string2:

Value Meaning
<0 string1 less than string2
=0 string1 equal to string2
>0 string1 greater than string2

See Also: memcmp, strcmp

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strncmp (void) {

char str1 [] = "Wrodanahan T.J.";
char str2 [] = "Wrodanaugh J.W.";

char i;

i = strncmp (str1, str2, 15);

if (i < 0) printf ("str1 < str2\n");

else if (i > 0) printf ("str1 > str2\n");
else printf ("str1 == str2\n");
}

8
 316 Chapter 8. Library Reference

strncpy
Summary: #include <string.h>
char *strncpy (
char *dest, /* destination string */
char *src, /* source string */
int len); /* max characters to
copy */

Description: The strncpy function copies at most len characters from

src to dest. If src contains fewer characters than len,
dest is padded out with null characters to len characters.

Return Value: The strncpy function returns dest.

See Also: strcat, strcpy, strlen, strncat

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strncpy ( char *s) {

char buf [21];

strncpy (buf, s, sizeof (buf));

buf [sizeof (buf)] = '\0';
}

8
Keil Software — Cx51 Compiler User’s Guide 317

strpbrk
Summary: #include <string.h>
char *strpbrk (
char *string, /* string to search */
char *set); /* characters to find */

Description: The strpbrk function searches string for the first

occurrence of any character from set. The null terminator is
not included in the search.

Return Value: The strpbrk function returns a pointer to the matching

character in string. If string contains no characters from
set, a null pointer is returned.

See Also: strchr, strcspn, strpos, strrchr, strrpbrk, strrpos, strspn

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strpbrk (void) {

char vowels [] ="AEIOUaeiou";
char text [] = "Seven years ago...";

char *p;

p = strpbrk (text, vowels);

if (p == NULL)
printf ("No vowels found in %s\n", text);

else
printf ("Found a vowel at %s\n", p);

8
 318 Chapter 8. Library Reference

strpos
Summary: #include <string.h>
int strpos (
const char *string, /* string to search */
char c); /* character to find */

Description: The strpos function searches string for the first occurrence
of c. The null character terminating string is included in
the search.

Return Value: The strpos function returns the index of the character
matching c in string or a value of -1 if no matching
character was found. The index of the first character in
string is 0.

See Also: strchr, strcspn, strpbrk, strrchr, strrpbrk, strrpos,

strspn

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strpos (void) {

char text [] = "Search this string for
blanks";

int i;

i = strpos (text, ' ');

if (i == -1)
printf ("No spaces found in %s\n", text);

else
printf ("Found a space at offset %d\n", i);

8
Keil Software — Cx51 Compiler User’s Guide 319

strrchr
Summary: #include <string.h>
char *strrchr (
const char *string, /* string to search */
char c); /* character to find */

Description: The strrchr function searches string for the last occurrence
of c. The null character terminating string is included in
the search.

Return Value: The strrchr function returns a pointer to the last character c
found in string or a null pointer if no matching character
was found.

See Also: strchr, strcspn, strpbrk, strpos, strrpbrk, strrpos, strspn

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strrchr (void) {

char *s;
char buf [] = "This is a test";

s = strrchr (buf, 't');

if (s != NULL)
printf ("found the last 't' at %s\n", s);

8
 320 Chapter 8. Library Reference

strrpbrk
Summary: #include <string.h>
char *strrpbrk (
char *string, /* string to search */
char *set); /* characters to find */

Description: The strrpbrk function searches string for the last

occurrence of any character from set. The null terminator is
not included in the search.

Return Value: The strrpbrk function returns a pointer to the last matching
character in string. If string contains no characters from
set, a null pointer is returned.

See Also: strchr, strcspn, strpbrk, strpos, strrchr, strrpos, strspn

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strrpbrk (void) {

char vowels [] ="AEIOUaeiou";
char text [] = "American National Standards
Institute";

char *p;

p = strpbrk (text, vowels);

if (p == NULL)
printf ("No vowels found in %s\n", text);

else
printf ("Last vowel is at %s\n", p);

8
Keil Software — Cx51 Compiler User’s Guide 321

strrpos
Summary: #include <string.h>
int strrpos (
const char *string, /* string to search */
char c); /* character to find */

Description: The strrpos function searches string for the last occurrence
of c. The null character terminating string is included in
the search.

Return Value: The strrpos function returns the index of the last character
matching c in string or a value of -1 if no matching
character was found. The index of the first character in
string is 0.

See Also: strchr, strcspn, strpbrk, strpos, strrchr, strrpbrk, strspn

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strrpos ( char *s) {

int i;

i = strpos (s, ' ');

if (i == -1)
printf ("No spaces found in %s\n", s);

else
printf ("Last space in %s is at offset %d\n",
s, i);
}

8
 322 Chapter 8. Library Reference

strspn
Summary: #include <string.h>
int strspn (
char *string, /* string to search */
char *set); /* characters to allow */

Description: The strspn function searches the src string for characters
not found in the set string.

Return Value: The strspn function returns the index of the first character
located in src that does not match a character in set. If the
first character in src does not match a character in set, a
value of 0 is returned. If all characters in src are found in
set, the length of src is returned.

See Also: strchr, strcspn, strpbrk, strpos, strrchr, strrpbrk,

strrpos

Example: #include <string.h>

#include <stdio.h> /* for printf */

void tst_strspn ( char *digit_str) {

char octd [] = "01234567";
int i;

i = strspn (digit_str, octd);

if (digit_str [i] != '\0')

printf ("%c is not an octal digit\n",
digit_str [i]);

8
Keil Software — Cx51 Compiler User’s Guide 323

strtod / strtod517
Summary: #include <stdlib.h>
unsigned long strtod (
const char *string, /* string to convert */
char **ptr); /* ptr to subsequent characters */

Description: The strtod function converts string into a floating-point

value. The input string is a sequence of characters that can
be interpreted as a floating-point number. Whitespace
characters at the beginning of string are skipped.

The strtod517 function is identical to atof, but uses the

arithmetic unit of the Infineon 80C517 to provide faster
execution. For using this function include the header file
80C517.H. Do not use this routine with a CPU that does not
support this feature.

The strtod function requires string to have the following

format:

{+ | -} digits
. digits
{e | E}
{+ | -} digits
where:

digits may be one or more decimal digits.

The value of ptr is set to point to the first character in

string immediately following the converted part of the
string. If ptr is NULL no value is assigned to ptr. If no
conversion is possible, ptr is set to the value of string and
the value 0 is returned by strtoul.

Return Value: The strtod function returns the floating-point value that is
produced by interpreting the characters in string as a
number.

See Also: atof, atoi, atol, strtol, strtoul

8
 324 Chapter 8. Library Reference

Example: #include <stdlib.h>

#include <stdio.h> /* for printf */

void tst_strtod (void) {

float f;
char s [] = "1.23";

f = strtod (s, NULL);

printf ("strtod(%s) = %f\n", s, f);
}

8
Keil Software — Cx51 Compiler User’s Guide 325

strtol
Summary: #include <stdlib.h>
long strtol (
const char *string, /* string to convert */
char **ptr, /* ptr to subsequent characters */
unsigned char base);/* number base for conversion */

Description: The strtol function converts string into a long value. The
input string is a sequence of characters that can be
interpreted as an integer number. Whitespace characters at
the beginning of string are skipped. An optional sign may
precede the number.

The strtol function requires string to have the following

format:

whitespace
{+ | -} digits

where:

digits may be one or more decimal digits.

If base is zero, the number should have the format of a

decimal-constant, octal-constant or hexadecimal-constant.
The radix of the number is deduced from its format. If the
value of base is between 2 and 36 the number must consist
of nonzero sequence of letters and digits representing an
integer in the specified base. The letters a through z (or A
through Z) represent the values 10 through 36, respectively.
Only those letters representing values less than base are
permitted. If base is 16 the number may begin with 0x or
0X, which is ignored.

The value of ptr is set to point to the first character in

string immediately following the converted part of the
string. If ptr is NULL no value is assigned to ptr. If no
conversion is possible, ptr is set to the value of string and
the value 0 is returned by strtol. 8
Return Value: The strtol function returns the integer value that is produced
by interpreting the characters in string as number. The
 326 Chapter 8. Library Reference

value LONG_MIN or LONG_MAX is returned in case of

overflow.

See Also: atof, atoi, atol, strtod, strtoul

Example: #include <stdlib.h>

#include <stdio.h> /* for printf */

char s [] = "-123456789";

void tst_strtol (void) {

long l;

l = strtol (s, NULL, 10);

printf ("strtol(%s) = %ld\n", s, l);
}

8
Keil Software — Cx51 Compiler User’s Guide 327

strtoul
Summary: #include <stdlib.h>
unsigned long strtoul (
const char *string, /* string to convert */
char **ptr, /* ptr to subsequent characters */
unsigned char base);/* number base for conversion */

Description: The strtoul function converts string into an unsigned long

value. The input string is a sequence of characters that can
be interpreted as an integer number. Whitespace characters
at the beginning of string are skipped. An optional sign may
precede the number.

The strtoul function requires string to have the following

format:

whitespace
{+ | -} digits

where:

digits may be one or more decimal digits.

If base is zero, the number should have the format of a

decimal-constant, octal-constant or hexadecimal-constant.
The radix of the number is deduced from its format. If the
value of base is between 2 and 36 the number must consist
of nonzero sequence of letters and digits representing an
integer in the specified base. The letters a through z (or A
through Z) represent the values 10 through 36, respectively.
Only those letters representing values less than base are
permitted. If base is 16 the number may begin with 0x or
0X, which is ignored.

The value of ptr is set to point to the first character in

string immediately following the converted part of the
string. If ptr is NULL no value is assigned to ptr. If no
conversion is possible, ptr is set to the value of string and
the value 0 is returned by strtoul. 8
Return Value: The strtoul function returns the integer value that is
produced by interpreting the characters in string as
 328 Chapter 8. Library Reference

number. The value ULONG_MAX is returned in case of

overflow.

See Also: atof, atoi, atol, strtod, strtol

Example: #include <stdlib.h>

#include <stdio.h> /* for printf */

char s [] = "12345AB";

void tst_strtoul (void) {

unsigned long ul;

ul = strtoul (s, NULL, 16);

printf ("strtoul(%s) = %lx\n", s, ul);
}

8
Keil Software — Cx51 Compiler User’s Guide 329

tan / tan517
Summary: #include <math.h>
float tan (
float x); /* value to calculate tangent of
*/

Description: The tan function calculates the tangent of the floating-point

value x. The value of x must be in the -65535 to +65535
range or an NaN error value is generated.

The tan517 function is identical to tan, but uses the

arithmetic unit of the Infineon C517x, C509 to provide
faster execution. When using this function, include the
header file 80C517.H. Do not use this routine with a CPU
that does not support this feature.

Return Value: The tan function returns the tangent of x.

See Also: cos, sin

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_tan (void) {

float x, y, pi;

pi = 3.14159;

for (x = -(pi/4); x < (pi/4); x += 0.1) {

y = tan (x);
printf ("TAN(%f) = %f\n", x, y);
}
}

8
 330 Chapter 8. Library Reference

tanh
Summary: #include <math.h>
float tanh (
float x); /* value to calc hyperbolic
tangent for */

Description: The tanh function calculates the hyperbolic tangent for the
floating-point value x.

Return Value: The tanh function returns the hyperbolic tangent of x.

See Also: cosh, sinh

Example: #include <math.h>

#include <stdio.h> /* for printf */

void tst_tanh (void) {

float x;
float y;
float pi;

pi = 3.14159;

for (x = -(pi/4); x < (pi/4); x += 0.1) {

y = tanh (x);
printf ("TANH(%f) = %f\n", x, y);
}

8
Keil Software — Cx51 Compiler User’s Guide 331

_testbit_
Summary: #include <intrins.h>
bit _testbit_ (
bit b); * bit to test and clear */

Description: The _testbit_ routine produces a JBC instruction in the

generated program code to simultaneously test the bit b and
clear it to 0. This routine can be used only on directly
addressable bit variables and is invalid on any type of
expression. This routine is implemented as an intrinsic
function. The code required is included in-line rather than
being called.

Return Value: The _testbit_ routine returns the value of b.

Example: #include <intrins.h>

#include <stdio.h> /* for printf */

void tst_testbit (void){

bit test_flag;

if (_testbit_ (test_flag))
printf ("Bit was set\n");

else
printf ("Bit was clear\n");
}

8
 332 Chapter 8. Library Reference

toascii
Summary: #include <ctype.h>
char toascii (
char c); /* character to convert */

Description: The toascii macro converts c to a 7-bit ASCII character.

This macro clears all but the lower 7 bits of c.

Return Value: The toascii macro returns the 7-bit ASCII character for c.

See Also: toint

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_toascii ( char c) {

char k;

k = toascii (c);

printf ("%c is an ASCII character\n", k);

}

8
Keil Software — Cx51 Compiler User’s Guide 333

toint
Summary: #include <ctype.h>
char toint (
char c); /* digit to convert */

Description: The toint function interprets c as a hexadecimal value.

ASCII characters ‘0’ through ‘9’ generate values of 0 to 9.
ASCII characters ‘A’ through ‘F’ and ‘a’ through ‘f’
generate values of 10 to 15. If the value of c is not a
hexadecimal digit, the function returns –1.

Return Value: The toint function returns the value of the ASCII
hexadecimal character c.

See Also: toascii

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_toint (void) {

unsigned long l;
char k;

for (l = 0; isdigit (k = getchar ());

l *= 10) {

l += toint (k);
}
}

8
 334 Chapter 8. Library Reference

tolower
Summary: #include <ctype.h>
char tolower (
char c); /* character to convert */

Description: The tolower function converts c to a lowercase character.

If c is not an alphabetic letter, the tolower function has no
effect.

Return Value: The tolower function returns the lowercase equivalent of c.

See Also: _tolower, toupper, _toupper

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_tolower (void) {

unsigned char i;

for (i = 0x20; i < 0x7F; i++) {

printf ("tolower(%c) = %c\n", i, tolower(i));
}
}

8
Keil Software — Cx51 Compiler User’s Guide 335

_tolower
Summary: #include <ctype.h>
char _tolower (
char c); /* character to convert */

Description: The _tolower macro is a version of tolower that can be used

when c is known to be an uppercase character.

Return Value: The _tolower macro returns a lowercase character.

See Also: tolower, toupper, _toupper

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst__tolower ( char k) {

if (isupper (k)) k = _tolower (k);

}

8
 336 Chapter 8. Library Reference

toupper
Summary: #include <ctype.h>
char toupper (
char c); /* character to convert */

Description: The toupper function converts c to an uppercase character.

If c is not an alphabetic letter, the toupper function has no
effect.

Return Value: The toupper function returns the uppercase equivalent of c.

See Also: tolower, _tolower, _toupper

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst_toupper (void) {

unsigned char i;

for (i = 0x20; i < 0x7F; i++) {

printf ("toupper(%c) = %c\n", i, toupper(i));
}
}

8
Keil Software — Cx51 Compiler User’s Guide 337

_toupper
Summary: #include <ctype.h>
char _toupper (
char c); /* character to convert */

Description: The _toupper macro is a version of toupper that can be

used when c is known to be a lowercase character.

Return Value: The _toupper macro returns an uppercase character.

See Also: tolower, _tolower, toupper

Example: #include <ctype.h>

#include <stdio.h> /* for printf */

void tst__toupper ( char k) {

if (islower (k)) k = _toupper (k);
}

8
 338 Chapter 8. Library Reference

ungetchar
Summary: #include <stdio.h>
char ungetchar (
char c); /* character to unget */

Description: The ungetchar function stores the character c back into the
input stream. Subsequent calls to getchar and other stream
input functions return c. Only one character may be passed
to unget between calls to getchar.

Return Value: The ungetchar function returns the character c if

successful. If ungetchar is called more than once between
function calls that read from the input stream, EOF is
returned indicating an error condition.

See Also: _getkey, putchar, ungetchar

Example: #include <stdio.h>

void tst_ungetchar (void) {

char k;

while (isdigit (k = getchar ())) {

/* stay in the loop as long as k is a digit */
}
ungetchar (k);
}

8
Keil Software — Cx51 Compiler User’s Guide 339

va_arg
Summary: #include <stdarg.h>
type va_arg (
argptr, /* optional argument list */
type); /* type of next argument */

Description: The va_arg macro is used to extract subsequent arguments

from a variable-length argument list referenced by argptr.
The type argument specifies the data type of the argument
to extract. This macro may be called only once for each
argument and must be called in the order of the parameters
in the argument list.

The first call to va_arg returns the first argument after the
prevparm argument specified in the va_start macro.
Subsequent calls to va_arg return the remaining arguments
in succession.

Return Value: The va_arg macro returns the value for the specified
argument type.

See Also: va_end, va_start

8
 340 Chapter 8. Library Reference

Example: #include <stdarg.h>

#include <stdio.h> /* for printf */

int varfunc (char *buf, int id, ...) {

va_list tag;

va_start (tag, id);

if (id == 0) {
int arg1;
char *arg2;
long arg3;

arg1 = va_arg (tag, int);

arg2 = va_arg (tag, char *);
arg3 = va_arg (tag, long);
}
else {
char *arg1;
char *arg2;
long arg3;

arg1 = va_arg (tag, char *);

arg2 = va_arg (tag, char *);
arg3 = va_arg (tag, long);
}
}

void caller (void) {

char tmp_buffer [10];

varfunc (tmp_buffer, 0, 27, "Test Code", 100L);

varfunc (tmp_buffer, 1, "Test", "Code", 348L);
}

8
Keil Software — Cx51 Compiler User’s Guide 341

va_end
Summary: #include <stdarg.h>
void va_end (
argptr); /* optional argument list */

Description: The va_end macro is used to terminate use of the

variable-length argument list pointer argptr that was
initialized using the va_start macro.

Return Value: None.

See Also: va_arg, va_start

Example: See va_arg.

8
 342 Chapter 8. Library Reference

va_start
Summary: #include <stdarg.h>
void va_start (
argptr, /* optional argument list */
prevparm); /* arg preceding optional args */

Description: The va_start macro, when used in a function with a

variable-length argument list, initializes argptr for
subsequent use by the va_arg and va_end macros. The
prevparm argument must be the name of the function
argument immediately preceding the optional arguments
specified by an ellipses (…). This function must be called
to initialize a variable-length argument list pointer before
any access using the va_arg macro is made.

Return Value: None.

See Also: va_arg, va_end

Example: See va_arg.

8
Keil Software — Cx51 Compiler User’s Guide 343

vprintf
Summary: #include <stdio.h>
void vprintf (
const char * fmtstr, /* pointer to format string */
char * argptr); /* pointer to argument list */

Description: The vprintf function formats a series of strings and numeric

values and builds a string to write to the output stream using
the putchar function. The function is similar to the
counterpart printf, but it accepts a pointer to a list of
arguments instead of an argument list.

The fmtstr argument is a pointer to a format string and has

the same form and function as the fmtstr argument for the
printf function. Refer to “printf / printf517” on page 285
for a description of the format string. The argptr argument
points to a list of arguments that are converted and output
according to the corresponding format specifications in the
format.

NOTE
This function is implementation-specific and is based on the
operation of the putchar function. This function, as
provided in the standard library, writes characters using the
serial port of the 8051. Custom functions may use other I/O
devices.

Return Value: The vprintf function returns the number of characters

actually written to the output stream.

See Also: gets, puts, printf, scanf, sprintf, sscanf, vsprintf

8
 344 Chapter 8. Library Reference

Example: #include <stdio.h>

#include <stdarg.h>

void error (char *fmt, ...) {

va_list arg_ptr;

va_start (arg_ptr, fmt); /* format string */

vprintf (fmt, arg_ptr);
va_end (arg_ptr);
}

void tst_vprintf (void) {

int i;
i = 1000;
/* call error with one parameter */
error ("Error: '%d' number too large\n", i);
/* call error with just a format string */
error ("Syntax Error\n");
}

8
Keil Software — Cx51 Compiler User’s Guide 345

vsprintf
Summary: #include <stdio.h>
void vsprintf (
char *buffer, /* pointer to storage buffer */
const char * fmtstr, /* pointer to format string */
char * argptr); /* pointer to argument list */

Description: The vsprintf function formats a series of strings and

numeric values and stores the string in buffer. The function
is similar to the counterpart sprintf, but it accepts a pointer
to a list of arguments instead of an argument list.

The fmtstr argument is a pointer to a format string and has

the same form and function as the fmtstr argument for the
printf function. Refer to “printf / printf517” on page 285
for a description of the format string. The argptr argument
points to a list of arguments that are converted and output
according the corresponding format specifications in the
format.

Return Value: The vsprintf function returns the number of characters

actually written to the output stream.

See Also: gets, puts, printf, scanf, sprintf, sscanf, vprintf

8
 346 Chapter 8. Library Reference

Example: #include <stdio.h>

#include <stdarg.h>

xdata char etxt[30]; /* text buffer */

void error (char *fmt, ...) {

va_list arg_ptr;

va_start (arg_ptr, fmt); /* format string */

vsprintf (etxt, fmt, arg_ptr);
va_end (arg_ptr);
}

void tst_vprintf (void) {

int i;
i = 1000;

/* call error with one parameter */

error ("Error: '%d' number too large\n", i);

/* call error with just a format string */

error ("Syntax Error\n");
}

8
Keil Software — Cx51 Compiler User’s Guide 347

Appendix A. Differences from ANSI C

The Cx51 compiler differs in only a few aspects from the ANSI C Standard.
These differences can be grouped into compiler-related differences and A
library-related differences.

Compiler-related Differences

Wide Characters
Wide 16-bit characters are not supported by Cx51. ANSI provides wide
characters for future support of an international character set.

Recursive Function Calls
Recursive function calls are not supported by default. Functions that are
recursive must be declared using the reentrant function attribute. Reentrant
functions can be called recursively because the local data and parameters are
stored in a reentrant stack. In comparison, functions which are not declared
using the reentrant attribute use static memory segments for the local data of
the function. A recursive call to these functions overwrites the local data of
the prior function call instance.

Library-related Differences
The ANSI C Standard Library includes a vast number of routines, most of which
are included in Cx51. Many, however, are not applicable to an embedded
application and are excluded from the Cx51 library.

The following ANSI Standard library routines are included in the Cx51 library:

abs cosh isdigit

acos exp isgraph
asin fabs islower
atan floor isprint
atan2 fmod ispunct
atof free isspace
atoi getchar isupper
atol gets isxdigit
calloc isalnum labs
ceil isalpha log
cos iscntrl log10
 348 Appendix A. Differences from ANSI C

longjmp sin strrchr

malloc sinh strspn
memchr sprintf strtod
A memcmp
memcpy
sqrt
srand
strtol
strtoul
memmove sscanf tan
memset strcat tanh
modf strchr tolower
pow strcmp toupper
printf strcpy va_arg
putchar strcspn va_end
puts strlen va_start
rand strncat vprintf
realloc strncmp vsprintf
scanf strncpy
setjmp strpbrk

The following ANSI Standard library routines are not included in the Cx51
library:

abort freopen remove

asctime frexp rename
atexit fscanf rewind
bsearch fseek setbuf
clearerr fsetpos setlocale
clock ftell setvbuf
ctime fwrite signal
difftime getc strcoll
div getenv strerror
exit gmtime strftime
fclose ldexp strstr
feof ldiv strtok
ferror localeconv strxfrm
fflush localtime system
fgetc mblen time
fgetpos mbstowcs tmpfile
fgets mbtowc tmpnam
fopen mktime ungetc
fprintf perror vfprintf
fputc putc wcstombs
fputs qsort wctomb
fread raise
Keil Software — Cx51 Compiler User’s Guide 349

The following routines are not found in the ANSI Standard Library but are
included in the Cx51 library.

acos517
asin517
_irol_
_iror_
sscanf517
strpos
A
atan517 log10517 strrpbrk
atof517 log517 strrpos
strtod517 _lrol_ tan517
cabs _lror_ _testbit_
_chkfloat_ memccpy toascii
cos517 _nop_ toint
_crol_ printf517 _tolower
_cror_ scanf517 _toupper
exp517 sin517 ungetchar
_getkey sprintf517
init_mempool sqrt517
 350 Appendix A. Differences from ANSI C

A
Keil Software — Cx51 Compiler User’s Guide 351

Appendix B. Version Differences

This following appendix lists an overview of major product enhancements and
differences between Version 6.10 and previous versions. The current version of
the Cx51 compiler contains all enhancements listed below:

Version 6.0 Differences B

OMF251 directive
Selects a new OMF file format that provides detailed symbol type checking
across moduls and eliminates historic limitations of the Intel OMF51 file
format. Constants can be located also into an xdata ROM to free code space
for program code.

STRING directive
When you are using the OMF2 format, Cx51 allows to locate constant strings
into const xdata or const far space. This gives you more code space for
program code.

Support for Philips 80C51MX and Dallas Contigouos Mode
Cx51 provides support for the Philips 80C51MX architecture and the Dallas
Contigouos Mode that is available on Dallas 390 and variants.

VARBANKING directive
Cx51 now uses a char variable to represent an enum, if the enum range
allows that.

NOTE
Support for the OMF2 output file format, Philips 80C51MX, Dallas Contigouos
Mode, and VARBANKING requires the PK51 Professional Developers Kit.
These options cannot be used in the CA51 and DK51 packages.

Version 5 Differences

Optimize Level 7, 8, and 9
C51 offers three new optimizer levels. These new optimizations focus
primarily on code density. Refer to “Optimizer” on page 155 for more
information.
 352 Appendix B. Version Differences

Directives for the dual DPTR support

C51 provides dual DPTR support for Atmel, Atmel WM, and, Philips with
the directives MODA2 and MODP2.

data, pdata, xdata automatic variables overlayable in all memory models
C51 now overlays all data, pdata, and xdata automatic variables regardless of
the selected memory model. In previous C51 versions only automatic
variables of the default memory type are overlaid. For example, C51 Version

B 5 did not overlay pdata or xdata variables if a function where compiled in the
SMALL memory model.

The data type enum adjusts automatically 8 or 16 bits.
C51 now uses a char variable to represent an enum, if the enum range allows
that.

modf, strtod, strtol, strtoul Library Functions
C51 includes now the ANSI standard library functions modf, strtod, strtol,
strtou

Directives BROWSE, INCDIR, ONEREGBANK, RET_XSTK,
RET_PSTK
C51 supports new directives for generating Browse Information, specifing
include directives, optimizing interrupt code, and using the reentrant stack for
return addresses. Refer to “Chapter 2. Compiling with Cx51” on page 19 for
more information.

Version 4 Differences

Byte Order of Floating-point Numbers
Floating-point numbers are now stored in the big endian order. Previous
releases of the C51 compiler stored floating-point numbers in little endian
format. Refer to “Floating-point Numbers” on page 177 for more
information.

_chkfloat_ Library Function
The intrinsic function _chkfloat_ allows for fast testing of floating-point
numbers for error (NaN), ±INF, zero and normal numbers. Refer to
“_chkfloat_” on page 240 for more information.

FLOATFUZZY Directive
C51 now supports the FLOATFUZZY directive. This directive controls the
number of bits ignored during the execution of a floating-point compare.
Refer to “FLOATFUZZY” on page 40 for more information.

Floating-point Arithmetic is Fully Reentrant
Intrinsic floating-point arithmetic operations (add, subtract, multiply, divide,
Keil Software — Cx51 Compiler User’s Guide 353

and compare) are now fully reentrant. The C library routines fpsave and
fprestore are no longer needed. Several library routines are also reentrant.
Refer to “Routines by Category” on page 212 for more information.

Long and Floating-point Operations no Longer use an Arithmetic Stack
The long and floating-point arithmetic is more efficient; the code generated
is now totally register-based and does not use a simulated arithmetic stack.
This also reduces the memory needs of the generated code.

Memory Types
The memory types have been changed to achieve better performance in the
B
run-time library and to reflect the memory map of the MCS® 251 architecture.

Memory Type Bytes for Generic Pointers
The memory type bytes used in generic pointers have changed. The
following table contains the memory type byte values and their associated
memory type.
Memory Type idata data bdata xdata pdata code
C51 V5 Value 0x00 0x00 0x00 0x01 0xFE 0xFF
C51 V4 Value 0x01 0x04 0x04 0x02 0x03 0x05

WARNINGLEVEL Directive
C51 now supports the WARNINGLEVEL directive which lets you specify
the strength of the warning detection for the C51 compiler. The C51
compiler now also checks for unused local variables, labels, and expressions.
Refer to “VARBANKING
Abbreviation: VB

Arguments: None.

Default: The standard C51 library set is used.

µVision2 Control: Options – C51 – Misc controls: enter the directive.

Description: The VARBANKING directive selects a different set of

library functions that is required to support variable banking
in the xdata space.

NOTE
For extended 8051 devices and variable banking with
classic 8051 devices several Keil Application Notes will
become available on www.keil.com or the Keil development
 354 Appendix B. Version Differences

tools CD-ROM that explain the banking and memory

configuration for these devices.

Example: C51 SAMPLE.C VARBANKING

#pragma VARBANKING

B
WARNINGLEVEL” on page 81 for more information.

Keil Software — Cx51 Compiler User’s Guide 355

Version 3.4 Differences

_at_Keyword
C51 supports variable location using the _at_ keyword. This new keyword
allows you to specify the address of a variable in a declaration. Refer to “The
_at_ Keyword” on page 182 for more information.

NOAMAKE Directive
C51 now supports the NOAMAKE directive. This directive causes C51 to
generate object modules without project information and register
B
optimization records. This is necessary only if you want to use object files
with older versions of C51 tools.

OH51 Hex File Converter
The OHS51 Object-Hex-Symbol Converter provided with prior versions of
C51 has been replaced with OH51.

Optimizer Level 6
C51 now supports optimizer level 6 which provides loop rotation. The
resulting code is more efficient and executes faster. Refer to “Error!
Reference source not found.” on page Error! Bookmark not defined. for
more information.

ORDER Directive
When you specify the ORDER directive, C51 locates variables in memory in
the order in which they are declared in your source file. Refer to “ORDER”
on page 64 for more information.

REGFILE Directive
C51 now supports the REGFILE directive which lets you specify the name
of the register definition file generated by the linker. This file contains
information that is used to optimize the use of registers between functions in
different modules. Refer to “REGFILE” on page 69 for more information.

vprintf and vsprint Library Functions
The vprintf and vsprintf library functions have been added. Refer to
“vprintf” on page 343 and “vsprintf” on page 345 for more information.
 356 Appendix B. Version Differences

Version 3.2 Differences

ANSI Standard Automatic Integer Promotion
The latest version of the ANSI C Standard requires that calculations use int
values if char or unsigned char values might overflow during the
calculation. This new requirement is based on the premise that int and char
operations are similar on 16-bit CPUs. C51 supports this feature as the
B default and provides you with two new control directives, INTPROMOTE
and NOINTPROMOTE, to enable or disable integer promotion.

There is a big difference between 8-bit and 16-bit operations on the 8-bit
8051 in terms of code size and execution speed. For this reason, you might
want to disable integer promotion by using the NOINTPROMOTE control
directive.

However, if you wish to retain maximum compatibility with other C

compilers and platforms, leave integer promotions enabled.

Assembly Source Generation with In-Line Assembly
You may use the new control directives ASM and ENDASM to include
source text to output to .SRC files generated using the SRC command
directive.

New Control Directives
The control directives ASM, ENDASM, INTERVAL, INTPROMOTE,
INTVECTOR, MAXARGS, and NOINTPROMOTE have been added or
enhanced.

Offset and Interval Can Now Be Specified for Interrupt Vectors
You may now specify the offset and interval for the interrupt vector table.
These features provide support for the SIECO-51 derivatives and allow you
to specify a different location for the interrupt vector in situations where the
interrupt table is not located at address 0000h.

Parameter Passing to Indirectly Called Functions
Function parameters may now be passed to indirectly called functions if all of
the parameters can be passed in CPU registers. These functions do not have
to be declared with the reentrant attribute.

Source Code Provided For Memory Allocation Functions
C source code for the memory allocation routines is now provided with the
C51 compiler. You may now more easily adapt these functions to the
hardware architecture of your embedded system.
Keil Software — Cx51 Compiler User’s Guide 357

Trigraphs
C51 now supports trigraph sequences.

Variable-length Argument Lists for All Functions
Variable-length argument lists are now supported for all function types.
Functions with a variable length argument list do not have to be declared
using the reentrant attribute. The new command line directive MAXARGS
determines the size of the parameter passing area.

Version 3.0 Differences

B

New Control Directive Added for Assembly Source File Output
The SRC control directive has been added to direct the compiler to generate
an assembly language source file instead of an object file.

New Library Functions
The library functions calloc, free, init_mempool, malloc, and realloc have
been added.
 358 Appendix B. Version Differences

Version 2 Differences

Absolute Register Addressing
C51 now generates code that performs absolute register addressing. This
improves execution speed. The control directives AREGS and NOAREGS,
respectively, enable or disable this feature.

Bit-addressable Memory Type
B Variable types of char and int can now be declared to reside in the
bit-addressable internal memory area by using the bdata memory specifier.

Intrinsic Functions
Intrinsic functions have been added to the library to support some of the
special instructions built in to the 8051.

Mixed Memory Models
Calls to and from functions of different memory models are now supported.

New Optimizer Levels
Two new levels of optimization have been added to the C51 compiler. These
new levels support register variables, local common subexpression
elimination, loop optimizations, and global common subexpression
elimination, to name a few.

New Predefined Macros
The macros _ _C51_ _ and _ _MODEL_ _ are now defined by the
preprocessor at compile time.

Reentrant and Recursive Functions
Individual functions may now be defined as being reentrant or recursive by
using the reentrant function attribute.

Registers Used for Parameter Passing
C51 now passes up to 3 function arguments using registers. The
REGPARMS and NOREGPARMS directives enable or disable this feature.

Support for Memory-specific Pointers
Pointers may now be defined to reference data in a particular memory area.

Support for PL/M-51 Functions
The alien keyword has been added to support PL/M-51 compatible functions
and function calls.

Volatile Type Specifier
The volatile variable attribute may be used to enforce variable access and to
prevent optimizations involving that variable.
Keil Software — Cx51 Compiler User’s Guide 359

Using C51 Version 5 with Previous

Versions
You may wish to use the C51 Version 5 with older versions of the 8051
development tools such as BL51, OHS51, or debugging tools and emulators.
The new compiler adds object file records for register optimization which makes
the object format incompatible with the old tools. However, you can direct the
compiler and linker to generate object modules that are compatible with the old
tools.
B
1. Invoke C51 with the control NOAMAKE and do not use REGFILE.

or

2. Invoke BL51 with the control NOAMAKE.

If you are using old debugging tools, you may have problems displaying
floating-point numbers and pointers. Make sure that you have current versions
of the debugging software.
Keil Software — Cx51 Compiler User’s Guide 361

Appendix C. Writing Optimum Code

This section lists a number of ways you can improve the efficiency (i.e., smaller
code and faster execution) of the 8051 code generated by the Cx51 compiler.
The following is by no means a complete list of things to try. These suggestions
in most cases, however, improve the speed and code size of your program.

Memory Model
The most significant impact on code size and execution speed is memory model.
Compiling in small model always generates the smallest, fastest code possible.
The SMALL control directive instructs the Cx51 compiler to use the small
memory model. In small model, all variables, unless declared otherwise, reside
C
in the internal memory of the 8051. Memory access to internal data memory is
fast (typically performed in 1 or 2 clock cycles), and the generated code is much
smaller than that generated with the compact or large models. For example, the
following loop:
for (i = 0; i < 100; i++) {
do_nothing ();
}

is compiled both in small model and in large model to demonstrate the difference
in generated code. The following is the small model translation:
stmt level source

1 #pragma small
2
3 void do_nothing (void);
4
5
6 void func (void)
7 {
8 1 unsigned char i;
9 1
10 1 for (i = 0; i < 100; i++)
11 1 {
12 2 do_nothing ();
13 2 }
14 1 }
; FUNCTION func (BEGIN)
; SOURCE LINE # 10
0000 E4 CLR A
0001 F500 R MOV i,A
0003 ?C0001:
0003 E500 R MOV A,i
0005 C3 CLR C
0006 9464 SUBB A,#064H
 362 Appendix C. Writing Optimum Code

0008 5007 JNC ?C0004

; SOURCE LINE # 12
000A 120000 E LCALL do_nothing
; SOURCE LINE # 13
000D 0500 R INC i
000F 80F2 SJMP ?C0001
; SOURCE LINE # 14
0011 ?C0004:
0011 22 RET
; FUNCTION func (END)

In small model, the variable i is maintained in internal data memory. The

instructions to access i, MOV A,i and INC i, require only two bytes each of
code space. In addition, each of these instructions executes in only one clock
cycle. The total size for the main function when compiled in small model is 11h
or 17 bytes.
C The following is the same code compiled using the large model:
; FUNCTION func (BEGIN)
; SOURCE LINE # 10
0000 E4 CLR A
0001 900000 R MOV DPTR,#i
0004 F0 MOVX @DPTR,A
0005 ?C0001:
0005 900000 R MOV DPTR,#i
0008 E0 MOVX A,@DPTR
0009 C3 CLR C
000A 9464 SUBB A,#064H
000C 500B JNC ?C0004
; SOURCE LINE # 12
000E 120000 E LCALL do_nothing
; SOURCE LINE # 13
0011 900000 R MOV DPTR,#i
0014 E0 MOVX A,@DPTR
0015 04 INC A
0016 F0 MOVX @DPTR,A
0017 80EC SJMP ?C0001
; SOURCE LINE # 14
0019 ?C0004:
0019 22 RET
; FUNCTION func (END)

In large model, the variable i is maintained in external data memory. To access

i, the compiler must first load the data pointer and then perform an external
memory access (see offset 0001h through 0004h in the above listing). These two
instructions alone take 4 clock cycles. The code to increment i is found from
offset 0011h to offset 0016h. This operation consumes 6 bytes of code space
and takes 7 clock cycles to execute. The total size for the main function when
compiled in small model is 19h or 25 bytes.
Keil Software — Cx51 Compiler User’s Guide 363

Variable Location
Frequently accessed data objects should be located in the internal data memory
of the 8051. Accessing the internal data memory is much more efficient than
accessing the external data memory. The internal data memory is shared among
register banks, the bit data area, the stack, and other user defined variables with
the memory type data.

Because of the limited amount of internal data memory (128 to 256 bytes), all
your program variables may not fit into this memory area. In this case, you must
locate some variables in other memory areas. There are two ways to do this.

One way is to change the memory model and let the compiler do all the work.
This is the simplest method, but it is also the most costly in terms of the amount C
of generated code and system performance. Refer to “Memory Model” on page
361 for more information.

Another way to locate variables in other memory areas is to manually select the
variables that can be moved into external data memory and declare them using
the xdata memory specifier. Usually, string buffers and other large arrays can
be declared with the xdata memory type without a significant degradation in
performance or increase in code size.

Variable Size
Members of the 8051 family are all 8-bit CPUs. Operations that use 8-bit types
(like char and unsigned char) are much more efficient than operations that use
int or long types. For this reason, always use the smallest data type possible.

The Cx51 compiler directly supports all byte operations. Byte types are not
promoted to integers unless required. See the INTPROMOTE directive for
more information.

An example can be illustrated by examining of multiplication operations. The

multiplication of two char objects is done inline with the 8051 instruction
MUL AB. To accomplish the same operation with int or long types would require
a call to a compiler library function.
 364 Appendix C. Writing Optimum Code

Unsigned Types
The 8051 family of processors does not specifically support operations with
signed numbers. The compiler must generate additional code to deal with sign
extensions. Far less code is produced if unsigned objects are used wherever
possible.

Local Variables
When possible, use local variables for loops and other temporary calculations.
As part of the optimization process, the compiler attempts to maintain local
C variables in registers. Register access is the fastest type of memory access. The
best effect is normally achieved with unsigned char and unsigned int variable
types.

Other Sources
The quality of the compiler generated code is more often than not directly
influenced by the algorithms implemented in the program. Sometimes, you can
improve the performance or reduce the code size simply by using a different
algorithm. For example, a heap sort algorithm always outperforms a bubble sort
algorithm.

For more information on how to write efficient programs, refer to the following
books:

The Elements of Programming Style, Second Edition

Kernighan & Plauger
McGraw-Hill
ISBN 0-07-034207-5

Writing Efficient Programs

Jon Louis Bentley
Prentice-Hall Software Series
ISBN 0-13-970244-X

Efficient C
Plum & Brodie
Plum Hall, Inc.
ISBN 0-911537-05-8
Keil Software — Cx51 Compiler User’s Guide 365

Appendix D. Compiler Limits

The Cx51 compiler embodies some known limitations that can be arranged into
two distinct categories:

Limitations of the compiler implementation

Limitations of the Intel Object Module Format (OMF-51)

For the most part, there are no limits placed on the compiler with respect to
components of the C language; for example, you may specify an unlimited
number of symbols or number of case statements in a switch block. If there is
enough address space, several thousand symbols could be defined. However if
you are using the Intel OMF51 file format, C51 is bound by a historical limit of
256 global symbols.

Limitations of the Cx51 Compiler D

Implementation

A maximum of 19 levels of indirection (access modifiers) to any standard
data type are supported. This includes array descriptors, indirection
operators, and function descriptors.

Number of functions in a module (see OMF-51 Limitation values).

Names can be up to 255 characters long. However, only the first 32 are
significant. The C language provides for case sensitivity in regard to function
and variable names. However, for compatibility reasons, all names in the
object file appear in capital letters. It is therefore irrelevant if an external
object name within the source program is written in capital or small letters.

The maximum number of case statements in a switch block is not fixed.
Limits are imposed only by the available memory size and the maximum size
of individual functions.

The maximum number of nested function calls in an invocation parameter list
is 10.

The maximum number of nested include files is 9. This value is independent
of list files, preprocessor files, or whether or not an object file is to be
generated.
 366 Appendix D. Compiler Limits

The maximum depth of directives for conditional compilation is 20. This is a

preprocessor limitation.

Instruction blocks ({…}) may be nested up to 15 levels deep.

Macros may be nested up to 8 levels deep.

A maximum of 32 parameters may be passed in a macro or function call.

The maximum length of a line or a macro definition is 2000 characters. Even
after a macro expansion, the result may not exceed 2000 characters.

Limitations of the Intel Object Module

Format

There may be a maximum of 255 segments. The number of functions that
may exist in a module is difficult to calculate. Each function definition in a
source program module receives a separate code segment. If local variables
D exist within the function, a separate data segment is also created. If bit
variables exist within the function, a separate bit segment is created too. For
these reasons, the number of functions that may exist within a module
depends upon the number of variables in the functions.

There may be a maximum of 256 external symbols. All module names which
have the memory class extern, are contained in this external symbol class.
The compiler produces external names for external functions, that are
dependent upon whether or not bit or data parameters are contained in the
function call. Thus, the reference name to the external data and bit segments
is produced in a manner analogous to the global functions.

NOTE
This limits do not apply if you are using the OMF2 file format or the CX51
compiler.
Keil Software — Cx51 Compiler User’s Guide 367

Appendix E. Byte Ordering

Most microprocessors have a memory architecture that is composed of 8-bit
address locations known as bytes. Many data items (addresses, numbers, and
strings) are too long to be stored using a single byte and must be stored in a
series of consecutive bytes.

When using data that are stored in multiple bytes, byte ordering becomes an
issue. Unfortunately, there is not just one standard for the order in which bytes
in multi-byte data are stored. There are two popular methods of byte ordering
currently in widespread use.

The first method is called little endian and is often referred to as Intel order. In
little endian, the least significant, or low-order byte is stored first. For example,
a 16-bit integer value of 0x1234 (4660 decimal) would be stored using the little
endian method in two consecutive bytes as follows:

Address +0 +1
Contents 0x34 0x12

A 32-bit integer value of 0x57415244 (1463898692 decimal) would be stored

using the little endian method as follows:
E
Address +0 +1 +2 +3
Contents 0x44 0x52 0x41 0x57

A second method of accessing multi-byte data is called big endian and is often
referred to as Motorola order. In big endian, the most significant, or high-order
byte is stored first, and the least significant, or low-order byte is stored last. For
example, a 16-bit integer value of 0x1234 would be stored using the big endian
method in two consecutive bytes as follows:

Address +0 +1
Contents 0x12 0x34

A 32-bit integer value of 0x004A4F4E would be stored using the big endian
method as follows:

Address +0 +1 +2 +3
Contents 0x00 0x4A 0x4F 0x4E
 368 Appendix E. Byte Ordering

The 8051 is an 8-bit machine and has no instructions for directly manipulating
data objects that are larger than 8 bits. Multi-byte data are stored according to
the following rules.

The 8051 LCALL instruction stores the address of the next instruction on the
stack. The address is pushed onto the stack low-order byte first. The address
is, therefore, stored in memory in little endian format.

All other 16-bit and 32-bit values are stored, contrary to other Intel
processors, in big endian format, with the high-order byte stored first. For
example, the LJMP and LCALL instructions expect 16-bit addresses that are
in big endian format.

Floating-point numbers are stored according to the IEEE-754 format and are
stored in big endian format with the high-order byte stored first.

If your 8051 embedded application performs data communications with other

microprocessors, it may be necessary to know the byte ordering method used by
the other CPU. This is certainly true when transmitting raw binary data.

E
Keil Software — Cx51 Compiler User’s Guide 369

Appendix F. Hints, Tips, and

Techniques
This section lists a number of illustrations and tips which commonly require
further explanation. Items in this section are listed in no particular order and are
merely intended to be referenced if you experience similar problems.

Recursive Code Reference Error

The following program example:
#pragma code symbols debug oe

void func1(unsigned char *msg ) { ; }

void func2( void ) {

unsigned char uc;
func1("xxxxxxxxxxxxxxx");
}

code void (*func_array[])() = { func2 };

void main( void ) {

(*func_array[0])();
}

when compiled and linked using the following command lines:

C51 EXAMPLE1.C

BL51 EXAMPLE1.OBJ IX F
fails and display the following error message.
*** WARNING 13: RECURSIVE CALL TO SEGMENT
SEGMENT: ?CO?EXAMPLE1
CALLER: ?PR?FUNC2?EXAMPLE1

In this program example, func2 defines a constant string (“xxx…xxx”) which is

directed into the constant code segment ?CO?EXAMPLE1. The definition code
void (*func_array[])() = { func2 }; yields a reference between segment
?CO?EXAMPLE1 (where the code table is located) and the executable code
segment ?PR?FUNC2?EXAMPLE1. Because func2 also refers to segment
?CO?EXAMPLE1, BL51 assumes that there is a recursive call.
 370 Appendix F. Hints, Tips, and Techniques

To avoid this problem, link using the following command line:

BL51 EXAMPLE1.OBJ IX OVERLAY &
(?CO?EXAMPLE1 ~ FUNC2, MAIN ! FUNC2)

?CO?EXAMPLE1 ~ FUNC2 deletes the implied call reference between func2 and
the code constant segment in the example. Then, MAIN ! FUNC2 adds an
additional call to the reference listing between MAIN and FUNC2 instead. Refer
to the 8051 Utilities User’s Guide for more information.

In summary, automatic overlay analysis cannot be successfully accomplished

when references are made via pointers to functions. References of this type must
be manually implemented, as in the example above.

Problems Using the printf Routines

The printf functions are implemented using a variable-length argument list.
Arguments specified after the format string are passed using their inherent data
type. This can cause problems when the format specification expects a data
object of a different type than was passed. For example, the following code:
printf ("%c %d %u %bu", 'A', 1, 2, 3);

does not print the string “A 1 2 3”. This is because the Cx51 compiler passes the
arguments 1, 2, and 3 all as 8-bit byte types. The format specifiers %d and
%u both expect 16-bit int types.

F To avoid this type of problem, you must explicitly define the data type to pass to
the printf function. To do this, you must type cast the above values. For
example:
printf ("%c %d %u %bu", 'A',(int) 1, (unsigned int) 2, (char) 3);

If you are uncertain of the size of the argument that is passed, you may cast the
value to the desired size.
Keil Software — Cx51 Compiler User’s Guide 371

Uncalled Functions
It is common practice during the development process to write but not call
additional functions. While the compiler permits this without error, the
Linker/Locator does not treat this code casually, because of the support for data
overlaying, and emits a warning message.

Interrupt functions are never called, they are invoked by the hardware. An
uncalled routine is treated as a potential interrupt routine by the linker. This
means that the function is assigned non-overlayable data space for its local
variables. This quickly exhausts all available data memory (depending upon the
memory model used).

If you unexpectedly run out of memory, be sure to check for linker warnings
relating to uncalled or unused routines. You can use the linker’s IXREF control
directive to include a cross reference list in the linker map (.M51) file.

F
 372 Appendix F. Hints, Tips, and Techniques

Trouble with the bdata Memory Type

Some users have reported difficulties in using the bdata memory type. Using
bdata is similar to using the sfr modifier. The most common error is
encountered when referencing a bdata variable defined in another module. For
example:
extern bdata char xyz_flag;

sbit xyz_bit1 = xyz_flag^1;

In order to generate the appropriate instructions, the compiler must have the
absolute value of the reference to be generated. In the above example, this
cannot be done, as this address of xyz_flag cannot be known until after the
linking phase has been completed. Follow the rules below to avoid this problem.

1. A bdata variable (defined and used in the same way as an sfr) must be
defined in global space; not within the scope of a procedure.

2. A bdata bit variable (defined and used in the same way as an sbit) must also
be defined in global space, and cannot be located within the scope of a
procedure.

3. The definition of the bdata variable and the creation of its sbit access
component name must be accomplished where the compiler has a “view” of
both the variable and the component.

For example, declare the bdata variable and the bit component in the same
F source module:
bdata char xyz_flag;
sbit xyz_bit1 = xyz_flag^1;

Then, declare the bit component external:

extern bit xyz_bit1;

As with any other declared and named C variable that reserves space, simply
define your bdata variable and its component sbits in a module. Then, use the
extern bit specifier to reference it as the need arises.
Keil Software — Cx51 Compiler User’s Guide 373

Using Monitor-51
If you want to test a C program with Monitor-51 and if the Monitor-51 is
installed at code address 0, consider the following rules (the specification refers
to a target system where the available code memory for user programs starts at
address 8000H):

All C modules which contain interrupt functions must be translated with the
control directive INTVECTOR (0x8000).

In the file STARTUP.A51 (directory: LIB) the statement CSEG AT 0 must be
replaced with CSEG AT 8000H. The this file must be assembled and added
to the linker/locator invocation according the specifications in the file header.

F
 374 Appendix F. Hints, Tips, and Techniques

Function Pointers
Function pointers are one of the most difficult aspects of C to understand and to
properly utilize. Most problems involving function pointers are caused by
improper declaration of the function pointer, improper assignment, and improper
dereferencing.

The following brief example demonstrates how to declare a function pointer (f),
how to assign function addresses to it, and how to call the functions through the
pointer. The printf routine is used for example purposes when running DS51 to
simulate program execution.
#pragma code symbols debug oe

#include <reg51.h> /* special function register declarations */

#include <stdio.h> /* prototype declarations for I/O functions */

void func1(int d) { /* function #1 */

printf("In FUNC1(%d)\n", d);
}

void func2(int i) { /* function #2 */

printf("In FUNC2(%d)\n", i);
}

void main(void) {
void (*f)(int i); /* Declaration of a function pointer */
/* that takes one integer arguments */
/* and returns nothing */

SCON = 0x50; /* SCON: mode 1, 8-bit UART, enable rcvr */

TMOD |= 0x20; /* TMOD: timer 1, mode 2, 8-bit reload */
TH1 = 0xf3; /* TH1: reload value for 2400 baud */

F TR1
TI
= 1;
= 1; /* TI:
/* TR1: timer 1 run
set TI to send first char of UART
*/
*/

while( 1 ) {
f = (void *)func1; /* f points to function #1 */
f(1);
f = (void *)func2; /* f points to function #2 */
f(2);
}
}

NOTE
Because of the limited stack space of the 8051, the linker overlays function
variables and arguments in memory. When you use a function pointer, the linker
cannot correctly create a call tree for your program. For this reason, you may
have to correct the call tree for the data overlaying. Use the OVERLAY
directive with the linker to do this. Refer to the 8051 Utilities User’s Guide for
more information.
Keil Software — Cx51 Compiler User’s Guide 375

Glossary
A51
The standard 8051 Macro Assembler.

AX51
The extended 8051 Macro Assembler.

ANSI
American National Standards Institute. The organization responsible for
defining the C language standard.

argument
The value that is passed to a macro or function.

arithmetic types
Data types that are integral, floating-point, or enumerations.

array
A set of elements, all of the same data type.

ASCII
American Standard Code for Information Interchange. This is a set of 256
codes used by computers to represent digits, characters, punctuation, and
other special symbols. The first 128 characters are standardized. The
remaining 128 are defined by the implementation.

batch file
An ASCII text file containing commands and programs that can be invoked
from the command line.

Binary-Coded Decimal (BCD)

A BCD (Binary-Coded Decimal) is a system used to encode decimal numbers
in binary form. Each decimal digit of a number is encoded as a binary value
4 bits long. A byte can hold 2 BCD digits – one in the upper 4 bits (or
nibble) and one in the lower 4 bits (or nibble).

BL51
The standard 8051 linker/locator.

block
A sequence of C statements, including definitions and declarations, enclosed
within braces ({ }).
376 Glossary

C51
The Optimizing C Compiler for classic 8051 and extended 8051 devices.

CX51
The Optimizing C Compiler for Philips 80C51MX architecture.

constant expression
Any expression that evaluates to a constant non-variable value. Constants
may include character and integer constant values.

control
Command line control switch to the compiler, assembler or linker.

declaration
A C construct that associates the attributes of a variable, type, or function
with a name.

definition
A C construct that specifies the name, formal parameters, body, and return
type of a function or that initializes and allocates storage for a variable.

directive
Instruction or control switch to the compiler, assembler or linker.

escape sequence
A backslash (‘\’) character followed by a single letter or a combination of
digits that specifies a particular character value in strings and character
constants.

expression
A combination of any number of operators and operands that produces a
constant value.

formal parameters
The variables that receive the value of arguments passed to a function.

function
A combination of declarations and statements that can be called by name to
perform an operation and/or return a value.

function body
A block containing the declarations and statements that make up a function.

function call
An expression that invokes and possibly passes arguments to a function.
Keil Software — Cx51 Compiler User’s Guide 377

function declaration
A declaration providing the name and return type of a function that is
explicitly defined elsewhere in the program.

function definition
A definition providing the name, formal parameters, return type, declarations,
and statements describing what a function does.

function prototype
A function declaration that includes a list of formal parameters in parentheses
following the function name.

in-circuit emulator (ICE)

A hardware device that aids in debugging embedded software by providing
hardware-level single-stepping, tracing, and break-pointing. Some ICEs
provide a trace buffer that stores the most recent CPU events.

include file
A text file that is incorporated into a source file.

keyword
A reserved word with a predefined meaning for the compiler or assembler.

L51
The old version of the 8051 linker/locator. L51 is replaced with the BL51
linker/locater.

LX51
The extended 8051 linker/locator.

LIB51, LIBX51
The commands to manipulate library files using the Library Manager.

library
A file that stores a number of possibly related object modules. The linker can
extract modules from the library to use in building a target object file.

LSB
Least significant bit or byte.

macro
An identifier that represents a series of keystrokes.

manifest constant
A macro that is defined to have a constant value.
378 Glossary

MCS® 51
The general name applied to the Intel family of 8051 compatible
microprocessors.

memory model
Any of the models that specifies which memory areas are used for function
arguments and local variables.

mnemonic
An ASCII string that represents a machine language opcode in an assembly
language instruction.

MON51
An 8051 program that can be loaded into your target CPU to aid in debugging
and rapid product development through rapid software downloading.

MSB
Most significant bit or byte.

newline character
A character used to mark the end of a line in a text file or the escape sequence
(‘\n’) to represent the newline character.

null character
ASCII character with the value 0 represented as the escape sequence (‘\0’).

null pointer
A pointer that references nothing. A null pointer has the integer value 0.

object
An area of memory that can be examined. Usually used when referring to the
memory area associated with a variable or function.

object file
A file, created by the compiler, that contains the program segment
information and relocatable machine code.

OH51, OHX51
The commands to convert absolute object files into Intel HEX file format.

opcode
Also referred to as operation code. An opcode is the first byte of a machine
code instruction and is usually represented as a 2–digit hexadecimal number.
The opcode indicates the type of machine language instruction and the type
of operation to perform.
Keil Software — Cx51 Compiler User’s Guide 379

operand
A variable or constant that is used in an expression.

operator
A symbol (e.g., +, -, *, /) that specifies how to manipulate the operands of an
expression.

parameter
The value that is passed to a macro or function.

PL/M-51
A high-level programming language introduced by Intel at the beginning of
the 80ths

pointer
A variable containing the address of another variable, function, or memory
area.

pragma
A statement that passes an instruction to the compiler at compile time.

preprocessor
The compiler’s first pass text processor that manipulates the contents of a C
file. The preprocessor defines and expands macros, reads include files, and
passes control directives to the compiler.

relocatable
Object code that can be relocated and is not at a fixed address.

RTX51 Full
An 8051 Real-time Executive that provides a multitasking operating system
kernel and library of routines for its use.

RTX51 Tiny
A limited version of RTX51.

scalar types
In C, integer, enumerated, floating-point, and pointer types.

scope
Sections of a program where an item (function or variable) can be referenced
by name. The scope of an item may be limited to file, function, or block.

Special Function Register (SFR)

An SFR or Special Function Register is a register in the 8051 internal data
memory space that is used to read an write to the hardware components of the
380 Glossary

8051. This includes the serial port, timers, counters, I/O ports, and other
hardware control registers.

source file
A text file containing C program or assembly program code.

stack
An area of memory, indirectly accessed by a stack pointer, that shrinks and
expands dynamically as items are pushed onto and popped off of the stack.
Items in the stack are removed on a LIFO (last-in first-out) basis.

static
A storage class that, when used with a variable declaration in a function,
causes variables to retain their value after exiting the block or function in
which they are declared.

stream functions
Routines in the library that read and write characters using the input and
output streams.

string
An array of characters that is terminated with a null character (‘\0’).

string literal
A string of characters enclosed within double quotes (“ ”).

structure
A set of elements of possibly different types grouped together under one
name.

structure member
One element of a structure.

token
A fundamental symbol that represents a name or entity in a programming
language.

two’s complement
A binary notation that is used to represent both positive and negative
numbers. Negative values are created by complementing all bits of a positive
value and adding 1.

type
A description of the range of values associated with a variable. For example,
an int type can have any value within its specified range (-32768 to 32767).
Keil Software — Cx51 Compiler User’s Guide 381

type cast
An operation in which an operand of one type is converted to another type by
specifying the desired type, enclosed within parentheses, immediately
preceding the operand.

µVision2
An integrated software development platform that supports the Keil Software
development tools. µVision2 combines Project Management, Source Code
Editing, and Program Debugging in one environment.

whitespace character
Characters used as delimiters in C programs such as space, tab, and newline.

wild card
One of the characters (? or *) that can be used in place of characters in a
filename.
382 Index

Index
_tolower 213,335
# _toupper 213,337

# 130
## 131
+
#define 129 +INF
#elif 129 described 179
#else 129
#endif 129 1
#error 129
#if 129 16-bit Binary Integer Operations 136
#ifdef 129
#ifndef 129 3
#include 129
32-bit Binary Integer Operations 136
#line 129
#pragma 129
#undef 129 8
8051 Derivatives 133
. 8051 Hardware Stack 114
8051 Memory Areas 86
.I files 21
8051 Variants 17
.LST files 21
8051-Specific Optimizations 156
.OBJ files 21
80C320/520 or variants 54
.SRC files 21
80C517 Routines
acos517 221
_ asin517 221
_ _C51_ _ 132,203 atan517 221
_ _DATE_ _ 132,203 atof517 221
_ _FILE_ _ 132,203 cos517 221
_ _LINE_ _ 132,203 exp517 221
_ _MODEL_ _ 132,203 log10517 221
_ _STDC_ _ 132,200,203 log517 221
_ _TIME_ _ 132,200,203 printf517 221
_at_ 99,182,355 scanf517 221
_chkfloat_ 215,240 sin517 221
_crol_ 205,215,243 sprintf517 221
_cror_ 205,215,244 sqrt517 221
_getkey 217,251 sscanf517 221
_irol_ 205,215,254 strtod517 221
_iror_ 205,215,255 tan517 221
_lrol_ 205,215,272 80C517.H 221
_lror_ 205,215,273 80C751.LIB 206
_nop_ 205,220,282 80x8252 or variants 53
_testbit_ 205,220,331
Keil Software — Cx51 Compiler User’s Guide 383

A AREGS 26
Argument lists, variable-length 50,220
A51 argument, defined 375
Interfacing 161 arithmetic types, defined 375
A51, defined 375 array, defined 375
abs 214,228 ASCII, defined 375
ABSACC.H 222 asin 214,230
Absolute Memory Access function timing 137
Macros 208 asin517 230
CBYTE 208 function timing 137
CWORD 208 ASM 28
DBYTE 209 Assembly code in-line 28
DWORD 209 Assembly listing 31
PBYTE 210 Assembly source file generation 78
PWORD 210 assert 231
XBYTE 211 ASSERt.H 223
XWORD 211 atan 214,232
Absolute Memory Locations 180 function timing 137
Absolute register addressing 26 atan2 214,233
Absolute value atan517 232
abs 228 function timing 137
cabs 237 Atmel
fabs 246 89x8252 and variants 134
labs 267 Atmel 80x8252 or variants 53
Abstract Pointers 109 Atmel WM
Access Optimizing 156 dual DPTR support 140
Accessing Absolute Memory atof 214,234
Locations 180 function timing 137
acos 214,229 atof517 234
function timing 137 function timing 137
acos517 229 atoi 214,235
function timing 137 atol 214,236
Additional items, notational AUTOEXEC.BAT 19
conventions 5 AX51, defined 375
Address of interrupts 120
Advanced Programming
Techniques 143 B
alien 127 batch file, defined 375
ANSI bdata 87
Differences 347 bdata, tips for 372
Include Files 221 big endian 367
Library 205 Binary Integer Operations 136
Standard C Constant 132 Binary-Coded Decimal (BCD),
ANSI, defined 375 defined 375
Arc bit
cosine 229 As first parameter in function
sine 230 call 115
tangent 232,233 Bit shifting functions
384 Index

_crol_ 215 Character Classification

_cror_ 215 Routines 213
_irol_ 215 isalnum 213
_iror_ 215 isalpha 213
_lrol_ 215 iscntrl 213
_lror_ 215 isdigit 213
Bit Types 93 isgraph 213
Bit-addressable objects 94 islower 213
BL51, defined 375 isprint 213
block, defined 375 ispunct 213
bold capital text, use of 5 isspace 213
bold type, use of 5 isupper 213
Books isxdigit 213
About the C Language 18 Character Conversion and
BR 30 Classification Routines 213
braces, use of 5 Character Conversion Routines 213
BROWSE 30 _tolower 213
Buffer manipulation routines _toupper 213
memccpy 212,275 toascii 213
memchr 212,276 toint 213
memcmp 212,277 tolower 213
memcpy 212,278 toupper 213
memmove 212,279 Choices, notational conventions 5
memset 212,280 CO 33
Buffer Manipulation Routines 212 code 86
CODE 31
C Code generation options 157
compact 116
C51 command 19 COMPACT 32
C51, defined 376 Compact memory model 32
C517 CPU 51 Compact Model 90
C51C.LIB 206 Compatibility
C51FPC.LIB 206 differences from standard C 347
C51FPL.LIB 206 Differences to previous
C51FPS.LIB 206 versions 351
C51INC 19 Differences to Version 2 358
C51L.LIB 206 Differences to Version 3.0 357
C51LIB 19 Differences to Version 3.2 356
C51S.LIB 206 Differences to Version 3.4 355
cabs 214,237 Differences to Version 4 352
calloc 216,238 Differences to Version 5 351
CALLOC.C 153 Differences to Version 6.0 351
Case/Switch Optimizing 156 standard C library differences
Categories of Cx51 directives 22 347
CBYTE 180,208 Compiling 19
CD 31 COND 33
ceil 214,239 Conditional compilation 33
Keil Software — Cx51 Compiler User’s Guide 385

constant expression, defined 376 DB 35

Constant Folding 156 DBYTE 180,209
Control directives 22 Dead Code Elimination 156
control, defined 376 DEBUG 35
cos 214,241 Debug information 35,59
function timing 137 Debugging 183
cos517 241 declaration, defined 376
function timing 137 define 129
cosh 214,242 DEFINE 36
courier typeface, use of 5 Defining macros on the
CP 32 command line 36
CTYPE.H 223 definition, defined 376
Customization Files 143 Derivatives 17
CWORD 180,208 DF 36
Cx51 Differences from Standard C 347
Control directives 22 Differences to Previous Versions 351
Errorlevel 21 Directive categories 22
Extensions 85 Directive reference 25
Output files 21 directive, defined 376
CX51 command 19 DISABLE 37
CX51, defined 376 Disabling interrupts 37
Displayed text, notational
D conventions 5
Document conventions 5
Dallas 80C320/520 or variants 54 double brackets, use of 5
Dallas Semiconductor DWORD 180,209
80C320 134
80C420 134
80C520 134
E
80C530 134 EJ 39
data 87 EJECT 39
Data Conversion Routines 214 elif 129
abs 214 ellipses, use of 5
atof 214 ellipses, vertical, use of 5
atoi 214 else 129
atol 214 ENDASM 28
cabs 214 endian 367
labs 214 endif 129
strtod 214 Environment Variables 19
strtol 214 EOF 225
strtoul 214 error 129
Data memory 87 ERRORLEVEL 21
Data Overlaying 156 escape sequence, defined 376
data pointers 134,135,140 Execution timings 136
Data sizes 92 exp 214,245
Data Storage Formats 174 function timing 137
Data type ranges 92 exp517 245
Data Types 92 function timing 137
386 Index

exponent 177 Reentrant 124

expression, defined 376 Register Bank 117
Extensions for Cx51 85 Stack & Parameters 114
Extensions to C 85
External Data Memory 88 G
General Optimizations 156
F getchar 217,250
fabs 214,246 GETKEY.C 153
Fatal Error Messages 185 gets 217,252
FF 40 Global Common Subexpression
Filename, notational conventions 5 Elimination 156
Files generated by Cx51 21 Global register optimization 69
FLOATFUZZY 40 Glossary 375
Floating-point
exponent 177 H
mantissa 177
storage format 177 High-Speed Arithmetic 136
Floating-point compare 40
Floating-Point Errors 179 I
+INF 179 IBPSTACK 144
-INF 179 IBPSTACKTOP 145
Nan 179 ICE, defined 377
Floating-point numbers 177 ID 41
Floating-point Operations 137 idata 87
floor 214,247 IDATALEN 144
fmod 214,248 IEEE-754 standard 177
Form feeds 39 if 129
formal parameters, defined 376 ifdef 129
free 216,249 ifndef 129
FREE.C 153 INCDIR 41
function body, defined 376 in-circuit emulator, defined 377
function call, defined 376 include 129
function declaration, defined 377 Include file listing 49
Function Declarations 113 include file, defined 377
function definition, defined 377 Include Files 41,221
Function extensions 113 80C517.H 221
Function Parameters 161 ABSACC.H 222
Function Pointers, tips for 374 ASSERT.H 223
function prototype, defined 377 CTYPE.H 223
Function return values 115 INTRINS.H 223
Function Return Values 163 MATH.H 223
function, defined 376 REGxxx.H 221
Functions 113 SETJMP.H 225
Interrupt 120 STDARG.H 225
Memory Models 116 STDDEF.H 225
Parameters in Registers 115 STDIO.H 225
Recursive 124
Keil Software — Cx51 Compiler User’s Guide 387

STDLIB.H 226 iscntrl 213,258

STRING.H 226 isdigit 213,259
-INF isgraph 213,260
described 179 islower 213,261
Infineon isprint 213,262
C517, C517A, C509 135 ispunct 213,263
Infineon C517 51 isspace 213,264
INIT.A51 150 isupper 213,265
INIT_MEM.C 153 isxdigit 213,266
init_mempool 216,253 italicized text, use of 5
INIT751.A51 151 IV 46
Initializing memory 144
Initializing the stream I/O J
routines 217
In-line assembly 28 jmp_buf 207
Integer Operations 136 Jump Optimizing 156
Integer promotion 43
Interfacing C Programs to A51 161 K
Interfacing C Programs to
Key names, notational
PL/M-51 173
conventions 5
Internal Data Memory 87
keyword, defined 377
interrupt 118,121
Keywords 85
Interrupt
Addresses 120
Description 120 L
Function rules 123 L51, defined 377
Functions 120 LA 48
Numbers 120 labs 214,267
Interrupt vector 46 Language elements, notational
Interrupt vector interval 42 conventions 5
Interrupt vector offset 46 Language Extensions 85
INTERVAL 42 large 116
INTPROMOTE 43 LARGE 48
INTRINS.H 223 Large memory model 48
Intrinsic Routines 205 Large Model 90
_crol_ 205 LC 49
_cror_ 205 LIB51, defined 377
_irol_ 205 Library Files 206
_iror_ 205 80C751.LIB 206
_lrol_ 205 C51C.LIB 206
_lror_ 205 C51FPC.LIB 206
_nop_ 205 C51FPL.LIB 206
_testbit_ 205 C51FPS.LIB 206
INTVECTOR 46 C51L.LIB 206
IP 43 C51S.LIB 206
isalnum 213,256 Library Reference 205
isalpha 213,257 Library Routines
388 Index

ANSI, excluded from Cx51 348 atan2 214

ANSI, included in Cx51 347 ceil 214
non-ANSI 349 cos 214
Library Routines by Category 212 cosh 214
library, defined 377 exp 214
LIBX51, defined 377 fabs 214
Limitations floor 214
Cx51 365 fmod 214
OMF-51 366 log 214
line 129 log10 214
Linker Location Controls 181 modf 214
LISTINCLUDE 49 pow 214
Listing file generation 68 rand 215
Listing file page length 65 sin 215
Listing file page width 66 sinh 215
Listing include files 49 sqrt 215
little endian 367 srand 215
log 214,268 tan 215
function timing 137 tanh 215
log10 214,269 MATH.H 223
function timing 137 MAXARGS 50
log10517 269 Maximum arguments in
function timing 137 variable-length argument lists 50
log517 268 MCS® 51, defined 378
function timing 137 memccpy 212,275
longjmp 220,270 memchr 212,276
LSB, defined 377 memcmp 212,277
LX51, defined 377 memcpy 212,278
memmove 212,279
M Memory Allocation Routines 216
calloc 216
macro, defined 377 free 216
malloc 216,274 init_mempool 216
MALLOC.C 154 malloc 216
manifest constant, defined 377 realloc 216
mantissa 177 Memory areas 86
Manual organization 4 external data 88
Math Routines 214 internal data 87
_chkfloat_ 215 program 86
_crol_ 215 special function register 89
_cror_ 215 Memory Model 89
_irol_ 215 Compact 90
_iror_ 215 Function 116
_lrol_ 215 Large 90
_lror_ 215 Small 89
acos 214 memory model, defined 378
asin 214 Memory Type 90
atan 214
Keil Software — Cx51 Compiler User’s Guide 389

bdata 87,91 null character, defined 378

code 86,91 null pointer, defined 378
data 87
idata 87,91 O
pdata 88,91
xdata 88,91 O2 61
Memory Typedata 91 OB 60
memset 212,280 OBJECT 58
Miscellaneous Routines 220 Object file generation 58
_nop_ 220 object file, defined 378
_testbit_ 220 object, defined 378
longjmp 220 OBJECTEXTEND 59
setjmp 220 OE 59
mnemonic, defined 378 offsetof 283
MOD517 51,135 OH51, defined 378
MODA2 53,134 OHS51 355
MODDP2 54,134 OHX51, defined 378
modf 214,281 OJ 58
MODP2 55,140 OMF2 61
monitor51, defined 378 Omitted text, notational
MSB, defined 378 conventions 5
ONEREGBANK 60
opcode, defined 378
N operand, defined 379
NaN 241,284,300,301,329 Operation timings 136
described 179 operator, defined 379
newline character, defined 378 OPTIMIZE 62
NOAMAKE 56 Optimizer 155
NOAREGS 26 Optimizing programs 62
NOAU 51 Optimum Code
NOCO 33 Local Variables 364
NOCOND 33 Memory Model 361
NODP8 52 Other Sources 364
NOEXTEND 57 Variable Location 363
NOINTPROMOTE 43 Variable Size 363
NOINTVECTOR 46 Variable Types 364
NOIP 43 Optional items, notational
NOIV 46 conventions 5
NOMOD517 51 Options for Code Generation 157
NOMODA2 53,134 OR 64
NOMODDP2 54,134 ORDER 64
NOMODP2 55,140 Order of variables 64
NOOBJECT 58 OT 62
NOOJ 58 Output files 21
NOPR 68 Overlaying Segments 166
NOPRINT 68
NOREGPARMS 71
NULL 226
390 Index

P _ _MODEL_ _ 132
_ _STDC_ _ 132
Page length in listing file 65 _ _TIME_ _ 132
Page width in listing file 66 Preface 3
PAGELENGTH 65 PREPRINT 67
PAGEWIDTH 66 Preprocessor 129
Parameter Passing in Fixed Preprocessor directives
Memory Locations 163 define 129
Parameter Passing in Registers 162 elif 129
Parameter Passing Via Registers 156 else 129
parameter, defined 379 endif 129
Passing arguments in registers 71 error 129
Passing Parameters in Registers 115 if 129
PATH 19 ifdef 129
PBPSTACK 145 ifndef 129
PBPSTACKTOP 145 include 129
PBYTE 180,210 line 129
pdata 88 pragma 129
PDATALEN 144 undef 129
PDATASTART 144 Preprocessor output file
Peephole Optimization 156 generation 67
Philips preprocessor, defined 379
8xC750 139 PRINT 68
8xC751 139 Printed text, notational
8xC752 139 conventions 5
dual DPTR support 140 printf 217,285
Philips dual DPTR 55 printf, tips for 370
PL 65 printf517 285
PL/M-51 127 Program Memory 86
Defined 379 Program memory size 75
Interfacing 173 putchar 217,291
Pointer Conversions 106 PUTCHAR.C 153
Pointers 101 puts 217,292
Generic 101 PW 66
Memory-specific 104 PWORD 180,210
pointers, defined 379
pow 214,284
PP 67 R
PPAGE 145 R0-R7 26
PPAGEENABLE 145 rand 215,293
PR 68 Range for data types 92
pragma 129 RB 70
pragma, defined 379 realloc 216,294
Predefined Macro Constants 132 REALLOC.C 154
_ _C51_ _ 132 Real-Time Function Tasks 128
_ _DATE_ _ 132 Recursive Code, tips for 369
_ _FILE_ _ 132 Recursive Functions 124
_ _LINE_ _ 132 reentrant 124
Keil Software — Cx51 Compiler User’s Guide 391

Reentrant Functions 124 small 116

REGFILE 69 SMALL 77
Register bank 26,70 Small memory model 77
Register Bank 117,119 Small Model 89
Register banks 26 source file, defined 380
Register Usage 166 Special Function Register (SFR),
Register Variables 156 defined 379
REGISTERBANK 70 Special Function Register
Registers used for parameters 71 Memory 89
Registers used for return values 115 Special Function Registers 96
REGPARMS 71 sprintf 217,302
relocatable, defined 379 sprintf517 302
RESTORE 76 sqrt 215,304
RET_PSTK 73 function timing 137
RET_XSTK 73 sqrt517 304
Return values 115 function timing 137
RF 69 srand 215,305
ROM 75 SRC 78
Routines by Category 212 sscanf 217,306
RTX51 Full, defined 379 sscanf517 306
RTX51 Tiny, defined 379 ST 79
Rules for interrupt functions 123 Stack 114
Stack usage 73
S stack, defined 380
Standard Types 207
sans serif typeface, use of 5 jmp_buf 207
SAVE 76 va_list 207
SB 80 START751.A51 148
sbit 97 STARTUP.A51 144
scalar types, defined 379 static, defined 380
scanf 217,295 STDARG.H 225
scanf517 295 STDDEF.H 225
scope, defined 379 STDIO.H 225
Segment Naming Conventions 158 STDLIB.H 226
Serial Port, initializing for Storage format
stream I/O 217 bit 174
setjmp 220,299 char 175
SETJMP.H 225 code pointer 175
sfr 96 data pointer 175
sfr16 97 enum 175
SIECO-51 356 far pointer 176
sin 215,300 float 177
function timing 137 generic pointer 176
sin517 300 idata pointer 175
function timing 137 int 175
sinh 215,301 long 175
Size of data types 92 pdata pointer 175
SM 77
392 Index

short 175 strncmp 219,315

xdata pointer 175 strncpy 219,316
Store return addresses 73 strpbrk 219,317
strcat 219,308 strpos 219,318
strchr 219,309 strrchr 219,319
strcmp 219,310 strrpbrk 219,320
strcpy 219,311 strrpos 219,321
strcspn 219,312 strspn 219,322
stream functions, defined 380 strtod 214,323
Stream I/O Routines 217 strtod517 323
_getkey 217 strtol 214,325
getchar 217 strtoul 214,327
gets 217 structure member, defined 380
Initializing 217 structure, defined 380
printf 217 Symbol table generation 80
putchar 217 SYMBOLS 80
puts 217 Syntax and Semantic Errors 189
scanf 217
sprintf 217 T
sscanf 217
ungetchar 217 tan 215,329
vprintf 217 function timing 137
vsprintf 217 tan517 329
Stream Input and Output 217 function timing 137
STRING 79 tanh 215,330
string literal, defined 380 Temic dual DPTR 55
String Manipulation Routines 219 Timing operation execution 136
strcat 219 TMP 19
strchr 219 toascii 213,332
strcmp 219 toint 213,333
strcpy 219 token, defined 380
strcspn 219 Token-pasting operator 131
strlen 219 tolower 213,334
strncat 219 toupper 213,336
strncmp 219 two’s complement, defined 380
strncpy 219 type cast, defined 381
strpbrk 219 type, defined 380
strpos 219
strrchr 219 U
strrpbrk 219
Uncalled Functions, tips for 371
strrpos 219
undef 129
strspn 219
ungetchar 217,338
string, defined 380
using 117,122
STRING.H 226
Using Monitor-51, tips for 373
Stringize Operator 130
strlen 219,313
strncat 219,314
Keil Software — Cx51 Compiler User’s Guide 393

V W
va_arg 220,339 Warning detection 82
va_end 220,341 WARNINGLEVEL 82
va_list 207 Warnings 201
va_start 220,342 WATCHDOG 150
VARBANKING 81,353 whitespace character, defined 381
Variable-length argument list wild card, defined 381
routines 220 WL 82
Variable-Length Argument List
Routines X
va_arg 220
va_end 220 XBPSTACK 145
va_start 220 XBPSTACKTOP 145
Variable-length argument lists 50 XBYTE 180,211
Variables, notational xdata 88
conventions 5 XDATALEN 144
VB 81,353 XDATASTART 144
vertical bar, use of 5 XOFF 153
vprintf 217,343 XON 153
vsprintf 217,345 XWORD 180,211