avr-libc-user-manual.pdf.zip

AVRSiDE i co dalej? Skąd pobrać dodatkowe biblioteki?

Większość wymienionych procedur przez ciebie znajduje się już AVR-GCC. W załaczniku masz pełny opis bibliotek zajdujących się w AVR-GCC - po angielsku. Co do wyboru srodowiska to jest to kwestja przyzwyczajenia i upodobania. AVRSIDE ma tą zaletę że jest po polsku. Możesz sprawdzić jeszcze WinAVR http://winavr.sourceforge.net Jeśli chodzi o STK-200 to jest to standard i tylko od ciebie zależy jakiego softu będziesz używał do programowania procka - możesz BASKOM-a, AVR ISP, yaap, PonyProg itp. lub po odpowiednim zkonfigurowaniu avrdude z AVR-GCC. Zajrzyj również na strony: http://www.freepgs.com/robkry/ ; http://www.mikrokontrolery.of.pl/ i oczywiście http://www.avrfreaks.net

  • avr-libc-user-manual.pdf.zip
    • avr-libc-user-manual.pdf


Pobierz plik - link do postu

avr-libc-user-manual.pdf.zip > avr-libc-user-manual.pdf

avr-libc Reference Manual
1.0.2

Generated by Doxygen 1.2.18
Wed Nov 26 21:10:21 2003

CONTENTS

i

Contents
1

AVR Libc
1.0.1

2

1
Supported Devices . . . . . . . . . . . . . . . . . . . . . . .

2
3

2.1
3

avr-libc Module Index

3

avr-libc Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4

3.1
4

avr-libc Data Structure Index

4

avr-libc Data Structures . . . . . . . . . . . . . . . . . . . . . . . . .

4

4.1
5

avr-libc Page Index

4

avr-libc Related Pages . . . . . . . . . . . . . . . . . . . . . . . . . .

avr-libc Module Documentation

5

5.1

Bootloader Support Utilities . . . . . . . . . . . . . . . . . . . . . .

5

5.1.1

Detailed Description . . . . . . . . . . . . . . . . . . . . . .

5

5.1.2

Define Documentation . . . . . . . . . . . . . . . . . . . . .

7

CRC Computations . . . . . . . . . . . . . . . . . . . . . . . . . . .

8

5.2.1

Detailed Description . . . . . . . . . . . . . . . . . . . . . .

8

5.2.2

Function Documentation . . . . . . . . . . . . . . . . . . . .

9

EEPROM handling . . . . . . . . . . . . . . . . . . . . . . . . . . .

10

5.3.1

Detailed Description . . . . . . . . . . . . . . . . . . . . . .

10

5.3.2

Define Documentation . . . . . . . . . . . . . . . . . . . . .

11

5.3.3

Function Documentation . . . . . . . . . . . . . . . . . . . .

12

5.4

AVR device-specific IO definitions . . . . . . . . . . . . . . . . . . .

12

5.5

Program Space String Utilities . . . . . . . . . . . . . . . . . . . . .

13

5.5.1

Detailed Description . . . . . . . . . . . . . . . . . . . . . .

13

5.5.2

Define Documentation . . . . . . . . . . . . . . . . . . . . .

15

5.5.3

Function Documentation . . . . . . . . . . . . . . . . . . . .

16

5.6

Additional notes from & lt; avr/sfr defs.h & gt; . . . . . . . . . . . . . . . .

19

5.7

Power Management and Sleep Modes . . . . . . . . . . . . . . . . .

21

5.7.1

21

5.2

5.3

Detailed Description . . . . . . . . . . . . . . . . . . . . . .

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

CONTENTS

ii

5.7.2

Define Documentation . . . . . . . . . . . . . . . . . . . . .

21

5.7.3

Function Documentation . . . . . . . . . . . . . . . . . . . .

22

Watchdog timer handling . . . . . . . . . . . . . . . . . . . . . . . .

22

5.8.1

Detailed Description . . . . . . . . . . . . . . . . . . . . . .

22

5.8.2

Define Documentation . . . . . . . . . . . . . . . . . . . . .

23

Character Operations . . . . . . . . . . . . . . . . . . . . . . . . . .

24

5.9.1

Detailed Description . . . . . . . . . . . . . . . . . . . . . .

24

5.9.2

Function Documentation . . . . . . . . . . . . . . . . . . . .

25

5.10 System Errors (errno) . . . . . . . . . . . . . . . . . . . . . . . . . .

27

5.10.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . .

27

5.10.2 Define Documentation . . . . . . . . . . . . . . . . . . . . .

27

5.11 Integer Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

28

5.11.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . .

28

5.11.2 Typedef Documentation . . . . . . . . . . . . . . . . . . . .

29

5.12 Mathematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

30

5.12.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . .

30

5.12.2 Define Documentation . . . . . . . . . . . . . . . . . . . . .

31

5.12.3 Function Documentation . . . . . . . . . . . . . . . . . . . .

31

5.13 Setjmp and Longjmp . . . . . . . . . . . . . . . . . . . . . . . . . .

34

5.13.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . .

34

5.13.2 Function Documentation . . . . . . . . . . . . . . . . . . . .

35

5.14 Standard IO facilities . . . . . . . . . . . . . . . . . . . . . . . . . .

36

5.14.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . .

36

5.14.2 Define Documentation . . . . . . . . . . . . . . . . . . . . .

39

5.14.3 Function Documentation . . . . . . . . . . . . . . . . . . . .

41

5.15 General utilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

50

5.15.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . .

50

5.15.2 Define Documentation . . . . . . . . . . . . . . . . . . . . .

52

5.15.3 Typedef Documentation . . . . . . . . . . . . . . . . . . . .

53

5.15.4 Function Documentation . . . . . . . . . . . . . . . . . . . .

53

5.15.5 Variable Documentation . . . . . . . . . . . . . . . . . . . .

60

5.8

5.9

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

CONTENTS

iii

5.16 Strings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.16.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . .

61

5.16.2 Function Documentation . . . . . . . . . . . . . . . . . . . .

62

5.17 Interrupts and Signals . . . . . . . . . . . . . . . . . . . . . . . . . .

69

5.17.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . .

69

5.17.2 Define Documentation . . . . . . . . . . . . . . . . . . . . .

72

5.17.3 Function Documentation . . . . . . . . . . . . . . . . . . . .

73

5.18 Special function registers . . . . . . . . . . . . . . . . . . . . . . . .

74

5.18.1 Detailed Description . . . . . . . . . . . . . . . . . . . . . .

74

5.18.2 Define Documentation . . . . . . . . . . . . . . . . . . . . .
6

61

76

avr-libc Data Structure Documentation

79

6.1

div t Struct Reference . . . . . . . . . . . . . . . . . . . . . . . . . .

79

6.1.1

Detailed Description . . . . . . . . . . . . . . . . . . . . . .

79

ldiv t Struct Reference . . . . . . . . . . . . . . . . . . . . . . . . .

79

6.2.1

79

6.2

7

Detailed Description . . . . . . . . . . . . . . . . . . . . . .

avr-libc Page Documentation

79

7.1

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . .

79

7.2

avr-libc and assembler programs . . . . . . . . . . . . . . . . . . . .

80

7.2.1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . .

80

7.2.2

Invoking the compiler . . . . . . . . . . . . . . . . . . . . .

81

7.2.3

Example program . . . . . . . . . . . . . . . . . . . . . . . .

81

Frequently Asked Questions . . . . . . . . . . . . . . . . . . . . . .

84

7.3.1

FAQ Index . . . . . . . . . . . . . . . . . . . . . . . . . . .

84

7.3.2

My program doesn't recognize a variable updated within an
interrupt routine . . . . . . . . . . . . . . . . . . . . . . . .

85

7.3.3

I get "undefined reference to..." for functions like "sin()" . . .

86

7.3.4

How to permanently bind a variable to a register? . . . . . . .

86

7.3.5

How to modify MCUCR or WDTCR early? . . . . . . . . . .

86

7.3.6

What is all this BV() stuff about? . . . . . . . . . . . . . . .

87

7.3.7

Can I use C++ on the AVR? . . . . . . . . . . . . . . . . . .

88

7.3

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

CONTENTS

iv

7.3.8

Shouldn't I initialize all my variables? . . . . . . . . . . . . .

89

7.3.9

Why do some 16-bit timer registers sometimes get trashed? .

89

7.3.10 How do I use a #define'd constant in an asm statement? . . . .

90

7.3.11 Why does the PC randomly jump around when single-stepping
through my program in avr-gdb? . . . . . . . . . . . . . . . .

91

7.3.12 How do I trace an assembler file in avr-gdb? . . . . . . . . . .

91

7.3.13 How do I pass an IO port as a parameter to a function? . . . .

93

7.3.14 What registers are used by the C compiler? . . . . . . . . . .

95

7.3.15 How do I put an array of strings completely in ROM? . . . . .

96

7.3.16 How to use external RAM? . . . . . . . . . . . . . . . . . . .

98

7.3.17 Which -O flag to use? . . . . . . . . . . . . . . . . . . . . .

98

7.3.18 How do I relocate code to a fixed address? . . . . . . . . . . .

99

7.3.19 My UART is generating nonsense! My ATmega128 keeps
crashing! Port F is completely broken! . . . . . . . . . . . . . 100
7.3.20 Why do all my "foo...bar" strings eat up the SRAM? . . . . .
7.3.21 Why does the compiler compile an 8-bit operation that uses
bitwise operators into a 16-bit operation in assembly? . . . . .

102

Inline Asm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

102

7.4.1

GCC asm Statement . . . . . . . . . . . . . . . . . . . . . .

103

7.4.2

Assembler Code . . . . . . . . . . . . . . . . . . . . . . . .

105

7.4.3

Input and Output Operands . . . . . . . . . . . . . . . . . . .

105

7.4.4

Clobbers . . . . . . . . . . . . . . . . . . . . . . . . . . . .

109

7.4.5

Assembler Macros . . . . . . . . . . . . . . . . . . . . . . .

111

7.4.6

C Stub Functions . . . . . . . . . . . . . . . . . . . . . . . .

112

7.4.7

C Names Used in Assembler Code . . . . . . . . . . . . . . .

113

7.4.8

Links . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

114

Using malloc() . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

114

7.5.1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . .

114

7.5.2

Internal vs. external RAM . . . . . . . . . . . . . . . . . . .

115

7.5.3

Tunables for malloc() . . . . . . . . . . . . . . . . . . . . . .

115

7.5.4

7.5

101

7.3.22 How to detect RAM memory and variable overlap problems? .
7.4

100

Implementation details . . . . . . . . . . . . . . . . . . . . .

117

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

CONTENTS

7.6

v

Release Numbering and Methodology . . . . . . . . . . . . . . . . .

118

7.6.1

Release Version Numbering Scheme . . . . . . . . . . . . . . 118

7.6.2

Releasing AVR Libc . . . . . . . . . . . . . . . . . . . . . .

119

Memory Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . .

121

7.7.1

The .text Section . . . . . . . . . . . . . . . . . . . . . . . .

122

7.7.2

The .data Section . . . . . . . . . . . . . . . . . . . . . . . .

122

7.7.3

The .bss Section . . . . . . . . . . . . . . . . . . . . . . . .

122

7.7.4

The .eeprom Section . . . . . . . . . . . . . . . . . . . . . .

122

7.7.5

The .noinit Section . . . . . . . . . . . . . . . . . . . . . . .

122

7.7.6

The .initN Sections . . . . . . . . . . . . . . . . . . . . . . .

123

7.7.7

The .finiN Sections . . . . . . . . . . . . . . . . . . . . . . .

124

7.7.8

Using Sections in Assembler Code . . . . . . . . . . . . . . .

125

7.7.9

Using Sections in C Code . . . . . . . . . . . . . . . . . . .

125

Installing the GNU Tool Chain . . . . . . . . . . . . . . . . . . . . .

126

7.8.1

Required Tools . . . . . . . . . . . . . . . . . . . . . . . . .

127

7.8.2

Optional Tools . . . . . . . . . . . . . . . . . . . . . . . . .

127

7.8.3

GNU Binutils for the AVR target . . . . . . . . . . . . . . . .

128

7.8.4

GCC for the AVR target . . . . . . . . . . . . . . . . . . . .

129

7.8.5

AVR Libc . . . . . . . . . . . . . . . . . . . . . . . . . . . .

130

7.8.6

UISP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

131

7.8.7

Avrdude . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

131

7.8.8

GDB for the AVR target . . . . . . . . . . . . . . . . . . . .

131

7.8.9

Simulavr . . . . . . . . . . . . . . . . . . . . . . . . . . . .

132

7.8.10 AVaRice . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

132

Using the avrdude program . . . . . . . . . . . . . . . . . . . . . . .

133

7.10 Using the GNU tools . . . . . . . . . . . . . . . . . . . . . . . . . .

134

7.10.1 Options for the C compiler avr-gcc . . . . . . . . . . . . . . .

135

7.10.2 Options for the assembler avr-as . . . . . . . . . . . . . . . .

139

7.10.3 Controlling the linker avr-ld . . . . . . . . . . . . . . . . . .

140

7.11 A simple project . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

142

7.11.1 The Project . . . . . . . . . . . . . . . . . . . . . . . . . . .

143

7.7

7.8

7.9

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

1 AVR Libc

1

7.11.2 The Source Code . . . . . . . . . . . . . . . . . . . . . . . .
7.11.3 Compiling and Linking . . . . . . . . . . . . . . . . . . . . .

146

7.11.4 Examining the Object File . . . . . . . . . . . . . . . . . . .

146

7.11.5 Linker Map Files . . . . . . . . . . . . . . . . . . . . . . . .

150

7.11.6 Intel Hex Files . . . . . . . . . . . . . . . . . . . . . . . . .

152

7.11.7 Make Build the Project . . . . . . . . . . . . . . . . . . . . .

152

7.12 Example using the two-wire interface (TWI) . . . . . . . . . . . . . .

154

7.12.1 Introduction into TWI . . . . . . . . . . . . . . . . . . . . .

154

7.12.2 The TWI example project . . . . . . . . . . . . . . . . . . .

155

7.12.3 The Source Code . . . . . . . . . . . . . . . . . . . . . . . .

155

7.13 Todo List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

167

7.14 Deprecated List . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

144

167

AVR Libc

The latest version of this document is always available
http://savannah.nongnu.org/projects/avr-libc/

from

The AVR Libc package provides a subset of the standard C library for Atmel AVR 8-bit
RISC microcontrollers. In addition, the library provides the basic startup code needed
by most applications.
There is a wealth of information in this document which goes beyond simply describing the interfaces and routines provided by the library. We hope that this document
provides enough information to get a new AVR developer up to speed quickly using
the freely available development tools: binutils, gcc avr-libc and many others.
If you find yourself stuck on a problem which this document doesn't quite address, you
may wish to post a message to the avr-gcc mailing list. Most of the developers of the
AVR binutils and gcc ports in addition to the devleopers of avr-libc subscribe to the
list, so you will usually be able to get your problem resolved. You can subscribe to
the list at http://www.avr1.org/mailman/listinfo/avr-gcc-list/.
Before posting to the list, you might want to try reading the Frequently Asked Questions chapter of this document.
Note:
This document is a work in progress.
As such, it may contain incorrect information.
If you find a mistake, please send an email to
avr-libc-dev@nongnu.org describing the mistake. Also, send us an email
if you find that a specific topic is missing from the document.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

1 AVR Libc

1.0.1

Supported Devices

The following is a list of AVR devices currently supported by the library.
AT90S Type Devices:
o at90s1200 [1]
o at90s2313
o at90s2323
o at90s2333
o at90s2343
o at90s4414
o at90s4433
o at90s4434
o at90s8515
o at90c8534
o at90s8535
ATmega Type Devices:
o atmega8
o atmega103
o atmega128
o atmega16
o atmega161
o atmega162
o atmega163
o atmega169
o atmega32
o atmega323
o atmega64 [untested]
o atmega8515 [untested]
o atmega8535 [untested]
ATtiny Type Devices:
o attiny11 [1]
o attiny12 [1]
o attiny15 [1]
o attiny22
o attiny26
o attiny28 [1]
Misc Devices:
o at94K [2]
o at76c711 [3]
o at43usb320
o at43usb355
o at86rf401

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

2

2 avr-libc Module Index

3

Note:
[1] Assembly only. There is no support for these devices to be programmed in C
since they do not have a ram based stack.

Note:
[2] The at94K devices are a combination of FPGA and AVR microcontroller.
[TRoth-2002/11/12: Not sure of the level of support for these. More information
would be welcomed.]

Note:
[3] The at76c711 is a USB to fast serial interface bridge chip using an AVR core.

2

avr-libc Module Index

2.1

avr-libc Modules

Here is a list of all modules:
Bootloader Support Utilities

5

CRC Computations

8

EEPROM handling

10

AVR device-specific IO definitions

12

Program Space String Utilities

13

Power Management and Sleep Modes

21

Watchdog timer handling

22

Character Operations

24

System Errors (errno)

27

Integer Types

28

Mathematics

30

Setjmp and Longjmp

34

Standard IO facilities

36

General utilities

50

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

3 avr-libc Data Structure Index

4

Strings

61

Interrupts and Signals

69

Special function registers

74

Additional notes from & lt; avr/sfr defs.h & gt;

3

19

avr-libc Data Structure Index

3.1

avr-libc Data Structures

Here are the data structures with brief descriptions:
div t

79

ldiv t

79

4

avr-libc Page Index

4.1

avr-libc Related Pages

Here is a list of all related documentation pages:
Acknowledgments

79

avr-libc and assembler programs

80

Frequently Asked Questions

84

Inline Asm

102

Using malloc()

114

Release Numbering and Methodology

118

Memory Sections

121

Installing the GNU Tool Chain

126

Using the avrdude program

133

Using the GNU tools

134

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5 avr-libc Module Documentation

5

A simple project

142

Example using the two-wire interface (TWI)

154

Todo List

167

Deprecated List

167

5

avr-libc Module Documentation

5.1

Bootloader Support Utilities

5.1.1

Detailed Description

#include & lt; avr/io.h & gt;
#include & lt; avr/boot.h & gt;

The macros in this module provide a C language interface to the bootloader support
functionality of certain AVR processors. These macros are designed to work with all
sizes of flash memory.
Note:
Not all AVR processors provide bootloader support. See your processor datasheet
to see if it provides bootloader support.
Todo:
From email with Marek: On smaller devices (all except ATmega64/128), SPM REG is in the I/O space, accessible with the shorter "in" and "out" instructions since the boot loader has a limited size, this could be an important optimization.
API Usage Example
The following code shows typical usage of the boot API.
#include & lt; avr/interrupt.h & gt;
#include & lt; avr/pgmspace.h & gt;
#define ADDRESS

0x1C000UL

void boot_test(void)
{
unsigned char buffer[8];
cli();
// Erase page.
boot_page_erase((unsigned long)ADDRESS);

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.1

Bootloader Support Utilities

while(boot_rww_busy())
{
boot_rww_enable();
}
// Write data to buffer a word at a time. Note incrementing address
// by 2. SPM_PAGESIZE is defined in the microprocessor IO header file.
for(unsigned long i = ADDRESS; i & lt; ADDRESS + SPM_PAGESIZE; i += 2)
{
boot_page_fill(i, (i-ADDRESS) + ((i-ADDRESS+1) & lt; & lt; 8));
}
// Write page.
boot_page_write((unsigned long)ADDRESS);
while(boot_rww_busy())
{
boot_rww_enable();
}
sei();
// Read back the values and display.
// (The show() function is undefined and is used here as an example
// only.)
for(unsigned long i = ADDRESS; i & lt; ADDRESS + 256; i++)
{
show(utoa(pgm_read_byte(i), buffer, 16));
}
return;
}

Defines
o
o
o
o
o
o
o
o
o
o
o
o

#define BOOTLOADER SECTION attribute ((section (".bootloader")))
#define boot spm interrupt enable() ( SPM REG |= (uint8 t) BV(SPMIE))
#define boot spm interrupt disable() ( SPM REG & = (uint8 t)~ BV(SPMIE))
#define boot is spm interrupt() ( SPM REG & (uint8 t) BV(SPMIE))
#define boot rww busy() ( SPM REG & (uint8 t) BV( COMMON ASB))
#define boot spm busy() ( SPM REG & (uint8 t) BV(SPMEN))
#define boot spm busy wait() do{}while(boot spm busy())
#define boot page fill(address, data) boot page fill normal(address, data)
#define boot page erase(address) boot page erase normal(address)
#define boot page write(address) boot page write normal(address)
#define boot rww enable() boot rww enable()
#define boot lock bits set(lock bits) boot lock bits set(lock bits)

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

6

5.1

Bootloader Support Utilities

5.1.2

7

Define Documentation

5.1.2.1 #define boot is spm interrupt() ( SPM REG & (uint8 t) BV(SPMIE))
Check if the SPM interrupt is enabled.
5.1.2.2 #define boot lock bits set(lock bits) boot lock bits set(lock bits)
Set the bootloader lock bits.
5.1.2.3

#define boot page erase(address) boot page erase normal(address)

Erase the flash page that contains address.
Note:
address is a byte address in flash, not a word address.

5.1.2.4
data)

#define boot page fill(address, data)

boot page fill normal(address,

Fill the bootloader temporary page buffer for flash address with data word.
Note:
The address is a byte address. The data is a word. The AVR writes data to the
buffer a word at a time, but addresses the buffer per byte! So, increment your
address by 2 between calls, and send 2 data bytes in a word format! The LSB of
the data is written to the lower address; the MSB of the data is written to the higher
address.

5.1.2.5

#define boot page write(address) boot page write normal(address)

Write the bootloader temporary page buffer to flash page that contains address.
Note:
address is a byte address in flash, not a word address.

5.1.2.6
ASB))

#define boot rww busy() ( SPM REG & (uint8 t) BV( COMMON -

Check if the RWW section is busy.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.2

CRC Computations

8

5.1.2.7 #define boot rww enable() boot rww enable()
Enable the Read-While-Write memory section.
5.1.2.8

#define boot spm busy() ( SPM REG & (uint8 t) BV(SPMEN))

Check if the SPM instruction is busy.
5.1.2.9

#define boot spm busy wait() do{}while(boot spm busy())

Wait while the SPM instruction is busy.
5.1.2.10 #define boot spm interrupt disable() ( SPM REG & = (uint8 t)~ BV(SPMIE))
Disable the SPM interrupt.
5.1.2.11 #define
BV(SPMIE))

boot spm interrupt enable()

( SPM REG

|=

(uint8 t) -

Enable the SPM interrupt.
5.1.2.12 #define BOOTLOADER SECTION
loader")))

attribute

((section (".boot-

Used to declare a function or variable to be placed into a new section called .bootloader. This section and its contents can then be relocated to any address (such as the
bootloader NRWW area) at link-time.

5.2
5.2.1

CRC Computations
Detailed Description

#include & lt; avr/crc16.h & gt;

This header file provides a optimized inline functions for calculating 16 bit cyclic redundancy checks (CRC) using common polynomials.
References:
See the Dallas Semiconductor app note 27 for 8051 assembler example and general
CRC optimization suggestions. The table on the last page of the app note is the
key to understanding these implementations.
Jack Crenshaw's "Impementing CRCs" article in the January 1992 isue of Embedded Systems Programming. This may be difficult to find, but it explains CRC's in
very clear and concise terms. Well worth the effort to obtain a copy.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.2

CRC Computations

Functions
o uint16 t crc16 update (uint16 t crc, uint8 t data)
o uint16 t crc xmodem update (uint16 t crc, uint8 t data)
o uint16 t crc ccitt update (uint16 t crc, uint8 t data)
5.2.2

Function Documentation

5.2.2.1 uint16 t crc16 update (uint16 t crc, uint8 t data) [static]
Optimized CRC-16 calcutation.
Polynomial: x? 16 + x? 15 + x? 2 + 1 (0xa001)
Initial value: 0xffff
This CRC is normally used in disk-drive controllers.
5.2.2.2 uint16 t crc ccitt update (uint16 t crc, uint8 t data) [static]
Optimized CRC-CCITT calculation.
Polynomial: x? 16 + x? 12 + x? 5 + 1 (0x8408)
Initial value: 0xffff
This is the CRC used by PPP and IrDA.
See RFC1171 (PPP protocol) and IrDA IrLAP 1.1
Note:
Although the CCITT polynomial is the same as that used by the Xmodem protocol,
they are quite different. The difference is in how the bits are shifted through the
alorgithm. Xmodem shifts the MSB of the CRC and the input first, while CCITT
shifts the LSB of the CRC and the input first.
The following is the equivalent functionality written in C.
uint16_t
crc_ccitt_update (uint16_t crc, uint8_t data)
{
data ^= lo8 (crc);
data ^= data & lt; & lt; 4;
return ((((uint16_t)data & lt; & lt; 8) | hi8 (crc)) ^ (uint8_t)(data & gt; & gt; 4)
^ ((uint16_t)data & lt; & lt; 3));
}

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

9

5.3

EEPROM handling

5.2.2.3 uint16 t
[static]

crc xmodem update

10

(uint16 t

crc,

uint8 t

data)

Optimized CRC-XMODEM calculation.
Polynomial: x? 16 + x? 12 + x? 5 + 1 (0x1021)
Initial value: 0x0
This is the CRC used by the Xmodem-CRC protocol.
The following is the equivalent functionality written in C.
uint16_t
crc_xmodem_update (uint16_t crc, uint8_t data)
{
int i;
crc = crc ^ ((uint16_t)data & lt; & lt; 8);
for (i=0; i & lt; 8; i++)
{
if (crc & 0x8000)
crc = (crc & lt; & lt; 1) ^ 0x1021;
else
crc & lt; & lt; = 1;
}
return crc;
}

5.3
5.3.1

EEPROM handling
Detailed Description

#include & lt; avr/eeprom.h & gt;

This header file declares the interface to some simple library routines suitable for handling the data EEPROM contained in the AVR microcontrollers. The implementation
uses a simple polled mode interface. Applications that require interrupt-controlled
EEPROM access to ensure that no time will be wasted in spinloops will have to deploy
their own implementation.
Note:
All of the read/write functions first make sure the EEPROM is ready to be accessed. Since this may cause long delays if a write operation is still pending, timecritical applications should first poll the EEPROM e. g. using eeprom is ready()
before attempting any actual I/O.
Note:
This library will not work with the ATmega169 since this device has the EEPROM
IO ports at different locations!

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.3

EEPROM handling

avr-libc declarations
o
o
o
o
o
o
o

#define eeprom is ready() bit is clear(EECR, EEWE)
uint8 t eeprom read byte (const uint8 t *addr)
uint16 t eeprom read word (const uint16 t *addr)
void eeprom read block (void *buf, const void *addr, size t n)
void eeprom write byte (uint8 t *addr, uint8 t val)
void eeprom write word (uint16 t *addr, uint16 t val)
void eeprom write block (const void *buf, void *addr, size t n)

Backwards compatibility defines
o #define eeprom rb(addr) eeprom read byte ((uint8 t *)(addr))
o #define eeprom rw(addr) eeprom read word ((uint16 t *)(addr))
o #define eeprom wb(addr, val) eeprom write byte ((uint8 t *)(addr), (uint8 t)(val))
IAR C compatibility defines
o #define EEPUT(addr, val) eeprom wb(addr, val)
o #define EEGET(var, addr) (var) = eeprom rb(addr)
5.3.2
5.3.2.1

Define Documentation
#define EEGET(var, addr) (var) = eeprom rb(addr)

Read a byte from EEPROM.
5.3.2.2

#define EEPUT(addr, val) eeprom wb(addr, val)

Write a byte to EEPROM.
5.3.2.3

#define eeprom is ready() bit is clear(EECR, EEWE)

Returns:
1 if EEPROM is ready for a new read/write operation, 0 if not.

5.3.2.4 #define eeprom rb(addr) eeprom read byte ((uint8 t *)(addr))
Deprecated:
Use eeprom read byte() in new programs.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

11

5.4

AVR device-specific IO definitions

5.3.2.5 #define eeprom rw(addr) eeprom read word ((uint16 t *)(addr))
Deprecated:
Use eeprom read word() in new programs.

5.3.2.6 #define eeprom wb(addr, val) eeprom write byte ((uint8 t *)(addr),
(uint8 t)(val))
Deprecated:
Use eeprom write byte() in new programs.
5.3.3

Function Documentation

5.3.3.1

void eeprom read block (void * buf, const void * addr, size t n)

Read a block of n bytes from EEPROM address addr to buf.
5.3.3.2 uint8 t eeprom read byte (const uint8 t * addr)
Read one byte from EEPROM address addr.
5.3.3.3 uint16 t eeprom read word (const uint16 t * addr)
Read one 16-bit word (little endian) from EEPROM address addr.
5.3.3.4

void eeprom write block (const void * buf, void * addr, size t n)

Write a block of n bytes to EEPROM address addr from buf.
5.3.3.5 void eeprom write byte (uint8 t * addr, uint8 t val)
Write a byte val to EEPROM address addr.
5.3.3.6

void eeprom write word (uint16 t * addr, uint16 t val)

Write a word val to EEPROM address addr.

5.4

AVR device-specific IO definitions

#include & lt; avr/io.h & gt;

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

12

5.5

Program Space String Utilities

This header file includes the apropriate IO definitions for the device that has been specified by the -mmcu= compiler command-line switch. This is done by diverting to the
appropriate file & lt; avr/ioXXXX.h & gt; which should never be included directly. Some
register names common to all AVR devices are defined directly within & lt; avr/io.h & gt; ,
but most of the details come from the respective include file.
Note that this file always includes
#include & lt; avr/sfr_defs.h & gt;

See Special function registers for the details.
Included are definitions of the IO register set and their respective bit values as specified
in the Atmel documentation. Note that Atmel is not very consistent in its naming
conventions, so even identical functions sometimes get different names on different
devices.
Also included are the specific names useable for interrupt function definitions as documented here.
Finally, the following macros are defined:
o RAMEND
A constant describing the last on-chip RAM location.
o XRAMEND
A constant describing the last possible location in RAM. This is equal to RAMEND for devices that do not allow for external RAM.
o E2END
A constant describing the address of the last EEPROM cell.
o FLASHEND
A constant describing the last byte address in flash ROM.
o SPM PAGESIZE
For devices with bootloader support, the flash pagesize (in bytes) to be used for
the SPM instruction.

5.5

Program Space String Utilities

5.5.1

Detailed Description

#include & lt; avr/io.h & gt;
#include & lt; avr/pgmspace.h & gt;

The functions in this module provide interfaces for a program to access data stored in
program space (flash memory) of the device. In order to use these functions, the target
device must support either the LPM or ELPM instructions.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

13

5.5

Program Space String Utilities

Note:
These functions are an attempt to provide some compatibility with header files
that come with IAR C, to make porting applications between different compilers
easier. This is not 100% compatibility though (GCC does not have full support for
multiple address spaces yet).
Note:
If you are working with strings which are completely based in ram, use the standard string functions described in Strings.
Note:
If possible, put your constant tables in the lower 64K and use pgm read byte near() or pgm read word near() instead of pgm read byte far() or pgm read word far() since it is more efficient that way, and you can still use the upper 64K
for executable code.
Backwards compatibility macros
o #define PRG RDB(addr) pgm read byte(addr)
Defines
o #define PSTR(s) ({static char c[ ] PROGMEM = (s); c;})
o #define pgm read byte near(address short) LPM((uint16 t)(address short))
o #define pgm read word near(address short) LPM word((uint16 t)(address short))
o #define pgm read byte far(address long) ELPM((uint32 t)(address long))
o #define pgm read word far(address long) ELPM word((uint32 t)(address long))
o #define pgm read byte(address short) pgm read byte near(address short)
o #define pgm read word(address short) pgm read word near(address short)
o #define PGM P const prog char *
o #define PGM VOID P const prog void *
Functions
o
o
o
o
o
o
o

void * memcpy P (void *, PGM VOID P, size t)
int strcasecmp P (const char *, PGM P) ATTR PURE
char * strcat P (char *, PGM P)
int strcmp P (const char *, PGM P) ATTR PURE
char * strcpy P (char *, PGM P)
size t strlcat P (char *, PGM P, size t)
size t strlcpy P (char *, PGM P, size t)

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

14

5.5

Program Space String Utilities

o
o
o
o
o

15

size t strlen P (PGM P) ATTR CONST
int strncasecmp P (const char *, PGM P, size t) ATTR PURE
char * strncat P (char *, PGM P, size t)
int strncmp P (const char *, PGM P, size t) ATTR PURE
char * strncpy P (char *, PGM P, size t)

5.5.2
5.5.2.1

Define Documentation
#define PGM P const prog char *

Used to declare a variable that is a pointer to a string in program space.
5.5.2.2
short)

#define pgm read byte(address short) pgm read byte near(address -

Read a byte from the program space with a 16-bit (near) address.
Note:
The address is a byte address. The address is in the program space.

5.5.2.3
long))

#define pgm read byte far(address long)

ELPM((uint32 t)(address -

Read a byte from the program space with a 32-bit (far) address.
Note:
The address is a byte address. The address is in the program space.

5.5.2.4 #define pgm read byte near(address short)
short))

LPM((uint16 t)(address -

Read a byte from the program space with a 16-bit (near) address.
Note:
The address is a byte address. The address is in the program space.

5.5.2.5
short)

#define pgm read word(address short) pgm read word near(address -

Read a word from the program space with a 16-bit (near) address.
Note:
The address is a byte address. The address is in the program space.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.5

Program Space String Utilities

5.5.2.6 #define
t)(address long))

pgm read word far(address long)

16

ELPM word((uint32 -

Read a word from the program space with a 32-bit (far) address.
Note:
The address is a byte address. The address is in the program space.

5.5.2.7 #define pgm read word near(address short)
t)(address short))

LPM word((uint16 -

Read a word from the program space with a 16-bit (near) address.
Note:
The address is a byte address. The address is in the program space.

5.5.2.8

#define PGM VOID P const prog void *

Used to declare a generic pointer to an object in program space.
5.5.2.9

#define PRG RDB(addr) pgm read byte(addr)

Deprecated:
Use pgm read byte() in new programs.

5.5.2.10

#define PSTR(s) ({static char c[ ] PROGMEM = (s); c;})

Used to declare a static pointer to a string in program space.
5.5.3 Function Documentation
5.5.3.1

void * memcpy P (void * dest, PGM VOID P src, size t n)

The memcpy P() function is similar to memcpy(), except the src string resides in program space.
Returns:
The memcpy P() function returns a pointer to dest.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.5

Program Space String Utilities

5.5.3.2 int strcasecmp P (const char * s1, PGM P s2)
Compare two strings ignoring case.
The strcasecmp P() function compares the two strings s1 and s2, ignoring the case of
the characters.
Parameters:
s1 A pointer to a string in the devices SRAM.
s2 A pointer to a string in the devices Flash.
Returns:
The strcasecmp P() function returns an integer less than, equal to, or greater than
zero if s1 is found, respectively, to be less than, to match, or be greater than s2.

5.5.3.3

char * strcat P (char * dest, PGM P src)

The strcat P() function is similar to strcat() except that the src string must be located in
program space (flash).
Returns:
The strcat() function returns a pointer to the resulting string dest.

5.5.3.4

int strcmp P (const char * s1, PGM P s2)

The strcmp P() function is similar to strcmp() except that s2 is pointer to a string in
program space.
Returns:
The strcmp P() function returns an integer less than, equal to, or greater than zero
if s1 is found, respectively, to be less than, to match, or be greater than s2.

5.5.3.5

char * strcpy P (char * dest, PGM P src)

The strcpy P() function is similar to strcpy() except that src is a pointer to a string in
program space.
Returns:
The strcpy P() function returns a pointer to the destination string dest.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

17

5.5

Program Space String Utilities

5.5.3.6 size t strlcat P (char * dst, PGM P, size t siz)
Concatenate two strings.
The strlcat P() function is similar to strlcat(), except that the src string must be located
in program space (flash).
Appends src to string dst of size siz (unlike strncat(), siz is the full size of dst, not space
left). At most siz-1 characters will be copied. Always NULL terminates (unless siz & lt; =
strlen(dst)).
Returns:
The strlcat P() function returns strlen(src) + MIN(siz, strlen(initial dst)). If retval
& gt; = siz, truncation occurred.

5.5.3.7

size t strlcpy P (char * dst, PGM P, size t siz)

Copy a string from progmem to RAM.
Copy src to string dst of size siz. At most siz-1 characters will be copied. Always
NULL terminates (unless siz == 0).
Returns:
The strlcpy P() function returns strlen(src). If retval & gt; = siz, truncation occurred.

5.5.3.8

size t strlen P (PGM P src)

The strlen P() function is similar to strlen(), except that src is a pointer to a string in
program space.
Returns:
The strlen() function returns the number of characters in src.

5.5.3.9

int strncasecmp P (const char * s1, PGM P s2, size t n)

Compare two strings ignoring case.
The strncasecmp P() function is similar to strcasecmp P(), except it only compares the
first n characters of s1.
Parameters:
s1 A pointer to a string in the devices SRAM.
s2 A pointer to a string in the devices Flash.
n The maximum number of bytes to compare.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

18

5.6

Additional notes from & lt; avr/sfr defs.h & gt;

Returns:
The strcasecmp P() function returns an integer less than, equal to, or greater than
zero if s1 (or the first n bytes thereof) is found, respectively, to be less than, to
match, or be greater than s2.

5.5.3.10

char * strncat P (char * dest, PGM P src, size t len)

Concatenate two strings.
The strncat P() function is similar to strncat(), except that the src string must be located
in program space (flash).
Returns:
The strncat P() function returns a pointer to the resulting string dest.

5.5.3.11 int strncmp P (const char * s1, PGM P s2, size t n)
The strncmp P() function is similar to strcmp P() except it only compares the first (at
most) n characters of s1 and s2.
Returns:
The strncmp P() function returns an integer less than, equal to, or greater than zero
if s1 (or the first n bytes thereof) is found, respectively, to be less than, to match,
or be greater than s2.

5.5.3.12

char * strncpy P (char * dest, PGM P src, size t n)

The strncpy P() function is similar to strcpy P() except that not more than n bytes of
src are copied. Thus, if there is no null byte among the first n bytes of src, the result
will not be null-terminated.
In the case where the length of src is less than that of n, the remainder of dest will be
padded with nulls.
Returns:
The strncpy P() function returns a pointer to the destination string dest.

5.6

Additional notes from & lt; avr/sfr defs.h & gt;

The & lt; avr/sfr defs.h & gt; file is included by all of the & lt; avr/ioXXXX.h & gt; files,
which use macros defined here to make the special function register definitions look
like C variables or simple constants, depending on the SFR ASM COMPAT define.
Some examples from & lt; avr/iom128.h & gt; to show how to define such macros:
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

19

5.6

Additional notes from & lt; avr/sfr defs.h & gt;

#define
#define
#define
#define

PORTA
TCNT1
PORTF
TCNT3

_SFR_IO8(0x1b)
_SFR_IO16(0x2c)
_SFR_MEM8(0x61)
_SFR_MEM16(0x88)

If SFR ASM COMPAT is not defined, C programs can use names like PORTA directly
in C expressions (also on the left side of assignment operators) and GCC will do the
right thing (use short I/O instructions if possible). The SFR OFFSET definition is
not used in any way in this case.
Define SFR ASM COMPAT as 1 to make these names work as simple constants (addresses of the I/O registers). This is necessary when included in preprocessed assembler (*.S) source files, so it is done automatically if ASSEMBLER is defined. By
default, all addresses are defined as if they were memory addresses (used in lds/sts
instructions). To use these addresses in in/out instructions, you must subtract 0x20
from them.
For more backwards compatibility, insert the following at the start of your old assembler source file:
#define __SFR_OFFSET 0

This automatically subtracts 0x20 from I/O space addresses, but it's a hack, so it is
recommended to change your source: wrap such addresses in macros defined here, as
shown below. After this is done, the SFR OFFSET definition is no longer necessary
and can be removed.
Real example - this code could be used in a boot loader that is portable between devices
with SPMCR at different addresses.
& lt; avr/iom163.h & gt; : #define SPMCR _SFR_IO8(0x37)
& lt; avr/iom128.h & gt; : #define SPMCR _SFR_MEM8(0x68)
#if _SFR_IO_REG_P(SPMCR)
out
_SFR_IO_ADDR(SPMCR), r24
#else
sts
_SFR_MEM_ADDR(SPMCR), r24
#endif

You can use the in/out/cbi/sbi/sbic/sbis instructions, without the SFR IO REG P test, if you know that the register is in the I/O space (as with SREG, for
example). If it isn't, the assembler will complain (I/O address out of range 0...0x3f),
so this should be fairly safe.
If you do not define SFR OFFSET (so it will be 0x20 by default), all special register
addresses are defined as memory addresses (so SREG is 0x5f), and (if code size and
speed are not important, and you don't like the ugly if above) you can always use lds/sts
to access them. But, this will not work if SFR OFFSET != 0x20, so use a different
macro (defined only if SFR OFFSET == 0x20) for safety:
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

20

5.7

Power Management and Sleep Modes

sts

_SFR_ADDR(SPMCR), r24

In C programs, all 3 combinations of SFR ASM COMPAT and SFR OFFSET are
supported - the SFR ADDR(SPMCR) macro can be used to get the address of the
SPMCR register (0x57 or 0x68 depending on device).
The old inp()/outp() macros are still supported, but not recommended to use in new
code. The order of outp() arguments is confusing.

5.7

Power Management and Sleep Modes

5.7.1

Detailed Description

#include & lt; avr/sleep.h & gt;

Use of the SLEEP instruction can allow your application to reduce it's power comsumption considerably. AVR devices can be put into different sleep modes by changing the SMn bits of the MCU Control Register ( MCUCR ). Refer to the datasheet for the
details relating to the device you are using.
Sleep Modes
Note:
Some of these modes are not available on all devices. See the datasheet for target
device for the available sleep modes.
o
o
o
o
o
o

#define SLEEP
#define SLEEP
#define SLEEP
#define SLEEP
#define SLEEP
#define SLEEP

MODE
MODE
MODE
MODE
MODE
MODE

IDLE 0
ADC BV(SM0)
PWR DOWN BV(SM1)
PWR SAVE ( BV(SM0) | BV(SM1))
STANDBY ( BV(SM1) | BV(SM2))
EXT STANDBY ( BV(SM0) | BV(SM1) | BV(SM2))

Sleep Functions
o void set sleep mode (uint8 t mode)
o void sleep mode (void)
5.7.2
5.7.2.1

Define Documentation
#define SLEEP MODE ADC BV(SM0)

ADC Noise Reduction Mode.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

21

5.8

Watchdog timer handling

5.7.2.2 #define SLEEP MODE EXT STANDBY ( BV(SM0) | BV(SM1) | BV(SM2))
Extended Standby Mode.
5.7.2.3

#define SLEEP MODE IDLE 0

Idle mode.
5.7.2.4

#define SLEEP MODE PWR DOWN BV(SM1)

Power Down Mode.
5.7.2.5

#define SLEEP MODE PWR SAVE ( BV(SM0) | BV(SM1))

Power Save Mode.
5.7.2.6

#define SLEEP MODE STANDBY ( BV(SM1) | BV(SM2))

Standby Mode.
5.7.3

Function Documentation

5.7.3.1

void set sleep mode (uint8 t mode)

Set the bits in the MCUCR to select a sleep mode.
5.7.3.2

void sleep mode (void)

Put the device in sleep mode. How the device is brought out of sleep mode depends on
the specific mode selected with the set sleep mode() function. See the data sheet for
your device for more details.

5.8
5.8.1

Watchdog timer handling
Detailed Description

#include & lt; avr/wdt.h & gt;

This header file declares the interface to some inline macros handling the watchdog
timer present in many AVR devices. In order to prevent the watchdog timer configuration from being accidentally altered by a crashing application, a special timed sequence
is required in order to change it. The macros within this header file handle the required
sequence automatically before changing any value. Interrupts will be disabled during
the manipulation.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

22

5.8

Watchdog timer handling

23

Note:
Depending on the fuse configuration of the particular device, further restrictions
might apply, in particular it might be disallowed to turn off the watchdog timer.
Defines
o
o
o
o
o
o
o
o
o
o
o

#define wdt reset() asm
volatile ("wdr")
#define wdt enable(timeout) wdt write((timeout) | BV(WDE))
#define wdt disable() wdt write(0)
#define WDTO 15MS 0
#define WDTO 30MS 1
#define WDTO 60MS 2
#define WDTO 120MS 3
#define WDTO 250MS 4
#define WDTO 500MS 5
#define WDTO 1S 6
#define WDTO 2S 7

5.8.2

Define Documentation

5.8.2.1 #define wdt disable() wdt write(0)
Disable the watchdog timer, if possible. This attempts to turn off the WDE bit in the
WDTCR register.
5.8.2.2

#define wdt enable(timeout) wdt write((timeout) | BV(WDE))

Enable the watchdog timer, configuring it for expiry after timeout (which is a combination of the WDP0 through WDP2 to write into the WDTCR register).
See also the symbolic constants WDTO 15MS et al.
5.8.2.3

#define wdt reset() asm

volatile ("wdr")

Reset the watchdog timer. When the watchdog timer is enabled, a call to this instruction
is required before the timer expires, otherwise a watchdog-initiated device reset will
occur.
5.8.2.4

#define WDTO 120MS 3

See WDT0 15MS

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.9

Character Operations

5.8.2.5 #define WDTO 15MS 0
Symbolic constants for the watchdog timeout. Since the watchdog timer is based on
a free-running RC oscillator, the times are approximate only and apply to a supply
voltage of 5 V. At lower supply voltages, the times will increase. For older devices,
the times will be as large as three times when operating at Vcc = 3 V, while the newer
devices (e. g. ATmega128, ATmega8) only experience a negligible change.
Possible timeout values are: 15 ms, 30 ms, 60 ms, 120 ms, 250 ms, 500 ms, 1 s, 2 s.
Symbolic constants are formed by the prefix WDTO , followed by the time.
Example that would select a watchdog timer expiry of approximately 500 ms:
wdt_enable(WDTO_500MS);

5.8.2.6

#define WDTO 1S 6

See WDT0 15MS
5.8.2.7

#define WDTO 250MS 4

See WDT0 15MS
5.8.2.8

#define WDTO 2S 7

See WDT0 15MS
5.8.2.9

#define WDTO 30MS 1

See WDT0 15MS
5.8.2.10

#define WDTO 500MS 5

See WDT0 15MS
5.8.2.11

#define WDTO 60MS 2

WDT0 15MS

5.9
5.9.1

Character Operations
Detailed Description

These functions perform various operations on characters.
#include & lt; ctype.h & gt;

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

24

5.9

Character Operations

25

Character classification routines
These functions perform character classification. They return true or false status depending whether the character passed to the function falls into the function's classification (i.e. isdigit() returns true if its argument is any value '0' though '9', inclusive.)
o
o
o
o
o
o
o
o
o
o
o
o
o

int isalnum (int
int isalpha (int
int isascii (int
int isblank (int
int iscntrl (int
int isdigit (int
int isgraph (int
int islower (int
int isprint (int
int ispunct (int
int isspace (int
int isupper (int
int isxdigit (int

c)
c)
c)
c)
c)
c)
c)
c)
c)
c)
c)
c)
c)

ATTR CONST
ATTR CONST
ATTR CONST
ATTR CONST
ATTR CONST
ATTR CONST
ATTR CONST
ATTR CONST
ATTR CONST
ATTR CONST
ATTR CONST
ATTR CONST
ATTR CONST

Character convertion routines
If c is not an unsigned char value, or EOF, the behaviour of these functions is undefined.
o int toascii (int c) ATTR CONST
o int tolower (int c) ATTR CONST
o int toupper (int c) ATTR CONST
5.9.2
5.9.2.1

Function Documentation
int isalnum (int c)

Checks for an alphanumeric character.
isdigit(c)).
5.9.2.2

int isalpha (int c)

Checks for an alphabetic character.
islower(c)).
5.9.2.3

It is equivalent to (isalpha(c) ||

It is equivalent to (isupper(c) ||

int isascii (int c)

Checks whether c is a 7-bit unsigned char value that fits into the ASCII character set.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.9

Character Operations

5.9.2.4 int isblank (int c)
Checks for a blank character, that is, a space or a tab.
5.9.2.5

int iscntrl (int c)

Checks for a control character.
5.9.2.6 int isdigit (int c)
Checks for a digit (0 through 9).
5.9.2.7

int isgraph (int c)

Checks for any printable character except space.
5.9.2.8

int islower (int c)

Checks for a lower-case character.
5.9.2.9

int isprint (int c)

Checks for any printable character including space.
5.9.2.10

int ispunct (int c)

Checks for any printable character which is not a space or an alphanumeric character.
5.9.2.11

int isspace (int c)

Checks for white-space characters. For the avr-libc library, these are: space, formfeed ('\f'), newline ('\n'), carriage return ('\r'), horizontal tab ('\t'), and vertical tab
('\v').
5.9.2.12

int isupper (int c)

Checks for an uppercase letter.
5.9.2.13

int isxdigit (int c)

Checks for a hexadecimal digits, i.e. one of 0 1 2 3 4 5 6 7 8 9 a b c d e f A B C D E F.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

26

5.10 System Errors (errno)

5.9.2.14

int toascii (int c)

Converts c to a 7-bit unsigned char value that fits into the ASCII character set, by
clearing the high-order bits.
Warning:
Many people will be unhappy if you use this function. This function will convert
accented letters into random characters.

5.9.2.15

int tolower (int c)

Converts the letter c to lower case, if possible.
5.9.2.16

int toupper (int c)

Converts the letter c to upper case, if possible.

5.10

System Errors (errno)

5.10.1

Detailed Description

#include & lt; errno.h & gt;

Some functions in the library set the global variable errno when an error occurs. The
file, & lt; errno.h & gt; , provides symbolic names for various error codes.
Warning:
The errno global variable is not safe to use in a threaded or multi-task system. A
race condition can occur if a task is interrupted between the call which sets error
and when the task examines errno. If another task changes errno during this
time, the result will be incorrect for the interrupted task.
Defines
o #define EDOM 33
o #define ERANGE 34
5.10.2
5.10.2.1

Define Documentation
#define EDOM 33

Domain error.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

27

5.11 Integer Types

5.10.2.2 #define ERANGE 34
Range error.

5.11

Integer Types

5.11.1

Detailed Description

#include & lt; inttypes.h & gt;

Use [u]intN t if you need exactly N bits.
Since these typedefs are mandated by the C99 standard, they are preferred over rolling
your own typedefs.
Note:
If avr-gcc's -mint8 option is used, no 32-bit types will be available.
Todo:
There is a pending patch that may go into gcc to change the behaviour of the mint8 option. The current (2003-09-17) situation for -mint8 is sizeof(int) == 1,
sizeof(long) == 2 and sizeof(long long) == 8. Note the absence of a 4-byte, 32-bit
type. The patch proposes to change sozeof(long long) to be 4 bytes (32 bits). When
and if the patch is included in gcc, we will need to change avr-libc accordingly.
8-bit types.
o typedef signed char int8 t
o typedef unsigned char uint8 t
16-bit types.
o typedef int int16 t
o typedef unsigned int uint16 t
32-bit types.
o typedef long int32 t
o typedef unsigned long uint32 t
64-bit types.
o typedef long long int64 t
o typedef unsigned long long uint64 t

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

28

5.11 Integer Types

Pointer types.
These allow you to declare variables of the same size as a pointer.
o typedef int16 t intptr t
o typedef uint16 t uintptr t
5.11.2

Typedef Documentation

5.11.2.1 typedef int int16 t
16-bit signed type.
5.11.2.2 typedef long int32 t
32-bit signed type.
5.11.2.3

typedef long long int64 t

64-bit signed type.
5.11.2.4

typedef signed char int8 t

8-bit signed type.
5.11.2.5

typedef int16 t intptr t

Signed pointer compatible type.
5.11.2.6 typedef unsigned int uint16 t
16-bit unsigned type.
5.11.2.7

typedef unsigned long uint32 t

32-bit unsigned type.
5.11.2.8

typedef unsigned long long uint64 t

64-bit unsigned type.
5.11.2.9

typedef unsigned char uint8 t

8-bit unsigned type.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

29

5.12 Mathematics

5.11.2.10

typedef uint16 t uintptr t

Unsigned pointer compatible type.

5.12

Mathematics

5.12.1

Detailed Description

#include & lt; math.h & gt;

This header file declares basic mathematics constants and functions.
Note:
In order to access the functions delcared herein, it is usually also required to additionally link against the library libm.a. See also the related FAQ entry.
Defines
o #define M PI 3.141592653589793238462643
o #define M SQRT2 1.4142135623730950488016887
Functions
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o

double cos (double x) ATTR CONST
double fabs (double x) ATTR CONST
double fmod (double x, double y) ATTR CONST
double modf (double value, double * iptr)
double sin (double x) ATTR CONST
double sqrt (double x) ATTR CONST
double tan (double x) ATTR CONST
double floor (double x) ATTR CONST
double ceil (double x) ATTR CONST
double frexp (double value, int * exp)
double ldexp (double x, int exp) ATTR CONST
double exp (double x) ATTR CONST
double cosh (double x) ATTR CONST
double sinh (double x) ATTR CONST
double tanh (double x) ATTR CONST
double acos (double x) ATTR CONST
double asin (double x) ATTR CONST
double atan (double x) ATTR CONST
double atan2 (double y, double x) ATTR CONST
double log (double x) ATTR CONST

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

30

5.12 Mathematics

o
o
o
o
o
o

double log10 (double x) ATTR CONST
double pow (double x, double y) ATTR CONST
int isnan (double x) ATTR CONST
int isinf (double x) ATTR CONST
double square (double x) ATTR CONST
double inverse (double) ATTR CONST

5.12.2

Define Documentation

5.12.2.1 #define M PI 3.141592653589793238462643
The constant pi.
5.12.2.2

#define M SQRT2 1.4142135623730950488016887

The square root of 2.
5.12.3
5.12.3.1

Function Documentation
double acos (double x)

The acos() function computes the principal value of the arc cosine of x. The returned
value is in the range [0, pi] radians. A domain error occurs for arguments not in the
range [-1, +1].
5.12.3.2 double asin (double x)
The asin() function computes the principal value of the arc sine of x. The returned
value is in the range [0, pi] radians. A domain error occurs for arguments not in the
range [-1, +1].
5.12.3.3

double atan (double x)

The atan() function computes the principal value of the arc tangent of x. The returned
value is in the range [0, pi] radians. A domain error occurs for arguments not in the
range [-1, +1].
5.12.3.4

double atan2 (double y, double x)

The atan2() function computes the principal value of the arc tangent of y / x, using
the signs of both arguments to determine the quadrant of the return value. The returned
value is in the range [-pi, +pi] radians. If both x and y are zero, the global variable
errno is set to EDOM.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

31

5.12 Mathematics

5.12.3.5 double ceil (double x)
The ceil() function returns the smallest integral value greater than or equal to x, expressed as a floating-point number.
5.12.3.6

double cos (double x)

The cos() function returns the cosine of x, measured in radians.
5.12.3.7

double cosh (double x)

The cosh() function returns the hyperbolic cosine of x.
5.12.3.8

double exp (double x)

The exp() function returns the exponential value of x.
5.12.3.9

double fabs (double x)

The fabs() function computes the absolute value of a floating-point number x.
5.12.3.10

double floor (double x)

The floor() function returns the largest integral value less than or equal to x, expressed
as a floating-point number.
5.12.3.11

double fmod (double x, double y)

The function fmod() returns the floating-point remainder of x / y.
5.12.3.12

double frexp (double value, int * exp)

The frexp() function breaks a floating-point number into a normalized fraction and an
integral power of 2. It stores the integer in the int object pointed to by exp.
The frexp() function returns the value x, such that x is a double with magnitude in the
interval [1/2, 1) or zero, and value equals x times 2 raised to the power *exp. If
value is zero, both parts of the result are zero.
5.12.3.13

double inverse (double)

The function inverse() returns 1 / x.
Note:
This function does not belong to the C standard definition.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

32

5.12 Mathematics

5.12.3.14

int isinf (double x)

The function isinf() returns 1 if the argument x is either positive or negative infinity,
otherwise 0.
5.12.3.15

int isnan (double x)

The function isnan() returns 1 if the argument x represents a "not-a-number" (NaN)
object, otherwise 0.
5.12.3.16

double ldexp (double x, int exp)

The ldexp() function multiplies a floating-point number by an integral power of 2.
The ldexp() function returns the value of x times 2 raised to the power exp.
If the resultant value would cause an overflow, the global variable errno is set to
ERANGE, and the value NaN is returned.
5.12.3.17

double log (double x)

The log() function returns the natural logarithm of argument x.
If the argument is less than or equal 0, a domain error will occur.
5.12.3.18

double log10 (double x)

The log() function returns the logarithm of argument x to base 10.
If the argument is less than or equal 0, a domain error will occur.
5.12.3.19

double modf (double value, double * iptr)

The modf() function breaks the argument value into integral and fractional parts,
each of which has the same sign as the argument. It stores the integral part as a double
in the object pointed to by iptr.
The modf() function returns the signed fractional part of value.
5.12.3.20

double pow (double x, double y)

The function pow() returns the value of x to the exponent y.
5.12.3.21

double sin (double x)

The sin() function returns the sine of x, measured in radians.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

33

5.13 Setjmp and Longjmp

5.12.3.22

double sinh (double x)

The sinh() function returns the hyperbolic sine of x.
5.12.3.23

double sqrt (double x)

The sqrt() function returns the non-negative square root of x.
5.12.3.24

double square (double x)

The function square() returns x * x.
Note:
This function does not belong to the C standard definition.

5.12.3.25

double tan (double x)

The tan() function returns the tangent of x, measured in radians.
5.12.3.26

double tanh (double x)

The tanh() function returns the hyperbolic tangent of x.

5.13

Setjmp and Longjmp

5.13.1

Detailed Description

While the C language has the dreaded goto statement, it can only be used to jump to
a label in the same (local) function. In order to jump directly to another (non-local)
function, the C library provides the setjmp() and longjmp() functions. setjmp() and
longjmp() are useful for dealing with errors and interrupts encountered in a low-level
subroutine of a program.
Note:
setjmp() and longjmp() make programs hard to understand and maintain. If possible, an alternative should be used.
For a very detailed discussion of setjmp()/longjmp(), see Chapter 7 of Advanced Programming in the UNIX Environment, by W. Richard Stevens.
Example:
#include & lt; setjmp.h & gt;

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

34

5.13 Setjmp and Longjmp

jmp_buf env;
int main (void)
{
if (setjmp (env))
{
... handle error ...
}
while (1)
{
... main processing loop which calls foo() some where ...
}
}
...
void foo (void)
{
... blah, blah, blah ...
if (err)
{
longjmp (env, 1);
}
}

Functions
o int setjmp (jmp buf jmpb)
o void longjmp (jmp buf jmpb, int ret) ATTR NORETURN
5.13.2

Function Documentation

5.13.2.1 void longjmp (jmp buf jmpb, int ret)
Non-local jump to a saved stack context.
#include & lt; setjmp.h & gt;

longjmp() restores the environment saved by the last call of setjmp() with the corresponding jmpb argument. After longjmp() is completed, program execution continues as if the corresponding call of setjmp() had just returned the value ret.
Note:
longjmp() cannot cause 0 to be returned. If longjmp() is invoked with a second
argument of 0, 1 will be returned instead.
Parameters:
jmpb Information saved by a previous call to setjmp().
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

35

5.14 Standard IO facilities

ret Value to return to the caller of setjmp().
Returns:
This function never returns.

5.13.2.2

int setjmp (jmp buf jmpb)

Save stack context for non-local goto.
#include & lt; setjmp.h & gt;

setjmp() saves the stack context/environment in jmpb for later use by longjmp(). The
stack context will be invalidated if the function which called setjmp() returns.
Parameters:
jmpb Variable of type jmp buf which holds the stack information such that the
environment can be restored.
Returns:
setjmp() returns 0 if returning directly, and non-zero when returning from
longjmp() using the saved context.

5.14

Standard IO facilities

5.14.1

Detailed Description

#include & lt; stdio.h & gt;

Warning:
This implementation of the standard IO facilities is new to avr-libc. It is not yet
expected to remain stable, so some aspects of the API might change in a future
release.
This file declares the standard IO facilities that are implemented in avr-libc. Due
to the nature of the underlying hardware, only a limited subset of standard IO is implemented. There is no actual file implementation available, so only device IO can be
performed. Since there's no operating system, the application needs to provide enough
details about their devices in order to make them usable by the standard IO facilities.
Due to space constraints, some functionality has not been implemented at all (like some
of the printf conversions that have been left out). Nevertheless, potential users of
this implementation should be warned: the printf and scanf families of functions,
although usually associated with presumably simple things like the famous "Hello,
world!" program, are actually fairly complex which causes their inclusion to eat up
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

36

5.14 Standard IO facilities

a fair amount of code space. Also, they are not fast due to the nature of interpreting
the format string at run-time. Whenever possible, resorting to the (sometimes nonstandard) predetermined conversion facilities that are offered by avr-libc will usually
cost much less in terms of speed and code size.
In order to allow programmers a code size vs. functionality tradeoff, the function
vfprintf() which is the heart of the printf family can be selected in different flavours
using linker options. See the documentation of vfprintf() for a detailed description.
The same applies to vfscanf() and the scanf family of functions.
The standard streams stdin, stdout, and stderr are provided, but contrary to the
C standard, since avr-libc has no knowledge about applicable devices, these streams are
not already pre-initialized at application startup. Also, since there is no notion of "file"
whatsoever to avr-libc, there is no function fopen() that could be used to associate a
stream to some device. (See note 1.) Instead, the function fdevopen() is provided
to associate a stream to a device, where the device needs to provide a function to send a
character, to receive a character, or both. There is no differentiation between "text" and
"binary" streams inside avr-libc. Character \\n is sent literally down to the device's
put() function. If the device requires a carriage return (\\r) character to be sent
before the linefeed, its put() routine must implement this (see note 2).
For convenience, the first call to fdevopen() that opens a stream for reading
will cause the resulting stream to be aliased to stdin. Likewise, the first call to
fdevopen() that opens a stream for writing will cause the resulting stream to be
aliased to both, stdout, and stderr. Thus, if the open was done with both, read
and write intent, all three standard streams will be identical. Note that these aliases are
indistinguishable from each other, thus calling fclose() on such a stream will also
effectively close all of its aliases (note 3).
All the printf and scanf family functions come in two flavours: the standard name,
where the format string is expected to be in SRAM, as well as a version with the suffix
" P" where the format string is expected to reside in the flash ROM. The macro PSTR
(explained in Program Space String Utilities) becomes very handy for declaring these
format strings.
Note 1:
It might have been possible to implement a device abstraction that is compatible
with fopen() but since this would have required to parse a string, and to take all
the information needed either out of this string, or out of an additional table that
would need to be provided by the application, this approach was not taken.
Note 2:
This basically follows the Unix approach: if a device such as a terminal needs
special handling, it is in the domain of the terminal device driver to provide this
functionality. Thus, a simple function suitable as put() for fdevopen() that
talks to a UART interface might look like this:
int

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

37

5.14 Standard IO facilities

uart_putchar(char c)
{
if (c == '\n')
uart_putchar('\r');
loop_until_bit_is_set(UCSRA, UDRE);
UDR = c;
return 0;
}

Note 3:
This implementation has been chosen because the cost of maintaining an alias
is considerably smaller than the cost of maintaining full copies of each stream.
Yet, providing an implementation that offers the complete set of standard
streams was deemed to be useful. Not only that writing printf() instead of
fprintf(mystream, ...) saves typing work, but since avr-gcc needs to
resort to pass all arguments of variadic functions on the stack (as opposed to passing them in registers for functions that take a fixed number of parameters), the
ability to pass one parameter less by implying stdin will also save some execution time.
Defines
o
o
o
o
o
o
o
o
o

#define getchar() fgetc(stdin)
#define FILE struct file
#define stdin ( iob[0])
#define stdout ( iob[1])
#define stderr ( iob[2])
#define EOF (-1)
#define putc( c, stream) fputc( c, stream)
#define putchar( c) fputc( c, stdout)
#define getc( stream) fgetc( stream)

Functions
o
o
o
o
o
o
o
o
o

int fclose (FILE * stream)
int vfprintf (FILE * stream, const char * fmt, va list ap)
int vfprintf P (FILE * stream, const char * fmt, va list ap)
int fputc (int c, FILE * stream)
int printf (const char * fmt,...)
int printf P (const char * fmt,...)
int sprintf (char * s, const char * fmt,...)
int sprintf P (char * s, const char * fmt,...)
int snprintf (char * s, size t n, const char * fmt,...)

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

38

5.14 Standard IO facilities

o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o

int snprintf P (char * s, size t n, const char * fmt,...)
int vsprintf (char * s, const char * fmt, va list ap)
int vsprintf P (char * s, const char * fmt, va list ap)
int vsnprintf (char * s, size t n, const char * fmt, va list ap)
int vsnprintf P (char * s, size t n, const char * fmt, va list ap)
int fprintf (FILE * stream, const char * fmt,...)
int fprintf P (FILE * stream, const char * fmt,...)
int fputs (const char * str, FILE * stream)
int fputs P (const char * str, FILE * stream)
int puts (const char * str)
int puts P (const char * str)
size t fwrite (const void * ptr, size t size, size t nmemb, FILE * stream)
int fgetc (FILE * stream)
int ungetc (int c, FILE * stream)
char * fgets (char * str, int size, FILE * stream)
char * gets (char * str)
size t fread (void * ptr, size t size, size t nmemb, FILE * stream)
void clearerr (FILE * stream)
int feof (FILE * stream)
int ferror (FILE * stream)
int vfscanf (FILE * stream, const char * fmt, va list ap)
int vfscanf P (FILE * stream, const char * fmt, va list ap)
int fscanf (FILE * stream, const char * fmt,...)
int fscanf P (FILE * stream, const char * fmt,...)
int scanf (const char * fmt,...)
int scanf P (const char * fmt,...)
int sscanf (const char * buf, const char * fmt,...)
int sscanf P (const char * buf, const char * fmt,...)
FILE * fdevopen (int(*put)(char), int(*get)(void), int opts attribute ((unused)))

5.14.2
5.14.2.1

Define Documentation
#define EOF (-1)

EOF declares the value that is returned by various standard IO functions in case of an
error. Since the AVR platform (currently) doesn't contain an abstraction for actual files,
its origin as "end of file" is somewhat meaningless here.
5.14.2.2

#define FILE struct file

FILE is the opaque structure that is passed around between the various standard IO
functions.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

39

5.14 Standard IO facilities

5.14.2.3 #define getc( stream) fgetc( stream)
The macro getc used to be a "fast" macro implementation with a functionality identical to fgetc(). For space constraints, in avr-libc, it is just an alias for fgetc.

5.14.2.4

int getchar(void) fgetc(stdin)

The macro getchar reads a character from stdin. Return values and error handling
is identical to fgetc().
5.14.2.5

#define putc( c, stream) fputc( c, stream)

The macro putc used to be a "fast" macro implementation with a functionality identical to fputc(). For space constraints, in avr-libc, it is just an alias for fputc.

5.14.2.6

#define putchar( c) fputc( c, stdout)

The macro putchar sends character c to stdout.
5.14.2.7

#define stderr ( iob[2])

Stream destined for error output. Unless specifically assigned, identical to stdout.
If stderr should point to another stream, the result of another fdevopen() must
be explicitly assigned to it without closing the previous stderr (since this would also
close stdout).
5.14.2.8

#define stdin ( iob[0])

Stream that will be used as an input stream by the simplified functions that don't take
a stream argument.
The first stream opened with read intent using fdevopen() will be assigned to
stdin.
5.14.2.9

#define stdout ( iob[1])

Stream that will be used as an output stream by the simplified functions that don't take
a stream argument.
The first stream opened with write intent using fdevopen() will be assigned to both,
stdin, and stderr.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

40

5.14 Standard IO facilities

5.14.3

Function Documentation

5.14.3.1 void clearerr (FILE * stream)
Clear the error and end-of-file flags of stream.
5.14.3.2

int fclose (FILE * stream)

This function closes stream, and disallows and further IO to and from it.
It currently always returns 0 (for success).
5.14.3.3 FILE* fdevopen (int(* put)(char), int(* get)(void), int opts attribute ((unused)))
This function is a replacement for fopen().
It opens a stream for a device where the actual device implementation needs to be
provided by the application. If successful, a pointer to the structure for the opened
stream is returned. Reasons for a possible failure currently include that neither the
put nor the get argument have been provided, thus attempting to open a stream with
no IO intent at all, or that insufficient dynamic memory is available to establish a new
stream.
If the put function pointer is provided, the stream is opened with write intent. The
function passed as put shall take one character to write to the device as argument ,
and shall return 0 if the output was successful, and a nonzero value if the character
could not be sent to the device.
If the get function pointer is provided, the stream is opened with read intent. The
function passed as get shall take no arguments, and return one character from the
device, passed as an int type. If an error occurs when trying to read from the device,
it shall return -1.
If both functions are provided, the stream is opened with read and write intent.
The first stream opened with read intent is assigned to stdin, and the first one opened
with write intent is assigned to both, stdout and stderr.
The third parameter opts is currently unused, but reserved for future extensions.
5.14.3.4

int feof (FILE * stream)

Test the end-of-file flag of stream. This flag can only be cleared by a call to clearerr().
Note:
Since there is currently no notion for end-of-file on a device, this function will
always return a false value.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

41

5.14 Standard IO facilities

5.14.3.5 int ferror (FILE * stream)
Test the error flag of stream. This flag can only be cleared by a call to clearerr().
5.14.3.6

int fgetc (FILE * stream)

The function fgetc reads a character from stream. It returns the character, or EOF
in case end-of-file was encountered or an error occurred. The routines feof() or ferror()
must be used to distinguish between both situations.
5.14.3.7 char* fgets (char * str, int size, FILE * stream)
Read at most size - 1 bytes from stream, until a newline character was encountered, and store the characters in the buffer pointed to by str. Unless an error was
encountered while reading, the string will then be terminated with a NUL character.
If an error was encountered, the function returns NULL and sets the error flag of
stream, which can be tested using ferror(). Otherwise, a pointer to the string will
be returned.
5.14.3.8

int fprintf (FILE * stream, const char * fmt, ...)

The function fprintf performs formatted output to stream. See vfprintf()
for details.
5.14.3.9 int fprintf P (FILE * stream, const char * fmt, ...)
Variant of fprintf() that uses a fmt string that resides in program memory.
5.14.3.10

int fputc (int c, FILE * stream)

The function fputc sends the character c (though given as type int) to stream. It
returns the character, or EOF in case an error occurred.
5.14.3.11

int fputs (const char * str, FILE * stream)

Write the string pointed to by str to stream stream.
Returns 0 on success and EOF on error.
5.14.3.12

int fputs P (const char * str, FILE * stream)

Variant of fputs() where str resides in program memory.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

42

5.14 Standard IO facilities

5.14.3.13

43

size t fread (void * ptr, size t size, size t nmemb, FILE * stream)

Read nmemb objects, size bytes each, from stream, to the buffer pointed to by
ptr.
Returns the number of objects successfully read, i. e. nmemb unless an input error
occured or end-of-file was encountered. feof() and ferror() must be used to distinguish
between these two conditions.
5.14.3.14

int fscanf (FILE * stream, const char * fmt, ...)

The function fscanf performs formatted input, reading the input data from stream.
See vfscanf() for details.
5.14.3.15

int fscanf P (FILE * stream, const char * fmt, ...)

Variant of fscanf() using a fmt string in program memory.
5.14.3.16 size t fwrite (const void *
stream)

ptr, size t

size, size t

nmemb, FILE *

Write nmemb objects, size bytes each, to stream. The first byte of the first object
is referenced by ptr.
Returns the number of objects successfully written, i. e. nmemb unless an output error
occured.
5.14.3.17

char* gets (char * str)

Similar to fgets() except that it will operate on stream stdin, and the trailing newline
(if any) will not be stored in the string. It is the caller's responsibility to provide enough
storage to hold the characters read.
5.14.3.18

int printf (const char * fmt, ...)

The function printf performs formatted output to stream stderr.
vfprintf() for details.
5.14.3.19

int printf P (const char * fmt, ...)

Variant of printf() that uses a fmt string that resides in program memory.
5.14.3.20

int puts (const char * str)

Write the string pointed to by str, and a trailing newline character, to stdout.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

See

5.14 Standard IO facilities

5.14.3.21

int puts P (const char * str)

Variant of puts() where str resides in program memory.
5.14.3.22

int scanf (const char * fmt, ...)

The function scanf performs formatted input from stream stdin.
See vfscanf() for details.
5.14.3.23

int scanf P (const char * fmt, ...)

Variant of scanf() where fmt resides in program memory.
5.14.3.24

int snprintf (char * s, size t n, const char * fmt, ...)

Like sprintf(), but instead of assuming s to be of infinite size, no more than n
characters (including the trailing NUL character) will be converted to s.
Returns the number of characters that would have been written to s if there were
enough space.
5.14.3.25

int snprintf P (char * s, size t n, const char * fmt, ...)

Variant of snprintf() that uses a fmt string that resides in program memory.
5.14.3.26

int sprintf (char * s, const char * fmt, ...)

Variant of printf() that sends the formatted characters to string s.
5.14.3.27

int sprintf P (char * s, const char * fmt, ...)

Variant of sprintf() that uses a fmt string that resides in program memory.
5.14.3.28

int sscanf (const char * buf, const char * fmt, ...)

The function sscanf performs formatted input, reading the input data from the buffer
pointed to by buf.
See vfscanf() for details.
5.14.3.29

int sscanf P (const char * buf, const char * fmt, ...)

Variant of sscanf() using a fmt string in program memory.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

44

5.14 Standard IO facilities

5.14.3.30

int ungetc (int c, FILE * stream)

The ungetc() function pushes the character c (converted to an unsigned char) back onto
the input stream pointed to by stream. The pushed-back character will be returned
by a subsequent read on the stream.
Currently, only a single character can be pushed back onto the stream.
The ungetc() function returns the character pushed back after the conversion, or EOF if
the operation fails. If the value of the argument c character equals EOF, the operation
will fail and the stream will remain unchanged.
5.14.3.31

int vfprintf (FILE * stream, const char * fmt, va list ap)

vfprintf is the central facility of the printf family of functions. It outputs values
to stream under control of a format string passed in fmt. The actual values to print
are passed as a variable argument list ap.
vfprintf returns the number of characters written to stream, or EOF in case of
an error. Currently, this will only happen if stream has not been opened with write
intent.
The format string is composed of zero or more directives: ordinary characters (not
%), which are copied unchanged to the output stream; and conversion specifications,
each of which results in fetching zero or more subsequent arguments. Each conversion
specification is introduced by the % character. The arguments must properly correspond
(after type promotion) with the conversion specifier. After the %, the following appear
in sequence:
o Zero or more of the following flags:
- # The value should be converted to an "alternate form". For c, d, i, s, and
u conversions, this option has no effect. For o conversions, the precision of
the number is increased to force the first character of the output string to
a zero (except if a zero value is printed with an explicit precision of zero).
For x and X conversions, a non-zero result has the string '0x' (or '0X' for
X conversions) prepended to it.
- 0 (zero) Zero padding. For all conversions, the converted value is padded
on the left with zeros rather than blanks. If a precision is given with a
numeric conversion (d, i, o, u, i, x, and X), the 0 flag is ignored.
- - A negative field width flag; the converted value is to be left adjusted on
the field boundary. The converted value is padded on the right with blanks,
rather than on the left with blanks or zeros. A - overrides a 0 if both are
given.
- ' ' (space) A blank should be left before a positive number produced by a
signed conversion (d, or i).
- + A sign must always be placed before a number produced by a signed
conversion. A + overrides a space if both are used.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

45

5.14 Standard IO facilities

- An optional decimal digit string specifying a minimum field width. If the
converted value has fewer characters than the field width, it will be padded with
spaces on the left (or right, if the left-adjust173 ment flag has been given) to fill
out the field width.
o An optional precision, in the form of a period . followed by an optional digit
string. If the digit string is omitted, the precision is taken as zero. This gives the
minimum number of digits to appear for d, i, o, u, x, and X conversions, or the
maximum number of characters to be printed from a string for s conversions.
o An optional l length modifier, that specifies that the argument for the d, i, o, u,
x, or X conversion is a " long int " rather than int.
o A character that specifies the type of conversion to be applied.
The conversion specifiers and their meanings are:
o diouxX The int (or appropriate variant) argument is converted to signed decimal (d and i), unsigned octal (o), unsigned decimal (u), or unsigned hexadecimal
(x and X) notation. The letters "abcdef" are used for x conversions; the letters
"ABCDEF" are used for X conversions. The precision, if any, gives the minimum number of digits that must appear; if the converted value requires fewer
digits, it is padded on the left with zeros.
o p The void * argument is taken as an unsigned integer, and converted similarly
as a %x command would do.
o c The int argument is converted to an " unsigned char " , and the resulting
character is written.
o s The " char * " argument is expected to be a pointer to an array of character
type (pointer to a string). Characters from the array are written up to (but not
including) a terminating NUL character; if a precision is specified, no more than
the number specified are written. If a precision is given, no null character need
be present; if the precision is not specified, or is greater than the size of the array,
the array must contain a terminating NUL character.
o % A % is written. No argument is converted. The complete conversion specification is "%%".
o eE The double argument is rounded and converted in the format
" [-]d.ddde177dd " where there is one digit before the decimal-point character and the number of digits after it is equal to the precision; if the precision
is missing, it is taken as 6; if the precision is zero, no decimal-point character
appears. An E conversion uses the letter 'E' (rather than 'e') to introduce the
exponent. The exponent always contains two digits; if the value is zero, the
exponent is 00.
o fF The double argument is rounded and converted to decimal notation in the
format " [-]ddd.ddd " , where the number of digits after the decimal-point
character is equal to the precision specification. If the precision is missing, it is
taken as 6; if the precision is explicitly zero, no decimal-point character appears.
If a decimal point appears, at least one digit appears before it.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

46

5.14 Standard IO facilities

o gG The double argument is converted in style f or e (or F or E for G conversions). The precision specifies the number of significant digits. If the precision
is missing, 6 digits are given; if the precision is zero, it is treated as 1. Style e is
used if the exponent from its conversion is less than -4 or greater than or equal to
the precision. Trailing zeros are removed from the fractional part of the result; a
decimal point appears only if it is followed by at least one digit.
In no case does a non-existent or small field width cause truncation of a numeric field;
if the result of a conversion is wider than the field width, the field is expanded to contain
the conversion result.
Since the full implementation of all the mentioned features becomes fairly large, three
different flavours of vfprintf() can be selected using linker options. The default vfprintf() implements all the mentioned functionality except floating point conversions.
A minimized version of vfprintf() is available that only implements the very basic integer and string conversion facilities, but none of the additional options that can be
specified using conversion flags (these flags are parsed correctly from the format specification, but then simply ignored). This version can be requested using the following
compiler options:
-Wl,-u,vfprintf -lprintf_min

If the full functionality including the floating point conversions is required, the following options should be used:
-Wl,-u,vfprintf -lprintf_flt -lm

Limitations:
o The specified width and precision can be at most 127.
o For floating-point conversions, trailing digits will be lost if a number close to
DBL MAX is converted with a precision & gt; 0.

5.14.3.32

int vfprintf P (FILE * stream, const char * fmt, va list ap)

Variant of vfprintf() that uses a fmt string that resides in program memory.
5.14.3.33

int vfscanf (FILE * stream, const char * fmt, va list ap)

Formatted input. This function is the heart of the scanf family of functions.
Characters are read from stream and processed in a way described by fmt. Conversion results will be assigned to the parameters passed via ap.
The format string fmt is scanned for conversion specifications. Anything that doesn't
comprise a conversion specification is taken as text that is matched literally against
the input. White space in the format string will match any white space in the data
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

47

5.14 Standard IO facilities

(including none), all other characters match only itself. Processing is aborted as soon as
the data and format string no longer match, or there is an error or end-of-file condition
on stream.
Most conversions skip leading white space before starting the actual conversion.
Conversions are introduced with the character %. Possible options can follow the %:
o a * indicating that the conversion should be performed but the conversion result
is to be discarded; no parameters will be processed from ap,
o the character h indicating that the argument is a pointer to short int (rather
than int),
o the character l indicating that the argument is a pointer to long int (rather
than int, for integer type conversions), or a pointer to double (for floating
point conversions).
In addition, a maximal field width may be specified as a nonzero positive decimal
integer, which will restrict the conversion to at most this many characters from the
input stream. This field width is limited to at most 127 characters which is also the
default value (except for the c conversion that defaults to 1).
The following conversion flags are supported:
o % Matches a literal % character. This is not a conversion.
o d Matches an optionally signed decimal integer; the next pointer must be a
pointer to int.
o i Matches an optionally signed integer; the next pointer must be a pointer to
int. The integer is read in base 16 if it begins with 0x or 0X, in base 8 if it
begins with 0, and in base 10 otherwise. Only characters that correspond to the
base are used.
o o Matches an octal integer; the next pointer must be a pointer to unsigned
int.
o u Matches an optionally signed decimal integer; the next pointer must be a
pointer to unsigned int.
o x Matches an optionally signed hexadecimal integer; the next pointer must be a
pointer to unsigned int.
o f Matches an optionally signed floating-point number; the next pointer must be
a pointer to float.
o e, g, E, G Equivalent to f.
o s Matches a sequence of non-white-space characters; the next pointer must be a
pointer to char, and the array must be large enough to accept all the sequence
and the terminating NUL character. The input string stops at white space or at
the maximum field width, whichever occurs first.
o c Matches a sequence of width count characters (default 1); the next pointer must
be a pointer to char, and there must be enough room for all the characters (no
terminating NUL is added). The usual skip of leading white space is suppressed.
To skip white space first, use an explicit space in the format.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

48

5.14 Standard IO facilities

o [ Matches a nonempty sequence of characters from the specified set of accepted
characters; the next pointer must be a pointer to char, and there must be enough
room for all the characters in the string, plus a terminating NUL character. The
usual skip of leading white space is suppressed. The string is to be made up
of characters in (or not in) a particular set; the set is defined by the characters
between the open bracket [ character and a close bracket ] character. The set
excludes those characters if the first character after the open bracket is a circumflex ? . To include a close bracket in the set, make it the first character after the
open bracket or the circumflex; any other position will end the set. The hyphen
character - is also special; when placed between two other characters, it adds all
intervening characters to the set. To include a hyphen, make it the last character
before the final close bracket. For instance, [? ]0-9-] means the set of everything except close bracket, zero through nine, and hyphen. The string ends with
the appearance of a character not in the (or, with a circumflex, in) set or when
the field width runs out.
o p Matches a pointer value (as printed by p in printf()); the next pointer must be
a pointer to void.
o n Nothing is expected; instead, the number of characters consumed thus far from
the input is stored through the next pointer, which must be a pointer to int. This
is not a conversion, although it can be suppressed with the * flag.
These functions return the number of input items assigned, which can be fewer than
provided for, or even zero, in the event of a matching failure. Zero indicates that, while
there was input available, no conversions were assigned; typically this is due to an
invalid input character, such as an alphabetic character for a d conversion. The value
EOF is returned if an input failure occurs before any conversion such as an end-of-file
occurs. If an error or end-of-file occurs after conversion has begun, the number of
conversions which were successfully completed is returned.
By default, all the conversions described above are available except the floating-point
conversions, and the %[ conversion. These conversions will be available in the extended version provided by the library libscanf flt.a. Note that either of these
conversions requires the availability of a buffer that needs to be obtained at run-time
using malloc(). If this buffer cannot be obtained, the operation is aborted, returning the
value EOF. To link a program against the extended version, use the following compiler
flags in the link stage:
-Wl,-u,vfscanf -lscanf_flt -lm

A third version is available for environments that are tight on space. This version is
provided in the library libscanf min.a, and can be requested using the following
options in the link stage:
-Wl,-u,vfscanf -lscanf_min -lm

In addition to the restrictions of the standard version, this version implements no field
width specification, no conversion assignment suppression flag (*), no n specification,
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

49

5.15 General utilities

and no general format character matching at all. All characters in fmt that do not
comprise a conversion specification will simply be ignored, including white space (that
is normally used to consume any amount of white space in the input stream). However,
the usual skip of initial white space in the formats that support it is implemented.
5.14.3.34

int vfscanf P (FILE * stream, const char * fmt, va list ap)

Variant of vfscanf() using a fmt string in program memory.
5.14.3.35

int vsnprintf (char * s, size t n, const char * fmt, va list ap)

Like vsprintf(), but instead of assuming s to be of infinite size, no more than n
characters (including the trailing NUL character) will be converted to s.
Returns the number of characters that would have been written to s if there were
enough space.
5.14.3.36

int vsnprintf P (char * s, size t n, const char * fmt, va list ap)

Variant of vsnprintf() that uses a fmt string that resides in program memory.
5.14.3.37

int vsprintf (char * s, const char * fmt, va list ap)

Like sprintf() but takes a variable argument list for the arguments.
5.14.3.38

int vsprintf P (char * s, const char * fmt, va list ap)

Variant of vsprintf() that uses a fmt string that resides in program memory.

5.15

General utilities

5.15.1

Detailed Description

#include & lt; stdlib.h & gt;

This file declares some basic C macros and functions as defined by the ISO standard,
plus some AVR-specific extensions.
Data Structures
o struct div t
o struct ldiv t

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

50

5.15 General utilities

51

Non-standard (i.e. non-ISO C) functions.
o
o
o
o
o
o
o
o

#define RANDOM MAX 0x7FFFFFFF
char * itoa (int val, char * s, int radix)
char * ltoa (long int val, char * s, int radix)
char * utoa (unsigned int val, char * s, int radix)
char * ultoa (unsigned long int val, char * s, int radix)
long random (void)
void srandom (unsigned long seed)
long random r (unsigned long *ctx)

Conversion functions for double arguments.
Note that these functions are not located in the default library, libc.a, but in the
mathematical library, libm.a. So when linking the application, the -lm option needs
to be specified.
o
o
o
o

#define DTOSTR ALWAYS SIGN 0x01
#define DTOSTR PLUS SIGN 0x02
#define DTOSTR UPPERCASE 0x04
char * dtostre (double val, char * s, unsigned char prec, unsigned char
flags)
o char * dtostrf (double val, char width, char prec, char * s)

-

Defines
o #define RAND MAX 0x7FFF
Typedefs
o typedef int(* compar fn t )(const void *, const void *)
Functions
o inline void abort (void) ATTR NORETURN
o int abs (int i) ATTR CONST
o long labs (long i) ATTR CONST
o void * bsearch (const void * key, const void * base, size t nmemb, size t
size, int(* compar)(const void *, const void *))
o div t div (int num, int denom) asm (" divmodhi4") ATTR CONST
o ldiv t ldiv (long num, long denom) asm (" divmodsi4") ATTR CONST
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.15 General utilities

o
o
o
o
o
o
o
o
o
o
o
o
o

void qsort (void * base, size t nmemb, size t size, compar fn t compar)
long strtol (const char * nptr, char ** endptr, int base)
unsigned long strtoul (const char * nptr, char ** endptr, int base)
inline long atol (const char * nptr) ATTR PURE
inline int atoi (const char * nptr) ATTR PURE
void exit (int status) ATTR NORETURN
void * malloc (size t size) ATTR MALLOC
void free (void * ptr)
void * calloc (size t nele, size t size) ATTR MALLOC
double strtod (const char * nptr, char ** endptr)
int rand (void)
void srand (unsigned int seed)
int rand r (unsigned long *ctx)

Variables
o size t malloc margin
o char * malloc heap start
o char * malloc heap end
5.15.2

Define Documentation

5.15.2.1 #define DTOSTR ALWAYS SIGN 0x01
Bit value that can be passed in flags to dtostre().
5.15.2.2

#define DTOSTR PLUS SIGN 0x02

Bit value that can be passed in flags to dtostre().
5.15.2.3

#define DTOSTR UPPERCASE 0x04

Bit value that can be passed in flags to dtostre().
5.15.2.4

#define RAND MAX 0x7FFF

Highest number that can be generated by rand().
5.15.2.5

#define RANDOM MAX 0x7FFFFFFF

Highest number that can be generated by random().

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

52

5.15 General utilities

5.15.3

Typedef Documentation

5.15.3.1 typedef int(* compar fn t)(const void *, const void *)
Comparision function type for qsort(), just for convenience.
5.15.4
5.15.4.1

Function Documentation
inline void abort (void)

The abort() function causes abnormal program termination to occur. In the limited
AVR environment, execution is effectively halted by entering an infinite loop.
5.15.4.2

int abs (int i)

The abs() function computes the absolute value of the integer i.
Note:
The abs() and labs() functions are builtins of gcc.

5.15.4.3 int atoi (const char * string)
Convert a string to an integer.
The atoi() function converts the initial portion of the string pointed to by nptr to
integer representation.
It is equivalent to:
(int)strtol(nptr, (char **)NULL, 10);

except that atoi() does not detect errors.
5.15.4.4

long int atol (const char * string)

Convert a string to a long integer.
The atol() function converts the initial portion of the string pointed to by stringp to
integer representation.
It is equivalent to:
strtol(nptr, (char **)NULL, 10);

except that atol() does not detect errors.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

53

5.15 General utilities

54

5.15.4.5 void* bsearch (const void * key, const void *
size t size, int(* compar)(const void *, const void *))

base, size t

nmemb,

The bsearch() function searches an array of nmemb objects, the initial member of
which is pointed to by base, for a member that matches the object pointed to by
key. The size of each member of the array is specified by size.
The contents of the array should be in ascending sorted order according to the comparison function referenced by compar. The compar routine is expected to have two
arguments which point to the key object and to an array member, in that order, and
should return an integer less than, equal to, or greater than zero if the key object is
found, respectively, to be less than, to match, or be greater than the array member.
The bsearch() function returns a pointer to a matching member of the array, or a null
pointer if no match is found. If two members compare as equal, which member is
matched is unspecified.
5.15.4.6

void* calloc (size t nele, size t size)

Allocate nele elements of size each. Identical to calling malloc() using nele
* size as argument, except the allocated memory will be cleared to zero.
5.15.4.7 div t div (int num, int denom)
The div() function computes the value num/denom and returns the quotient and remainder in a structure named div t that contains two int members named quot and
rem.
5.15.4.8 char* dtostre (double
char flags)

val, char *

s, unsigned char

prec, unsigned

The dtostre() function converts the double value passed in val into an ASCII representation that will be stored under s. The caller is responsible for providing sufficient
storage in s.
Conversion is done in the format " [-]d.ddde177dd " where there is one digit before the decimal-point character and the number of digits after it is equal to the precision prec; if the precision is zero, no decimal-point character appears. If flags has
the DTOSTRE UPPERCASE bit set, the letter 'E' (rather than 'e' ) will be used to
introduce the exponent. The exponent always contains two digits; if the value is zero,
the exponent is " 00 " .
If flags has the DTOSTRE ALWAYS SIGN bit set, a space character will be placed
into the leading position for positive numbers.
If flags has the DTOSTRE PLUS SIGN bit set, a plus sign will be used instead of a
space character in this case.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.15 General utilities

5.15.4.9 char* dtostrf (double val, char width, char prec, char * s)
The dtostrf() function converts the double value passed in val into an ASCII representationthat will be stored under s. The caller is responsible for providing sufficient
storage in s.
Conversion is done in the format " [-]d.ddd " . The minimum field width of the
output string (including the '.' and the possible sign for negative values) is given in
width, and prec determines the number of digits after the decimal sign.
5.15.4.10

void exit (int status)

The exit() function terminates the application. Since there is no environment to return to, status is ignored, and code execution will eventually reach an infinite loop,
thereby effectively halting all code processing.
In a C++ context, global destructors will be called before halting execution.
5.15.4.11

void free (void * ptr)

The free() function causes the allocated memory referenced by ptr to be made available for future allocations. If ptr is NULL, no action occurs.
5.15.4.12

char* itoa (int val, char * s, int radix)

Convert an integer to a string.
The function itoa() converts the integer value from val into an ASCII representation
that will be stored under s. The caller is responsible for providing sufficient storage in
s.
Note:
The minimal size of the buffer s depends on the choice of radix. For example, if
the radix is 2 (binary), you need to supply a buffer with a minimal length of 8 *
sizeof (int) + 1 characters, i.e. one character for each bit plus one for the string
terminator. Using a larger radix will require a smaller minimal buffer size.
Warning:
If the buffer is too small, you risk a buffer overflow.
Conversion is done using the radix as base, which may be a number between 2
(binary conversion) and up to 36. If radix is greater than 10, the next digit after
'9' will be the letter 'a'.
If radix is 10 and val is negative, a minus sign will be prepended.
The itoa() function returns the pointer passed as s.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

55

5.15 General utilities

5.15.4.13

long labs (long i)

The labs() function computes the absolute value of the long integer i.
Note:
The abs() and labs() functions are builtins of gcc.

5.15.4.14 ldiv t ldiv (long num, long denom)
The ldiv() function computes the value num/denom and returns the quotient and remainder in a structure named ldiv t that contains two long integer members named
quot and rem.
5.15.4.15

char* ltoa (long int val, char * s, int radix)

Convert a long integer to a string.
The function ltoa() converts the long integer value from val into an ASCII representation that will be stored under s. The caller is responsible for providing sufficient
storage in s.
Note:
The minimal size of the buffer s depends on the choice of radix. For example,
if the radix is 2 (binary), you need to supply a buffer with a minimal length of 8
* sizeof (long int) + 1 characters, i.e. one character for each bit plus one for the
string terminator. Using a larger radix will require a smaller minimal buffer size.
Warning:
If the buffer is too small, you risk a buffer overflow.
Conversion is done using the radix as base, which may be a number between 2
(binary conversion) and up to 36. If radix is greater than 10, the next digit after
'9' will be the letter 'a'.
If radix is 10 and val is negative, a minus sign will be prepended.
The ltoa() function returns the pointer passed as s.
5.15.4.16

void* malloc (size t size)

The malloc() function allocates size bytes of memory. If malloc() fails, a NULL
pointer is returned.
Note that malloc() does not initialize the returned memory to zero bytes.
See the chapter about malloc() usage for implementation details.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

56

5.15 General utilities

5.15.4.17 void qsort (void *
compar)

57

base, size t

nmemb, size t

size,

compar fn t

The qsort() function is a modified partition-exchange sort, or quicksort.
The qsort() function sorts an array of nmemb objects, the initial member of which is
pointed to by base. The size of each object is specified by size. The contents of the
array base are sorted in ascending order according to a comparison function pointed to
by compar, which requires two arguments pointing to the objects being compared.
The comparison function must return an integer less than, equal to, or greater than zero
if the first argument is considered to be respectively less than, equal to, or greater than
the second.
5.15.4.18 int rand (void)
The rand() function computes a sequence of pseudo-random integers in the range of 0
to RAND MAX (as defined by the header file & lt; stdlib.h & gt; ).
The srand() function sets its argument seed as the seed for a new sequence of pseudorandom numbers to be returned by rand(). These sequences are repeatable by calling
srand() with the same seed value.
If no seed value is provided, the functions are automatically seeded with a value of 1.
In compliance with the C standard, these functions operate on int arguments. Since
the underlying algorithm already uses 32-bit calculations, this causes a loss of precision. See random() for an alternate set of functions that retains full 32-bit precision.

5.15.4.19

int rand r (unsigned long * ctx)

Variant of rand() that stores the context in the user-supplied variable located at ctx
instead of a static library variable so the function becomes re-entrant.
5.15.4.20

long random (void)

The random() function computes a sequence of pseudo-random integers in the range of
0 to RANDOM MAX (as defined by the header file & lt; stdlib.h & gt; ).
The srandom() function sets its argument seed as the seed for a new sequence of
pseudo-random numbers to be returned by rand(). These sequences are repeatable by
calling srandom() with the same seed value.
If no seed value is provided, the functions are automatically seeded with a value of 1.
5.15.4.21 long random r (unsigned long * ctx)

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.15 General utilities

Variant of random() that stores the context in the user-supplied variable located at ctx
instead of a static library variable so the function becomes re-entrant.
5.15.4.22

void srand (unsigned int seed)

Pseudo-random number generator seeding; see rand().
5.15.4.23

void srandom (unsigned long seed)

Pseudo-random number generator seeding; see random().
5.15.4.24

double strtod (const char * nptr, char ** endptr)

The strtod() function converts the initial portion of the string pointed to by nptr to
double representation.
The expected form of the string is an optional plus ( '+' ) or minus sign ( '-' ) followed
by a sequence of digits optionally containing a decimal-point character, optionally followed by an exponent. An exponent consists of an 'E' or 'e', followed by an optional
plus or minus sign, followed by a sequence of digits.
Leading white-space characters in the string are skipped.
The strtod() function returns the converted value, if any.
If endptr is not NULL, a pointer to the character after the last character used in the
conversion is stored in the location referenced by endptr.
If no conversion is performed, zero is returned and the value of nptr is stored in the
location referenced by endptr.
If the correct value would cause overflow, plus or minus HUGE VAL is returned (according to the sign of the value), and ERANGE is stored in errno. If the correct value
would cause underflow, zero is returned and ERANGE is stored in errno.
FIXME: HUGE VAL needs to be defined somewhere. The bit pattern is 0x7fffffff, but
what number would this be?
5.15.4.25

long strtol (const char * nptr, char ** endptr, int base)

The strtol() function converts the string in nptr to a long value. The conversion is
done according to the given base, which must be between 2 and 36 inclusive, or be the
special value 0.
The string may begin with an arbitrary amount of white space (as determined by isspace()) followed by a single optional '+' or '-' sign. If base is zero or 16, the string
may then include a " 0x " prefix, and the number will be read in base 16; otherwise, a
zero base is taken as 10 (decimal) unless the next character is '0', in which case it is
taken as 8 (octal).

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

58

5.15 General utilities

The remainder of the string is converted to a long value in the obvious manner, stopping
at the first character which is not a valid digit in the given base. (In bases above 10, the
letter 'A' in either upper or lower case represents 10, 'B' represents 11, and so forth,
with 'Z' representing 35.)
If endptr is not NULL, strtol() stores the address of the first invalid character in
*endptr. If there were no digits at all, however, strtol() stores the original value of
nptr in endptr. (Thus, if *nptr is not '\\0' but **endptr is '\\0' on return, the
entire string was valid.)
The strtol() function returns the result of the conversion, unless the value would underflow or overflow. If no conversion could be performed, 0 is returned. If an overflow or
underflow occurs, errno is set to ERANGE and the function return value is clamped
to LONG MIN or LONG MAX, respectively.
5.15.4.26 unsigned long strtoul (const char * nptr, char ** endptr, int base)
The strtoul() function converts the string in nptr to an unsigned long value. The conversion is done according to the given base, which must be between 2 and 36 inclusive,
or be the special value 0.
The string may begin with an arbitrary amount of white space (as determined by isspace()) followed by a single optional '+' or '-' sign. If base is zero or 16, the string
may then include a " 0x " prefix, and the number will be read in base 16; otherwise, a
zero base is taken as 10 (decimal) unless the next character is '0', in which case it is
taken as 8 (octal).
The remainder of the string is converted to an unsigned long value in the obvious
manner, stopping at the first character which is not a valid digit in the given base. (In
bases above 10, the letter 'A' in either upper or lower case represents 10, 'B' represents
11, and so forth, with 'Z' representing 35.)
If endptr is not NULL, strtoul() stores the address of the first invalid character in
*endptr. If there were no digits at all, however, strtoul() stores the original value of
nptr in endptr. (Thus, if *nptr is not '\\0' but **endptr is '\\0' on return, the
entire string was valid.)
The strtoul() function return either the result of the conversion or, if there was a leading minus sign, the negation of the result of the conversion, unless the original (nonnegated) value would overflow; in the latter case, strtoul() returns ULONG MAX, and
errno is set to ERANGE. If no conversion could be performed, 0 is returned.
5.15.4.27

char* ultoa (unsigned long int val, char * s, int radix)

Convert an unsigned long integer to a string.
The function ultoa() converts the unsigned long integer value from val into an ASCII
representation that will be stored under s. The caller is responsible for providing suf-

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

59

5.15 General utilities

ficient storage in s.
Note:
The minimal size of the buffer s depends on the choice of radix. For example, if
the radix is 2 (binary), you need to supply a buffer with a minimal length of 8 *
sizeof (unsigned long int) + 1 characters, i.e. one character for each bit plus one
for the string terminator. Using a larger radix will require a smaller minimal buffer
size.
Warning:
If the buffer is too small, you risk a buffer overflow.
Conversion is done using the radix as base, which may be a number between 2
(binary conversion) and up to 36. If radix is greater than 10, the next digit after
'9' will be the letter 'a'.
The ultoa() function returns the pointer passed as s.
5.15.4.28

char* utoa (unsigned int val, char * s, int radix)

Convert an unsigned integer to a string.
The function utoa() converts the unsigned integer value from val into an ASCII representation that will be stored under s. The caller is responsible for providing sufficient
storage in s.
Note:
The minimal size of the buffer s depends on the choice of radix. For example, if
the radix is 2 (binary), you need to supply a buffer with a minimal length of 8 *
sizeof (unsigned int) + 1 characters, i.e. one character for each bit plus one for the
string terminator. Using a larger radix will require a smaller minimal buffer size.
Warning:
If the buffer is too small, you risk a buffer overflow.
Conversion is done using the radix as base, which may be a number between 2
(binary conversion) and up to 36. If radix is greater than 10, the next digit after
'9' will be the letter 'a'.
The utoa() function returns the pointer passed as s.
5.15.5
5.15.5.1

Variable Documentation
char* malloc heap end

malloc() tunable.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

60

5.16 Strings

5.15.5.2 char* malloc heap start
malloc() tunable.
5.15.5.3

size t malloc margin

malloc() tunable.

5.16

Strings

5.16.1

Detailed Description

#include & lt; string.h & gt;

The string functions perform string operations on NULL terminated strings.
Note:
If the strings you are working on resident in program space (flash), you will need
to use the string functions described in Program Space String Utilities.
Functions
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o

void * memccpy (void *, const void *, int, size t)
void * memchr (const void *, int, size t) ATTR PURE
int memcmp (const void *, const void *, size t) ATTR PURE
void * memcpy (void *, const void *, size t)
void * memmove (void *, const void *, size t)
void * memset (void *, int, size t)
int strcasecmp (const char *, const char *) ATTR PURE
char * strcat (char *, const char *)
char * strchr (const char *, int) ATTR PURE
int strcmp (const char *, const char *) ATTR PURE
char * strcpy (char *, const char *)
size t strlcat (char *, const char *, size t)
size t strlcpy (char *, const char *, size t)
size t strlen (const char *) ATTR PURE
char * strlwr (char *)
int strncasecmp (const char *, const char *, size t) ATTR PURE
char * strncat (char *, const char *, size t)
int strncmp (const char *, const char *, size t)
char * strncpy (char *, const char *, size t)
size t strnlen (const char *, size t) ATTR PURE
char * strrchr (const char *, int) ATTR PURE

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

61

5.16 Strings

o
o
o
o
o

char * strrev (char *)
char * strsep (char **, const char *) ATTR PURE
char * strstr (const char *, const char *) ATTR PURE
char * strtok r (char *, const char *, char **) ATTR PURE
char * strupr (char *)

5.16.2
5.16.2.1

Function Documentation
void * memccpy (void * dest, const void * src, int val, size t len)

Copy memory area.
The memccpy() function copies no more than len bytes from memory area src to memory area dest, stopping when the character val is found.
Returns:
The memccpy() function returns a pointer to the next character in dest after val, or
NULL if val was not found in the first len characters of src.

5.16.2.2

void * memchr (const void * src, int val, size t len)

Scan memory for a character.
The memchr() function scans the first len bytes of the memory area pointed to by src
for the character val. The first byte to match val (interpreted as an unsigned character)
stops the operation.
Returns:
The memchr() function returns a pointer to the matching byte or NULL if the
character does not occur in the given memory area.

5.16.2.3

int memcmp (const void * s1, const void * s2, size t len)

Compare memory areas.
The memcmp() function compares the first len bytes of the memory areas s1 and s2.
The comparision is performed using unsigned char operations.
Returns:
The memcmp() function returns an integer less than, equal to, or greater than zero
if the first len bytes of s1 is found, respectively, to be less than, to match, or be
greater than the first len bytes of s2.
Note:
Be sure to store the result in a 16 bit variable since you may get incorrect results if
you use an unsigned char or char due to truncation.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

62

5.16 Strings

Warning:
This function is not -mint8 compatible, although if you only care about testing for
equality, this function should be safe to use.

5.16.2.4

void * memcpy (void * dest, const void * src, size t len)

Copy a memory area.
The memcpy() function copies len bytes from memory area src to memory area dest.
The memory areas may not overlap. Use memmove() if the memory areas do overlap.
Returns:
The memcpy() function returns a pointer to dest.

5.16.2.5

void * memmove (void * dest, const void * src, size t len)

Copy memory area.
The memmove() function copies len bytes from memory area src to memory area dest.
The memory areas may overlap.
Returns:
The memmove() function returns a pointer to dest.

5.16.2.6

void * memset (void * dest, int val, size t len)

Fill memory with a constant byte.
The memset() function fills the first len bytes of the memory area pointed to by dest
with the constant byte val.
Returns:
The memset() function returns a pointer to the memory area dest.

5.16.2.7

int strcasecmp (const char * s1, const char * s2)

Compare two strings ignoring case.
The strcasecmp() function compares the two strings s1 and s2, ignoring the case of the
characters.
Returns:
The strcasecmp() function returns an integer less than, equal to, or greater than
zero if s1 is found, respectively, to be less than, to match, or be greater than s2.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

63

5.16 Strings

5.16.2.8 char * strcat (char * dest, const char * src)
Concatenate two strings.
The strcat() function appends the src string to the dest string overwriting the '\0' character at the end of dest, and then adds a terminating '\0' character. The strings may not
overlap, and the dest string must have enough space for the result.
Returns:
The strcat() function returns a pointer to the resulting string dest.

5.16.2.9

char * strchr (const char * src, int val)

Locate character in string.
The strchr() function returns a pointer to the first occurrence of the character val in the
string src.
Here "character" means "byte" - these functions do not work with wide or multi-byte
characters.
Returns:
The strchr() function returns a pointer to the matched character or NULL if the
character is not found.

5.16.2.10 int strcmp (const char * s1, const char * s2)
Compare two strings.
The strcmp() function compares the two strings s1 and s2.
Returns:
The strcmp() function returns an integer less than, equal to, or greater than zero if
s1 is found, respectively, to be less than, to match, or be greater than s2.

5.16.2.11

char * strcpy (char * dest, const char * src)

Copy a string.
The strcpy() function copies the string pointed to by src (including the terminating
'\0' character) to the array pointed to by dest. The strings may not overlap, and the
destination string dest must be large enough to receive the copy.
Returns:
The strcpy() function returns a pointer to the destination string dest.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

64

5.16 Strings

Note:
If the destination string of a strcpy() is not large enough (that is, if the programmer
was stupid/lazy, and failed to check the size before copying) then anything might
happen. Overflowing fixed length strings is a favourite cracker technique.

5.16.2.12

size t strlcat (char * dst, const char * src, size t siz)

Concatenate two strings.
Appends src to string dst of size siz (unlike strncat(), siz is the full size of dst, not space
left). At most siz-1 characters will be copied. Always NULL terminates (unless siz & lt; =
strlen(dst)).
Returns:
The strlcat() function returns strlen(src) + MIN(siz, strlen(initial dst)). If retval & gt; =
siz, truncation occurred.

5.16.2.13

size t strlcpy (char * dst, const char * src, size t siz)

Copy a string.
Copy src to string dst of size siz. At most siz-1 characters will be copied. Always
NULL terminates (unless siz == 0).
Returns:
The strlcpy() function returns strlen(src). If retval & gt; = siz, truncation occurred.

5.16.2.14

size t strlen (const char * src)

Calculate the length of a string.
The strlen() function calculates the length of the string src, not including the terminating '\0' character.
Returns:
The strlen() function returns the number of characters in src.

5.16.2.15

char * strlwr (char * string)

Convert a string to lower case.
The strlwr() function will convert a string to lower case. Only the upper case alphabetic
characters [A .. Z] are converted. Non-alphabetic characters will not be changed.
Returns:
The strlwr() function returns a pointer to the converted string.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

65

5.16 Strings

5.16.2.16

int strncasecmp (const char * s1, const char * s2, size t len)

Compare two strings ignoring case.
The strncasecmp() function is similar to strcasecmp(), except it only compares the first
n characters of s1.
Returns:
The strncasecmp() function returns an integer less than, equal to, or greater than
zero if s1 (or the first n bytes thereof) is found, respectively, to be less than, to
match, or be greater than s2.

5.16.2.17

char * strncat (char * dest, const char * src, size t len)

Concatenate two strings.
The strncat() function is similar to strcat(), except that only the first n characters of src
are appended to dest.
Returns:
The strncat() function returns a pointer to the resulting string dest.

5.16.2.18

int strncmp (const char * s1, const char * s2, size t len)

Compare two strings.
The strncmp() function is similar to strcmp(), except it only compares the first (at most)
n characters of s1 and s2.
Returns:
The strncmp() function returns an integer less than, equal to, or greater than zero
if s1 (or the first n bytes thereof) is found, respectively, to be less than, to match,
or be greater than s2.

5.16.2.19

char * strncpy (char * dest, const char * src, size t len)

Copy a string.
The strncpy() function is similar to strcpy(), except that not more than n bytes of src
are copied. Thus, if there is no null byte among the first n bytes of src, the result will
not be null-terminated.
In the case where the length of src is less than that of n, the remainder of dest will be
padded with nulls.
Returns:
The strncpy() function returns a pointer to the destination string dest.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

66

5.16 Strings

5.16.2.20

size t strnlen (const char * src, size t len)

Determine the length of a fixed-size string.
The strnlen function returns the number of characters in the string pointed to by src, not
including the terminating '\0' character, but at most len. In doing this, strnlen looks
only at the first len characters at src and never beyond src+len.
Returns:
The strnlen function returns strlen(src), if that is less than len, or len if there is no
'\0' character among the first len characters pointed to by src.

5.16.2.21

char * strrchr (const char * src, int val)

Locate character in string.
The strrchr() function returns a pointer to the last occurrence of the character val in the
string src.
Here "character" means "byte" - these functions do not work with wide or multi-byte
characters.
Returns:
The strrchr() function returns a pointer to the matched character or NULL if the
character is not found.

5.16.2.22

char * strrev (char * string)

Reverse a string.
The strrev() function reverses the order of the string.
Returns:
The strrev() function returns a pointer to the beginning of the reversed string.

5.16.2.23

char * strsep (char ** string, const char * delim)

Parse a string into tokens.
The strsep() function locates, in the string referenced by *string, the first occurrence
of any character in the string delim (or the terminating '\0' character) and replaces it
with a '\0'. The location of the next character after the delimiter character (or NULL,
if the end of the string was reached) is stored in *string. An "empty" field, i.e. one
caused by two adjacent delimiter characters, can be detected by comparing the location
referenced by the pointer returned in *string to '\0'.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

67

5.16 Strings

Returns:
The strtok r() function returns a pointer to the original value of *string. If *stringp
is initially NULL, strsep() returns NULL.

5.16.2.24

char * strstr (const char * s1, const char * s2)

Locate a substring.
The strstr() function finds the first occurrence of the substring s2 in the string s1. The
terminating '\0' characters are not compared.
Returns:
The strstr() function returns a pointer to the beginning of the substring, or NULL
if the substring is not found. If s2 points to a string of zero length, the function
returns s1.

5.16.2.25

char * strtok r (char * string, const char * delim, char ** last)

Parses the string s into tokens.
strtok r parses the string s into tokens. The first call to strtok r should have string as
its first argument. Subsequent calls should have the first argument set to NULL. If a
token ends with a delimiter, this delimiting character is overwritten with a '\0' and a
pointer to the next character is saved for the next call to strtok r. The delimiter string
delim may be different for each call. last is a user allocated char* pointer. It must be
the same while parsing the same string. strtok r is a reentrant version of strtok().
Returns:
The strtok r() function returns a pointer to the next token or NULL when no more
tokens are found.

5.16.2.26

char * strupr (char * string)

Convert a string to upper case.
The strupr() function will convert a string to upper case. Only the lower case alphabetic
characters [a .. z] are converted. Non-alphabetic characters will not be changed.
Returns:
The strupr() function returns a pointer to the converted string. The pointer is the
same as that passed in since the operation is perform in place.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

68

5.17 Interrupts and Signals

5.17

Interrupts and Signals

5.17.1

Detailed Description

Note:
This discussion of interrupts and signals was taken from Rich Neswold's document. See Acknowledgments.
It's nearly impossible to find compilers that agree on how to handle interrupt code.
Since the C language tries to stay away from machine dependent details, each compiler
writer is forced to design their method of support.
In the AVR-GCC environment, the vector table is predefined to point to interrupt routines with predetermined names. By using the appropriate name, your routine will be
called when the corresponding interrupt occurs. The device library provides a set of
default interrupt routines, which will get used if you don't define your own.
Patching into the vector table is only one part of the problem. The compiler uses, by
convention, a set of registers when it's normally executing compiler-generated code.
It's important that these registers, as well as the status register, get saved and restored.
The extra code needed to do this is enabled by tagging the interrupt function with attribute ((interrupt)).
These details seem to make interrupt routines a little messy, but all these details are
handled by the Interrupt API. An interrupt routine is defined with one of two macros,
INTERRUPT() and SIGNAL(). These macros register and mark the routine as an interrupt handler for the specified peripheral. The following is an example definition of
a handler for the ADC interrupt.
#include & lt; avr/signal.h & gt;
INTERRUPT(SIG_ADC)
{
// user code here
}

Refer to the chapter explaining assembler programming for an explanation about interrupt routines written solely in assembler language.
If an unexpected interrupt occurs (interrupt is enabled and no handler is installed, which
usually indicates a bug), then the default action is to reset the device by jumping to
the reset vector. You can override this by supplying a function named vector default which should be defined with either SIGNAL() or INTERRUPT() as such.
#include & lt; avr/signal.h & gt;
SIGNAL(__vector_default)
{
// user code here
}

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

69

5.17 Interrupts and Signals

70

The interrupt is chosen by supplying one of the symbols in following table. Note that
every AVR device has a different interrupt vector table so some signals might not be
available. Check the data sheet for the device you are using.
[FIXME: Fill in the blanks! Gotta read those durn data sheets ;-)]
Note:
The SIGNAL() and INTERRUPT() macros currently cannot spell-check the argument passed to them. Thus, by misspelling one of the names below in a call to
SIGNAL() or INTERRUPT(), a function will be created that, while possibly being
usable as an interrupt function, is not actually wired into the interrupt vector table.
No warning will be given about this situation.
Signal Name
SIG 2WIRE SERIAL
SIG ADC
SIG COMPARATOR
SIG EEPROM READY
SIG FPGA INTERRUPT0
SIG FPGA INTERRUPT1
SIG FPGA INTERRUPT2
SIG FPGA INTERRUPT3
SIG FPGA INTERRUPT4
SIG FPGA INTERRUPT5
SIG FPGA INTERRUPT6
SIG FPGA INTERRUPT7
SIG FPGA INTERRUPT8
SIG FPGA INTERRUPT9
SIG FPGA INTERRUPT10
SIG FPGA INTERRUPT11
SIG FPGA INTERRUPT12
SIG FPGA INTERRUPT13
SIG FPGA INTERRUPT14
SIG FPGA INTERRUPT15
SIG INPUT CAPTURE1
SIG INPUT CAPTURE3
SIG INTERRUPT0
SIG INTERRUPT1
SIG INTERRUPT2
SIG INTERRUPT3
SIG INTERRUPT4
SIG INTERRUPT5
SIG INTERRUPT6
SIG INTERRUPT7
SIG OUTPUT COMPARE0
SIG OUTPUT COMPARE1A
SIG OUTPUT COMPARE1B
SIG OUTPUT COMPARE1C
SIG OUTPUT COMPARE2
SIG OUTPUT COMPARE3A
SIG OUTPUT COMPARE3B
SIG OUTPUT COMPARE3C
SIG OVERFLOW0

Description
2-wire serial interface (aka. I178C [tm])
ADC Conversion complete
Analog Comparator Interrupt
Eeprom ready

Input Capture1 Interrupt
Input Capture3 Interrupt
External Interrupt0
External Interrupt1
External Interrupt2
External Interrupt3
External Interrupt4
External Interrupt5
External Interrupt6
External Interrupt7
Output Compare0 Interrupt
Output Compare1(A) Interrupt
Output Compare1(B) Interrupt
Output Compare1(C) Interrupt
Output Compare2 Interrupt
Output Compare3(A) Interrupt
Output Compare3(B) Interrupt
Output Compare3(C) Interrupt
Overflow0 Interrupt

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.17 Interrupts and Signals

Signal Name
SIG OVERFLOW1
SIG OVERFLOW2
SIG OVERFLOW3
SIG PIN
SIG PIN CHANGE0
SIG PIN CHANGE1
SIG RDMAC
SIG SPI
SIG SPM READY
SIG SUSPEND RESUME
SIG TDMAC
SIG UART0
SIG UART0 DATA
SIG UART0 RECV
SIG UART0 TRANS
SIG UART1
SIG UART1 DATA
SIG UART1 RECV
SIG UART1 TRANS
SIG UART DATA
SIG UART RECV
SIG UART TRANS
SIG USART0 DATA
SIG USART0 RECV
SIG USART0 TRANS
SIG USART1 DATA
SIG USART1 RECV
SIG USART1 TRANS
SIG USB HW

71

Description
Overflow1 Interrupt
Overflow2 Interrupt
Overflow3 Interrupt

SPI Interrupt
Store program memory ready

UART(0) Data Register Empty Interrupt
UART(0) Receive Complete Interrupt
UART(0) Transmit Complete Interrupt
UART(1) Data Register Empty Interrupt
UART(1) Receive Complete Interrupt
UART(1) Transmit Complete Interrupt
UART Data Register Empty Interrupt
UART Receive Complete Interrupt
UART Transmit Complete Interrupt
USART(0) Data Register Empty Interrupt
USART(0) Receive Complete Interrupt
USART(0) Transmit Complete Interrupt
USART(1) Data Register Empty Interrupt
USART(1) Receive Complete Interrupt
USART(1) Transmit Complete Interrupt

Global manipulation of the interrupt flag
The global interrupt flag is maintained in the I bit of the status register (SREG).
o #define sei() asm
o #define cli() asm

volatile ("sei" ::)
volatile ("cli" ::)

Macros for writing interrupt handler functions
o #define SIGNAL(signame)
o #define INTERRUPT(signame)
o #define EMPTY INTERRUPT(signame)
Allowing specific system-wide interrupts
In addition to globally enabling interrupts, each device's particular interrupt needs to
be enabled separately if interrupts for this device are desired. While some devices

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.17 Interrupts and Signals

72

maintain their interrupt enable bit inside the device's register set, external and timer
interrupts have system-wide configuration registers.
Example:
// Enable timer 1 overflow interrupts.
timer_enable_int(_BV(TOIE1));
// Do some work...
// Disable all timer interrupts.
timer_enable_int(0);

Note:
Be careful when you use these functions. If you already have a different interrupt
enabled, you could inadvertantly disable it by enabling another intterupt.
o void timer enable int (unsigned char ints)
5.17.2
5.17.2.1

Define Documentation
#define cli() asm

volatile ("cli" ::)

#include & lt; avr/interrupt.h & gt;

Disables all interrupts by clearing the global interrupt mask. This function actually
compiles into a single line of assembly, so there is no function call overhead.
5.17.2.2 #define EMPTY INTERRUPT(signame)
Value:
void signame (void) __attribute__ ((naked));
\
void signame (void) { __asm__ __volatile__ ( " reti " ::); }
#include & lt; avr/signal.h & gt;

Defines an empty interrupt handler function. This will not generate any prolog or
epilog code and will only return from the ISR. Do not define a function body as this
will define it for you. Example:
EMPTY_INTERRUPT(SIG_ADC);

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.17 Interrupts and Signals

73

5.17.2.3 #define INTERRUPT(signame)
Value:
void signame (void) __attribute__ ((interrupt));
void signame (void)

\

#include & lt; avr/signal.h & gt;

Introduces an interrupt handler function that runs with global interrupts initially enabled. This allows interrupt handlers to be interrupted.
5.17.2.4 #define sei() asm

volatile ("sei" ::)

#include & lt; avr/interrupt.h & gt;

Enables interrupts by clearing the global interrupt mask. This function actually compiles into a single line of assembly, so there is no function call overhead.
5.17.2.5 #define SIGNAL(signame)
Value:
void signame (void) __attribute__ ((signal));
void signame (void)

\

#include & lt; avr/signal.h & gt;

Introduces an interrupt handler function that runs with global interrupts initially disabled.
5.17.3

Function Documentation

5.17.3.1 void timer enable int (unsigned char ints)
#include & lt; avr/interrupt.h & gt;

This function modifies the timsk register. The value you pass via ints is device
specific.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

5.18 Special function registers

5.18

Special function registers

5.18.1

Detailed Description

When working with microcontrollers, many of the tasks usually consist of controlling
the peripherals that are connected to the device, respectively programming the subsystems that are contained in the controller (which by itself communicate with the circuitry
connected to the controller).
The AVR series of microcontrollers offers two different paradigms to perform this task.
There's a separate IO address space available (as it is known from some high-level
CISC CPUs) that can be addressed with specific IO instructions that are applicable to
some or all of the IO address space (in, out, sbi etc.). The entire IO address space
is also made available as memory-mapped IO, i. e. it can be accessed using all the
MCU instructions that are applicable to normal data memory. The IO register space is
mapped into the data memory address space with an offset of 0x20 since the bottom
of this space is reserved for direct access to the MCU registers. (Actual SRAM is
available only behind the IO register area, starting at either address 0x60, or 0x100
depending on the device.)
AVR Libc supports both these paradigms. While by default, the implementation uses
memory-mapped IO access, this is hidden from the programmer. So the programmer
can access IO registers either with a special function like outb():
#include & lt; avr/io.h & gt;
outb(PORTA, 0x33);

or they can assign a value directly to the symbolic address:
PORTA = 0x33;

The compiler's choice of which method to use when actually accessing the IO port is
completely independent of the way the programmer chooses to write the code. So even
if the programmer uses the memory-mapped paradigm and writes
PORTA |= 0x40;

the compiler can optimize this into the use of an sbi instruction (of course, provided
the target address is within the allowable range for this instruction, and the right-hand
side of the expression is a constant value known at compile-time).
The advantage of using the memory-mapped paradigm in C programs is that it makes
the programs more portable to other C compilers for the AVR platform. Some people
might also feel that this is more readable. For example, the following two statements
would be equivalent:

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

74

5.18 Special function registers

outb(DDRD, inb(DDRD) & ~LCDBITS);
DDRD & = ~LCDBITS;

The generated code is identical for both. Whitout optimization, the compiler strictly
generates code following the memory-mapped paradigm, while with optimization
turned on, code is generated using the (faster and smaller) in/out MCU instructions.
Note that special care must be taken when accessing some of the 16-bit timer IO registers where access from both the main program and within an interrupt context can
happen. See Why do some 16-bit timer registers sometimes get trashed?.
Modules
o Additional notes from & lt; avr/sfr defs.h & gt;
Bit manipulation
o #define BV(bit) (1 & lt; & lt; (bit))
IO register bit manipulation
o
o
o
o

#define bit is set(sfr, bit) (sfr & BV(bit))
#define bit is clear(sfr, bit) (!(sfr & BV(bit)))
#define loop until bit is set(sfr, bit) do { } while (bit is clear(sfr, bit))
#define loop until bit is clear(sfr, bit) do { } while (bit is set(sfr, bit))

Deprecated Macros
o
o
o
o
o
o
o
o
o

#define cbi(sfr, bit) ( SFR BYTE(sfr) & = ~ BV(bit))
#define sbi(sfr, bit) ( SFR BYTE(sfr) |= BV(bit))
#define inb(sfr) SFR BYTE(sfr)
#define outb(sfr, val) ( SFR BYTE(sfr) = (val))
#define inw(sfr) SFR WORD(sfr)
#define outw(sfr, val) ( SFR WORD(sfr) = (val))
#define outp(val, sfr) outb(sfr, val)
#define inp(sfr) inb(sfr)
#define BV(bit) BV(bit)

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

75

5.18 Special function registers

5.18.2

Define Documentation

5.18.2.1 #define BV(bit) (1 & lt; & lt; (bit))
#include & lt; avr/io.h & gt;

Converts a bit number into a byte value.
Note:
The bit shift is performed by the compiler which then inserts the result into the
code. Thus, there is no run-time overhead when using BV().

5.18.2.2

#define bit is clear(sfr, bit) (!(sfr & BV(bit)))

#include & lt; avr/io.h & gt;

Test whether bit bit in IO register sfr is clear. This will return non-zero if the bit is
clear, and a 0 if the bit is set.
5.18.2.3 #define bit is set(sfr, bit) (sfr & BV(bit))
#include & lt; avr/io.h & gt;

Test whether bit bit in IO register sfr is set. This will return a 0 if the bit is clear,
and non-zero if the bit is set.
5.18.2.4

#define BV(bit) BV(bit)

Deprecated:
For backwards compatibility only. This macro will eventually be removed.
Use BV() in new programs.
5.18.2.5

#define cbi(sfr, bit) ( SFR BYTE(sfr) & = ~ BV(bit))

Deprecated:
#include & lt; avr/io.h & gt;

For backwards compatibility only. This macro will eventually be removed.
Clear bit bit in IO register sfr.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

76

5.18 Special function registers

5.18.2.6 #define inb(sfr) SFR BYTE(sfr)
Deprecated:
#include & lt; avr/io.h & gt;

For backwards compatibility only. This macro will eventually be removed.
Use direct access in new programs.
5.18.2.7

#define inp(sfr) inb(sfr)

Deprecated:
For backwards compatibility only. This macro will eventually be removed.
Use direct access in new programs.
5.18.2.8

#define inw(sfr) SFR WORD(sfr)

Deprecated:
#include & lt; avr/io.h & gt;

For backwards compatibility only. This macro will eventually be removed.
Read a 16-bit word from IO register pair sfr.
Use direct access in new programs.
5.18.2.9

#define loop until bit is clear(sfr, bit) do { } while (bit is set(sfr, bit))

#include & lt; avr/io.h & gt;

Wait until bit bit in IO register sfr is clear.
5.18.2.10

#define loop until bit is set(sfr, bit) do { } while (bit is clear(sfr, bit))

#include & lt; avr/io.h & gt;

Wait until bit bit in IO register sfr is set.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

77

5.18 Special function registers

5.18.2.11

78

#define outb(sfr, val) ( SFR BYTE(sfr) = (val))

Deprecated:
\end{Desc}

For backwards compatibility only. This macro will eventually be removed.
{\bf Use direct access in new programs}.

\begin{Desc}
\item[Note: ]\par
The order of the arguments was switched in older versions of avr-libc (versions $ & lt; $= 20020203). \
\hypertarget{group__avr__sfr_a11}{
\index{avr_sfr@{avr\_\-sfr}!outp@{outp}}
\index{outp@{outp}!avr_sfr@{avr\_\-sfr}}
\paragraph[outp]{\setlength{\rightskip}{0pt plus 5cm}\#define outp(val, sfr)\ outb(sfr, val)}\hfi
\label{group__avr__sfr_a11}

\begin{Desc}
\item[\hyperlink{deprecated__deprecated000011}{Deprecated: }]\par
For backwards compatibility only. This macro will eventually be removed.\end{Desc}

{\bf Use direct access in new programs}. \hypertarget{group__avr__sfr_a10}{
\index{avr_sfr@{avr\_\-sfr}!outw@{outw}}
\index{outw@{outw}!avr_sfr@{avr\_\-sfr}}
\paragraph[outw]{\setlength{\rightskip}{0pt plus 5cm}\#define outw(sfr, val)\ (\_\-SFR\_\-WORD(sf
\label{group__avr__sfr_a10}

\begin{Desc}
\item[\hyperlink{deprecated__deprecated000010}{Deprecated: }]\par

\footnotesize\begin{verbatim}#include & lt; avr/io.h & gt;

For backwards compatibility only. This macro will eventually be removed.
Write the 16-bit value val to IO register pair sfr. Care will be taken to write the
lower register first. When used to update 16-bit registers where the timing is critical
and the operation can be interrupted, the programmer is the responsible for disabling
interrupts before accessing the register pair.
Use direct access in new programs.
Note:
The order of the arguments was switched in older versions of avr-libc (versions
& lt; = 20020203).

5.18.2.12

#define sbi(sfr, bit) ( SFR BYTE(sfr) |= BV(bit))

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

6 avr-libc Data Structure Documentation

Deprecated:
#include & lt; avr/io.h & gt;

For backwards compatibility only. This macro will eventually be removed.
Set bit bit in IO register sfr.

6

avr-libc Data Structure Documentation

6.1

div t Struct Reference

6.1.1

Detailed Description

Result type for function div().
The documentation for this struct was generated from the following file:
o stdlib.h

6.2

ldiv t Struct Reference

6.2.1

Detailed Description

Result type for function ldiv().
The documentation for this struct was generated from the following file:
o stdlib.h

7
7.1

avr-libc Page Documentation
Acknowledgments

This document tries to tie together the labors of a large group of people. Without
these individuals' efforts, we wouldn't have a terrific, free set of tools to develop AVR
projects. We all owe thanks to:
o The GCC Team, which produced a very capable set of development tools for an
amazing number of platforms and processors.
o Denis Chertykov [ denisc@overta.ru ] for making the AVR-specific
changes to the GNU tools.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

79

7.2

avr-libc and assembler programs

o Denis Chertykov and Marek Michalkiewicz [ marekm@linux.org.pl ] for
developing the standard libraries and startup code for AVR-GCC.
o Uros Platise for developing the AVR programmer tool, uisp.
o Joerg Wunsch [ joerg@FreeBSD.ORG ] for adding all the AVR development
tools to the FreeBSD [ http://www.freebsd.org ] ports tree and for providing the basics for the demo project.
o Brian Dean [ bsd@bsdhome.com ] for developing avrdude (an alternative to
uisp) and for contributing documentation which describes how to use it. Avrdude was previously called avrprog.
o Eric Weddington [ eric@ecentral.com ] for maintaining the WinAVR
package and thus making the continued improvements to the Opensource AVR
toolchain available to many users.
o Rich Neswold for writing the original avr-tools document (which he graciously
allowed to be merged into this document) and his improvements to the demo
project.
o All the people who have submitted suggestions, patches and bug reports. (See
the AUTHORS files of the various tools.)
o And lastly, all the users who use the software. If nobody used the software, we
would probably not be very motivated to continue to develop it. Keep those bug
reports coming. ;-)

7.2

avr-libc and assembler programs

7.2.1

Introduction

There might be several reasons to write code for AVR microcontrollers using plain
assembler source code. Among them are:
o Code for devices that do not have RAM and are thus not supported by the C
compiler.
o Code for very time-critical applications.
o Special tweaks that cannot be done in C.
Usually, all but the first could probably be done easily using the inline assembler facility
of the compiler.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

80

7.2

avr-libc and assembler programs

81

Although avr-libc is primarily targeted to support programming AVR microcontrollers
using the C (and C++) language, there's limited support for direct assembler usage as
well. The benefits of it are:
o Use of the C preprocessor and thus the ability to use the same symbolic constants
that are available to C programs, as well as a flexible macro concept that can use
any valid C identifier as a macro (whereas the assembler's macro concept is
basically targeted to use a macro in place of an assembler instruction).
o Use of the runtime framework like automatically assigning interrupt vectors. For
devices that have RAM, initializing the RAM variables can also be utilized.
7.2.2

Invoking the compiler

For the purpose described in this document, the assembler and linker are usually not
invoked manually, but rather using the C compiler frontend (avr-gcc) that in turn
will call the assembler and linker as required.
This approach has the following advantages:
o There is basically only one program to be called directly, avr-gcc, regardless
of the actual source language used.
o The invokation of the C preprocessor will be automatic, and will include the
appropriate options to locate required include files in the filesystem.
o The invokation of the linker will be automatic, and will include the appropriate options to locate additional libraries as well as the application start-up code
(crtXXX.o) and linker script.
Note that the invokation of the C preprocessor will be automatic when the filename
provided for the assembler file ends in .S (the capital letter "s"). This would even apply
to operating systems that use case-insensitive filesystems since the actual decision is
made based on the case of the filename suffix given on the command-line, not based on
the actual filename from the file system.
Alternatively, the language can
assembler-with-cpp option.
7.2.3

explicitly

be

specified

using

the

-x

Example program

The following annotated example features a simple 100 kHz square wave generator
using an AT90S1200 clocked with a 10.7 MHz crystal. Pin PD6 will be used for the
square wave output.
#include & lt; avr/io.h & gt;

; Note [1]

work

; Note [2]

=

16

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.2

avr-libc and assembler programs

tmp

=

17

inttmp

=

19

intsav

=

0

SQUARE

=

PD6

82

tmconst= 10700000 / 200000
fuzz=
8

; Note [3]
; Note [4]:
; 100 kHz = & gt; 200000 edges/s
; # clocks in ISR until TCNT0 is set

.section .text
.global main

; Note [5]

main:
rcall

ioinit

rjmp

1b

1:

.global SIG_OVERFLOW0
SIG_OVERFLOW0:
ldi
inttmp, 256 - tmconst + fuzz
out
_SFR_IO_ADDR(TCNT0), inttmp
in
sbic
rjmp
sbi
rjmp
cbi

_SFR_IO_ADDR(DDRD), SQUARE

ldi
out

work, _BV(TOIE0)
_SFR_IO_ADDR(TIMSK), work

ldi
out

work, _BV(CS00)
; tmr0:
_SFR_IO_ADDR(TCCR0), work

ldi
out

; Note [8]

_SFR_IO_ADDR(SREG), intsav

sbi

; Note [7]

_SFR_IO_ADDR(PORTD), SQUARE
1f
_SFR_IO_ADDR(PORTD), SQUARE
2f
_SFR_IO_ADDR(PORTD), SQUARE

out
reti

1:
2:

intsav, _SFR_IO_ADDR(SREG)

; Note [6]

; Note [9]

work, 256 - tmconst
_SFR_IO_ADDR(TCNT0), work

ioinit:

CK/1

sei
ret
.global __vector_default
__vector_default:

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

; Note [10]

7.2

avr-libc and assembler programs

reti
.end

Note [1]
As in C programs, this includes the central processor-specific file containing the
IO port definitions for the device. Note that not all include files can be included
into assembler sources.
Note [2]
Assignment of registers to symbolic names used locally. Another option would be
to use a C preprocessor macro instead:
#define work 16

Note [3]
Our bit number for the square wave output. Note that the right-hand side consists of a CPP macro which will be substituted by its value (6 in this case) before
actually being passed to the assembler.
Note [4]
The assembler uses integer operations in the host-defined integer size (32 bits or
longer) when evaluating expressions. This is in contrast to the C compiler that
uses the C type int by default in order to calculate constant integer expressions.
In order to get a 100 kHz output, we need to toggle the PD6 line 200000 times
per second. Since we use timer 0 without any prescaling options in order to get
the desired frequency and accuracy, we already run into serious timing considerations: while accepting and processing the timer overflow interrupt, the timer
already continues to count. When pre-loading the TCCNT0 register, we therefore
have to account for the number of clock cycles required for interrupt acknowledge
and for the instructions to reload TCCNT0 (4 clock cycles for interrupt acknowledge, 2 cycles for the jump from the interrupt vector, 2 cycles for the 2 instructions
that reload TCCNT0). This is what the constant fuzz is for.
Note [5]
External functions need to be declared to be .global. main is the application entry point that will be jumped to from the ininitalization routine in crts1200.o.
Note [6]
The main loop is just a single jump back to itself. Square wave generation itself is
completely handled by the timer 0 overflow interrupt service. A sleep instruction
(using idle mode) could be used as well, but probably would not conserve much
energy anyway since the interrupt service is executed quite frequently.
Note [7]
Interrupt functions can get the usual names that are also available to C programs.
The linker will then put them into the appropriate interrupt vector slots. Note
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

83

7.3

Frequently Asked Questions

that they must be declared .global in order to be acceptable for this purpose.
This will only work if & lt; avr/io.h & gt; has been included. Note that the assembler
or linker have no chance to check the correct spelling of an interrupt function,
so it should be double-checked. (When analyzing the resulting object file using
avr-objdump or avr-nm, a name like vector N should appear, with N
being a small integer number.)
Note [8]
As explained in the section about special function registers, the actual IO port
address should be obtained using the macro SFR IO ADDR. (The AT90S1200
does not have RAM thus the memory-mapped approach to access the IO registers
is not available. It would be slower than using in / out instructions anyway.)
Since the operation to reload TCCNT0 is time-critical, it is even performed before
saving SREG. Obviously, this requires that the instructions involved would not
change any of the flag bits in SREG.
Note [9]
Interrupt routines must not clobber the global CPU state. Thus, it is usually necessary to save at least the state of the flag bits in SREG. (Note that this serves as an
example here only since actually, all the following instructions would not modify
SREG either, but that's not commonly the case.)
Also, it must be made sure that registers used inside the interrupt routine do
not conflict with those used outside. In the case of a RAM-less device like the
AT90S1200, this can only be done by agreeing on a set of registers to be used
exclusively inside the interrupt routine; there would not be any other chance to
"save" a register anywhere.
If the interrupt routine is to be linked together with C modules, care must be taken
to follow the register usage guidelines imposed by the C compiler. Also, any
register modified inside the interrupt sevice needs to be saved, usually on the stack.
Note [10]
As explained in Interrupts and Signals, a global "catch-all" interrupt handler that
gets all unassigned interrupt vectors can be installed using the name vector default. This must be .global, and obviously, should end in a reti instruction. (By default, a jump to location 0 would be implied instead.)

7.3
7.3.1

Frequently Asked Questions
FAQ Index

1. My program doesn't recognize a variable updated within an interrupt routine
2. I get "undefined reference to..." for functions like "sin()"
3. How to permanently bind a variable to a register?
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

84

7.3

Frequently Asked Questions

4. How to modify MCUCR or WDTCR early?
5. What is all this BV() stuff about?
6. Can I use C++ on the AVR?
7. Shouldn't I initialize all my variables?
8. Why do some 16-bit timer registers sometimes get trashed?
9. How do I use a #define'd constant in an asm statement?
10. Why does the PC randomly jump around when single-stepping through my program in avr-gdb?
11. How do I trace an assembler file in avr-gdb?
12. How do I pass an IO port as a parameter to a function?
13. What registers are used by the C compiler?
14. How do I put an array of strings completely in ROM?
15. How to use external RAM?
16. Which -O flag to use?
17. How do I relocate code to a fixed address?
18. My UART is generating nonsense! My ATmega128 keeps crashing! Port F is
completely broken!
19. Why do all my "foo...bar" strings eat up the SRAM?
20. Why does the compiler compile an 8-bit operation that uses bitwise operators
into a 16-bit operation in assembly?
21. How to detect RAM memory and variable overlap problems?
7.3.2

My program doesn't recognize a variable updated within an interrupt routine

When using the optimizer, in a loop like the following one:
uint8_t flag;
...
while (flag == 0) {
...
}

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

85

7.3

Frequently Asked Questions

the compiler will typically optimize the access to flag completely away, since its
code path analysis shows that nothing inside the loop could change the value of flag
anyway. To tell the compiler that this variable could be changed outside the scope of
its code path analysis (e. g. from within an interrupt routine), the variable needs to be
declared like:
volatile uint8_t flag;

Back to FAQ Index.

7.3.3

I get "undefined reference to..." for functions like "sin()"

In order to access the mathematical functions that are declared in & lt; math.h & gt; , the
linker needs to be told to also link the mathematical library, libm.a.
Typically, system libraries like libm.a are given to the final C compiler command
line that performs the linking step by adding a flag -lm at the end. (That is, the initial
lib and the filename suffix from the library are written immediately after a -l flag. So
for a libfoo.a library, -lfoo needs to be provided.) This will make the linker
search the library in a path known to the system.
An alternative would be to specify the full path to the libm.a file at the same place
on the command line, i. e. after all the object files (*.o). However, since this requires knowledge of where the build system will exactly find those library files, this is
deprecated for system libraries.
Back to FAQ Index.

7.3.4

How to permanently bind a variable to a register?

This can be done with
register unsigned char counter asm( " r3 " );

See C Names Used in Assembler Code for more details.
Back to FAQ Index.

7.3.5

How to modify MCUCR or WDTCR early?

The method of early initialization (MCUCR, WDTCR or anything else) is different (and
more flexible) in the current version. Basically, write a small assembler file which
looks like this:
;; begin xram.S

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

86

7.3

Frequently Asked Questions

#include & lt; avr/io.h & gt;
.section .init1, " ax " ,@progbits
ldi r16,_BV(SRE) | _BV(SRW)
out _SFR_IO_ADDR(MCUCR),r16
;; end xram.S

Assemble it, link the resulting xram.o with other files in your program, and this piece
of code will be inserted in initialization code, which is run right after reset. See the
linker script for comments about the new .initN sections (which one to use, etc.).
The advantage of this method is that you can insert any initialization code you want
(just remember that this is very early startup - no stack and no zero reg yet), and
no program memory space is wasted if this feature is not used.
There should be no need to modify linker scripts anymore, except for some very special cases. It is best to leave stack at its default value (end of internal SRAM
- faster, and required on some devices like ATmega161 because of errata), and add
-Wl,-Tdata,0x801100 to start the data section above the stack.
For more information on using sections, including how to use them from C code, see
Memory Sections.
Back to FAQ Index.

7.3.6

What is all this BV() stuff about?

When performing low-level output work, which is a very central point in microcontroller programming, it is quite common that a particular bit needs to be set or cleared
in some IO register. While the device documentation provides mnemonic names for
the various bits in the IO registers, and the AVR device-specific IO definitions reflect
these names in definitions for numerical constants, a way is needed to convert a bit
number (usually within a byte register) into a byte value that can be assigned directly
to the register. However, sometimes the direct bit numbers are needed as well (e. g. in
an sbi() call), so the definitions cannot usefully be made as byte values in the first
place.
So in order to access a particular bit number as a byte value, use the BV() macro. Of
course, the implementation of this macro is just the usual bit shift (which is done by the
compiler anyway, thus doesn't impose any run-time penalty), so the following applies:
_BV(3) = & gt; 1 & lt; & lt; 3 = & gt; 0x08

However, using the macro often makes the program better readable.
"BV" stands for "bit value", in case someone might ask you. :-)
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

87

7.3

Frequently Asked Questions

Example: clock timer 2 with full IO clock (CS2x = 0b001), toggle OC2 output on
compare match (COM2x = 0b01), and clear timer on compare match (CTC2 = 1). Make
OC2 (PD7) an output.
TCCR2 = _BV(COM20)|_BV(CTC2)|_BV(CS20);
DDRD = _BV(PD7);
Back to FAQ Index.

7.3.7

Can I use C++ on the AVR?

Basically yes, C++ is supported (assuming your compiler has been configured and
compiled to support it, of course). Source files ending in .cc, .cpp or .C will automatically cause the compiler frontend to invoke the C++ compiler. Alternatively, the C++
compiler could be explicitly called by the name avr-c++.
However, there's currently no support for libstdc++, the standard support library
needed for a complete C++ implementation. This imposes a number of restrictions on
the C++ programs that can be compiled. Among them are:
o Obviously, none of the C++ related standard functions, classes, and template
classes are available.
o The operators new and delete are not implemented, attempting to use them
will cause the linker to complain about undefined external references. (This
could perhaps be fixed.)
o Some of the supplied include files are not C++ safe, i. e. they need to be wrapped
into
extern " C " { . . . }

(This could certainly be fixed, too.)
o Exceptions are not supported. Since exceptions are enabled by default in the
C++ frontend, they explicitly need to be turned off using -fno-exceptions
in the compiler options. Failing this, the linker will complain about an undefined
external reference to gxx personality sj0.
Constructors and destructors are supported though, including global ones.
When programming C++ in space- and runtime-sensitive environments like microcontrollers, extra care should be taken to avoid unwanted side effects of the C++ calling
conventions like implied copy constructors that could be called upon function invocation etc. These things could easily add up into a considerable amount of time and
program memory wasted. Thus, casual inspection of the generated assembler code
(using the -S compiler option) seems to be warranted.
Back to FAQ Index.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

88

7.3

7.3.8

Frequently Asked Questions

Shouldn't I initialize all my variables?

Global and static variables are guaranteed to be initialized to 0 by the C standard.
avr-gcc does this by placing the appropriate code into section .init4 (see The
.initN Sections). With respect to the standard, this sentence is somewhat simplified
(because the standard allows for machines where the actual bit pattern used differs
from all bits being 0), but for the AVR target, in general, all integer-type variables are
set to 0, all pointers to a NULL pointer, and all floating-point variables to 0.0.
As long as these variables are not initialized (i. e. they don't have an equal sign and
an initialization expression to the right within the definition of the variable), they go
into the .bss section of the file. This section simply records the size of the variable,
but otherwise doesn't consume space, neither within the object file nor within flash
memory. (Of course, being a variable, it will consume space in the target's SRAM.)
In contrast, global and static variables that have an initializer go into the .data section
of the file. This will cause them to consume space in the object file (in order to record
the initializing value), and in the flash ROM of the target device. The latter is needed
since the flash ROM is the only way that the compiler can tell the target device the
value this variable is going to be initialized to.
Now if some programmer "wants to make doubly sure" their variables really get a 0
at program startup, and adds an initializer just containing 0 on the right-hand side,
they waste space. While this waste of space applies to virtually any platform C is
implemented on, it's usually not noticeable on larger machines like PCs, while the
waste of flash ROM storage can be very painful on a small microcontroller like the
AVR.
So in general, variables should only be explicitly initialized if the initial value is nonzero.
Back to FAQ Index.

7.3.9

Why do some 16-bit timer registers sometimes get trashed?

Some of the timer-related 16-bit IO registers use a temporary register (called TEMP in
the Atmel datasheet) to guarantee an atomic access to the register despite the fact that
two separate 8-bit IO transfers are required to actually move the data. Typically, this
includes access to the current timer/counter value register (TCNTn), the input capture
register (ICRn), and write access to the output compare registers (OCRnM). Refer to
the actual datasheet for each device's set of registers that involves the TEMP register.
When accessing one of the registers that use TEMP from the main application, and
possibly any other one from within an interrupt routine, care must be taken that no
access from within an interrupt context could clobber the TEMP register data of an
in-progress transaction that has just started elsewhere.
To protect interrupt routines against other interrupt routines, it's usually best to use the

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

89

7.3

Frequently Asked Questions

90

SIGNAL() macro when declaring the interrupt function, and to ensure that interrupts
are still disabled when accessing those 16-bit timer registers.
Within the main program, access to those registers could be encapsulated in calls to the
cli() and sei() macros. If the status of the global interrupt flag before accessing one of
those registers is uncertain, something like the following example code can be used.
uint16_t
read_timer1(void)
{
uint8_t sreg;
uint16_t val;
sreg = SREG;
cli();
val = TCNT1;
SREG = sreg;
return val;
}

Back to FAQ Index.

7.3.10

How do I use a #define'd constant in an asm statement?

So you tried this:
asm volatile( " sbi 0x18,0x07; " );

Which works. When you do the same thing but replace the address of the port by its
macro name, like this:
asm volatile( " sbi PORTB,0x07; " );

you get a compilation error: " Error:

constant value required " .

PORTB is a precompiler definition included in the processor specific file included in
avr/io.h. As you may know, the precompiler will not touch strings and PORTB,
instead of 0x18, gets passed to the assembler. One way to avoid this problem is:
asm volatile( " sbi %0, 0x07 " : " I " (PORTB):);

Note:
avr/io.h already provides a sbi() macro definition, which can be used in C
programs.
Back to FAQ Index.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.3

Frequently Asked Questions

7.3.11

Why does the PC randomly jump around when single-stepping through
my program in avr-gdb?

When compiling a program with both optimization (-O) and debug information (-g)
which is fortunately possible in avr-gcc, the code watched in the debugger is optimized code. While it is not guaranteed, very often this code runs with the exact same
optimizations as it would run without the -g switch.
This can have unwanted side effects. Since the compiler is free to reorder code execution as long as the semantics do not change, code is often rearranged in order to
make it possible to use a single branch instruction for conditional operations. Branch
instructions can only cover a short range for the target PC (-63 through +64 words from
the current PC). If a branch instruction cannot be used directly, the compiler needs to
work around it by combining a skip instruction together with a relative jump (rjmp)
instruction, which will need one additional word of ROM.
Another side effect of optimzation is that variable usage is restricted to the area of code
where it is actually used. So if a variable was placed in a register at the beginning of
some function, this same register can be re-used later on if the compiler notices that the
first variable is no longer used inside that function, even though the variable is still in
lexical scope. When trying to examine the variable in avr-gdb, the displayed result
will then look garbled.
So in order to avoid these side effects, optimization can be turned off while debugging.
However, some of these optimizations might also have the side effect of uncovering
bugs that would otherwise not be obvious, so it must be noted that turning off optimization can easily change the bug pattern. In most cases, you are better off leaving
optimizations enabled while debugging.
Back to FAQ Index.

7.3.12

How do I trace an assembler file in avr-gdb?

When using the -g compiler option, avr-gcc only generates line number and other
debug information for C (and C++) files that pass the compiler. Functions that don't
have line number information will be completely skipped by a single step command
in gdb. This includes functions linked from a standard library, but by default also
functions defined in an assembler source file, since the -g compiler switch does not
apply to the assembler.
So in order to debug an assembler input file (possibly one that has to be passed through
the C preprocessor), it's the assembler that needs to be told to include line-number
information into the output file. (Other debug information like data types and variable
allocation cannot be generated, since unlike a compiler, the assembler basically doesn't
know about this.) This is done using the (GNU) assembler option --gstabs.
Example:

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

91

7.3

Frequently Asked Questions

92

$ avr-as -mmcu=atmega128 --gstabs -o foo.o foo.s

When the assembler is not called directly but through the C compiler frontend
(either implicitly by passing a source file ending in .S, or explicitly using -x
assembler-with-cpp), the compiler frontend needs to be told to pass the
--gstabs option down to the assembler. This is done using -Wa,--gstabs.
Please take care to only pass this option when compiling an assembler input file. Otherwise, the assembler code that results from the C compilation stage will also get line
number information, which confuses the debugger.
Note:
You can also use -Wa,-gstabs since the compiler will add the extra '-' for you.
Example:
$ EXTRA_OPTS= " -Wall -mmcu=atmega128 -x assembler-with-cpp "
$ avr-gcc -Wa,--gstabs ${EXTRA_OPTS} -c -o foo.o foo.S

Also note that the debugger might get confused when entering a piece of code that has
a non-local label before, since it then takes this label as the name of a new function that
appears to have been entered. Thus, the best practice to avoid this confusion is to only
use non-local labels when declaring a new function, and restrict anything else to local
labels. Local labels consist just of a number only. References to these labels consist
of the number, followed by the letter b for a backward reference, or f for a forward
reference. These local labels may be re-used within the source file, references will pick
the closest label with the same number and given direction.
Example:
myfunc: push
push
push
push
push
...
eor
ldi
ldi
rjmp
1:
ld
...
breq
...
inc
2:
cmp
brlo

r16
r17
r18
YL
YH

1:

YH
YL

pop
pop

r16, r16
; start loop
YL, lo8(sometable)
YH, hi8(sometable)
2f
; jump to loop test at end
r17, Y+
; loop continues here
1f

; return from myfunc prematurely

r16
r16, r18
1b

; jump back to top of loop

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.3

Frequently Asked Questions

pop
pop
pop
ret

93

r18
r17
r16

Back to FAQ Index.

7.3.13

How do I pass an IO port as a parameter to a function?

Consider this example code:
#include & lt; inttypes.h & gt;
#include & lt; avr/io.h & gt;
void
set_bits_func_wrong (volatile uint8_t port, uint8_t mask)
{
port |= mask;
}
void
set_bits_func_correct (volatile uint8_t *port, uint8_t mask)
{
*port |= mask;
}
#define set_bits_macro(port,mask) ((port) |= (mask))
int main (void)
{
set_bits_func_wrong (PORTB, 0xaa);
set_bits_func_correct ( & PORTB, 0x55);
set_bits_macro (PORTB, 0xf0);
return (0);
}

The first function will generate object code which is not even close to what is intended.
The major problem arises when the function is called. When the compiler sees this call,
it will actually pass the value of the PORTB register (using an IN instruction), instead
of passing the address of PORTB (e.g. memory mapped io addr of 0x38, io port 0x18
for the mega128). This is seen clearly when looking at the disassembly of the call:
set_bits_func_wrong
10a:
6a ea
10c:
88 b3
10e:
0e 94 65 00

(PORTB,
ldi
in
call

0xaa);
r22, 0xAA
r24, 0x18
0xca

; 170
; 24

So, the function, once called, only sees the value of the port register and knows nothing
about which port it came from. At this point, whatever object code is generated for
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.3

Frequently Asked Questions

94

the function by the compiler is irrelevant. The interested reader can examine the full
disassembly to see that the function's body is completely fubar.
The second function shows how to pass (by reference) the memory mapped address of
the io port to the function so that you can read and write to it in the function. Here's
the object code generated for the function call:
set_bits_func_correct ( & PORTB, 0x55);
112:
65 e5
ldi
r22, 0x55
114:
88 e3
ldi
r24, 0x38
116:
90 e0
ldi
r25, 0x00
118:
0e 94 7c 00
call
0xf8

; 85
; 56
;0

You can clearly see that 0x0038 is correctly passed for the address of the io port.
Looking at the disassembled object code for the body of the function, we can see that
the function is indeed performing the operation we intended:
void
set_bits_func_correct (volatile uint8_t *port, uint8_t mask)
{
f8:
fc 01
movw
r30, r24
*port |= mask;
fa:
80 81
ld
r24, Z
fc:
86 2b
or
r24, r22
fe:
80 83
st
Z, r24
}
100:
08 95
ret

Notice that we are accessing the io port via the LD and ST instructions.
The port parameter must be volatile to avoid a compiler warning.
Note:
Because of the nature of the IN and OUT assembly instructions, they can not be
used inside the function when passing the port in this way. Readers interested in
the details should consult the Instruction Set data sheet.
Finally we come to the macro version of the operation. In this contrived example, the
macro is the most efficient method with respect to both execution speed and code size:
set_bits_macro (PORTB, 0xf0);
11c:
88 b3
in
r24, 0x18
11e:
80 6f
ori
r24, 0xF0
120:
88 bb
out
0x18, r24

; 24
; 240
; 24

Of course, in a real application, you might be doing a lot more in your function which
uses a passed by reference io port address and thus the use of a function over a macro
could save you some code space, but still at a cost of execution speed.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.3

Frequently Asked Questions

95

Care should be taken when such an indirect port access is going to one of the 16-bit
IO registers where the order of write access is critical (like some timer registers). All
versions of avr-gcc up to 3.3 will generate instructions that use the wrong access order
in this situation (since with normal memory operands where the order doesn't matter,
this sometimes yields shorter code).
See http://mail.nongnu.org/archive/html/avr-libc-dev/2003-01/msg00044.html
for a possible workaround.
avr-gcc versions after 3.3 have been fixed in a way where this optimization will be
disabled if the respective pointer variable is declared to be volatile, so the correct
behaviour for 16-bit IO ports can be forced that way.
Back to FAQ Index.

7.3.14

What registers are used by the C compiler?

o Data types:
char is 8 bits, int is 16 bits, long is 32 bits, long long is 64 bits, float and
double are 32 bits (this is the only supported floating point format), pointers
are 16 bits (function pointers are word addresses, to allow addressing the whole
128K program memory space on the ATmega devices with & gt; 64 KB of flash
ROM). There is a -mint8 option (see Options for the C compiler avr-gcc) to
make int 8 bits, but that is not supported by avr-libc and violates C standards
(int must be at least 16 bits). It may be removed in a future release.
o Call-used registers (r18-r27, r30-r31):
May be allocated by gcc for local data. You may use them freely in assembler subroutines. Calling C subroutines can clobber any of them - the caller is
responsible for saving and restoring.
o Call-saved registers (r2-r17, r28-r29):
May be allocated by gcc for local data. Calling C subroutines leaves them unchanged. Assembler subroutines are responsible for saving and restoring these
registers, if changed. r29:r28 (Y pointer) is used as a frame pointer (points to
local data on stack) if necessary.
o Fixed registers (r0, r1):
Never allocated by gcc for local data, but often used for fixed purposes:
r0 - temporary register, can be clobbered by any C code (except interrupt handlers which save it), may be used to remember something for a while within one
piece of assembler code
r1 - assumed to be always zero in any C code, may be used to remember something for a while within one piece of assembler code, but must then be cleared
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.3

Frequently Asked Questions

after use (clr r1). This includes any use of the [f]mul[s[u]] instructions,
which return their result in r1:r0. Interrupt handlers save and clear r1 on entry,
and restore r1 on exit (in case it was non-zero).
o Function call conventions:
Arguments - allocated left to right, r25 to r8. All arguments are aligned to start in
even-numbered registers (odd-sized arguments, including char, have one free
register above them). This allows making better use of the movw instruction on
the enhanced core.
If too many, those that don't fit are passed on the stack.
Return values: 8-bit in r24 (not r25!), 16-bit in r25:r24, up to 32 bits in r22-r25,
up to 64 bits in r18-r25. 8-bit return values are zero/sign-extended to 16 bits by
the caller (unsigned char is more efficient than signed char - just clr
r25). Arguments to functions with variable argument lists (printf etc.) are all
passed on stack, and char is extended to int.
Warning:
There was no such alignment before 2000-07-01, including the old patches for
gcc-2.95.2. Check your old assembler subroutines, and adjust them accordingly.
Back to FAQ Index.

7.3.15

How do I put an array of strings completely in ROM?

There are times when you may need an array of strings which will never be modified.
In this case, you don't want to waste ram storing the constant strings. The most obvious
(and incorrect) thing to do is this:
#include & lt; avr/pgmspace.h & gt;
PGM_P array[2] PROGMEM = {
" Foo " ,
" Bar "
};
int main (void)
{
char buf[32];
strcpy_P (buf, array[1]);
return 0;
}

The result is not want you want though. What you end up with is the array stored in
ROM, while the individual strings end up in RAM (in the .data section).
To work around this, you need to do something like this:
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

96

7.3

Frequently Asked Questions

97

#include & lt; avr/pgmspace.h & gt;
const char foo[] PROGMEM = " Foo " ;
const char bar[] PROGMEM = " Bar " ;
PGM_P array[2] PROGMEM = {
foo,
bar
};
int main (void)
{
char buf[32];
strcpy_P (buf, array[1]);
return 0;
}

Looking at the disassembly of the resulting object file we see that array is in flash as
such:
0000008c & lt; foo & gt; :
8c:
46 6f
8e:
6f 00

ori r20, 0xF6
.word
0x006f

; 246
; ????

00000090 & lt; bar & gt; :
90:
42 61
92:
72 00

ori r20, 0x12
.word
0x0072

; 18
; ????

00000094 & lt; array & gt; :
94:
8c 00
96:
90 00

.word
.word

; ????
; ????

0x008c
0x0090

foo is at addr 0x008c.
bar is at addr 0x0090.
array is at addr 0x0094.
Then in main we see this:
strcpy_P (buf, array[1]);
/* copy bar into buf
de:
60 e9
ldi r22, 0x90
; 144
e0:
70 e0
ldi r23, 0x00
;0
e2:
ce 01
movw
r24, r28
e4:
01 96
adiw
r24, 0x01
;1
e6:
0e 94 79 00
call
0xf2

The addr of bar (0x0090) is loaded into the r23:r22 pair which is the second parameter
passed to strcpy P. The r25:r24 pair is the addr of buf.
Back to FAQ Index.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.3

Frequently Asked Questions

7.3.16

How to use external RAM?

Well, there is no universal answer to this question; it depends on what the external
RAM is going to be used for.
Basically, the bit SRE (SRAM enable) in the MCUCR register needs to be set in order
to enable the external memory interface. Depending on the device to be used, and
the application details, further registers affecting the external memory operation like
XMCRA and XMCRB, and/or further bits in MCUCR might be configured. Refer to the
datasheet for details.
If the external RAM is going to be used to store the variables from the C program (i.
e., the .data and/or .bss segment) in that memory area, it is essential to set up the
external memory interface early during the device initialization so the initialization of
these variable will take place. Refer to How to modify MCUCR or WDTCR early? for
a description how to do this using few lines of assembler code, or to the chapter about
memory sections for an example written in C.
The explanation of malloc() contains a discussion about the use of internal RAM vs.
external RAM in particular with respect to the various possible locations of the heap
(area reserved for malloc()). It also explains the linker command-line options that are
required to move the memory regions away from their respective standard locations in
internal RAM.
Finally, if the application simply wants to use the additional RAM for private data
storage kept outside the domain of the C compiler (e. g. through a char * variable
initialized directly to a particular address), it would be sufficient to defer the initialization of the external RAM interface to the beginning of main(), so no tweaking of the
.init1 section is necessary. The same applies if only the heap is going to be located
there, since the application start-up code does not affect the heap.
It is not recommended to locate the stack in external RAM. In general, accessing external RAM is slower than internal RAM, and errata of some AVR devices even prevent
this configuration from working properly at all.
Back to FAQ Index.

7.3.17

Which -O flag to use?

There's a common misconception that larger numbers behind the -O option might automatically cause "better" optimization. First, there's no universal definition for "better",
with optimization often being a speed vs. code size tradeoff. See the detailed discussion for which option affects which part of the code generation.
A test case was run on an ATmega128 to judge the effect of compiling the library itself
using different optimization levels. The following table lists the results. The test case
consisted of around 2 KB of strings to sort. Test #1 used qsort() using the standard
library strcmp(), test #2 used a function that sorted the strings by their size (thus had

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

98

7.3

Frequently Asked Questions

99

two calls to strlen() per invocation).
When comparing the resulting code size, it should be noted that a floating point version
of fvprintf() was linked into the binary (in order to print out the time elapsed) which
is entirely not affected by the different optimization levels, and added about 2.5 KB to
the code.
Optimization
flags
-O3
-O2
-Os
-Os
-mcall-prologues

Size of .text

Time for test #1

Time for test #2

6898
6666
6618
6474

903 us
972 us
955 us
972 us

19.7 ms
20.1 ms
20.1 ms
20.1 ms

(The difference between 955 us and 972 us was just a single timer-tick, so take this
with a grain of salt.)
So generally, it seems -Os -mcall-prologues is the most universal "best" optimization level. Only applications that need to get the last few percent of speed benefit
from using -O3.
Back to FAQ Index.

7.3.18

How do I relocate code to a fixed address?

First, the code should be put into a new named section. This is done with a section
attribute:
__attribute__ ((section ( " .bootloader " )))

In this example, .bootloader is the name of the new section. This attribute needs to
be placed after the prototype of any function to force the function into the new section.
void boot(void) __attribute__ ((section ( " .bootloader " )));

To relocate the section to a fixed address the linker flag --section-start is used.
This option can be passed to the linker using the -Wl compiler option:
-Wl,--section-start=.bootloader=0x1E000

The name after section-start is the name of the section to be relocated. The number
after the section name is the beginning address of the named section.
Back to FAQ Index.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.3

Frequently Asked Questions

7.3.19

My UART is generating nonsense! My ATmega128 keeps crashing! Port
F is completely broken!

Well, certain odd problems arise out of the situation that the AVR devices as shipped
by Atmel often come with a default fuse bit configuration that doesn't match the user's
expectations. Here is a list of things to care for:
o All devices that have an internal RC oscillator ship with the fuse enabled that
causes the device to run off this oscillator, instead of an external crystal. This
often remains unnoticed until the first attempt is made to use something critical
in timing, like UART communication.
o The ATmega128 ships with the fuse enabled that turns this device into ATmega103 compatibility mode. This means that some ports are not fully usable,
and in particular that the internal SRAM is located at lower addresses. Since by
default, the stack is located at the top of internal SRAM, a program compiled for
an ATmega128 running on such a device will immediately crash upon the first
function call (or rather, upon the first function return).
o Devices with a JTAG interface have the JTAGEN fuse programmed by default.
This will make the respective port pins that are used for the JTAG interface unavailable for regular IO.
Back to FAQ Index.

7.3.20

Why do all my "foo...bar" strings eat up the SRAM?

By default, all strings are handled as all other initialized variables: they occupy RAM
(even though the compiler might warn you when it detects write attempts to these RAM
locations), and occupy the same amount of flash ROM so they can be initialized to the
actual string by startup code. The compiler can optimize multiple identical strings into
a single one, but obviously only for one compilation unit (i. e., a single C source file).
That way, any string literal will be a valid argument to any C function that expects a
const char * argument.
Of course, this is going to waste a lot of SRAM. In Program Space String Utilities, a
method is described how such constant data can be moved out to flash ROM. However, a constant string located in flash ROM is no longer a valid argument to pass to a
function that expects a const char *-type string, since the AVR processor needs
the special instruction LPM to access these strings. Thus, separate functions are needed
that take this into account. Many of the standard C library functions have equivalents
available where one of the string arguments can be located in flash ROM. Private functions in the applications need to handle this, too. For example, the following can be
used to implement simple debugging messages that will be sent through a UART:
#include & lt; inttypes.h & gt;

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

100

7.3

Frequently Asked Questions

#include & lt; avr/io.h & gt;
#include & lt; avr/pgmspace.h & gt;
void
uart_putchar(char c)
{
if (c == '\n')
uart_putchar('\r');
loop_until_bit_is_set(USR, UDRE);
UDR = c;
}
void
debug_P(const char *addr)
{
char c;
while ((c = pgm_read_byte(addr++)))
uart_putchar(c);
}
int
main(void)
{
debug_P(PSTR( " foo was here\n " ));
return 0;
}

Note:
By convention, the suffix P to the function name is used as an indication that
this function is going to accept a "program-space string". Note also the use of the
PSTR() macro.
Back to FAQ Index.

7.3.21

Why does the compiler compile an 8-bit operation that uses bitwise operators into a 16-bit operation in assembly?

Bitwise operations in Standard C will automatically promote their operands to an int,
which is (by default) 16 bits in avr-gcc.
To work around this use typecasts on the operands, including literals, to declare that
the values are to be 8 bit operands.
This may be especially important when clearing a bit:
var & = ~mask;

/* wrong way!

The bitwise "not" operator (~) will also promote the value in mask to an int. To keep
it an 8-bit value, typecast before the "not" operator:
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

101

7.4

Inline Asm

var & = (unsigned char)~mask;

Back to FAQ Index.

7.3.22

How to detect RAM memory and variable overlap problems?

You can simply run avr-nm on your output (ELF) file. Run it with the -n option, and
it will sort the symbols numerically (by default, they are sorted alphabetically).
Look for the symbol end, that's the first address in RAM that is not allocated by
a variable. (avr-gcc internally adds 0x800000 to all data/bss variable addresses, so
please ignore this offset.) Then, the run-time initialization code initializes the stack
pointer (by default) to point to the last avaialable address in (internal) SRAM. Thus,
the region between end and the end of SRAM is what is available for stack. (If your
application uses malloc(), which e. g. also can happen inside printf(), the heap for
dynamic memory is also located there. See Using malloc().)
The amount of stack required for your application cannot be determined that easily.
For example, if you recursively call a function and forget to break that recursion, the
amount of stack required is infinite. :-) You can look at the generated assembler code
(avr-gcc ... -S), there's a comment in each generated assembler file that tells
you the frame size for each generated function. That's the amount of stack required for
this function, you have to add up that for all functions where you know that the calls
could be nested.
Back to FAQ Index.

7.4

Inline Asm

AVR-GCC
Inline Assembler Cookbook
About this Document
The GNU C compiler for Atmel AVR RISC processors offers, to embed assembly
language code into C programs. This cool feature may be used for manually optimizing
time critical parts of the software or to use specific processor instruction, which are not
available in the C language.
Because of a lack of documentation, especially for the AVR version of the compiler, it
may take some time to figure out the implementation details by studying the compiler
and assembler source code. There are also a few sample programs available in the net.
Hopefully this document will help to increase their number.
It's assumed, that you are familiar with writing AVR assembler programs, because this
is not an AVR assembler programming tutorial. It's not a C language tutorial either.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

102

7.4

Inline Asm

Note that this document does not cover file written completely in assembler language,
refer to avr-libc and assembler programs for this.
Copyright (C) 2001-2002 by egnite Software GmbH
Permission is granted to copy and distribute verbatim copies of this manual provided
that the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual provided that
the entire resulting derived work is distributed under the terms of a permission notice
identical to this one.
This document describes version 3.3 of the compiler. There may be some parts, which
hadn't been completely understood by the author himself and not all samples had been
tested so far. Because the author is German and not familiar with the English language,
there are definitely some typos and syntax errors in the text. As a programmer the
author knows, that a wrong documentation sometimes might be worse than none. Anyway, he decided to offer his little knowledge to the public, in the hope to get enough
response to improve this document. Feel free to contact the author via e-mail. For the
latest release check http://www.ethernut.de/.
Herne, 17th of May 2002 Harald Kipp harald.kipp-at-egnite.de
Note:
As of 26th of July 2002, this document has been merged into the
documentation for avr-libc.
The latest version is now available at
http://savannah.nongnu.org/projects/avr-libc/.
7.4.1

GCC asm Statement

Let's start with a simple example of reading a value from port D:
asm( " in %0, %1 " : " =r " (value) : " I " (PORTD) : );

Each asm statement is devided by colons into four parts:
1. The assembler instructions, defined as a single string constant:
" in %0, %1 "

2. A list of output operands, separated by commas. Our example uses just one:
" =r " (value)

3. A comma separated list of input operands. Again our example uses one operand
only:
" I " (PORTD)

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

103

7.4

Inline Asm

4. Clobbered registers, left empty in our example.
You can write assembler instructions in much the same way as you would write assembler programs. However, registers and constants are used in a different way if they refer
to expressions of your C program. The connection between registers and C operands is
specified in the second and third part of the asm instruction, the list of input and output
operands, respectively. The general form is
asm(code : output operand list : input operand list : clobber list);

In the code section, operands are referenced by a percent sign followed by a single
digit. %0 refers to the first %1 to the second operand and so forth. From the above
example:
%0 refers to " =r " (value) and
%1 refers to " I " (PORTD).
This may still look a little odd now, but the syntax of an operand list will be explained
soon. Let us first examine the part of a compiler listing which may have been generated
from our example:
lds r24,value
/* #APP
in r24, 12
/* #NOAPP
sts value,r24

The comments have been added by the compiler to inform the assembler that the included code was not generated by the compilation of C statements, but by inline assembler statements. The compiler selected register r24 for storage of the value read
from PORTD. The compiler could have selected any other register, though. It may not
explicitely load or store the value and it may even decide not to include your assembler
code at all. All these decisions are part of the compiler's optimization strategy. For
example, if you never use the variable value in the remaining part of the C program,
the compiler will most likely remove your code unless you switched off optimization.
To avoid this, you can add the volatile attribute to the asm statement:
asm volatile( " in %0, %1 " : " =r " (value) : " I " (PORTD) : );

The last part of the asm instruction, the clobber list, is mainly used to tell the compiler
about modifications done by the assembler code. This part may be omitted, all other
parts are required, but may be left empty. If your assembler routine won't use any
input or output operand, two colons must still follow the assembler code string. A
good example is a simple statement to disable interrupts:
asm volatile( " cli " ::);

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

104

7.4

7.4.2

Inline Asm

105

Assembler Code

You can use the same assembler instruction mnemonics as you'd use with any other
AVR assembler. And you can write as many assembler statements into one code string
as you like and your flash memory is able to hold.
Note:
The available assembler directives vary from one assembler to another.
To make it more readable, you should put each statement on a seperate line:
asm volatile( " nop\n\t "
" nop\n\t "
" nop\n\t "
" nop\n\t "
::);

The linefeed and tab characters will make the assembler listing generated by the compiler more readable. It may look a bit odd for the first time, but that's the way the
compiler creates it's own assembler code.
You may also make use of some special registers.
Symbol
SREG
SP H
SP L
tmp reg
zero reg

Register
Status register at address 0x3F
Stack pointer high byte at address 0x3E
Stack pointer low byte at address 0x3D
Register r0, used for temporary storage
Register r1, always zero

Register r0 may be freely used by your assembler code and need not be restored at the
end of your code. It's a good idea to use tmp reg and zero reg instead of
r0 or r1, just in case a new compiler version changes the register usage definitions.
7.4.3

Input and Output Operands

Each input and output operand is described by a constraint string followed by a C
expression in parantheses. AVR-GCC 3.3 knows the following constraint characters:
Note:
The most up-to-date and detailed information on contraints for the avr can be found
in the gcc manual.
Note:
The x register is r27:r26, the y register is r29:r28, and the z register is
r31:r30

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.4

Inline Asm

Constraint
a
b
d
e
G
I
J
K
L
l
M
N
O
P
q
r
t
w
x
y
z

106

Used for
Simple upper registers
Base pointer registers
pairs
Upper register
Pointer register pairs
Floating point constant
6-bit positive integer
constant
6-bit negative integer
constant
Integer constant
Integer constant
Lower registers
8-bit integer constant
Integer constant
Integer constant
Integer constant
Stack pointer register
Any register
Temporary register
Special upper register
pairs
Pointer register pair X
Pointer register pair Y
Pointer register pair Z

Range
r16 to r23
y, z
r16 to r31
x, y, z
0.0
0 to 63
-63 to 0
2
0
r0 to r15
0 to 255
-1
8, 16, 24
1
SPH:SPL
r0 to r31
r0
r24, r26, r28, r30
x (r27:r26)
y (r29:r28)
z (r31:r30)

These definitions seem not to fit properly to the AVR instruction set. The author's assumption is, that this part of the compiler has never been really finished in this version,
but that assumption may be wrong. The selection of the proper contraint depends on
the range of the constants or registers, which must be acceptable to the AVR instruction
they are used with. The C compiler doesn't check any line of your assembler code. But
it is able to check the constraint against your C expression. However, if you specify
the wrong constraints, then the compiler may silently pass wrong code to the assembler. And, of course, the assembler will fail with some cryptic output or internal errors.
For example, if you specify the constraint " r " and you are using this register with an
" ori " instruction in your assembler code, then the compiler may select any register.
This will fail, if the compiler chooses r2 to r15. (It will never choose r0 or r1,
because these are uses for special purposes.) That's why the correct constraint in that
case is " d " . On the other hand, if you use the constraint " M " , the compiler will make
sure that you don't pass anything else but an 8-bit value. Later on we will see how to
pass multibyte expression results to the assembler code.
The following table shows all AVR assembler mnemonics which require operands, and
the related contraints. Because of the improper constraint definitions in version 3.3,
they aren't strict enough. There is, for example, no constraint, which restricts integer

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.4

Inline Asm

107

constants to the range 0 to 7 for bit set and bit clear operations.
Mnemonic
adc
adiw
andi
bclr
brbc
bset
cbi
com
cpc
cpse
elpm
in
ld
ldi
lpm
lsr
movw
neg
ori
pop
rol
sbc
sbi
sbiw
sbrc
ser
std
sub
swap

Constraints
r,r
w,I
d,M
I
I,label
I
I,I
r
r,r
r,r
t,z
r,I
r,e
d,M
t,z
r
r,r
r
d,M
r
r
r,r
I,I
w,I
r,I
d
b,r
r,r
r

Mnemonic
add
and
asr
bld
brbs
bst
cbr
cp
cpi
dec
eor
inc
ldd
lds
lsl
mov
mul
or
out
push
ror
sbci
sbic
sbr
sbrs
st
sts
subi

Constraints
r,r
r,r
r
r,I
I,label
r,I
d,I
r,r
d,M
r
r,r
r
r,b
r,label
r
r,r
r,r
r,r
I,r
r
r
d,M
I,I
d,M
r,I
e,r
label,r
d,M

Constraint characters may be prepended by a single constraint modifier. Contraints
without a modifier specify read-only operands. Modifiers are:
Modifier
=
+
&

Specifies
Write-only operand, usually used for all
output operands.
Read-write operand (not supported by
inline assembler)
Register should be used for output only

Output operands must be write-only and the C expression result must be an lvalue,
which means that the operands must be valid on the left side of assignments. Note,
that the compiler will not check if the operands are of reasonable type for the kind of
operation used in the assembler instructions.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.4

Inline Asm

Input operands are, you guessed it, read-only. But what if you need the same operand
for input and output? As stated above, read-write operands are not supported in inline
assembler code. But there is another solution. For input operators it is possible to use
a single digit in the constraint string. Using digit n tells the compiler to use the same
register as for the n-th operand, starting with zero. Here is an example:
asm volatile( " swap %0 " : " =r " (value) : " 0 " (value));

This statement will swap the nibbles of an 8-bit variable named value. Constraint " 0 "
tells the compiler, to use the same input register as for the first operand. Note however,
that this doesn't automatically imply the reverse case. The compiler may choose the
same registers for input and output, even if not told to do so. This is not a problem in
most cases, but may be fatal if the output operator is modified by the assembler code
before the input operator is used. In the situation where your code depends on different
registers used for input and output operands, you must add the & constraint modifier to
your output operand. The following example demonstrates this problem:
asm volatile( " in %0,%1 "
" \n\t "
" out %1, %2 " " \n\t "
: " = & r " (input)
: " I " (port), " r " (output)
);

In this example an input value is read from a port and then an output value is written to
the same port. If the compiler would have choosen the same register for input and output, then the output value would have been destroyed on the first assembler instruction.
Fortunately, this example uses the & constraint modifier to instruct the compiler not to
select any register for the output value, which is used for any of the input operands.
Back to swapping. Here is the code to swap high and low byte of a 16-bit value:
asm volatile( " mov __tmp_reg__, %A0 " " \n\t "
" mov %A0, %B0 "
" \n\t "
" mov %B0, __tmp_reg__ " " \n\t "
: " =r " (value)
: " 0 " (value)
);

First you will notice the usage of register tmp reg , which we listed among other
special registers in the Assembler Code section. You can use this register without
saving its contents. Completely new are those letters A and B in %A0 and %B0. In fact
they refer to two different 8-bit registers, both containing a part of value.
Another example to swap bytes of a 32-bit value:
asm volatile( " mov __tmp_reg__, %A0 " " \n\t "
" mov %A0, %D0 "
" \n\t "
" mov %D0, __tmp_reg__ " " \n\t "

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

108

7.4

Inline Asm

" mov __tmp_reg__, %B0 " " \n\t "
" mov %B0, %C0 "
" \n\t "
" mov %C0, __tmp_reg__ " " \n\t "
: " =r " (value)
: " 0 " (value)
);

If operands do not fit into a single register, the compiler will automatically assign
enough registers to hold the entire operand. In the assembler code you use %A0 to refer
to the lowest byte of the first operand, %A1 to the lowest byte of the second operand
and so on. The next byte of the first operand will be %B0, the next byte %C0 and so on.
This also implies, that it is often neccessary to cast the type of an input operand to the
desired size.
A final problem may arise while using pointer register pairs. If you define an input
operand
" e " (ptr)

and the compiler selects register Z (r30:r31), then
%A0 refers to r30 and
%B0 refers to r31.
But both versions will fail during the assembly stage of the compiler, if you explicitely
need Z, like in
ld r24,Z

If you write
ld r24, %a0

with a lower case a following the percent sign, then the compiler will create the proper
assembler line.
7.4.4

Clobbers

As stated previously, the last part of the asm statement, the list of clobbers, may be
omitted, including the colon seperator. However, if you are using registers, which
had not been passed as operands, you need to inform the compiler about this. The
following example will do an atomic increment. It increments an 8-bit value pointed
to by a pointer variable in one go, without being interrupted by an interrupt routine
or another thread in a multithreaded environment. Note, that we must use a pointer,
because the incremented value needs to be stored before interrupts are enabled.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

109

7.4

Inline Asm

asm volatile(
" cli "
" ld r24, %a0 "
" inc r24 "
" st %a0, r24 "
" sei "
:
: " e " (ptr)
: " r24 "
);

110

" \n\t "
" \n\t "
" \n\t "
" \n\t "
" \n\t "

The compiler might produce the following code:
cli
ld r24, Z
inc r24
st Z, r24
sei

One easy solution to avoid clobbering register r24 is, to make use of the special temporary register tmp reg defined by the compiler.
asm volatile(
" cli "
" ld __tmp_reg__, %a0 "
" inc __tmp_reg__ "
" st %a0, __tmp_reg__ "
" sei "
:
: " e " (ptr)
);

" \n\t "
" \n\t "
" \n\t "
" \n\t "
" \n\t "

The compiler is prepared to reload this register next time it uses it. Another problem
with the above code is, that it should not be called in code sections, where interrupts
are disabled and should be kept disabled, because it will enable interrupts at the end.
We may store the current status, but then we need another register. Again we can solve
this without clobbering a fixed, but let the compiler select it. This could be done with
the help of a local C variable.
{
uint8_t s;
asm volatile(
" in %0, __SREG__ "
" cli "
" ld __tmp_reg__, %a1 "
" inc __tmp_reg__ "
" st %a1, __tmp_reg__ "
" out __SREG__, %0 "
: " = & r " (s)
: " e " (ptr)
);

" \n\t "
" \n\t "
" \n\t "
" \n\t "
" \n\t "
" \n\t "

}

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.4

Inline Asm

111

Now every thing seems correct, but it isn't really. The assembler code modifies the
variable, that ptr points to. The compiler will not recognize this and may keep its
value in any of the other registers. Not only does the compiler work with the wrong
value, but the assembler code does too. The C program may have modified the value
too, but the compiler didn't update the memory location for optimization reasons. The
worst thing you can do in this case is:
{
uint8_t s;
asm volatile(
" in %0, __SREG__ "
" cli "
" ld __tmp_reg__, %a1 "
" inc __tmp_reg__ "
" st %a1, __tmp_reg__ "
" out __SREG__, %0 "
: " = & r " (s)
: " e " (ptr)
: " memory "
);

" \n\t "
" \n\t "
" \n\t "
" \n\t "
" \n\t "
" \n\t "

}

The special clobber "memory" informs the compiler that the assembler code may modify any memory location. It forces the compiler to update all variables for which the
contents are currently held in a register before executing the assembler code. And of
course, everything has to be reloaded again after this code.
In most situations, a much better solution would be to declare the pointer destination
itself volatile:
volatile uint8_t *ptr;

This way, the compiler expects the value pointed to by ptr to be changed and will
load it whenever used and store it whenever modified.
Situations in which you need clobbers are very rare. In most cases there will be better
ways. Clobbered registers will force the compiler to store their values before and reload
them after your assembler code. Avoiding clobbers gives the compiler more freedom
while optimizing your code.
7.4.5

Assembler Macros

In order to reuse your assembler language parts, it is useful to define them as macros
and put them into include files. AVR Libc comes with a bunch of them, which could
be found in the directory avr/include. Using such include files may produce compiler warnings, if they are used in modules, which are compiled in strict ANSI mode.
To avoid that, you can write asm instead of asm and volatile instead of
volatile. These are equivalent aliases.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.4

Inline Asm

112

Another problem with reused macros arises if you are using labels. In such
cases you may make use of the special pattern %=, which is replaced by a
unique number on each asm statement. The following code had been taken from
avr/include/iomacros.h:
#define loop_until_bit_is_clear(port,bit)
__asm__ __volatile__ (
" L_%=: " " sbic %0, %1 " " \n\t "
" rjmp L_%= "
: /* no outputs
\
: " I " ((uint8_t)(port)),
" I " ((uint8_t)(bit))
)

\
\
\
\
\
\

When used for the first time, L %= may be translated to L 1404, the next usage might
create L 1405 or whatever. In any case, the labels became unique too.
7.4.6

C Stub Functions

Macro definitions will include the same assembler code whenever they are referenced.
This may not be acceptable for larger routines. In this case you may define a C stub
function, containing nothing other than your assembler code.
void delay(uint8_t ms)
{
uint16_t cnt;
asm volatile (
" \n "
" L_dl1%=: " " \n\t "
" mov %A0, %A2 " " \n\t "
" mov %B0, %B2 " " \n "
" L_dl2%=: " " \n\t "
" sbiw %A0, 1 " " \n\t "
" brne L_dl2%= " " \n\t "
" dec %1 " " \n\t "
" brne L_dl1%= " " \n\t "
: " = & w " (cnt)
: " r " (ms), " r " (delay_count)
);
}

The purpose of this function is to delay the program execution by a specified number
of milliseconds using a counting loop. The global 16 bit variable delay count must
contain the CPU clock frequency in Hertz divided by 4000 and must have been set
before calling this routine for the first time. As described in the clobber section, the
routine uses a local variable to hold a temporary value.
Another use for a local variable is a return value. The following function returns a 16
bit value read from two successive port addresses.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.4

Inline Asm

uint16_t inw(uint8_t port)
{
uint16_t result;
asm volatile (
" in %A0,%1 " " \n\t "
" in %B0,(%1) + 1 "
: " =r " (result)
: " I " (port)
);
return result;
}

Note:
inw() is supplied by avr-libc.
7.4.7

C Names Used in Assembler Code

By default AVR-GCC uses the same symbolic names of functions or variables in C and
assembler code. You can specify a different name for the assembler code by using a
special form of the asm statement:
unsigned long value asm( " clock " ) = 3686400;

This statement instructs the compiler to use the symbol name clock rather than value.
This makes sense only for external or static variables, because local variables do not
have symbolic names in the assembler code. However, local variables may be held in
registers.
With AVR-GCC you can specify the use of a specific register:
void Count(void)
{
register unsigned char counter asm( " r3 " );
... some code...
asm volatile( " clr r3 " );
... more code...
}

The assembler instruction, " clr r3 " , will clear the variable counter. AVR-GCC will
not completely reserve the specified register. If the optimizer recognizes that the variable will not be referenced any longer, the register may be re-used. But the compiler
is not able to check wether this register usage conflicts with any predefined register. If
you reserve too many registers in this way, the compiler may even run out of registers
during code generation.
In order to change the name of a function, you need a prototype declaration, because
the compiler will not accept the asm keyword in the function definition:

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

113

7.5

Using malloc()

extern long Calc(void) asm ( " CALCULATE " );

Calling the function Calc() will create assembler instructions to call the function
CALCULATE.
7.4.8

Links

For a more thorough discussion of inline assembly usage, see the gcc user
manual.
The latest version of the gcc manual is always available here:
http://gcc.gnu.org/onlinedocs/

7.5
7.5.1

Using malloc()
Introduction

On a simple device like a microcontroller, implementing dynamic memory allocation
is quite a challenge.
Many of the devices that are possible targets of avr-libc have a minimal amount of
RAM. The smallest parts supported by the C environment come with 128 bytes of
RAM. This needs to be shared between initialized and uninitialized variables (sections
.data and .bss), the dynamic memory allocator, and the stack that is used for calling
subroutines and storing local (automatic) variables.
Also, unlike larger architectures, there is no hardware-supported memory management
which could help in separating the mentioned RAM regions from being overwritten by
each other.
The standard RAM layout is to place .data variables first, from the beginning of the
internal RAM, followed by .bss. The stack is started from the top of internal RAM,
growing downwards. The so-called "heap" available for the dynamic memory allocator
will be placed beyond the end of .bss. Thus, there's no risk that dynamic memory will
ever collide with the RAM variables (unless there were bugs in the implementation of
the allocator). There is still a risk that the heap and stack could collide if there are large
requirements for either dynamic memory or stack space. The former can even happen
if the allocations aren't all that large but dynamic memory allocations get fragmented
over time such that new requests don't quite fit into the "holes" of previously freed
regions. Large stack space requirements can arise in a C function containing large
and/or numerous local variables or when recursively calling function.
Note:
The pictures shown in this document represent typical situations where the RAM
locations refer to an ATmega128. The memory addresses used are not displayed
in a linear scale.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

114

.data

on-board RAM

!

.bss

variables variables

heap

external RAM

0xFFFF

115

0x10FF
0x1100

Using malloc()

0x0100

7.5

stack

SP

RAMEND

brkval ( & lt; = *SP - __malloc_margin)
__malloc_heap_start == __heap_start
__bss_end
__data_end == __bss_start
__data_start

Figure 1: RAM map of a device with internal RAM

Finally, there's a challenge to make the memory allocator simple enough so the code
size requirements will remain low, yet powerful enough to avoid unnecessary memory
fragmentation and to get it all done with reasonably few CPU cycles since microcontrollers aren't only often low on space, but also run at much lower speeds than the
typical PC these days.
The memory allocator implemented in avr-libc tries to cope with all of these constraints, and offers some tuning options that can be used if there are more resources
available than in the default configuration.
7.5.2

Internal vs. external RAM

Obviously, the constraints are much harder to satisfy in the default configuration where
only internal RAM is available. Extreme care must be taken to avoid a stack-heap
collision, both by making sure functions aren't nesting too deeply, and don't require
too much stack space for local variables, as well as by being cautious with allocating
too much dynamic memory.
If external RAM is available, it is strongly recommended to move the heap into the
external RAM, regardless of whether or not the variables from the .data and .bss
sections are also going to be located there. The stack should always be kept in internal
RAM. Some devices even require this, and in general, internal RAM can be accessed
faster since no extra wait states are required. When using dynamic memory allocation
and stack and heap are separated in distinct memory areas, this is the safest way to
avoid a stack-heap collision.
7.5.3

Tunables for malloc()

There are a number of variables that can be tuned to adapt the behavior of malloc()
to the expected requirements and constraints of the application. Any changes to these

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.5

Using malloc()

116

tunables should be made before the very first call to malloc(). Note that some library
functions might also use dynamic memory (notably those from the Standard IO facilities), so make sure the changes will be done early enough in the startup sequence.
The variables malloc heap start and malloc heap end can be used to restrict the malloc() function to a certain memory region. These variables are statically
initialized to point to heap start and heap end, respectively, where heap start is filled in by the linker to point just beyond .bss, and heap end is set to 0
which makes malloc() assume the heap is below the stack.
If the heap is going to be moved to external RAM, malloc heap end must be
adjusted accordingly. This can either be done at run-time, by writing directly to this
variable, or it can be done automatically at link-time, by adjusting the value of the
symbol heap end.
The following example shows a linker command to relocate the entire .data and .bss
segments, and the heap to location 0x1100 in external RAM. The heap will extend up
to address 0xffff.
avr-gcc ... -Wl,-Tdata=0x801100,--defsym=__heap_end=0x80ffff ...

stack

external RAM

.data
.bss
variables variables

0xFFFF

on-board RAM

0x10FF
0x1100

0x0100

Note:
See explanation for offset 0x800000. See the chapter about using gcc for the -Wl
options.

heap

SP
RAMEND

__malloc_heap_end == __heap_end
brkval
__malloc_heap_start == __heap_start
__bss_end
__data_end == __bss_start
__data_start

Figure 2: Internal RAM: stack only, external RAM: variables and heap

If dynamic memory should be placed in external RAM, while keeping the variables in
internal RAM, something like the following could be used. Note that for demonstration
purposes, the assignment of the various regions has not been made adjacent in this
example, so there are "holes" below and above the heap in external RAM that remain
completely unaccessible by regular variables or dynamic memory allocations (shown
in light bisque color in the picture below).
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.5

Using malloc()

117

avr-gcc ... -Wl,--defsym=__heap_start=0x802000,--defsym=__heap_end=0x803fff ...

.data

0xFFFF

0x3FFF

on-board RAM

0x2000

0x10FF
0x1100

0x0100

external RAM

.bss
stack

variables variables

heap

SP
RAMEND
__bss_end
__data_end == __bss_start

__malloc_heap_end == __heap_end
brkval
__malloc_heap_start == __heap_start

__data_start

Figure 3: Internal RAM: variables and stack, external RAM: heap

If malloc heap end is 0, the allocator attempts to detect the bottom of stack in order to prevent a stack-heap collision when extending the actual size of the heap to gain
more space for dynamic memory. It will not try to go beyond the current stack limit,
decreased by malloc margin bytes. Thus, all possible stack frames of interrupt
routines that could interrupt the current function, plus all further nested function calls
must not require more stack space, or they will risk colliding with the data segment.
The default value of malloc margin is set to 32.
7.5.4

Implementation details

Dynamic memory allocation requests will be returned with a two-byte header
prepended that records the size of the allocation. This is later used by free(). The
returned address points just beyond that header. Thus, if the application accidentally
writes before the returned memory region, the internal consistency of the memory allocator is compromised.
The implementation maintains a simple freelist that accounts for memory blocks that
have been returned in previous calls to free(). Note that all of this memory is considered
to be successfully added to the heap already, so no further checks against stack-heap
collisions are done when recycling memory from the freelist.
The freelist itself is not maintained as a separate data structure, but rather by modifying
the contents of the freed memory to contain pointers chaining the pieces together. That
way, no additional memory is reqired to maintain this list except for a variable that
keeps track of the lowest memory segment available for reallocation. Since both, a
chain pointer and the size of the chunk need to be recorded in each chunk, the minimum
chunk size on the freelist is four bytes.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.6

Release Numbering and Methodology

When allocating memory, first the freelist is walked to see if it could satisfy the request.
If there's a chunk available on the freelist that will fit the request exactly, it will be
taken, disconnected from the freelist, and returned to the caller. If no exact match could
be found, the closest match that would just satisfy the request will be used. The chunk
will normally be split up into one to be returned to the caller, and another (smaller)
one that will remain on the freelist. In case this chunk was only up to two bytes larger
than the request, the request will simply be altered internally to also account for these
additional bytes since no separate freelist entry could be split off in that case.
If nothing could be found on the freelist, heap extension is attempted. This is where
malloc margin will be considered if the heap is operating below the stack, or
where malloc heap end will be verified otherwise.
If the remaining memory is insufficient to satisfy the request, NULL will eventually be
returned to the caller.
When calling free(), a new freelist entry will be prepared. An attempt is then made to
aggregate the new entry with possible adjacent entries, yielding a single larger entry
available for further allocations. That way, the potential for heap fragmentation is
hopefully reduced.

7.6
7.6.1

Release Numbering and Methodology
Release Version Numbering Scheme

7.6.1.1 Stable Versions A stable release will always have a minor number that is
an even number. This implies that you should be able to upgrade to a new version of
the library with the same major and minor numbers without fear that any of the APIs
have changed. The only changes that should be made to a stable branch are bug fixes
and under some circumstances, additional functionality (e.g. adding support for a new
device).
If major version number has changed, this implies that the required versions of gcc and
binutils have changed. Consult the README file in the toplevel directory of the AVR
Libc source for which versions are required.
7.6.1.2 Development Versions The major version number of a development series
is always the same as the last stable release.
The minor version number of a development series is always an odd number and is 1
more than the last stable release.
The patch version number of a development series is always 0 until a new branch is cut
at which point the patch number is changed to 90 to denote the branch is approaching
a release and the date appended to the version to denote that it is still in development.
All versions in development in cvs will also always have the date appended as a fourth
version number. The format of the date will be YYYYMMDD.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

118

7.6

Release Numbering and Methodology

So, the development version number will look like this:
1.1.0.20030825

While a pre-release version number on a branch (destined to become either 1.2 or 2.0)
will look like this:
1.1.90.20030828

7.6.2

Releasing AVR Libc

The information in this section is only relevant to AVR Libc developers and can be
ignored by end users.
Note:
In what follows, I assume you know how to use cvs and how to checkout multiple
source trees in a single directory without having them clobber each other. If you
don't know how to do this, you probably shouldn't be making releases or cutting
branches.
7.6.2.1 Creating a cvs branch
in cvs:

The following steps should be taken to cut a branch

1. Check out a fresh source tree from cvs HEAD.
2. Update the NEWS file with pending release number and commit to cvs HEAD:
Change "Changes since avr-libc- & lt; last release & gt; :" to "Changes in avr-libc & lt; this relelase & gt; :".
3. Set the branch-point tag (setting & lt; major & gt; and & lt; minor & gt; accordingly):
'cvs tag avr-libc- & lt; major & gt; & lt; minor & gt; -branchpoint'
4. Create the branch:
'cvs tag -b avr-lib- & lt; major & gt; & lt; minor & gt; -branch'
5. Update the package version in configure.in and commit configure.in to cvs
HEAD:
Change minor number to next odd value.
6. Update the NEWS file and commit to cvs HEAD:
Add "Changes since avr-libc- & lt; this release & gt; :"
7. Check out a new tree for the branch:
'cvs co -r avr-lib- & lt; major & gt; & lt; minor & gt; -branch'
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

119

7.6

Release Numbering and Methodology

8. Update the package version in configure.in and commit configure.in to cvs
branch:
Change the patch number to 90 to denote that this now a branch leading up to a
release. Be sure to leave the & lt; date & gt; part of the version.
9. Bring the build system up to date by running reconf and doconf.
10. Perform a 'make distcheck' and make sure it succeeds. This will create the
snapshot source tarball. This should be considered the first release candidate.
11. Upload the snapshot tarball to savannah.
12. Announce the branch and the branch tag to the avr-libc-dev list so other developers can checkout the branch.
Note:
CVS tags do not allow the use of periods ('.').
7.6.2.2 Making a release
the cvs HEAD.

A stable release will only be done on a branch, not from

The following steps should be taken when making a release:
1. Make sure the source tree you are working from is on the correct branch:
'cvs update -r avr-lib- & lt; major & gt; & lt; minor & gt; -branch'
2. Update the package version in configure.in and commit it to cvs.
3. Update the gnu tool chain version requirements in the README and commit to
cvs.
4. Update the ChangeLog file to note the release and commit to cvs on the branch:
Add "Released avr-libc- & lt; this release & gt; ."
5. Bring the build system up to date by running reconf and doconf.
6. Perform a 'make distcheck' and make sure it succeeds. This will create the
source tarball.
7. Tag the release ( & lt; patch & gt; is not given if this is the first release on this branch):
'cvs tag avr-lib- & lt; major & gt; & lt; minor & gt; & lt; patch & gt; -release'
8. Upload the tarball to savannah.
9. Generate the latest documentation and upload to savannah.
10. Announce the release.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

120

7.7

Memory Sections

121

The following hypothetical diagram should help clarify version and branch relationships.

HEAD

1.0 Branch

1.2 Branch

cvs tag avr-libc-1_0-branchpoint

set version to 1.1.0. & lt; date & gt;

cvs tag -b avr-libc-1_0-branch

set version to 0.90.90. & lt; date & gt;
set version to 1.0
cvs tag avr-libc-1_0-release

set version to 1.0.0. & lt; date & gt;
set version to 1.0.1
cvs tag avr-libc-1_0_1-release
cvs tag avr-libc-1_2-branchpoint

set version to 1.3.0. & lt; date & gt;

cvs tag -b avr-libc-1_2-branch

set version to 1.1.90. & lt; date & gt;
set version to 1.2
cvs tag avr-libc-1_2-release

cvs tag avr-libc-2.0-branchpoint

set version to 2.1.0. & lt; date & gt;

Figure 4: Release tree

7.7

Memory Sections

Remarks:
Need to list all the sections which are available to the avr.
Weak Bindings
FIXME: need to discuss the .weak directive.
The following describes the various sections available.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.7

7.7.1

Memory Sections

The .text Section

The .text section contains the actual machine instructions which make up your program.
This section is further subdivided by the .initN and .finiN sections dicussed below.
Note:
The avr-size program (part of binutils), coming from a Unix background,
doesn't account for the .data initialization space added to the .text section, so in
order to know how much flash the final program will consume, one needs to add
the values for both, .text and .data (but not .bss), while the amount of pre-allocated
SRAM is the sum of .data and .bss.
7.7.2 The .data Section
This section contains static data which was defined in your code. Things like the following would end up in .data:
char err_str[] = " Your program has died a horrible death! " ;
struct point pt = { 1, 1 };

It is possible to tell the linker the SRAM address of the beginning of the .data section.
This is accomplished by adding -Wl,-Tdata,addr to the avr-gcc command
used to the link your program. Not that addr must be offset by adding 0x800000
the to real SRAM address so that the linker knows that the address is in the SRAM
memory space. Thus, if you want the .data section to start at 0x1100, pass 0x801100
at the address to the linker. [offset explained]
Note:
When using malloc() in the application (which could even happen inside library
calls), additional adjustments are required.
7.7.3

The .bss Section

Uninitialized global or static variables end up in the .bss section.
7.7.4

The .eeprom Section

This is where eeprom variables are stored.
7.7.5

The .noinit Section

This sections is a part of the .bss section. What makes the .noinit section special is that
variables which are defined as such:
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

122

7.7

Memory Sections

int foo __attribute__ ((section ( " .noinit " )));

will not be initialized to zero during startup as would normal .bss data.
Only uninitialized variables can be placed in the .noinit section. Thus, the following
code will cause avr-gcc to issue an error:
int bar __attribute__ ((section ( " .noinit " ))) = 0xaa;

It is possible to tell the linker explicitly where to place the .noinit section by adding
-Wl,--section-start=.noinit=0x802000 to the avr-gcc command line
at the linking stage. For example, suppose you wish to place the .noinit section at
SRAM address 0x2000:
$ avr-gcc ... -Wl,--section-start=.noinit=0x802000 ...

Note:
Because of the Harvard architecture of the AVR devices, you must manually add
0x800000 to the address you pass to the linker as the start of the section. Otherwise, the linker thinks you want to put the .noinit section into the .text section
instead of .data/.bss and will complain.
Alternatively, you can write your own linker script to automate this. [FIXME: need an
example or ref to dox for writing linker scripts.]
7.7.6 The .initN Sections
These sections are used to define the startup code from reset up through the start of
main(). These all are subparts of the .text section.
The purpose of these sections is to allow for more specific placement of code within
your program.
Note:
Sometimes, it is convenient to think of the .initN and .finiN sections as functions,
but in reality they are just symbolic names which tell the linker where to stick a
chunk of code which is not a function. Notice that the examples for asm and C can
not be called as functions and should not be jumped into.
The .initN sections are executed in order from 0 to 9.
.init0:
Weakly bound to init(). If user defines init(), it will be jumped into immediately
after a reset.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

123

7.7

Memory Sections

.init1:
Unused. User definable.
.init2:
In C programs, weakly bound to initialize the stack.
.init3:
Unused. User definable.
.init4:
Copies the .data section from flash to SRAM. Also sets up and zeros out the .bss
section. In Unix-like targets, .data is normally initialized by the OS directly from
the executable file. Since this is impossible in an MCU environment, avr-gcc
instead takes care of appending the .data variables after .text in the flash ROM
image. .init4 then defines the code (weakly bound) which takes care of copying
the contents of .data from the flash to SRAM.
.init5:
Unused. User definable.
.init6:
Unused for C programs, but used for constructors in C++ programs.
.init7:
Unused. User definable.
.init8:
Unused. User definable.
.init9:
Jumps into main().
7.7.7 The .finiN Sections
These sections are used to define the exit code executed after return from main() or a
call to exit(). These all are subparts of the .text section.
The .finiN sections are executed in descending order from 9 to 0.
.finit9:
Unused. User definable. This is effectively where exit() starts.
.fini8:
Unused. User definable.
.fini7:
Unused. User definable.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

124

7.7

Memory Sections

.fini6:
Unused for C programs, but used for destructors in C++ programs.
.fini5:
Unused. User definable.
.fini4:
Unused. User definable.
.fini3:
Unused. User definable.
.fini2:
Unused. User definable.
.fini1:
Unused. User definable.
.fini0:
Goes into an infinite loop after program termination and completion of any exit()
code (execution of code in the .fini9 - & gt; .fini1 sections).
7.7.8

Using Sections in Assembler Code

Example:
#include & lt; avr/io.h & gt;
.section .init1, " ax " ,@progbits
ldi
r0, 0xff
out
_SFR_IO_ADDR(PORTB), r0
out
_SFR_IO_ADDR(DDRB), r0

Note:
The , " ax " ,@progbits tells the assembler that the section is allocatable ("a"),
executable ("x") and contains data ("@progbits"). For more detailed information
on the .section directive, see the gas user manual.
7.7.9 Using Sections in C Code
Example:
#include & lt; avr/io.h & gt;
void my_init_portb (void) __attribute__ ((naked)) \
__attribute__ ((section ( " .init1 " )));

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

125

7.8

Installing the GNU Tool Chain

void
my_init_portb (void)
{
outb (PORTB, 0xff);
outb (DDRB, 0xff);
}

7.8

Installing the GNU Tool Chain

Note:
This discussion was taken directly from Rich Neswold's document. (See Acknowledgments).
Note:
This discussion is Unix specific. [FIXME: troth/2002-08-13: we need a volunteer
to add windows specific notes to these instructions.]
This chapter shows how to build and install a complete development environment for
the AVR processors using the GNU toolset.
The default behaviour for most of these tools is to install every thing under the
/usr/local directory. In order to keep the AVR tools separate from the base
system, it is usually better to install everything into /usr/local/avr. If the
/usr/local/avr directory does not exist, you should create it before trying to
install anything. You will need root access to install there. If you don't have root
access to the system, you can alternatively install in your home directory, for example, in $HOME/local/avr. Where you install is a completely arbitrary decision, but
should be consistent for all the tools.
You specify the installation directory by using the --prefix=dir option with the
configure script. It is important to install all the AVR tools in the same directory
or some of the tools will not work correctly. To ensure consistency and simplify the
discussion, we will use $PREFIX to refer to whatever directory you wish to install in.
You can set this as an environment variable if you wish as such (using a Bourne-like
shell):
$ PREFIX=$HOME/local/avr
$ export PREFIX

Note:
Be sure that you have your PATH environment variable set to search the directory you install everything in before you start installing anything. For example, if
you use --prefix=$PREFIX, you must have $PREFIX/bin in your exported
PATH. As such:

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

126

7.8

Installing the GNU Tool Chain

$ PATH=$PATH:$PREFIX/bin
$ export PATH

Warning:
If you have CC set to anything other than avr-gcc in your environment, this will
cause the configure script to fail. It is best to not have CC set at all.
Note:
The versions for the packages listed below are known to work together. If you mix
and match different versions, you may have problems.
7.8.1 Required Tools
o GNU Binutils (2.14)
http://sources.redhat.com/binutils/
Installation
o GCC (3.3.1)
http://gcc.gnu.org/
Installation
o AVR Libc (1.0)
http://savannah.gnu.org/projects/avr-libc/
Installation
Note 2003-08-15: The 1.0 branch was created today.
7.8.2

Optional Tools

You can develop programs for AVR devices without the following tools. They may or
may not be of use for you.
Note:
The following programs are in constant development and the stated versions may
be less than current when you read this. Check the given websites for the latest
versions.
o uisp (20030618)
http://savannah.gnu.org/projects/uisp/
Installation

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

127

7.8

Installing the GNU Tool Chain

o avrdude (4.1.0)
http://savannah.nongnu.org/projects/avrdude/
Installation
Usage Notes
o GDB (6.0)
http://sources.redhat.com/gdb/
Installation
Note 2003-08-15: gdb-6.0 should be released soon.
o Simulavr (0.1.1)
http://savannah.gnu.org/projects/simulavr/
Installation
o AVaRice (2.0)
http://avarice.sourceforge.net/
Installation
Note 2003-08-15: avarice-2.0 only exists in cvs, there should be 2.1 release soon.
7.8.3

GNU Binutils for the AVR target

The binutils package provides all the low-level utilities needed in building and manipulating object files. Once installed, your environment will have an AVR assembler
(avr-as), linker (avr-ld), and librarian (avr-ar and avr-ranlib). In addition, you get tools which extract data from object files (avr-objcopy), dissassemble object file information (avr-objdump), and strip information from object files
(avr-strip). Before we can build the C compiler, these tools need to be in place.
Download and unpack the source files:
$ bunzip2 -c binutils- & lt; version & gt; .tar.bz2 | tar xf $ cd binutils- & lt; version & gt;

Note:
Replace & lt; version & gt; with the version of the package you downloaded.
Note:
If you obtained a gzip compressed file (.gz), use gunzip instead of bunzip2.
It is usually a good idea to configure and build binutils in a subdirectory so as not
to pollute the source with the compiled files. This is recommended by the binutils
developers.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

128

7.8

Installing the GNU Tool Chain

$ mkdir obj-avr
$ cd obj-avr

The next step is to configure and build the tools. This is done by supplying arguments
to the configure script that enable the AVR-specific options.
$ ../configure --prefix=$PREFIX --target=avr --disable-nls

If you don't specify the --prefix option, the tools will get installed in the
/usr/local hierarchy (i.e. the binaries will get installed in /usr/local/bin,
the info pages get installed in /usr/local/info, etc.) Since these tools are changing frequently, It is preferrable to put them in a location that is easily removed.
When configure is run, it generates a lot of messages while it determines what
is available on your operating system. When it finishes, it will have created several
Makefiles that are custom tailored to your platform. At this point, you can build the
project.
$ make

Note:
BSD users should note that the project's Makefile uses GNU make syntax.
This means FreeBSD users may need to build the tools by using gmake.
If the tools compiled cleanly, you're ready to install them. If you specified a destination
that isn't owned by your account, you'll need root access to install them. To install:
$ make install

You should now have the programs from binutils installed into $PREFIX/bin. Don't
forget to set your PATH environment variable before going to build avr-gcc.
7.8.4

GCC for the AVR target

Warning:
You must install avr-binutils and make sure your path is set properly before installing avr-gcc.
The steps to build avr-gcc are essentially same as for binutils:
$
$
$
$
$

bunzip2 -c gcc- & lt; version & gt; .tar.bz2 | tar xf cd gcc- & lt; version & gt;
mkdir obj-avr
cd obj-avr
../configure --prefix=$PREFIX --target=avr --enable-languages=c,c++ \
--disable-nls
$ make
$ make install

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

129

7.8

Installing the GNU Tool Chain

To save your self some download time, you can alternatively download only the
gcc-core- & lt; version & gt; .tar.bz2 and gcc-c++- & lt; version & gt; .tar.bz2
parts of the gcc. Also, if you don't need C++ support, you only need the core part
and should only enable the C language support.
Note:
Early versions of these tools did not support C++.
Note:
The stdc++ libs are not included with C++ for AVR due to the size limitations of
the devices.
7.8.5 AVR Libc
Warning:
You must install avr-binutils, avr-gcc and make sure your path is set properly
before installing avr-libc.
Note:
If you have obtained the latest avr-libc from cvs, you will have to run the reconf
script before using either of the build methods described below.
To build and install avr-libc:
$
$
$
$
$
$

gunzip -c avr-libc- & lt; version & gt; .tar.gz
cd avr-libc- & lt; version & gt;
./doconf
./domake
cd build
make install

Note:
The doconf script will automatically use the $PREFIX environment variable if
you have set and exported it.
Alternatively, you could do this (shown for consistency with binutils and gcc):
$
$
$
$
$
$
$

gunzip -c avr-libc- & lt; version & gt; .tar.gz | tar xf cd avr-libc- & lt; version & gt;
mkdir obj-avr
cd obj-avr
../configure --prefix=$PREFIX
make
make install

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

130

7.8

7.8.6

Installing the GNU Tool Chain

UISP

Uisp also uses the configure system, so to build and install:
$
$
$
$
$
$
$

gunzip -c uisp- & lt; version & gt; .tar.gz | tar xf cd uisp- & lt; version & gt;
mkdir obj-avr
cd obj-avr
../configure --prefix=$PREFIX
make
make install

7.8.7

Avrdude

Note:
It has been ported to windows (via cygwin) and linux. Other unix systems should
be trivial to port to.
avrdude is part of the FreeBSD ports system. To install it, simply do the following:
# cd /usr/ports/devel/avrdude
# make install

Note:
Installation into the default location usually requires root permissions. However,
running the program only requires access permissions to the appropriate ppi(4)
device.
Building and installing on other systems should use the configure system, as such:
$
$
$
$
$
$
$

gunzip -c avrdude- & lt; version & gt; .tar.gz | tar xf cd avrdude- & lt; version & gt;
mkdir obj-avr
cd obj-avr
../configure --prefix=$PREFIX
make
make install

7.8.8

GDB for the AVR target

Gdb also uses the configure system, so to build and install:
$
$
$
$
$
$
$

bunzip2 -c gdb- & lt; version & gt; .tar.bz2 | tar xf cd gdb- & lt; version & gt;
mkdir obj-avr
cd obj-avr
../configure --prefix=$PREFIX --target=avr
make
make install

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

131

7.8

Installing the GNU Tool Chain

Note:
If you are planning on using avr-gdb, you will probably want to install either
simulavr or avarice since avr-gdb needs one of these to run as a a remote target
backend.
7.8.9

Simulavr

Simulavr also uses the configure system, so to build and install:
$
$
$
$
$
$
$

gunzip -c simulavr- & lt; version & gt; .tar.gz | tar xf cd simulavr- & lt; version & gt;
mkdir obj-avr
cd obj-avr
../configure --prefix=$PREFIX
make
make install

Note:
You might want to have already installed avr-binutils, avr-gcc and avr-libc if you
want to have the test programs built in the simulavr source.
7.8.10

AVaRice

Note:
These install notes are not applicable to avarice-1.5 or older. You probably don't
want to use anything that old anyways since there have been many improvements
and bug fixes since the 1.5 release.
AVaRice also uses the configure system, so to build and install:
$
$
$
$
$
$
$

gunzip -c avarice- & lt; version & gt; .tar.gz | tar xf cd avarice- & lt; version & gt;
mkdir obj-avr
cd obj-avr
../configure --prefix=$PREFIX
make
make install

Note:
AVaRice uses the bfd library for accessing various binary file formats. You may
need to tell the configure script where to find the lib and headers for the link to
work. This is usually done by invoking the configure script like this (Replace
& lt; hdr path & gt; with the path to the bfd.h file on your system. Replace & lt; lib path & gt; with the path to libbfd.a on your system.):
$ CPPFLAGS=-I & lt; hdr_path & gt; LDFLAGS=-L & lt; lib_path & gt; ../configure --prefix=$PREFIX

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

132

7.9

Using the avrdude program

Note:
As of 2003-08-15, no offical AVaRice release works like this. Use a 2.0 snapshot
until the 2.1 release is made, or use obtain the source from cvs.

7.9

Using the avrdude program

Note:
This section was contributed by Brian Dean [ bsd@bsdhome.com ].
Note:
The avrdude program was previously called avrprog. The name was changed to
avoid confusion with the avrprog program that Atmel ships with AvrStudio.
avrdude is a program that is used to update or read the flash and EEPROM memories
of Atmel AVR microcontrollers on FreeBSD Unix. It supports the Atmel serial programming protocol using the PC's parallel port and can upload either a raw binary file
or an Intel Hex format file. It can also be used in an interactive mode to individually
update EEPROM cells, fuse bits, and/or lock bits (if their access is supported by the
Atmel serial programming protocol.) The main flash instruction memory of the AVR
can also be programmed in interactive mode, however this is not very useful because
one can only turn bits off. The only way to turn flash bits on is to erase the entire
memory (using avrdude's -e option).
avrdude is part of the FreeBSD ports system. To install it, simply do the following:
# cd /usr/ports/devel/avrdude
# make install

Once installed, avrdude can program processors using the contents of the .hex file
specified on the command line. In this example, the file main.hex is burned into the
flash memory:
# avrdude -p 2313 -e -m flash -i main.hex
avrdude: AVR device initialized and ready to accept instructions
avrdude: Device signature = 0x1e9101
avrdude:
avrdude:
avrdude:
avrdude:

erasing chip
done.
reading input file " main.hex "
input file main.hex auto detected as Intel Hex

avrdude: writing flash:
1749 0x00
avrdude: 1750 bytes of flash written
avrdude: verifying flash memory against main.hex:
avrdude: reading on-chip flash data:

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

133

7.10 Using the GNU tools

1749 0x00
avrdude: verifying ...
avrdude: 1750 bytes of flash verified
avrdude done.

Thank you.

The -p 2313 option lets avrdude know that we are operating on an AT90S2313
chip. This option specifies the device id and is matched up with the device of the same
id in avrdude's configuration file ( /usr/local/etc/avrdude.conf ). To list
valid parts, specify the -v option. The -e option instructs avrdude to perform a
chip-erase before programming; this is almost always necessary before programming
the flash. The -m flash option indicates that we want to upload data into the flash
memory, while -i main.hex specifies the name of the input file.
The EEPROM is uploaded in the same way, the only difference is that you would use
-m eeprom instead of -m flash.
To use interactive mode, use the -t option:
# avrdude -p 2313 -t
avrdude: AVR device initialized and ready to accept instructions
avrdude: Device signature = 0x1e9101
avrdude & gt;
The '?' command displays a list of valid
commands:
avrdude & gt; ?
& gt; & gt; & gt; ?
Valid commands:
dump
read
write
erase
sig
part
send
help
?
quit

:
:
:
:
:
:
:
:
:
:

dump memory : dump & lt; memtype & gt; & lt; addr & gt; & lt; N-Bytes & gt;
alias for dump
write memory : write & lt; memtype & gt; & lt; addr & gt; & lt; b1 & gt; & lt; b2 & gt; ... & lt; bN & gt;
perform a chip erase
display device signature bytes
display the current part information
send a raw command : send & lt; b1 & gt; & lt; b2 & gt; & lt; b3 & gt; & lt; b4 & gt;
help
help
quit

Use the 'part' command to display valid memory types for use with the
'dump' and 'write' commands.
avrdude & gt;

7.10

Using the GNU tools

This is a short summary of the AVR-specific aspects of using the GNU tools. Normally,
the generic documentation of these tools is fairly large and maintained in texinfo
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

134

7.10 Using the GNU tools

135

files. Command-line options are explained in detail in the manual page.
7.10.1

Options for the C compiler avr-gcc

7.10.1.1 Machine-specific options for the AVR The following machine-specific
options are recognized by the C compiler frontend.
o -mmcu=architecture
Compile code for architecture. Currently known architectures are
avr1

Simple CPU core, only assembler
support
"Classic" CPU core, up to 8 KB of
ROM
"Classic" CPU core, more than 8
KB of ROM
"Enhanced" CPU core, up to 8 KB
of ROM
"Enhanced" CPU core, more than 8
KB of ROM

avr2
avr3
avr4
avr5

By default, code is generated for the avr2 architecture.
Note that when only using -mmcu=architecture but no -mmcu=MCU type, including the file & lt; avr/io.h & gt; cannot work since it cannot decide which device's definitions to select.
o -mmcu=MCU type
The following MCU types are currently understood by avr-gcc. The table
matches them against the corresponding avr-gcc architecture name, and shows
the preprocessor symbol declared by the -mmcu option.
Architecture
avr1
avr1
avr1
avr1
avr1
avr2
avr2
avr2
avr2
avr2
avr2
avr2

MCU name
at90s1200
attiny11
attiny12
attiny15
attiny28
at90s2313
at90s2323
at90s2333
at90s2343
attiny22
attiny26
at90s4414

Macro
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

AT90S1200
ATtiny11
ATtiny12
ATtiny15
ATtiny28
AT90S2313
AT90S2323
AT90S2333
AT90S2343
ATtiny22
ATtiny26
AT90S4414

7.10 Using the GNU tools

Architecture
avr2
avr2
avr2
avr2
avr2
avr2
avr3
avr3
avr3
avr3
avr3
avr4
avr4
avr4
avr5
avr5
avr5
avr5
avr5
avr5
avr5
avr5
avr5
avr5

136

MCU name
at90s4433
at90s4434
at90s8515
at90c8534
at90s8535
at86rf401
atmega103
atmega603
at43usb320
at43usb355
at76c711
atmega8
atmega8515
atmega8535
atmega16
atmega161
atmega162
atmega163
atmega169
atmega32
atmega323
atmega64
atmega128
at94k

Macro
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR
AVR

AT90S4433
AT90S4434
AT90S8515
AT90C8534
AT90S8535
AT86RF401
ATmega103
ATmega603
AT43USB320
AT43USB355
AT76C711
ATmega8
ATmega8515
ATmega8535
ATmega16
ATmega161
ATmega162
ATmega163
ATmega169
ATmega32
ATmega323
ATmega64
ATmega128
AT94K

o -morder1
o -morder2
Change the order of register assignment. The default is
r24, r25, r18, r19, r20, r21, r22, r23, r30, r31, r26, r27, r28, r29, r17, r16, r15,
r14, r13, r12, r11, r10, r9, r8, r7, r6, r5, r4, r3, r2, r0, r1
Order 1 uses
r18, r19, r20, r21, r22, r23, r24, r25, r30, r31, r26, r27, r28, r29, r17, r16, r15,
r14, r13, r12, r11, r10, r9, r8, r7, r6, r5, r4, r3, r2, r0, r1
Order 2 uses
r25, r24, r23, r22, r21, r20, r19, r18, r30, r31, r26, r27, r28, r29, r17, r16, r15,
r14, r13, r12, r11, r10, r9, r8, r7, r6, r5, r4, r3, r2, r1, r0
o -mint8
Assume int to be an 8-bit integer. Note that this is not really supported by
avr-libc, so it should normally not be used. The default is to use 16-bit
integers.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.10 Using the GNU tools

o -mno-interrupts
Generates code that changes the stack pointer without disabling interrupts. Normally, the state of the status register SREG is saved in a temporary register, interrupts are disabled while changing the stack pointer, and SREG is restored.
o -mcall-prologues
Use subroutines for function prologue/epilogue. For complex functions that use
many registers (that needs to be saved/restored on function entry/exit), this saves
some space at the cost of a slightly increased execution time.
o -minit-stack=nnnn
Set the initial stack pointer to nnnn. By default, the stack pointer is initialized
to the symbol stack, which is set to RAMEND by the run-time initialization
code.
o -mtiny-stack
Change only the low 8 bits of the stack pointer.
o -mno-tablejump
Do not generate tablejump instructions. By default, jump tables can be used to
optimize switch statements. When turned off, sequences of compare statements are used instead. Jump tables are usually faster to execute on average, but
in particular for switch statements where most of the jumps would go to the
default label, they might waste a bit of flash memory.
o -mshort-calls
Use rjmp/rcall (limited range) on & gt; 8K devices. On avr2 and avr4 architectures (less than 8 KB or flash memory), this is always the case. On avr3 and
avr5 architectures, calls and jumps to targets outside the current function will
by default use jmp/call instructions that can cover the entire address range,
but that require more flash ROM and execution time.
o -mrtl
Dump the internal compilation result called "RTL" into comments in the generated assembler code. Used for debugging avr-gcc.
o -msize
Dump the address, size, and relative cost of each statement into comments in the
generated assembler code. Used for debugging avr-gcc.
o -mdeb
Generate lots of debugging information to stderr.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

137

7.10 Using the GNU tools

7.10.1.2 Selected general compiler options The following general gcc options
might be of some interest to AVR users.
o -On
Optimization level n. Increasing n is meant to optimize more, an optimization
level of 0 means no optimization at all, which is the default if no -O option is
present. The special option -Os is meant to turn on all -O2 optimizations that
are not expected to increase code size.
Note that at -O3, gcc attempts to inline all "simple" functions. For the
AVR target, this will normally constitute a large pessimization due to the
code increasement. The only other optimization turned on with -O3 is
-frename-registers, which could rather be enabled manually instead.
A simple -O option is equivalent to -O1.
Note also that turning off all optimizations will prevent some warnings from being issued since the generation of those warnings depends on code analysis steps
that are only performed when optimizing (unreachable code, unused variables).
See also the appropriate FAQ entry for issues regarding debugging optimized
code.
o -Wa,assembler-options
o -Wl,linker-options
Pass the listed options to the assembler, or linker, respectively.
o -g
Generate debugging information that can be used by avr-gdb.
o -ffreestanding
Assume a "freestanding" environment as per the C standard. This turns off automatic builtin functions (though they can still be reached by prepending builtin to the actual function name). It also makes the compiler not complain when main() is declared with a void return type which makes some
sense in a microcontroller environment where the application cannot meaningfully provide a return value to its environment (in most cases, main() won't
even return anyway). However, this also turns off all optimizations normally
done by the compiler which assume that functions known by a certain name behave as described by the standard. E. g., applying the function strlen() to a literal
string will normally cause the compiler to immediately replace that call by the
actual length of the string, while with -ffreestanding, it will always call
strlen() at run-time.
o -funsigned-char
Make any unqualfied char type an unsigned char. Without this option, they
default to a signed char.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

138

7.10 Using the GNU tools

o -funsigned-bitfields
Make any unqualified bitfield type unsigned. By default, they are signed.
o -fshort-enums
Allocate to an enum type only as many bytes as it needs for the declared range
of possible values. Specifically, the enum type will be equivalent to the smallest
integer type which has enough room.
o -fpack-struct
Pack all structure members together without holes.
7.10.2

Options for the assembler avr-as

7.10.2.1

Machine-specific assembler options

o -mmcu=architecture
o -mmcu=MCU name
avr-as understands the same -mmcu= options as avr-gcc. By default, avr2 is assumed, but this can be altered by using the appropriate .arch pseudo-instruction
inside the assembler source file.
o -mall-opcodes
Turns off opcode checking for the actual MCU type, and allows any possible
AVR opcode to be assembled.
o -mno-skip-bug
Don't emit a warning when trying to skip a 2-word instruction with a
CPSE/SBIC/SBIS/SBRC/SBRS instruction. Early AVR devices suffered
from a hardware bug where these instructions could not be properly skipped.
o -mno-wrap
For RJMP/RCALL instructions, don't allow the target address to wrap around
for devices that have more than 8 KB of memory.
o --gstabs
Generate .stabs debugging symbols for assembler source lines. This enables
avr-gdb to trace through assembler source files. This option must not be used
when assembling sources that have been generated by the C compiler; these files
already contain the appropriate line number information from the C source files.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

139

7.10 Using the GNU tools

o -a[cdhlmns=file]
Turn on the assembler listing. The sub-options are:
-
-
-
-
-
-
-
-

c omit false conditionals
d omit debugging directives
h include high-level source
l include assembly
m include macro expansions
n omit forms processing
s include symbols
=file set the name of the listing file

The various sub-options can be combined into a single -a option list; =file must
be the last one in that case.
7.10.2.2 Examples for assembler options passed through the C compiler Remember that assembler options can be passed from the C compiler frontend using -Wa
(see above), so in order to include the C source code into the assembler listing in
file foo.lst, when compiling foo.c, the following compiler command-line can be
used:
$ avr-gcc -c -O foo.c -o foo.o -Wa,-ahls=foo.lst

In order to pass an assembler file through the C preprocessor first, and have the assembler generate line number debugging information for it, the following command can be
used:
$ avr-gcc -c -x assembler-with-cpp -o foo.o foo.S -Wa,--gstabs

Note that on Unix systems that have case-distinguishing file systems, specifying a file
name with the suffix .S (upper-case letter S) will make the compiler automatically
assume -x assembler-with-cpp, while using .s would pass the file directly to
the assembler (no preprocessing done).
7.10.3

Controlling the linker avr-ld

7.10.3.1 Selected linker options While there are no machine-specific options for
avr-ld, a number of the standard options might be of interest to AVR users.
o -lname
Locate the archive library named libname.a, and use it to resolve currently
unresolved symbols from it. The library is searched along a path that consists of builtin pathname entries that have been specified at compile time (e.
g. /usr/local/avr/lib on Unix systems), possibly extended by pathname entries as specified by -L options (that must precede the -l options on
the command-line).
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

140

7.10 Using the GNU tools

o -Lpath
Additional location to look for archive libraries requested by -l options.
o --defsym symbol=expr
Define a global symbol symbol using expr as the value.
o -M
Print a linker map to stdout.
o -Map mapfile
Print a linker map to mapfile.
o --cref
Output a cross reference table to the map file (in case -Map is also present), or
to stdout.
o --section-start sectionname=org
Start section sectionname at absolute address org.
o -Tbss org
o -Tdata org
o -Ttext org
Start the bss, data, or text section at org, respectively.
o -T scriptfile
Use scriptfile as the linker script, replacing the default linker script. Default linker scripts are stored in a system-specific location (e. g. under
/usr/local/avr/lib/ldscripts on Unix systems), and consist of the
AVR architecture name (avr2 through avr5) with the suffix .x appended. They
describe how the various memory sections will be linked together.
7.10.3.2 Passing linker options from the C compiler By default, all unknown
non-option arguments on the avr-gcc command-line (i. e., all filename arguments that
don't have a suffix that is handled by avr-gcc) are passed straight to the linker. Thus,
all files ending in .o (object files) and .a (object libraries) are provided to the linker.
System libraries are usually not passed by their explicit filename but rather using the
-l option which uses an abbreviated form of the archive filename (see above). avrlibc ships two system libraries, libc.a, and libm.a. While the standard library
libc.a will always be searched for unresolved references when the linker is started
using the C compiler frontend (i. e., there's always at least one implied -lc option),
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

141

7.11 A simple project

the mathematics library libm.a needs to be explicitly requested using -lm. See also
the entry in the FAQ explaining this.
Conventionally, Makefiles use the make macro LDLIBS to keep track of -l (and
possibly -L) options that should only be appended to the C compiler command-line
when linking the final binary. In contrast, the macro LDFLAGS is used to store other
command-line options to the C compiler that should be passed as options during the
linking stage. The difference is that options are placed early on the command-line,
while libraries are put at the end since they are to be used to resolve global symbols
that are still unresolved at this point.
Specific linker flags can be passed from the C compiler command-line using the -Wl
compiler option, see above. This option requires that there be no spaces in the appended
linker option, while some of the linker options above (like -Map or --defsym)
would require a space. In these situations, the space can be replaced by an equal sign
as well. For example, the following command-line can be used to compile foo.c into
an executable, and also produce a link map that contains a cross-reference list in the
file foo.map:
$ avr-gcc -O -o foo.out -Wl,-Map=foo.map -Wl,--cref foo.c

Alternatively, a comma as a placeholder will be replaced by a space before passing the
option to the linker. So for a device with external SRAM, the following command-line
would cause the linker to place the data segment at address 0x2000 in the SRAM:
$ avr-gcc -mmcu=atmega128 -o foo.out -Wl,-Tdata,0x802000

See the explanation of the data section for why 0x800000 needs to be added to the actual value. Note that unless a -minit-stack option has been given when compiling
the C source file that contains the function main(), the stack will still remain in internal RAM, through the symbol stack that is provided by the run-time startup code.
This is probably a good idea anyway (since internal RAM access is faster), and even
required for some early devices that had hardware bugs preventing them from using
a stack in external RAM. Note also that the heap for malloc() will still be placed
after all the variables in the data section, so in this situation, no stack/heap collision
can occur.

7.11

A simple project

At this point, you should have the GNU tools configured, built, and installed on your
system. In this chapter, we present a simple example of using the GNU tools in an AVR
project. After reading this chapter, you should have a better feel as to how the tools are
used and how a Makefile can be configured.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

142

7.11 A simple project

7.11.1

143

The Project

This project will use the pulse-width modulator ( PWM ) to ramp an LED on and off
every two seconds. An AT90S2313 processor will be used as the controller. The circuit
for this demonstration is shown in the schematic diagram. If you have a development
kit, you should be able to use it, rather than build the circuit, for this project.
VCC
IC1

.1uf

C4

18pf

GND
GND

C2
18pf

4mhz

C1

(SCK)PB7
(MISO)PB6
(MOSI)PB5
PB4
(OCI)PB3
PB2
(AIN1)PB1
(AIN0)PB0

19
18
17
16
15
14
13
12

(ICP)PD6
(T1)PD5
(T0)PD4
(INT1)PD3
(INT0)PD2
(TXD)PD1
(RXD)PD0
AT90S2313P

11
9
8
7
6
3
2

1
Q1

.01uf

20K

C3

R1

RESET

4

XTAL2

5

XTAL1

20 VCC
10 GND

R2*

LED5MM
D1

See note [7]
GND

Figure 5: Schematic of circuit for demo project

The source code is given in demo.c. For the sake of this example, create a file called
demo.c containing this source code. Some of the more important parts of the code
are:
Note [1]:
The PWM is being used in 10-bit mode, so we need a 16-bit variable to remember
the current value.
Note [2]:
SIGNAL() is a macro that marks the function as an interrupt routine. In this case,
the function will get called when the timer overflows. Setting up interrupts is
explained in greater detail in Interrupts and Signals.
Note [3]:
This section determines the new value of the PWM.
Note [4]:
Here's where the newly computed value is loaded into the PWM register. Since
we are in an interrupt routine, it is safe to use a 16-bit assignment to the register.
Outside of an interrupt, the assignment should only be performed with interrupts
disabled if there's a chance that an interrupt routine could also access this register
(or another register that uses TEMP), see the appropriate FAQ entry.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.11 A simple project

Note [5]:
This routine gets called after a reset. It initializes the PWM and enables interrupts.
Note [6]:
The main loop of the program does nothing - all the work is done by the interrupt
routine! If this was a real product, we'd probably put a SLEEP instruction in this
loop to conserve power.
Note [7]:
Early AVR devices saturate their outputs at rather low currents when sourcing current, so the LED can be connected directly, the resulting current through the LED
will be about 15 mA. For modern parts (at least for the ATmega 128), however
Atmel has drastically increased the IO source capability, so when operating at 5
V Vcc, R2 is needed. Its value should be about 150 Ohms. When operating the
circuit at 3 V, it can still be omitted though.
7.11.2

The Source Code

/*
* ---------------------------------------------------------------------------* " THE BEER-WARE LICENSE " (Revision 42):
* & lt; joerg@FreeBSD.ORG & gt; wrote this file. As long as you retain this notice you
* can do whatever you want with this stuff. If we meet some day, and you think
* this stuff is worth it, you can buy me a beer in return.
Joerg Wunsch
* ---------------------------------------------------------------------------*
* Simple AVR demonstration. Controls a LED that can be directly
* connected from OC1/OC1A to GND. The brightness of the LED is
* controlled with the PWM. After each period of the PWM, the PWM
* value is either incremented or decremented, that's all.
*
* $Id: demo.c,v 1.1 2002/09/30 18:16:07 troth Exp $
*/
#include
#include
#include
#include

& lt; inttypes.h & gt;
& lt; avr/io.h & gt;
& lt; avr/interrupt.h & gt;
& lt; avr/signal.h & gt;

#if defined(__AVR_AT90S2313__)
# define OC1 PB3
# define OCR OCR1
# define DDROC DDRB
#elif defined(__AVR_AT90S2333__) || defined(__AVR_AT90S4433__)
# define OC1 PB1
# define DDROC DDRB
# define OCR OCR1
#elif defined(__AVR_AT90S4414__) || defined(__AVR_AT90S8515__) || \
defined(__AVR_AT90S4434__) || defined(__AVR_AT90S8535__) || \
defined(__AVR_ATmega163__)
# define OC1 PD5

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

144

7.11 A simple project

# define DDROC DDRD
# define OCR OCR1A
#else
# error " Don't know what kind of MCU you are compiling for "
#endif
#if defined(COM11)
# define XCOM11 COM11
#elif defined(COM1A1)
# define XCOM11 COM1A1
#else
# error " need either COM1A1 or COM11 "
#endif
enum { UP, DOWN };
volatile uint16_t pwm; /* Note [1] */
volatile uint8_t direction;
SIGNAL (SIG_OVERFLOW1) /* Note [2] */
{
switch (direction) /* Note [3] */
{
case UP:
if (++pwm == 1023)
direction = DOWN;
break;
case DOWN:
if (--pwm == 0)
direction = UP;
break;
}
OCR = pwm; /* Note [4] */
}
void
ioinit (void) /* Note [5] */
{
/* tmr1 is 10-bit PWM */
TCCR1A = _BV (PWM10) | _BV (PWM11) | _BV (XCOM11);
/* tmr1 running on full MCU clock */
TCCR1B = _BV (CS10);
/* set PWM value to 0 */
OCR = 0;
/* enable OC1 and PB2 as output */
DDROC = _BV (OC1);
timer_enable_int (_BV (TOIE1));
/* enable interrupts */

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

145

7.11 A simple project

sei ();
}
int
main (void)
{
ioinit ();
/* loop forever, the interrupts are doing the rest */
for (;;) /* Note [6] */
;
return (0);
}

7.11.3

Compiling and Linking

This first thing that needs to be done is compile the source. When compiling, the
compiler needs to know the processor type so the -mmcu option is specified. The
-Os option will tell the compiler to optimize the code for efficient space usage (at the
possible expense of code execution speed). The -g is used to embed debug info. The
debug info is useful for disassemblies and doesn't end up in the .hex files, so I usually
specify it. Finally, the -c tells the compiler to compile and stop - don't link. This
demo is small enough that we could compile and link in one step. However, real-world
projects will have several modules and will typically need to break up the building of
the project into several compiles and one link.
$ avr-gcc -g -Os -mmcu=at90s2333 -c demo.c

The compilation will create a demo.o file. Next we link it into a binary called
demo.elf.
$ avr-gcc -g -mmcu=at90s2333 -o demo.elf demo.o

It is important to specify the MCU type when linking. The compiler uses the -mmcu
option to choose start-up files and run-time libraries that get linked together. If this
option isn't specified, the compiler defaults to the 8515 processor environment, which
is most certainly what you didn't want.
7.11.4

Examining the Object File

Now we have a binary file. Can we do anything useful with it (besides put it into the
processor?) The GNU Binutils suite is made up of many useful tools for manipulating
object files that get generated. One tool is avr-objdump, which takes information
from the object file and displays it in many useful ways. Typing the command by itself
will cause it to list out its options.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

146

7.11 A simple project

147

For instance, to get a feel of the application's size, the -h option can be used. The
output of this option shows how much space is used in each of the \sections (the .stab
and .stabstr sections hold the debugging information and won't make it into the
ROM file).
An even more useful option is -S. This option disassembles the binary file and intersperses the source code in the output! This method is much better, in my opinion, than
using the -S with the compiler because this listing includes routines from the libraries
and the vector table contents. Also, all the "fix-ups" have been satisfied. In other words,
the listing generated by this option reflects the actual code that the processor will run.
$ avr-objdump -h -S demo.elf & gt; demo.lst

Here's the output as saved in the demo.lst file:
demo.elf:

file format elf32-avr

Sections:
Idx Name
0 .text

Size
VMA
LMA
File off
000000ca 00000000 00000000 00000094
CONTENTS, ALLOC, LOAD, READONLY, CODE
1 .data
00000000 00800060 000000ca 0000015e
CONTENTS, ALLOC, LOAD, DATA
2 .bss
00000003 00800060 00800060 0000015e
ALLOC
3 .noinit
00000000 00800063 00800063 0000015e
CONTENTS
4 .eeprom
00000000 00810000 00810000 0000015e
CONTENTS
5 .stab
00000684 00000000 00000000 00000160
CONTENTS, READONLY, DEBUGGING
6 .stabstr
0000063d 00000000 00000000 000007e4
CONTENTS, READONLY, DEBUGGING
Disassembly of section .text:
00000000 & lt; __vectors & gt; :
0:
0a c0
2:
62 c0
4:
61 c0
6:
60 c0
8:
5f c0
a:
0a c0
c:
5d c0
e:
5c c0
10:
5b c0
12:
5a c0
14:
59 c0

rjmp
rjmp
rjmp
rjmp
rjmp
rjmp
rjmp
rjmp
rjmp
rjmp
rjmp

00000016 & lt; __ctors_end & gt; :
16:
11 24
eor
18:
1f be
out
1a:
cf ed
ldi

.+20
.+196
.+194
.+192
.+190
.+20
.+186
.+184
.+182
.+180
.+178

;
;
;
;
;
;
;
;
;
;
;

r1, r1
0x3f, r1
r28, 0xDF

; 63
; 223

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

0x16
0xc8
0xc8
0xc8
0xc8
0x20
0xc8
0xc8
0xc8
0xc8
0xc8

Algn
2**0
2**0
2**0
2**0
2**0
2**2
2**0

7.11 A simple project

1c:
1e:

cd bf
4e c0

148

out
rjmp

0x3d, r28
.+156

; 61
; 0xbc

00000020 & lt; __vector_5 & gt; :
volatile uint16_t pwm; /* Note [1] */
volatile uint8_t direction;
SIGNAL (SIG_OVERFLOW1) /* Note [2] */
{
20:
1f 92
push
r1
22:
0f 92
push
r0
24:
0f b6
in
r0, 0x3f
26:
0f 92
push
r0
28:
11 24
eor
r1, r1
2a:
2f 93
push
r18
2c:
8f 93
push
r24
2e:
9f 93
push
r25
switch (direction) /* Note [3] */
30:
80 91 60 00
lds
r24, 0x0060
34:
99 27
eor
r25, r25
36:
00 97
sbiw
r24, 0x00
38:
a1 f0
breq
.+40
3a:
01 97
sbiw
r24, 0x01
3c:
29 f5
brne
.+74
{
case UP:
if (++pwm == 1023)
direction = DOWN;
break;

3e:
42:
46:
48:
4c:
50:
54:
58:
5a:
5c:
60:
62:
66:
6a:
6c:
70:
74:
78:
7c:
7e:
80:
82:

case DOWN:
if (--pwm == 0)
80 91 61 00
lds
90 91 62 00
lds
01 97
sbiw
90 93 62 00
sts
80 93 61 00
sts
80 91 61 00
lds
90 91 62 00
lds
89 2b
or
b1 f4
brne
direction = UP;
10 92 60 00
sts
13 c0
rjmp
80 91 61 00
lds
90 91 62 00
lds
01 96
adiw
90 93 62 00
sts
80 93 61 00
sts
80 91 61 00
lds
90 91 62 00
lds
8f 5f
subi
93 40
sbci
19 f4
brne
81 e0
ldi

r24, 0x0061
r25, 0x0062
r24, 0x01
0x0062, r25
0x0061, r24
r24, 0x0061
r25, 0x0062
r24, r25
.+44
0x0060, r1
.+38
r24, 0x0061
r25, 0x0062
r24, 0x01
0x0062, r25
0x0061, r24
r24, 0x0061
r25, 0x0062
r24, 0xFF
r25, 0x03
.+6
r24, 0x01

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

; 63

;
;
;
;

0
0x62
1
0x88

;1

; 0x88

; 0x88

;1

;
;
;
;

255
3
0x88
1

7.11 A simple project

84:

80 93 60 00
break;

149

sts

0x0060, r24

}
OCR
88:
8c:
90:
92:

= pwm; /* Note [4] */
80 91 61 00
lds
90 91 62 00
lds
9b bd
out
8a bd
out

94:
96:
98:
9a:
9c:
9e:
a0:
a2:

9f
8f
2f
0f
0f
0f
1f
18

r24, 0x0061
r25, 0x0062
0x2b, r25
0x2a, r24

; 43
; 42

}
91
91
91
90
be
90
90
95

pop
pop
pop
pop
out
pop
pop
reti

r25
r24
r18
r0
0x3f, r0
r0
r1

; 63

000000a4 & lt; ioinit & gt; :
void
ioinit (void) /* Note [5] */
{
/* tmr1 is 10-bit PWM */
TCCR1A = _BV (PWM10) | _BV (PWM11) | _BV (XCOM11);
a4:
83 e8
ldi
r24, 0x83
; 131
a6:
8f bd
out
0x2f, r24
; 47
/* tmr1 running on full MCU clock */
TCCR1B = _BV (CS10);
a8:
81 e0
ldi
r24, 0x01
aa:
8e bd
out
0x2e, r24

;1
; 46

/* set PWM value to 0 */
OCR = 0;
ac:
1b bc
out
ae:
1a bc
out

; 43
; 42

0x2b, r1
0x2a, r1

/* enable OC1 and PB2 as output */
DDROC = _BV (OC1);
b0:
88 e0
ldi
r24, 0x08
b2:
87 bb
out
0x17, r24

;8
; 23

extern inline void timer_enable_int (unsigned char ints)
{
#ifdef TIMSK
TIMSK = ints;
b4:
80 e8
ldi
r24, 0x80
; 128
b6:
89 bf
out
0x39, r24
; 57
timer_enable_int (_BV (TOIE1));
/* enable interrupts */
sei ();

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.11 A simple project

150

b8:

78 94

sei

ba:

08 95

ret

}

000000bc & lt; main & gt; :
int
main (void)
{
bc:
cf ed
be:
d0 e0
c0:
de bf
c2:
cd bf
ioinit ();
c4:
ef df

ldi
ldi
out
out

r28, 0xDF
r29, 0x00
0x3e, r29
0x3d, r28

;
;
;
;

223
0
62
61

rcall

.-34

; 0xa4

/* loop forever, the interrupts are doing the rest */
for (;;) /* Note [6] */
c6:
ff cf
rjmp
000000c8 & lt; __bad_interrupt & gt; :
c8:
9b cf
rjmp

7.11.5

.-2

; 0xc6

.-202

; 0x0

Linker Map Files

avr-objdump is very useful, but sometimes it's necessary to see information about
the link that can only be generated by the linker. A map file contains this information.
A map file is useful for monitoring the sizes of your code and data. It also shows where
modules are loaded and which modules were loaded from libraries. It is yet another
view of your application. To get a map file, I usually add -Wl,-Map,demo.map to
my link command. Relink the application using the following command to generate
demo.map (a portion of which is shown below).
$ avr-gcc -g -mmcu=at90s2313 -Wl,-Map,demo.map -o demo.elf demo.o

Some points of interest in the demo.map file are:
.rela.plt
*(.rela.plt)
.text
*(.vectors)
.vectors

0x00000000

0xca

0x00000000
0x00000000
0x00000000
0x00000016

0x16 ../../../obj-avr/crt1/crts2313.o
__vectors
__vector_default
__ctors_start = .

The .text segment (where program instructions are stored) starts at location 0x0.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.11 A simple project

151

*(.fini2)
*(.fini1)
*(.fini0)
.data

0x000000ca
0x00800060
0x00800060

*(.data)
*(.gnu.linkonce.d*)
0x00800060
0x00800060
0x00800060
.bss
0x00800060
0x00800060
*(.bss)
*(COMMON)
COMMON
0x00800060

_etext = .
0x0 load address 0x000000ca
PROVIDE (__data_start, .)

. = ALIGN (0x2)
_edata = .
PROVIDE (__data_end, .)
0x3
PROVIDE (__bss_start, .)

0x00800063
0x00800063
0x00800063
0x00810000

PROVIDE (__noinit_end, .)
_end = .
PROVIDE (__heap_start, .)
0x0 load address 0x000000ca

0x00810000

.noinit

0x00800060
0x00800061
0x00800063
0x000000ca
0x000000ca
0x00800063
0x00800063

0x3 demo.o
0x0 (size before relaxing)
direction
pwm
PROVIDE (__bss_end, .)
__data_load_start = LOADADDR (.data)
__data_load_end = (__data_load_start + SIZEOF (.data))
0x0
PROVIDE (__noinit_start, .)

__eeprom_end = .

*(.noinit*)

.eeprom
*(.eeprom*)

The last address in the .text segment is location 0xf2 ( denoted by etext ), so the
instructions use up 242 bytes of FLASH.
The .data segment (where initialized static variables are stored) starts at location
0x60, which is the first address after the register bank on a 2313 processor.
The next available address in the .data segment is also location 0x60, so the application has no initialized data.
The .bss segment (where uninitialized data is stored) starts at location 0x60.
The next available address in the .bss segment is location 0x63, so the application
uses 3 bytes of uninitialized data.
The .eeprom segment (where EEPROM variables are stored) starts at location 0x0.
The next available address in the .eeprom segment is also location 0x0, so there aren't
any EEPROM variables.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.11 A simple project

7.11.6

152

Intel Hex Files

We have a binary of the application, but how do we get it into the processor? Most
(if not all) programmers will not accept a GNU executable as an input file, so we
need to do a little more processing. The next step is to extract portions of the binary
and save the information into .hex files. The GNU utility that does this is called
avr-objcopy.
The ROM contents can be pulled from our project's binary and put into the file
demo.hex using the following command:
$ avr-objcopy -j .text -j .data -O ihex demo.elf demo.hex

The resulting demo.hex file contains:
:100000000AC062C061C060C05FC00AC05DC05CC0A1
:100010005BC05AC059C011241FBECFEDCDBF4EC02A
:100020001F920F920FB60F9211242F938F939F93CD
:100030008091600099270097A1F0019729F58091A0
:10004000610090916200019790936200809361003B
:100050008091610090916200892BB1F41092600050
:1000600013C08091610090916200019690936200AC
:100070008093610080916100909162008F5F934056
:1000800019F481E08093600080916100909162009A
:100090009BBD8ABD9F918F912F910F900FBE0F90A6
:1000A0001F90189583E88FBD81E08EBD1BBC1ABCE4
:1000B00088E087BB80E889BF78940895CFEDD0E0D1
:0A00C000DEBFCDBFEFDFFFCF9BCF07
:00000001FF

The -j option indicates that we want the information from the .text and .data
segment extracted. If we specify the EEPROM segment, we can generate a .hex file
that can be used to program the EEPROM:
$ avr-objcopy -j .eeprom --change-section-lma .eeprom=0 -O ihex demo.elf demo_eeprom.hex

The resulting demo eeprom.hex file contains:
:00000001FF

which is an empty .hex file (which is expected, since we didn't define any EEPROM
variables).
7.11.7

Make Build the Project

Rather than type these commands over and over, they can all be placed in a make file.
To build the demo project using make, save the following in a file called Makefile.
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.11 A simple project

153

Note:
This Makefile can only be used as input for the GNU version of make.
PRG
OBJ
MCU_TARGET
OPTIMIZE

=
=
=
=

DEFS
LIBS

demo
demo.o
at90s2313
-O2

=
=

# You should not have to change anything below here.
CC

= avr-gcc

# Override is only needed by avr-lib build system.
override CFLAGS
override LDFLAGS
OBJCOPY
OBJDUMP

= -g -Wall $(OPTIMIZE) -mmcu=$(MCU_TARGET) $(DEFS)
= -Wl,-Map,$(PRG).map

= avr-objcopy
= avr-objdump

all: $(PRG).elf lst text eeprom
$(PRG).elf: $(OBJ)
$(CC) $(CFLAGS) $(LDFLAGS) -o $@ $^ $(LIBS)
clean:
rm -rf *.o $(PRG).elf *.eps *.png *.pdf *.bak
rm -rf *.lst *.map $(EXTRA_CLEAN_FILES)
lst:

$(PRG).lst

%.lst: %.elf
$(OBJDUMP) -h -S $ & lt; & gt; $@
# Rules for building the .text rom images
text: hex bin srec
hex: $(PRG).hex
bin: $(PRG).bin
srec: $(PRG).srec
%.hex: %.elf
$(OBJCOPY) -j .text -j .data -O ihex $ & lt; $@
%.srec: %.elf
$(OBJCOPY) -j .text -j .data -O srec $ & lt; $@
%.bin: %.elf
$(OBJCOPY) -j .text -j .data -O binary $ & lt; $@
# Rules for building the .eeprom rom images

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.12 Example using the two-wire interface (TWI)

154

eeprom: ehex ebin esrec
ehex: $(PRG)_eeprom.hex
ebin: $(PRG)_eeprom.bin
esrec: $(PRG)_eeprom.srec
%_eeprom.hex: %.elf
$(OBJCOPY) -j .eeprom --change-section-lma .eeprom=0 -O ihex $ & lt; $@
%_eeprom.srec: %.elf
$(OBJCOPY) -j .eeprom --change-section-lma .eeprom=0 -O srec $ & lt; $@
%_eeprom.bin: %.elf
$(OBJCOPY) -j .eeprom --change-section-lma .eeprom=0 -O binary $ & lt; $@
# Every thing below here is used by avr-libc's build system and can be ignored
# by the casual user.
FIG2DEV
EXTRA_CLEAN_FILES

= fig2dev
= *.hex *.bin *.srec

dox: eps png pdf
eps: $(PRG).eps
png: $(PRG).png
pdf: $(PRG).pdf
%.eps: %.fig
$(FIG2DEV) -L eps $ & lt; $@
%.pdf: %.fig
$(FIG2DEV) -L pdf $ & lt; $@
%.png: %.fig
$(FIG2DEV) -L png $ & lt; $@

7.12

Example using the two-wire interface (TWI)

Some newer devices of the ATmega series contain builtin support for interfacing the microcontroller to a two-wire bus, called TWI. This is essentially the same called I178C
by Philips, but that term is avoided in Atmel's documentation due to patenting issues.

For the original Philips documentation, see http://www.semiconductors.philips.com/buses/i2c/inde
7.12.1

Introduction into TWI

The two-wire interface consists of two signal lines named SDA (serial data) and SCL
(serial clock) (plus a ground line, of course). All devices participating in the bus are
connected together, using open-drain driver circuitry, so the wires must be terminated

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

7.12 Example using the two-wire interface (TWI)

using appropriate pullup resistors. The pullups must be small enough to recharge
the line capacity in short enough time compared to the desired maximal clock frequency, yet large enough so all drivers will not be overloaded. There are formulas in
the datasheet that help selecting the pullups.
Devices can either act as a master to the bus (i. e., they initiate a transfer), or as a slave
(they only act when being called by a master). The bus is multi-master capable, and
a particular device implementation can act as either master or slave at different times.
Devices are addressed using a 7-bit address (coordinated by Philips) transfered as the
first byte after the so-called start condition. The LSB of that byte is R/172W, i. e.
it determines whether the request to the slave is to read or write data during the next
cycles. (There is also an option to have devices using 10-bit addresses but that is not
covered by this example.)
7.12.2

The TWI example project

The ATmega TWI hardware supports both, master and slave operation. This example
will only demonstrate how to use an AVR microcontroller as TWI master. The implementation is kept simple in order to concentrate on the steps that are required to talk to
a TWI slave, so all processing is done in polled-mode, waiting for the TWI interface to
indicate that the next processing step is due (by setting the TWINT interrupt bit). If it
is desired to have the entire TWI communication happen in "background", all this can
be implemented in an interrupt-controlled way, where only the start condition needs to
be triggered from outside the interrupt routine.
There is a variety of slave devices available that can be connected to a TWI bus. For the
purpose of this example, an EEPROM device out of the industry-standard 24Cxx series
has been chosen (where xx can be one of 01, 02, 04, 08, or 16) which are available from
various vendors. The choice was almost arbitrary, mainly triggered by the fact that an
EEPROM device is being talked to in both directions, reading and writing the slave
device, so the example will demonstrate the details of both.
Usually, there is probably not much need to add more EEPROM to an ATmega system
that way: the smallest possible AVR device that offers hardware TWI support is the
ATmega8 which comes with 512 bytes of EEPROM, which is equivalent to an 24C04
device. The ATmega128 already comes with twice as much EEPROM as the 24C16
would offer. One exception might be to use an externally connected EEPROM device
that is removable; e. g. SDRAM PC memory comes with an integrated TWI EEPROM
that carries the RAM configuration information.
7.12.3
/*
*
*
*
*

The Source Code

--------------------------------------------------------------------------- " THE BEER-WARE LICENSE " (Revision 42):
& lt; joerg@FreeBSD.ORG & gt; wrote this file. As long as you retain this notice you
can do whatever you want with this stuff. If we meet some day, and you think

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

155

7.12 Example using the two-wire interface (TWI)

* this stuff is worth it, you can buy me a beer in return.
Joerg Wunsch
* ---------------------------------------------------------------------------*/
/* $Id: twitest.c,v 1.1 2002/12/18 22:35:38 joerg_wunsch Exp $ */
/*
* Simple demo program that talks to a 24Cxx IC EEPROM using the
* builtin TWI interface of an ATmega device.
*/
#include & lt; inttypes.h & gt;
#include & lt; stdio.h & gt;
#include & lt; stdlib.h & gt;
#include & lt; avr/io.h & gt;
#include & lt; avr/twi.h & gt;

/* Note [1] */

#define DEBUG 1
/*
* System clock in Hz.
*/
#define SYSCLK 14745600UL

/* Note [2] */

/*
* Compatibility defines. This should work on ATmega8, ATmega16,
* ATmega163, ATmega323 and ATmega128 (IOW: on all devices that
* provide a builtin TWI interface).
*
* On the 128, it defaults to USART 1.
*/
#ifndef UCSRB
# ifdef UCSR1A
/* ATmega128 */
# define UCSRA UCSR1A
# define UCSRB UCSR1B
# define UBRR UBRR1L
# define UDR UDR1
# else /* ATmega8 */
# define UCSRA USR
# define UCSRB UCR
# endif
#endif
#ifndef UBRR
# define UBRR UBRRL
#endif
/*
*
*
*
*
*
*
*

Note [3]
TWI address for 24Cxx EEPROM:
1
1
1
1

0
0
0
0

1
1
1
1

0
0
0
0

E2 E1 E0 R/~W
E2 E1 A8 R/~W
E2 A9 A8 R/~W
A10 A9 A8 R/~W

24C01/24C02
24C04
24C08
24C16

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

156

7.12 Example using the two-wire interface (TWI)

*/
#define TWI_SLA_24CXX

0xa0

/* E2 E1 E0 = 0 0 0 */

/*
* Maximal number of iterations to wait for a device to respond for a
* selection. Should be large enough to allow for a pending write to
* complete, but low enough to properly abort an infinite loop in case
* a slave is broken or not present at all. With 100 kHz TWI clock,
* transfering the start condition and SLA+R/W packet takes about 10
* s. The longest write period is supposed to not exceed ~ 10 ms.
* Thus, normal operation should not require more than 100 iterations
* to get the device to respond to a selection.
*/
#define MAX_ITER
200
/*
* Number of bytes that can be written in a row, see comments for
* ee24xx_write_page() below. Some vendor's devices would accept 16,
* but 8 seems to be the lowest common denominator.
*
* Note that the page size must be a power of two, this simplifies the
* page boundary calculations below.
*/
#define PAGE_SIZE 8
/*
* Saved TWI status register, for error messages only. We need to
* save it in a variable, since the datasheet only guarantees the TWSR
* register to have valid contents while the TWINT bit in TWCR is set.
*/
uint8_t twst;
/*
* Do all the startup-time peripheral initializations: UART (for our
* debug/test output), and TWI clock.
*/
void
ioinit(void)
{
UCSRB = _BV(TXEN);
/* tx enable */
UBRR = (SYSCLK / (16 * 9600UL)) - 1; /* 9600 Bd */
/* initialize TWI clock: 100 kHz clock, TWPS = 0 = & gt; prescaler = 1 */
#if defined(TWPS0)
/* has prescaler (mega128 & newer) */
TWSR = 0;
#endif
TWBR = (SYSCLK / 100000UL - 16) / 2;
}
/*
* Note [4]
* Send character c down the UART Tx, wait until tx holding register
* is empty.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

157

7.12 Example using the two-wire interface (TWI)

*/
int
uart_putchar(char c)
{
if (c == '\n')
uart_putchar('\r');
loop_until_bit_is_set(UCSRA, UDRE);
UDR = c;
return 0;
}
/*
* Note [5]
*
* Read " len " bytes from EEPROM starting at " eeaddr " into " buf " .
*
* This requires two bus cycles: during the first cycle, the device
* will be selected (master transmitter mode), and the address
* transfered. Address bits exceeding 256 are transfered in the
* E2/E1/E0 bits (subaddress bits) of the device selector.
*
* The second bus cycle will reselect the device (repeated start
* condition, going into master receiver mode), and transfer the data
* from the device to the TWI master. Multiple bytes can be
* transfered by ACKing the client's transfer. The last transfer will
* be NACKed, which the client will take as an indication to not
* initiate further transfers.
*/
int
ee24xx_read_bytes(uint16_t eeaddr, int len, uint8_t *buf)
{
uint8_t sla, twcr, n = 0;
int rv = 0;
/* patch high bits of EEPROM address into SLA */
sla = TWI_SLA_24CXX | (((eeaddr & gt; & gt; 8) & 0x07) & lt; & lt; 1);
/*
* Note [6]
* First cycle: master transmitter mode
*/
restart:
if (n++ & gt; = MAX_ITER)
return -1;
begin:
TWCR = _BV(TWINT) | _BV(TWSTA) | _BV(TWEN); /* send start condition */
while ((TWCR & _BV(TWINT)) == 0) ; /* wait for transmission */
switch ((twst = TW_STATUS))
{
case TW_REP_START:
/* OK, but should not happen */
case TW_START:
break;

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

158

7.12 Example using the two-wire interface (TWI)

case TW_MT_ARB_LOST:
goto begin;
default:
return -1;

/* Note [7] */

/* error: not in start condition */
/* NB: do /not/ send stop condition */

}
/* Note [8] */
/* send SLA+W */
TWDR = sla | TW_WRITE;
TWCR = _BV(TWINT) | _BV(TWEN); /* clear interrupt to start transmission */
while ((TWCR & _BV(TWINT)) == 0) ; /* wait for transmission */
switch ((twst = TW_STATUS))
{
case TW_MT_SLA_ACK:
break;
case TW_MT_SLA_NACK:

/* nack during select: device busy writing */
/* Note [9] */

goto restart;
case TW_MT_ARB_LOST:
goto begin;
default:
goto error;
}

/* re-arbitrate */

/* must send stop condition */

TWDR = eeaddr;
/* low 8 bits of addr */
TWCR = _BV(TWINT) | _BV(TWEN); /* clear interrupt to start transmission */
while ((TWCR & _BV(TWINT)) == 0) ; /* wait for transmission */
switch ((twst = TW_STATUS))
{
case TW_MT_DATA_ACK:
break;
case TW_MT_DATA_NACK:
goto quit;
case TW_MT_ARB_LOST:
goto begin;
default:
goto error;
}

/* must send stop condition */

/*
* Note [10]
* Next cycle(s): master receiver mode
*/
TWCR = _BV(TWINT) | _BV(TWSTA) | _BV(TWEN); /* send (rep.) start condition */
while ((TWCR & _BV(TWINT)) == 0) ; /* wait for transmission */
switch ((twst = TW_STATUS))
{

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

159

7.12 Example using the two-wire interface (TWI)

case TW_START:
case TW_REP_START:
break;

/* OK, but should not happen */

case TW_MT_ARB_LOST:
goto begin;
default:
goto error;
}
/* send SLA+R */
TWDR = sla | TW_READ;
TWCR = _BV(TWINT) | _BV(TWEN); /* clear interrupt to start transmission */
while ((TWCR & _BV(TWINT)) == 0) ; /* wait for transmission */
switch ((twst = TW_STATUS))
{
case TW_MR_SLA_ACK:
break;
case TW_MR_SLA_NACK:
goto quit;
case TW_MR_ARB_LOST:
goto begin;
default:
goto error;
}
for (twcr = _BV(TWINT) | _BV(TWEN) | _BV(TWEA) /* Note [11] */;
len & gt; 0;
len--)
{
if (len == 1)
twcr = _BV(TWINT) | _BV(TWEN); /* send NAK this time */
TWCR = twcr;
/* clear int to start transmission */
while ((TWCR & _BV(TWINT)) == 0) ; /* wait for transmission */
switch ((twst = TW_STATUS))
{
case TW_MR_DATA_NACK:
len = 0;
/* force end of loop */
/* FALLTHROUGH */
case TW_MR_DATA_ACK:
*buf++ = TWDR;
rv++;
break;
default:
goto error;
}
}
quit:
/* Note [12] */
TWCR = _BV(TWINT) | _BV(TWSTO) | _BV(TWEN); /* send stop condition */

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

160

7.12 Example using the two-wire interface (TWI)

return rv;
error:
rv = -1;
goto quit;
}
/*
* Write " len " bytes into EEPROM starting at " eeaddr " from " buf " .
*
* This is a bit simpler than the previous function since both, the
* address and the data bytes will be transfered in master transmitter
* mode, thus no reselection of the device is necessary. However, the
* EEPROMs are only capable of writing one " page " simultaneously, so
* care must be taken to not cross a page boundary within one write
* cycle. The amount of data one page consists of varies from
* manufacturer to manufacturer: some vendors only use 8-byte pages
* for the smaller devices, and 16-byte pages for the larger devices,
* while other vendors generally use 16-byte pages. We thus use the
* smallest common denominator of 8 bytes per page, declared by the
* macro PAGE_SIZE above.
*
* The function simply returns after writing one page, returning the
* actual number of data byte written. It is up to the caller to
* re-invoke it in order to write further data.
*/
int
ee24xx_write_page(uint16_t eeaddr, int len, uint8_t *buf)
{
uint8_t sla, n = 0;
int rv = 0;
uint16_t endaddr;
if (eeaddr + len & lt; (eeaddr | (PAGE_SIZE - 1)))
endaddr = eeaddr + len;
else
endaddr = (eeaddr | (PAGE_SIZE - 1)) + 1;
len = endaddr - eeaddr;
/* patch high bits of EEPROM address into SLA */
sla = TWI_SLA_24CXX | (((eeaddr & gt; & gt; 8) & 0x07) & lt; & lt; 1);
restart:
if (n++ & gt; = MAX_ITER)
return -1;
begin:
/* Note 13 */
TWCR = _BV(TWINT) | _BV(TWSTA) | _BV(TWEN); /* send start condition */
while ((TWCR & _BV(TWINT)) == 0) ; /* wait for transmission */
switch ((twst = TW_STATUS))
{
case TW_REP_START:
/* OK, but should not happen */
case TW_START:

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

161

7.12 Example using the two-wire interface (TWI)

break;
case TW_MT_ARB_LOST:
goto begin;
default:
return -1;

/* error: not in start condition */
/* NB: do /not/ send stop condition */

}
/* send SLA+W */
TWDR = sla | TW_WRITE;
TWCR = _BV(TWINT) | _BV(TWEN); /* clear interrupt to start transmission */
while ((TWCR & _BV(TWINT)) == 0) ; /* wait for transmission */
switch ((twst = TW_STATUS))
{
case TW_MT_SLA_ACK:
break;
case TW_MT_SLA_NACK:
goto restart;

/* nack during select: device busy writing */

case TW_MT_ARB_LOST:
goto begin;

/* re-arbitrate */

default:
goto error;
}

/* must send stop condition */

TWDR = eeaddr;
/* low 8 bits of addr */
TWCR = _BV(TWINT) | _BV(TWEN); /* clear interrupt to start transmission */
while ((TWCR & _BV(TWINT)) == 0) ; /* wait for transmission */
switch ((twst = TW_STATUS))
{
case TW_MT_DATA_ACK:
break;
case TW_MT_DATA_NACK:
goto quit;
case TW_MT_ARB_LOST:
goto begin;
default:
goto error;
}

/* must send stop condition */

for (; len & gt; 0; len--)
{
TWDR = *buf++;
TWCR = _BV(TWINT) | _BV(TWEN); /* start transmission */
while ((TWCR & _BV(TWINT)) == 0) ; /* wait for transmission */
switch ((twst = TW_STATUS))
{
case TW_MT_DATA_NACK:

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

162

7.12 Example using the two-wire interface (TWI)

goto error;

/* device write protected -- Note [14] */

case TW_MT_DATA_ACK:
rv++;
break;
default:
goto error;
}
}
quit:
TWCR = _BV(TWINT) | _BV(TWSTO) | _BV(TWEN); /* send stop condition */
return rv;
error:
rv = -1;
goto quit;
}
/*
* Wrapper around ee24xx_write_page() that repeats calling this
* function until either an error has been returned, or all bytes
* have been written.
*/
int
ee24xx_write_bytes(uint16_t eeaddr, int len, uint8_t *buf)
{
int rv, total;
total = 0;
do
{
#if DEBUG
printf( " Calling ee24xx_write_page(%d, %d, %p) " ,
eeaddr, len, buf);
#endif
rv = ee24xx_write_page(eeaddr, len, buf);
#if DEBUG
printf( " = & gt; %d\n " , rv);
#endif
if (rv == -1)
return -1;
eeaddr += rv;
len -= rv;
buf += rv;
total += rv;
}
while (len & gt; 0);
return total;
}
void
error(void)

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

163

7.12 Example using the two-wire interface (TWI)

{
printf( " error: TWI status %#x\n " , twst);
exit(0);
}
void
main(void)
{
uint16_t a;
int rv;
uint8_t b[16];
uint8_t x;
ioinit();
fdevopen(uart_putchar, NULL, 0);
for (a = 0; a & lt; 256;)
{
printf( " %#04x: " , a);
rv = ee24xx_read_bytes(a, 16, b);
if (rv & lt; = 0)
error();
if (rv & lt; 16)
printf( " warning: short read %d\n " , rv);
a += rv;
for (x = 0; x & lt; rv; x++)
printf( " %02x " , b[x]);
putchar('\n');
}
#define EE_WRITE(addr, str) ee24xx_write_bytes(addr, sizeof(str)-1, str)
rv = EE_WRITE(55, " The quick brown fox jumps over the lazy dog. " );
if (rv & lt; 0)
error();
printf( " Wrote %d bytes.\n " , rv);
for (a = 0; a & lt; 256;)
{
printf( " %#04x: " , a);
rv = ee24xx_read_bytes(a, 16, b);
if (rv & lt; = 0)
error();
if (rv & lt; 16)
printf( " warning: short read %d\n " , rv);
a += rv;
for (x = 0; x & lt; rv; x++)
printf( " %02x " , b[x]);
putchar('\n');
}
printf( " done.\n " );
}

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

164

7.12 Example using the two-wire interface (TWI)

Note [1]
The header file & lt; avr/io.h & gt; contains some macro definitions for symbolic constants used in the TWI status register. These definitions match the names used in
the Atmel datasheet except that all names have been prefixed with TW .
Note [2]
The clock is used in timer calculations done by the compiler, for the UART baud
rate and the TWI clock rate.
Note [3]
The address assigned for the 24Cxx EEPROM consists of 1010 in the upper four
bits. The following three bits are normally available as slave sub-addresses, allowing to operate more than one device of the same type on a single bus, where the
actual subaddress used for each device is configured by hardware strapping. However, since the next data packet following the device selection only allows for 8 bits
that are used as an EEPROM address, devices that require more than 8 address bits
(24C04 and above) "steal" subaddress bits and use them for the EEPROM cell address bits 9 to 11 as required. This example simply assumes all subaddress bits
are 0 for the smaller devices, so the E0, E1, and E2 inputs of the 24Cxx must be
grounded.
Note [4]
This function is used by the standard output facilities that are utilized in this example for debugging and demonstration purposes.
Note [5]
In order to shorten the data to be sent over the TWI bus, the 24Cxx EEPROMs
support multiple data bytes transfered within a single request, maintaining an internal address counter that is updated after each data byte transfered successfully.
When reading data, one request can read the entire device memory if desired (the
counter would wrap around and start back from 0 when reaching the end of the
device).
Note [6]
When reading the EEPROM, a first device selection must be made with write intent (R/172W bit set to 0 indicating a write operation) in order to transfer the
EEPROM address to start reading from. This is called master transmitter mode.
Each completion of a particular step in TWI communication is indicated by an
asserted TWINT bit in TWCR. (An interrupt would be generated if allowed.) After performing any actions that are needed for the next communication step, the
interrupt condition must be manually cleared by setting the TWINT bit. Unlike
with many other interrupt sources, this would even be required when using a true
interrupt routine, since as soon as TWINT is re-asserted, the next bus transaction
will start.
Note [7]
Since the TWI bus is multi-master capable, there is potential for a bus contention
when one master starts to access the bus. Normally, the TWI bus interface unit will
Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

165

7.12 Example using the two-wire interface (TWI)

detect this situation, and will not initiate a start condition while the bus is busy.
However, in case two masters were starting at exactly the same time, the way bus
arbitration works, there is always a chance that one master could lose arbitration
of the bus during any transmit operation. A master that has lost arbitration is
required by the protocol to immediately cease talking on the bus; in particular it
must not initiate a stop condition in order to not corrupt the ongoing transfer from
the active master. In this example, upon detecting a lost arbitration condition, the
entire transfer is going to be restarted. This will cause a new start condition to
be initiated, which will normally be delayed until the currently active master has
released the bus.
Note [8]
Next, the device slave is going to be reselected (using a so-called repeated start
condition which is meant to guarantee that the bus arbitration will remain at the
current master) using the same slave address (SLA), but this time with read intent
(R/172W bit set to 1) in order to request the device slave to start transfering data
from the slave to the master in the next packet.
Note [9]
If the EEPROM device is still busy writing one or more cells after a previous write
request, it will simply leave its bus interface drivers at high impedance, and does
not respond to a selection in any way at all. The master selecting the device will
see the high level at SDA after transfering the SLA+R/W packet as a NACK to
its selection request. Thus, the select process is simply started over (effectively
causing a repeated start condition), until the device will eventually respond. This
polling procedure is recommended in the 24Cxx datasheet in order to minimize
the busy wait time when writing. Note that in case a device is broken and never
responds to a selection (e. g. since it is no longer present at all), this will cause
an infinite loop. Thus the maximal number of iterations made until the device is
declared to be not responding at all, and an error is returned, will be limited to
MAX ITER.
Note [10]
This is called master receiver mode: the bus master still supplies the SCL clock,
but the device slave drives the SDA line with the appropriate data. After 8 data
bits, the master responds with an ACK bit (SDA driven low) in order to request
another data transfer from the slave, or it can leave the SDA line high (NACK),
indicating to the slave that it is going to stop the transfer now. Assertion of ACK
is handled by setting the TWEA bit in TWCR when starting the current transfer.
Note [11]
The control word sent out in order to initiate the transfer of the next data packet
is initially set up to assert the TWEA bit. During the last loop iteration, TWEA is
de-asserted so the client will get informed that no further transfer is desired.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

166

7.13 Todo List

Note [12]
Except in the case of lost arbitration, all bus transactions must properly be terminated by the master initiating a stop condition.
Note [13]
Writing to the EEPROM device is simpler than reading, since only a master transmitter mode transfer is needed. Note that the first packet after the SLA+W selection is always considered to be the EEPROM address for the next operation.
(This packet is exactly the same as the one above sent before starting to read the
device.) In case a master transmitter mode transfer is going to send more than one
data packet, all following packets will be considered data bytes to write at the indicated address. The internal address pointer will be incremented after each write
operation.
Note [14]
24Cxx devices can become write-protected by strapping their 172WC pin to logic
high. (Leaving it unconnected is explicitly allowed, and constitutes logic low level,
i. e. no write protection.) In case of a write protected device, all data transfer
attempts will be NACKed by the device. Note that some devices might not implement this.

7.13

Todo List

Group avr boot From email with Marek: On smaller devices (all except ATmega64/128), SPM REG is in the I/O space, accessible with the shorter "in"
and "out" instructions - since the boot loader has a limited size, this could be an
important optimization.

Group avr inttypes There is a pending patch that may go into gcc to change the
behaviour of the -mint8 option. The current (2003-09-17) situation for -mint8 is
sizeof(int) == 1, sizeof(long) == 2 and sizeof(long long) == 8. Note the absence
of a 4-byte, 32-bit type. The patch proposes to change sozeof(long long) to be 4
bytes (32 bits). When and if the patch is included in gcc, we will need to change
avr-libc accordingly.

7.14

Deprecated List

Global eeprom rb(addr) Use eeprom read byte() in new programs.

Global eeprom rw(addr) Use eeprom read word() in new programs.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

167

7.14 Deprecated List

Global eeprom wb(addr, val) Use eeprom write byte() in new programs.

Global PRG RDB(addr) Use pgm read byte() in new programs.

Global cbi(sfr, bit)

#include & lt; avr/io.h & gt;

For backwards compatibility only. This macro will eventually be removed.

Global sbi(sfr, bit)

#include & lt; avr/io.h & gt;

For backwards compatibility only. This macro will eventually be removed.

Global inb(sfr)

#include & lt; avr/io.h & gt;

For backwards compatibility only. This macro will eventually be removed.

Global outb(sfr, val)

Global inw(sfr)

#include & lt; avr/io.h & gt;

#include & lt; avr/io.h & gt;

For backwards compatibility only. This macro will eventually be removed.

Global outw(sfr, val)

#include & lt; avr/io.h & gt;

For backwards compatibility only. This macro will eventually be removed.

Global outp(val, sfr) For backwards compatibility only. This macro will eventually
be removed.

Global inp(sfr) For backwards compatibility only. This macro will eventually be
removed.

Global BV(bit) For backwards compatibility only. This macro will eventually be
removed.

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

168

Index
$PATH, 126
$PREFIX, 126
-prefix, 126
BV
avr sfr, 75
EEGET
avr eeprom, 11
EEPUT
avr eeprom, 11
compar fn t
avr stdlib, 52
malloc heap end
avr stdlib, 60
malloc heap start
avr stdlib, 60
malloc margin
avr stdlib, 60
crc16 update
avr crc, 8
crc ccitt update
avr crc, 8
crc xmodem update
avr crc, 9
abort
avr stdlib, 52
abs
avr stdlib, 52
acos
avr math, 30
Additional notes from & lt; avr/sfr defs.h & gt; , 19
asin
avr math, 30
atan
avr math, 31
atan2
avr math, 31
atoi
avr stdlib, 52
atol
avr stdlib, 53
AVR device-specific IO definitions, 12

avr boot
boot is spm interrupt, 6
boot lock bits set, 6
boot page erase, 6
boot page fill, 6
boot page write, 7
boot rww busy, 7
boot rww enable, 7
boot spm busy, 7
boot spm busy wait, 7
boot spm interrupt disable, 7
boot spm interrupt enable, 7
BOOTLOADER SECTION, 7
avr crc
crc16 update, 8
crc ccitt update, 8
crc xmodem update, 9
avr eeprom
EEGET, 11
EEPUT, 11
eeprom is ready, 11
eeprom rb, 11
eeprom read block, 11
eeprom read byte, 11
eeprom read word, 12
eeprom rw, 11
eeprom wb, 11
eeprom write block, 12
eeprom write byte, 12
eeprom write word, 12
avr errno
EDOM, 27
ERANGE, 27
avr interrupts
cli, 72
EMPTY INTERRUPT, 72
INTERRUPT, 72
sei, 72
SIGNAL, 72
timer enable int, 73
avr inttypes
int16 t, 28

INDEX

int32 t, 28
int64 t, 28
int8 t, 28
intptr t, 28
uint16 t, 28
uint32 t, 29
uint64 t, 29
uint8 t, 29
uintptr t, 29
avr math
acos, 30
asin, 30
atan, 31
atan2, 31
ceil, 31
cos, 31
cosh, 31
exp, 31
fabs, 31
floor, 31
fmod, 31
frexp, 32
inverse, 32
isinf, 32
isnan, 32
ldexp, 32
log, 32
log10, 32
M PI, 30
M SQRT2, 30
modf, 33
pow, 33
sin, 33
sinh, 33
sqrt, 33
square, 33
tan, 33
tanh, 33
avr pgmspace
memcpy P, 16
PGM P, 14
pgm read byte, 14
pgm read byte far, 14
pgm read byte near, 15
pgm read word, 15
pgm read word far, 15

170

pgm read word near, 15
PGM VOID P, 15
PRG RDB, 15
PSTR, 16
strcasecmp P, 16
strcat P, 16
strcmp P, 16
strcpy P, 17
strlcat P, 17
strlcpy P, 17
strlen P, 17
strncasecmp P, 18
strncat P, 18
strncmp P, 18
strncpy P, 18
avr sfr
BV, 75
bit is clear, 75
bit is set, 75
BV, 75
cbi, 76
inb, 76
inp, 76
inw, 76
loop until bit is clear, 76
loop until bit is set, 77
outb, 77
sbi, 78
avr sleep
set sleep mode, 21
sleep mode, 21
SLEEP MODE ADC, 21
SLEEP MODE EXT STANDBY, 21
SLEEP MODE IDLE, 21
SLEEP MODE PWR DOWN,
21
SLEEP MODE PWR SAVE, 21
SLEEP MODE STANDBY, 21
avr stdio
clearerr, 40
EOF, 39
fclose, 40
fdevopen, 40
feof, 41
ferror, 41

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

INDEX

fgetc, 41
fgets, 41
FILE, 39
fprintf, 41
fprintf P, 42
fputc, 42
fputs, 42
fputs P, 42
fread, 42
fscanf, 42
fscanf P, 42
fwrite, 42
getc, 39
getchar, 39
gets, 43
printf, 43
printf P, 43
putc, 39
putchar, 39
puts, 43
puts P, 43
scanf, 43
scanf P, 43
snprintf, 43
snprintf P, 43
sprintf, 44
sprintf P, 44
sscanf, 44
sscanf P, 44
stderr, 39
stdin, 40
stdout, 40
ungetc, 44
vfprintf, 44
vfprintf P, 47
vfscanf, 47
vfscanf P, 49
vsnprintf, 49
vsnprintf P, 49
vsprintf, 49
vsprintf P, 49
avr stdlib
compar fn t, 52
malloc heap end, 60
malloc heap start, 60
malloc margin, 60

171

abort, 52
abs, 52
atoi, 52
atol, 53
bsearch, 53
calloc, 53
div, 53
DTOSTR ALWAYS SIGN, 51
DTOSTR PLUS SIGN, 51
DTOSTR UPPERCASE, 52
dtostre, 53
dtostrf, 54
exit, 54
free, 54
itoa, 54
labs, 55
ldiv, 55
ltoa, 55
malloc, 56
qsort, 56
rand, 56
RAND MAX, 52
rand r, 56
random, 57
RANDOM MAX, 52
random r, 57
srand, 57
srandom, 57
strtod, 57
strtol, 58
strtoul, 58
ultoa, 59
utoa, 59
avr string
memccpy, 61
memchr, 61
memcmp, 61
memcpy, 62
memmove, 62
memset, 62
strcasecmp, 63
strcat, 63
strchr, 63
strcmp, 63
strcpy, 64
strlcat, 64

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

INDEX

strlcpy, 64
strlen, 64
strlwr, 65
strncasecmp, 65
strncat, 65
strncmp, 65
strncpy, 66
strnlen, 66
strrchr, 66
strrev, 67
strsep, 67
strstr, 67
strtok r, 67
strupr, 68
avr watchdog
wdt disable, 22
wdt enable, 22
wdt reset, 23
WDTO 120MS, 23
WDTO 15MS, 23
WDTO 1S, 23
WDTO 250MS, 23
WDTO 2S, 23
WDTO 30MS, 23
WDTO 500MS, 23
WDTO 60MS, 24
avrdude, usage, 133
avrprog, usage, 133
bit is clear
avr sfr, 75
bit is set
avr sfr, 75
boot is spm interrupt
avr boot, 6
boot lock bits set
avr boot, 6
boot page erase
avr boot, 6
boot page fill
avr boot, 6
boot page write
avr boot, 7
boot rww busy
avr boot, 7
boot rww enable

172

avr boot, 7
boot spm busy
avr boot, 7
boot spm busy wait
avr boot, 7
boot spm interrupt disable
avr boot, 7
boot spm interrupt enable
avr boot, 7
Bootloader Support Utilities, 4
BOOTLOADER SECTION
avr boot, 7
bsearch
avr stdlib, 53
BV
avr sfr, 75
calloc
avr stdlib, 53
cbi
avr sfr, 76
ceil
avr math, 31
Character Operations, 24
clearerr
avr stdio, 40
cli
avr interrupts, 72
cos
avr math, 31
cosh
avr math, 31
CRC Computations, 8
ctype
isalnum, 25
isalpha, 25
isascii, 25
isblank, 25
iscntrl, 25
isdigit, 25
isgraph, 25
islower, 25
isprint, 25
ispunct, 25
isspace, 25
isupper, 26

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

INDEX

isxdigit, 26
toascii, 26
tolower, 26
toupper, 26
disassembling, 146
div
avr stdlib, 53
div t, 78
DTOSTR ALWAYS SIGN
avr stdlib, 51
DTOSTR PLUS SIGN
avr stdlib, 51
DTOSTR UPPERCASE
avr stdlib, 52
dtostre
avr stdlib, 53
dtostrf
avr stdlib, 54
EDOM
avr errno, 27
EEPROM handling, 10
eeprom is ready
avr eeprom, 11
eeprom rb
avr eeprom, 11
eeprom read block
avr eeprom, 11
eeprom read byte
avr eeprom, 11
eeprom read word
avr eeprom, 12
eeprom rw
avr eeprom, 11
eeprom wb
avr eeprom, 11
eeprom write block
avr eeprom, 12
eeprom write byte
avr eeprom, 12
eeprom write word
avr eeprom, 12
EMPTY INTERRUPT
avr interrupts, 72
EOF

173

avr stdio, 39
ERANGE
avr errno, 27
exit
avr stdlib, 54
exp
avr math, 31
fabs
avr math, 31
FAQ, 84
fclose
avr stdio, 40
fdevopen
avr stdio, 40
feof
avr stdio, 41
ferror
avr stdio, 41
fgetc
avr stdio, 41
fgets
avr stdio, 41
FILE
avr stdio, 39
floor
avr math, 31
fmod
avr math, 31
fprintf
avr stdio, 41
fprintf P
avr stdio, 42
fputc
avr stdio, 42
fputs
avr stdio, 42
fputs P
avr stdio, 42
fread
avr stdio, 42
free
avr stdlib, 54
frexp
avr math, 32
fscanf

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

INDEX

avr stdio, 42
fscanf P
avr stdio, 42
fwrite
avr stdio, 42
General utilities, 50
getc
avr stdio, 39
getchar
avr stdio, 39
gets
avr stdio, 43
inb
avr sfr, 76
inp
avr sfr, 76
installation, 126
installation, avarice, 132
installation, avr-libc, 130
installation, avrdude, 131
installation, avrprog, 131
installation, binutils, 128
installation, gcc, 129
Installation, gdb, 131
installation, simulavr, 132
installation, uisp, 131
int16 t
avr inttypes, 28
int32 t
avr inttypes, 28
int64 t
avr inttypes, 28
int8 t
avr inttypes, 28
Integer Types, 27
INTERRUPT
avr interrupts, 72
Interrupts and Signals, 68
intptr t
avr inttypes, 28
inverse
avr math, 32
inw
avr sfr, 76

174

isalnum
ctype, 25
isalpha
ctype, 25
isascii
ctype, 25
isblank
ctype, 25
iscntrl
ctype, 25
isdigit
ctype, 25
isgraph
ctype, 25
isinf
avr math, 32
islower
ctype, 25
isnan
avr math, 32
isprint
ctype, 25
ispunct
ctype, 25
isspace
ctype, 25
isupper
ctype, 26
isxdigit
ctype, 26
itoa
avr stdlib, 54
labs
avr stdlib, 55
ldexp
avr math, 32
ldiv
avr stdlib, 55
ldiv t, 78
log
avr math, 32
log10
avr math, 32
longjmp
setjmp, 35

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

INDEX

loop until bit is clear
avr sfr, 76
loop until bit is set
avr sfr, 77
ltoa
avr stdlib, 55
M PI
avr math, 30
M SQRT2
avr math, 30
malloc
avr stdlib, 56
Mathematics, 29
memccpy
avr string, 61
memchr
avr string, 61
memcmp
avr string, 61
memcpy
avr string, 62
memcpy P
avr pgmspace, 16
memmove
avr string, 62
memset
avr string, 62
modf
avr math, 33
outb
avr sfr, 77
PGM P
avr pgmspace, 14
pgm read byte
avr pgmspace, 14
pgm read byte far
avr pgmspace, 14
pgm read byte near
avr pgmspace, 15
pgm read word
avr pgmspace, 15
pgm read word far
avr pgmspace, 15

175

pgm read word near
avr pgmspace, 15
PGM VOID P
avr pgmspace, 15
pow
avr math, 33
Power Management and Sleep Modes,
20
PRG RDB
avr pgmspace, 15
printf
avr stdio, 43
printf P
avr stdio, 43
Program Space String Utilities, 13
PSTR
avr pgmspace, 16
putc
avr stdio, 39
putchar
avr stdio, 39
puts
avr stdio, 43
puts P
avr stdio, 43
qsort
avr stdlib, 56
rand
avr stdlib, 56
RAND MAX
avr stdlib, 52
rand r
avr stdlib, 56
random
avr stdlib, 57
RANDOM MAX
avr stdlib, 52
random r
avr stdlib, 57
sbi
avr sfr, 78
scanf
avr stdio, 43

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

INDEX

scanf P
avr stdio, 43
sei
avr interrupts, 72
set sleep mode
avr sleep, 21
setjmp
longjmp, 35
setjmp, 35
Setjmp and Longjmp, 34
SIGNAL
avr interrupts, 72
sin
avr math, 33
sinh
avr math, 33
sleep mode
avr sleep, 21
SLEEP MODE ADC
avr sleep, 21
SLEEP MODE EXT STANDBY
avr sleep, 21
SLEEP MODE IDLE
avr sleep, 21
SLEEP MODE PWR DOWN
avr sleep, 21
SLEEP MODE PWR SAVE
avr sleep, 21
SLEEP MODE STANDBY
avr sleep, 21
snprintf
avr stdio, 43
snprintf P
avr stdio, 43
Special function registers, 73
sprintf
avr stdio, 44
sprintf P
avr stdio, 44
sqrt
avr math, 33
square
avr math, 33
srand
avr stdlib, 57
srandom

176

avr stdlib, 57
sscanf
avr stdio, 44
sscanf P
avr stdio, 44
Standard IO facilities, 36
stderr
avr stdio, 39
stdin
avr stdio, 40
stdout
avr stdio, 40
strcasecmp
avr string, 63
strcasecmp P
avr pgmspace, 16
strcat
avr string, 63
strcat P
avr pgmspace, 16
strchr
avr string, 63
strcmp
avr string, 63
strcmp P
avr pgmspace, 16
strcpy
avr string, 64
strcpy P
avr pgmspace, 17
Strings, 60
strlcat
avr string, 64
strlcat P
avr pgmspace, 17
strlcpy
avr string, 64
strlcpy P
avr pgmspace, 17
strlen
avr string, 64
strlen P
avr pgmspace, 17
strlwr
avr string, 65
strncasecmp

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

INDEX

avr string, 65
strncasecmp P
avr pgmspace, 18
strncat
avr string, 65
strncat P
avr pgmspace, 18
strncmp
avr string, 65
strncmp P
avr pgmspace, 18
strncpy
avr string, 66
strncpy P
avr pgmspace, 18
strnlen
avr string, 66
strrchr
avr string, 66
strrev
avr string, 67
strsep
avr string, 67
strstr
avr string, 67
strtod
avr stdlib, 57
strtok r
avr string, 67
strtol
avr stdlib, 58
strtoul
avr stdlib, 58
strupr
avr string, 68
supported devices, 1
System Errors (errno), 26
tan
avr math, 33
tanh
avr math, 33
timer enable int
avr interrupts, 73
toascii
ctype, 26

177

tolower
ctype, 26
tools, optional, 127
tools, required, 127
toupper
ctype, 26
uint16 t
avr
uint32 t
avr
uint64 t
avr
uint8 t
avr
uintptr t
avr
ultoa
avr
ungetc
avr
utoa
avr

inttypes, 28
inttypes, 29
inttypes, 29
inttypes, 29
inttypes, 29
stdlib, 59
stdio, 44
stdlib, 59

vfprintf
avr stdio, 44
vfprintf P
avr stdio, 47
vfscanf
avr stdio, 47
vfscanf P
avr stdio, 49
vsnprintf
avr stdio, 49
vsnprintf P
avr stdio, 49
vsprintf
avr stdio, 49
vsprintf P
avr stdio, 49
Watchdog timer handling, 22
wdt disable
avr watchdog, 22
wdt enable
avr watchdog, 22

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

INDEX

wdt reset
avr watchdog, 23
WDTO 120MS
avr watchdog, 23
WDTO 15MS
avr watchdog, 23
WDTO 1S
avr watchdog, 23
WDTO 250MS
avr watchdog, 23
WDTO 2S
avr watchdog, 23
WDTO 30MS
avr watchdog, 23
WDTO 500MS
avr watchdog, 23
WDTO 60MS
avr watchdog, 24

Generated on Wed Nov 26 21:10:23 2003 for avr-libc by Doxygen

178