EWAVR CompilerGuide PDF

IAR C/C++ Compiler
User Guide
for Microchip Technology’s

AVR Microcontroller Family
CAVR-10b
COPYRIGHT NOTICE
© 1996–2017 IAR Systems AB.
No part of this document may be reproduced without the prior written consent of IAR
Systems AB. The software described in this document is furnished under a license and
may only be used or copied in accordance with the terms of such a license.
DISCLAIMER
The information in this document is subject to change without notice and does not
represent a commitment on any part of IAR Systems. While the information contained
herein is assumed to be accurate, IAR Systems assumes no responsibility for any errors
or omissions.
In no event shall IAR Systems, its employees, its contractors, or the authors of this
document be liable for special, direct, indirect, or consequential damage, losses, costs,
charges, claims, demands, claim for lost profits, fees, or expenses of any nature or kind.
TRADEMARKS
IAR Systems, IAR Embedded Workbench, IAR Connect, C-SPY, C-RUN, C-STAT,
IAR Visual State, visualSTATE, IAR KickStart Kit, I-jet, I-jet Trace, I-scope, IAR
Academy, IAR, and the logotype of IAR Systems are trademarks or registered
trademarks owned by IAR Systems AB.
Microsoft and Windows are registered trademarks of Microsoft Corporation.
Atmel and AVR are registered trademarks of Microchip Technology.
Adobe and Acrobat Reader are registered trademarks of Adobe Systems Incorporated.
All other product names are trademarks or registered trademarks of their respective
owners.
EDITION NOTICE
Tenth edition: April 2017
Part number: CAVR-10b
This guide applies to version 7.1x of IAR Embedded Workbench® for Microchip
Technology’s AVR microcontroller family.
Internal reference: M23, Mym8.0, IMAE.
IAR C/C++ Compiler User Guide

2 for AVR
AFE1_AFE2-1:1
Brief contents
Tables ..................................................................................................................... 21
Preface .................................................................................................................... 23
Part 1. Using the compiler ....................................................... 31
Introduction to the IAR build tools ........................................................ 33
Developing embedded applications ........................................................ 39
Data storage ...................................................................................................... 53
Functions ............................................................................................................... 63
Linking overview ............................................................................................. 71
Linking your application ................................................................................ 87
The DLIB runtime environment ............................................................. 99
The CLIB runtime environment ............................................................ 141
Assembler language interface ................................................................. 149
Using C ................................................................................................................ 167
Using C++ .......................................................................................................... 175
Application-related considerations ....................................................... 193
Efficient coding for embedded applications ...................................... 205
Part 2. Reference information ........................................... 225
External interface details ............................................................................ 227
Compiler options ........................................................................................... 233
Data representation ...................................................................................... 281
Extended keywords ...................................................................................... 295
AFE1_AFE2-1:1
Pragma directives ........................................................................................... 319
Intrinsic functions ........................................................................................... 341
The preprocessor .......................................................................................... 351
C/C++ standard library functions .......................................................... 361

Segment reference ......................................................................................... 375
The stack usage control file ...................................................................... 391
Implementation-defined behavior for Standard C ........................ 399
Implementation-defined behavior for C89 ....................................... 415
Index ..................................................................................................................... 429

4 for AVR
AFE1_AFE2-1:1
Contents
Tables ..................................................................................................................... 21
Preface .................................................................................................................... 23
Who should read this guide ............................................................... 23
How to use this guide .......................................................................... 23
What this guide contains ................................................................... 24
Other documentation ........................................................................... 25
Document conventions ........................................................................ 28
Part 1. Using the compiler ....................................................... 31
Introduction to the IAR build tools ........................................................ 33
The IAR build tools—an overview ................................................... 33

IAR language overview ......................................................................... 34
Device support ......................................................................................... 35
Special support for embedded systems ........................................ 36
Developing embedded applications ........................................................ 39
Developing embedded software using IAR build tools .......... 39

The build process—an overview ...................................................... 41
Application execution—an overview ............................................. 44
Building applications—an overview ................................................ 48
Basic project configuration ................................................................. 48
Processor configuration ...................................................................... 49
Data storage ...................................................................................................... 53
Introduction ............................................................................................. 53
Memory types ......................................................................................... 54
Introduction to memory types ............................................................ 54
Using data memory attributes ............................................................ 54
Pointers and memory types ................................................................ 56
Structures and memory types ............................................................ 57
AFE1_AFE2-1:1
More examples ................................................................................... 57
C++ and memory types ..................................................................... 58
Memory models ...................................................................................... 59
Specifying a memory model .............................................................. 59
Storage of auto variables and parameters .................................. 60
Dynamic memory on the heap ........................................................ 61
Functions ............................................................................................................... 63
Function-related extensions .............................................................. 63

Function storage .................................................................................... 63
Using function memory attributes ...................................................... 63
Primitives for interrupts, concurrency, and OS-related
programming ............................................................................................ 64
Inlining functions ..................................................................................... 67
Linking overview ............................................................................................. 71
Linking—an overview ............................................................................ 71

Segments and memory ........................................................................ 72
The linking process in detail .............................................................. 73
Placing code and data—the linker configuration file .............. 74
Initialization at system startup ......................................................... 75
Stack usage analysis ............................................................................... 79
Linking your application ................................................................................ 87
Linking considerations .......................................................................... 87

Verifying the linked result of code and data placement ........ 95
The DLIB runtime environment ............................................................. 99
Introduction to the runtime environment .................................. 99

Setting up the runtime environment ........................................... 103
Additional information on the runtime environment .......... 110
Runtime library configurations ....................................................... 110
Prebuilt runtime libraries ................................................................. 111
Formatters for printf ......................................................................... 113
Formatters for scanf ........................................................................ 114

6 for AVR
AFE1_AFE2-1:1
Contents
The C-SPY emulated I/O mechanism .............................................. 116

Math functions ................................................................................. 116
System startup and termination ........................................................ 118
System initialization ......................................................................... 122
The DLIB low-level I/O interface .................................................... 123
abort ................................................................................................. 124
clock ................................................................................................. 124
__close ............................................................................................. 124
__exit ................................................................................................ 125
getenv ............................................................................................... 125
__getzone ......................................................................................... 126
__lseek ............................................................................................. 126
__open .............................................................................................. 126
raise .................................................................................................. 127
__read ............................................................................................... 127
remove .............................................................................................. 128
rename .............................................................................................. 129
_ReportAssert ................................................................................... 129
signal ................................................................................................ 130
system ............................................................................................... 130
__time32, __time64 ......................................................................... 130
__write ............................................................................................. 131
Configuration symbols for printf and scanf ..................................... 132
Configuration symbols for file input and output .............................. 133
Locale ............................................................................................... 133
Strtod ................................................................................................ 135
Managing a multithreaded environment .................................... 136
The CLIB runtime environment ............................................................ 141
Using a prebuilt runtime library .................................................... 141

Runtime library filename syntax ...................................................... 142
Input and output ................................................................................... 142
System startup and termination .................................................... 145
Overriding default library modules .............................................. 146
AFE1_AFE2-1:1
Customizing system initialization ................................................. 147
C-SPY emulated I/O ........................................................................... 147
Assembler language interface ................................................................. 149
Mixing C and assembler ..................................................................... 149

Inline assembler .............................................................................. 151
Calling assembler routines from C ............................................... 152
Calling assembler routines from C++ .......................................... 154
Calling convention ................................................................................ 155
Choosing a calling convention ......................................................... 156
Preserved versus scratch registers ................................................... 158
Function entrance ............................................................................ 159
Function exit ................................................................................... 162
Restrictions for special function types ............................................. 163
Examples .......................................................................................... 163
Call frame information ....................................................................... 165
Using C ................................................................................................................ 167
C language overview ........................................................................... 167

Extensions overview ............................................................................ 168
IAR C language extensions ............................................................... 169
Using C++ .......................................................................................................... 175
Overview—EC++ and EEC++ .......................................................... 175

Enabling support for C++ .................................................................. 177
EC++ feature descriptions ................................................................ 177
New and Delete operators ................................................................ 181
EEC++ feature description ............................................................... 183
Templates ......................................................................................... 183
C++ language extensions ................................................................... 189
Application-related considerations ....................................................... 193
Stack considerations ........................................................................... 193

Available stacks ............................................................................... 193
Heap considerations ............................................................................ 194

8 for AVR
AFE1_AFE2-1:1
Contents
Interaction between the tools and your application ............. 195

Checksum calculation for verifying image integrity .............. 196
Efficient coding for embedded applications ...................................... 205
Selecting data types ............................................................................. 205

Locating strings in ROM, RAM or flash ......................................... 205
Memory model and memory attributes for data ............................... 207
Controlling data and function placement in memory .......... 210
Controlling compiler optimizations ............................................. 214
Facilitating good code generation ................................................. 218
Accessing special function registers ................................................ 222
Part 2. Reference information ........................................... 225
External interface details ............................................................................ 227
Invocation syntax ................................................................................. 227

Include file search procedure .......................................................... 228
Compiler output ................................................................................... 229
Diagnostics .............................................................................................. 231
Compiler options ........................................................................................... 233
Options syntax ....................................................................................... 233

Summary of compiler options ........................................................ 235
Descriptions of compiler options ................................................... 240
--64bit_doubles ................................................................................ 240
--64k_flash ....................................................................................... 240
--c89 ................................................................................................. 240
--char_is_signed ............................................................................... 241
--char_is_unsigned ........................................................................... 241
--clib ................................................................................................. 241
--cpu ................................................................................................. 242
--cross_call_passes ........................................................................... 242
-D ..................................................................................................... 243
--debug, -r ......................................................................................... 243
--dependencies ................................................................................. 244
AFE1_AFE2-1:1
--diag_error ...................................................................................... 245
--diag_remark ................................................................................... 245
--diag_suppress ................................................................................ 246
--diag_warning ................................................................................. 246
--diagnostics_tables .......................................................................... 246
--disable_all_program_memory_load_instructions ......................... 247
--disable_direct_mode ...................................................................... 247
--disable_library_knowledge ........................................................... 247
--disable_mul ................................................................................... 247
--disable_spm ................................................................................... 248
--discard_unused_publics ................................................................. 248
--dlib ................................................................................................. 248
--dlib_config ..................................................................................... 249
--do_cross_call ................................................................................. 250
-e ...................................................................................................... 250
--ec++ ............................................................................................... 250
--eec++ ............................................................................................. 251
--eecr_address .................................................................................. 251
--eeprom_size ................................................................................... 251
--enable_external_bus ...................................................................... 252
--enable_multibytes .......................................................................... 252
--enable_restrict ............................................................................... 253
--enhanced_core ............................................................................... 253
--error_limit ...................................................................................... 253
-f ....................................................................................................... 254
--force_switch_type ......................................................................... 254
--guard_calls ..................................................................................... 255
--header_context ............................................................................... 255
-I ....................................................................................................... 255
--initializers_in_flash ....................................................................... 256
-l ....................................................................................................... 256
--library_module .............................................................................. 257
--lock_regs ....................................................................................... 258
--macro_positions_in_diagnostics ................................................... 258

10 for AVR
AFE1_AFE2-1:1
Contents
--memory_model, -m ....................................................................... 258

--mfc ................................................................................................. 259
--module_name ................................................................................ 260
--no_call_frame_info ....................................................................... 260
--no_clustering ................................................................................. 260
--no_code_motion ............................................................................ 261
--no_cross_call ................................................................................. 261
--no_cse ............................................................................................ 261
--no_inline ........................................................................................ 262
--no_path_in_file_macros ................................................................ 262
--no_rampd ....................................................................................... 262
--no_size_constraints ....................................................................... 262
--no_static_destruction ..................................................................... 263
--no_system_include ........................................................................ 263
--no_tbaa .......................................................................................... 263
--no_typedefs_in_diagnostics .......................................................... 264
--no_ubrof_messages ....................................................................... 264
--no_unroll ....................................................................................... 265
--no_warnings .................................................................................. 265
--no_wrap_diagnostics ..................................................................... 265
-O ..................................................................................................... 265
--omit_types ..................................................................................... 266
--only_stdout .................................................................................... 266
--output, -o ....................................................................................... 267
--pending_instantiations ................................................................... 267
--predef_macros ............................................................................... 267
--preinclude ..................................................................................... 268
--preprocess ...................................................................................... 268
--public_equ ..................................................................................... 268
--relaxed_fp ...................................................................................... 269
--remarks .......................................................................................... 269
--require_prototypes ......................................................................... 270
--root_variables ................................................................................ 270
-s ....................................................................................................... 271
11
AFE1_AFE2-1:1
--segment .......................................................................................... 271
--separate_cluster_for_initialized_variables .................................... 272
--silent .............................................................................................. 272
--spmcr_address ............................................................................... 273
--strict ............................................................................................... 273
--string_literals_in_flash .................................................................. 273
--system_include_dir ........................................................................ 274
--use_c++_inline .............................................................................. 274
-v ...................................................................................................... 275
--version ........................................................................................... 276
--version1_calls ................................................................................ 276
--version2_calls ................................................................................ 277
--vla .................................................................................................. 277
--warn_about_c_style_casts ............................................................. 277
--warnings_affect_exit_code ............................................................ 278
--warnings_are_errors ...................................................................... 278
--xmcra_address ............................................................................... 278
-y ...................................................................................................... 279
-z ....................................................................................................... 279
--zero_register .................................................................................. 280
Data representation ...................................................................................... 281
Alignment ................................................................................................ 281

Byte order ................................................................................................ 282
Basic data types—integer types .................................................... 282
Basic data types—floating-point types ....................................... 285
Pointer types .......................................................................................... 287
Data pointers .................................................................................... 287
Structure types ..................................................................................... 290
Type qualifiers ........................................................................................ 291
Data types in C++ ................................................................................. 293
Extended keywords ...................................................................................... 295
General syntax rules for extended keywords ........................... 295

12 for AVR
AFE1_AFE2-1:1
Contents
Summary of extended keywords ................................................... 299

Descriptions of extended keywords ............................................. 300
__eeprom ......................................................................................... 300
__ext_io ........................................................................................... 301
__far ................................................................................................. 302
__farflash ......................................................................................... 302
__farfunc .......................................................................................... 303
__flash .............................................................................................. 303
__generic .......................................................................................... 305
__huge .............................................................................................. 305
__hugeflash ...................................................................................... 306
__interrupt ........................................................................................ 307
__intrinsic ........................................................................................ 307
__io .................................................................................................. 307
__monitor ......................................................................................... 308
__near .............................................................................................. 308
__nearfunc ....................................................................................... 308
__nested ........................................................................................... 309
__no_alloc, __no_alloc16 ............................................................... 310
__no_alloc_str, __no_alloc_str16 ................................................... 310
__no_init .......................................................................................... 311
__no_runtime_init ............................................................................ 311
__noreturn ........................................................................................ 312
__raw ............................................................................................... 312
__regvar ........................................................................................... 312
__root ............................................................................................... 313
__ro_placement ................................................................................ 313
__task ............................................................................................... 314
__tiny ............................................................................................... 314
__tinyflash ....................................................................................... 315
__version_1 ...................................................................................... 315
__version_2 ...................................................................................... 315
__version_4 ...................................................................................... 316
__x ................................................................................................... 316
13
AFE1_AFE2-1:1
__x_z ................................................................................................ 316
__z .................................................................................................... 317
__z_x ................................................................................................ 317
Pragma directives ........................................................................................... 319
Summary of pragma directives ...................................................... 319

Descriptions of pragma directives ................................................ 321
basic_template_matching ................................................................. 321
bitfields ............................................................................................. 321
calls .................................................................................................. 322
call_graph_root ................................................................................ 322
constseg ............................................................................................ 323
data_alignment ................................................................................. 323
dataseg .............................................................................................. 324
default_function_attributes .............................................................. 324
default_variable_attributes ............................................................... 325
diag_default ...................................................................................... 326
diag_error ......................................................................................... 327
diag_remark ..................................................................................... 327
diag_suppress ................................................................................... 327
diag_warning .................................................................................... 328
error .................................................................................................. 328
include_alias ..................................................................................... 329
inline ................................................................................................. 329
language ........................................................................................... 330
location ............................................................................................. 331
message ............................................................................................ 331
object_attribute ................................................................................. 332
optimize ............................................................................................ 332
__printf_args .................................................................................... 334
public_equ ........................................................................................ 334
required ............................................................................................ 334
rtmodel ............................................................................................. 335
__scanf_args .................................................................................... 336

14 for AVR
AFE1_AFE2-1:1
Contents
segment ............................................................................................ 336

STDC CX_LIMITED_RANGE ....................................................... 337
STDC FENV_ACCESS ................................................................... 337
STDC FP_CONTRACT .................................................................. 338
type_attribute ................................................................................... 338
vector ................................................................................................ 339
weak ................................................................................................. 339
Intrinsic functions ........................................................................................... 341
Summary of intrinsic functions ....................................................... 341

Descriptions of intrinsic functions ................................................. 342
__delay_cycles ................................................................................. 342
__DES_decryption ........................................................................... 342
__DES_encryption ........................................................................... 343
__disable_interrupt .......................................................................... 343
__enable_interrupt ........................................................................... 343
__extended_load_program_memory ............................................... 343
__fractional_multiply_signed .......................................................... 344
__fractional_multiply_signed_with_unsigned ................................ 344
__fractional_multiply_unsigned ...................................................... 344
__get_interrupt_state ....................................................................... 344
__indirect_jump_to .......................................................................... 345
__insert_opcode ............................................................................... 345
__lac ................................................................................................. 345
__las ................................................................................................. 346
__lat ................................................................................................. 346
__load_program_memory ................................................................ 346
__multiply_signed ........................................................................... 346
__multiply_signed_with_unsigned .................................................. 346
__multiply_unsigned ....................................................................... 346
__no_operation ................................................................................ 347
__require .......................................................................................... 347
__restore_interrupt ........................................................................... 347
__reverse .......................................................................................... 348
15
AFE1_AFE2-1:1
__save_interrupt .............................................................................. 348
__set_interrupt_state ........................................................................ 348
__sleep ............................................................................................ 349
__swap_nibbles ................................................................................ 349
__watchdog_reset ............................................................................ 349
__xch ................................................................................................ 349
The preprocessor .......................................................................................... 351
Overview of the preprocessor ........................................................ 351

Description of predefined preprocessor symbols ................... 352
__BASE_FILE__ ............................................................................ 352
__BUILD_NUMBER__ .................................................................. 352
__CORE__ ...................................................................................... 352
__COUNTER__ .............................................................................. 352
__cplusplus ...................................................................................... 352
__CPU__ ......................................................................................... 352
__DATE__ ...................................................................................... 353
__device__ ....................................................................................... 353
__DOUBLE__ ................................................................................. 353
__embedded_cplusplus .................................................................... 353
__FILE__ ......................................................................................... 353
__func__ .......................................................................................... 354
__FUNCTION__ ............................................................................. 354
__HAS_EEPROM__ ....................................................................... 354
__HAS_EIND__ ............................................................................. 354
__HAS_ELPM__ ............................................................................ 354
__HAS_ENHANCED_CORE__ .................................................... 355
__HAS_FISCR__ ............................................................................ 355
__HAS_MUL__ .............................................................................. 355
__HAS_RAMPD__ ......................................................................... 355
__HAS_RAMPX__ ......................................................................... 355
__HAS_RAMPY__ ......................................................................... 355
__HAS_RAMPZ__ ......................................................................... 356
__IAR_SYSTEMS_ICC__ ............................................................. 356

16 for AVR
AFE1_AFE2-1:1
Contents
__ICC avr __
..................................................................................... 356
__LINE__ ........................................................................................ 356
__LITTLE_ENDIAN__ .................................................................. 356
__MEMORY_MODEL__ ............................................................... 356
__PRETTY_FUNCTION__ ............................................................ 356
__STDC__ ....................................................................................... 357
__STDC_VERSION__ ................................................................... 357
__SUBVERSION__ ........................................................................ 357
__TID__ .......................................................................................... 357
__TIME__ ....................................................................................... 358
__TIMESTAMP__ .......................................................................... 358
__TINY_AVR__ ............................................................................. 358
__VER__ ......................................................................................... 358
__VERSION_1_CALLS__ ............................................................. 358
__XMEGA_CORE__ ...................................................................... 359
__XMEGA_USB__ ......................................................................... 359
Descriptions of miscellaneous preprocessor extensions ..... 359
NDEBUG ........................................................................................ 359
#warning message ........................................................................... 360
C/C++ standard library functions .......................................................... 361

C/C++ standard library overview ................................................. 361
DLIB runtime environment—implementation details ......... 363
CLIB runtime environment—implementation details ......... 369
AVR-specific library functions ......................................................... 370
memcmp_G ...................................................................................... 371
memcpy_G ....................................................................................... 371
memcpy_P ........................................................................................ 371
printf_P ............................................................................................. 371
puts_G .............................................................................................. 371
puts_P ............................................................................................... 372
scanf_P ............................................................................................. 372
sprintf_P ........................................................................................... 372
sscanf_P ........................................................................................... 372
17
AFE1_AFE2-1:1
strcat_G ............................................................................................ 372
strcmp_G .......................................................................................... 372
strcmp_P ........................................................................................... 373
strcpy_G ........................................................................................... 373
strcpy_P ............................................................................................ 373
strerror_P .......................................................................................... 373
strlen_G ............................................................................................ 373
strlen_P ............................................................................................. 373
strncat_G .......................................................................................... 374
strncmp_G ........................................................................................ 374
strncmp_P ......................................................................................... 374
strncpy_G ......................................................................................... 374
strncpy_P .......................................................................................... 374
Segment reference ......................................................................................... 375
Summary of segments ...................................................................... 375

Descriptions of segments .................................................................. 377
CHECKSUM ................................................................................... 377
CODE ............................................................................................... 378
CSTACK ......................................................................................... 378
DIFUNCT ........................................................................................ 378
EEPROM_I ...................................................................................... 379
EEPROM_N ..................................................................................... 379
FARCODE ....................................................................................... 379
FAR_C ............................................................................................. 380
FAR_F .............................................................................................. 380
FAR_HEAP ..................................................................................... 380
FAR_I ............................................................................................... 381
FAR_ID ............................................................................................ 381
FAR_N ............................................................................................. 381
FAR_Z ............................................................................................. 382
HEAP ............................................................................................... 382
HUGE_C .......................................................................................... 382
HUGE_F .......................................................................................... 383

18 for AVR
AFE1_AFE2-1:1
Contents
HUGE_HEAP .................................................................................. 383

HUGE_I ........................................................................................... 383
HUGE_ID ........................................................................................ 384
HUGE_N .......................................................................................... 384
HUGE_Z .......................................................................................... 384
INITTAB .......................................................................................... 385
INTVEC .......................................................................................... 385
NEAR_C .......................................................................................... 385
NEAR_F ........................................................................................... 386
NEAR_HEAP .................................................................................. 386
NEAR_I ........................................................................................... 386
NEAR_ID ......................................................................................... 387
NEAR_N .......................................................................................... 387
NEAR_Z .......................................................................................... 387
RSTACK ......................................................................................... 388
SWITCH .......................................................................................... 388
TINY_F ............................................................................................ 388
TINY_HEAP .................................................................................... 389
TINY_I ............................................................................................. 389
TINY_ID .......................................................................................... 389
TINY_N ........................................................................................... 390
TINY_Z ............................................................................................ 390
The stack usage control file ...................................................................... 391
Overview .................................................................................................. 391

Stack usage control directives ........................................................ 391
call graph root directive ................................................................... 391
check that directive .......................................................................... 392
exclude directive .............................................................................. 393
function directive ............................................................................. 393
max recursion depth directive .......................................................... 393
no calls from directive ...................................................................... 394
possible calls directive ..................................................................... 394
19
AFE1_AFE2-1:1
Syntactic components ........................................................................ 395
category ............................................................................................ 395
func-spec .......................................................................................... 395
module-spec ..................................................................................... 396
name ................................................................................................. 396
call-info ............................................................................................ 396
stack-size .......................................................................................... 397
size .................................................................................................... 397
Implementation-defined behavior for Standard C ........................ 399
Descriptions of implementation-defined behavior ................ 399

Implementation-defined behavior for C89 ....................................... 415
Descriptions of implementation-defined behavior ................ 415

Index ..................................................................................................................... 429

20 for AVR
AFE1_AFE2-1:1
Tables
1: Typographic conventions used in this guide ......................................................... 28
2: Naming conventions used in this guide ................................................................ 29
3: Summary of processor configuration .................................................................... 50
4: Memory types and their corresponding memory attributes .................................. 55
5: Memory model characteristics .............................................................................. 59
6: Heaps supported in memory types ........................................................................ 62
7: Function memory attributes .................................................................................. 63
8: XLINK segment memory types ............................................................................ 72
9: segments holding initialized data .......................................................................... 76
10: Debug information and levels of C-SPY emulated I/O .................................... 104
11: Library configurations ....................................................................................... 110
12: Prebuilt libraries ................................................................................................ 113
13: Formatters for printf .......................................................................................... 113
14: Formatters for scanf .......................................................................................... 114
15: DLIB low-level I/O interface functions ............................................................ 123
16: Descriptions of printf configuration symbols ................................................... 132
17: Descriptions of scanf configuration symbols .................................................... 132
18: Library objects using TLS ................................................................................. 136
19: Macros for implementing TLS allocation ......................................................... 139
20: Runtime libraries ............................................................................................... 142
21: Registers used for passing parameters .............................................................. 160
22: Passing parameters in registers ......................................................................... 161
23: Registers used for returning values ................................................................... 162
24: Language extensions ......................................................................................... 169
25: Compiler optimization levels ............................................................................ 215
26: Compiler environment variables ....................................................................... 228
27: Error return codes .............................................................................................. 230
28: Compiler options summary ............................................................................... 235
29: Accessing variables with aggregate initializers ................................................ 256
30: Integer types ...................................................................................................... 282
31: Floating-point types .......................................................................................... 285
21
AFE1_AFE2-1:1
32: Function pointers ............................................................................................... 287
33: Data pointers ..................................................................................................... 287
34: size_t typedef .................................................................................................... 289
35: ptrdif_t typedef .................................................................................................. 290
36: Type of volatile accesses treated in a special way ............................................ 292
37: Extended keywords summary ........................................................................... 299
38: Pragma directives summary .............................................................................. 319
39: Intrinsic functions summary .............................................................................. 341
40: Traditional Standard C header files—DLIB ..................................................... 364
41: C++ header files ................................................................................................ 365
42: Standard template library header files ............................................................... 366
43: New Standard C header files—DLIB ............................................................... 366
44: CLIB runtime environment header files ........................................................... 370
45: Segment summary ............................................................................................. 375
46: Message returned by strerror()—DLIB runtime environment .......................... 414
47: Message returned by strerror()—DLIB runtime environment .......................... 425
48: Message returned by strerror()—CLIB runtime environment .......................... 428

22 for AVR
AFE1_AFE2-1:1
Preface
Welcome to the IAR C/C++ Compiler User Guide for AVR. The purpose of this
guide is to provide you with detailed reference information that can help you
to use the compiler to best suit your application requirements. This guide also
gives you suggestions on coding techniques so that you can develop
applications with maximum efficiency.
Who should read this guide

Read this guide if you plan to develop an application using the C or C++ language for
the AVR microcontroller and need detailed reference information on how to use the
compiler.
REQUIRED KNOWLEDGE
To use the tools in IAR Embedded Workbench, you should have working knowledge of:
● The architecture and instruction set of the Microchip AVR microcontroller (refer to
the chip manufacturer's documentation)
● The C or C++ programming language
● Application development for embedded systems
● The operating system of your host computer.
For more information about the other development tools incorporated in the IDE, refer
to their respective documentation, see Other documentation, page 25.
How to use this guide

When you start using the IAR C/C++ Compiler for AVR, you should read Part 1. Using
the compiler in this guide.
When you are familiar with the compiler and have already configured your project, you
can focus more on Part 2. Reference information.
If you are new to using this product, we suggest that you first read the guide Getting
Started with IAR Embedded Workbench® for an overview of the tools and the features
that the IDE offers. The tutorials, which you can find in IAR Information Center, will
help you get started using IAR Embedded Workbench.
23
AFE1_AFE2-1:1
What this guide contains
What this guide contains

Below is a brief outline and summary of the chapters in this guide.
PART 1. USING THE COMPILER

● Introduction to the IAR build tools gives an introduction to the IAR build tools,
which includes an overview of the tools, the programming languages, the available
device support, and extensions provided for supporting specific features of the AVR
microcontroller.
● Developing embedded applications gives the information you need to get started
developing your embedded software using the IAR build tools.
● Data storage describes how to store data in memory.
● Functions gives a brief overview of function-related extensions—mechanisms for
controlling functions—and describes some of these mechanisms in more detail.
● Linking overview describes the linking process using the IAR XLINK Linker and
the related concepts.
● Linking your application lists aspects that you must consider when linking your
application, including using XLINK options and tailoring the linker configuration
file.
● The DLIB runtime environment describes the DLIB runtime environment in which
an application executes. It covers how you can modify it by setting options,
overriding default library modules, or building your own library. The chapter also
describes system initialization introducing the file cstartup, how to use modules
for locale, and file I/O.
● The CLIB runtime environment gives an overview of the CLIB runtime libraries and
how to customize them. The chapter also describes system initialization and
introduces the file cstartup.
● Assembler language interface contains information required when parts of an
application are written in assembler language. This includes the calling convention.
● Using C gives an overview of the two supported variants of the C language and an
overview of the compiler extensions, such as extensions to Standard C.
● Using C++ gives an overview of the two levels of C++ support: The
industry-standard EC++ and IAR Extended EC++.
● Application-related considerations discusses a selected range of application issues
related to using the compiler and linker.
● Efficient coding for embedded applications gives hints about how to write code that
compiles to efficient code for an embedded application.

24 for AVR
AFE1_AFE2-1:1
Preface
PART 2. REFERENCE INFORMATION

● External interface details provides reference information about how the compiler
interacts with its environment—the invocation syntax, methods for passing options
to the compiler, environment variables, the include file search procedure, and the
different types of compiler output. The chapter also describes how the compiler’s
diagnostic system works.
● Compiler options explains how to set options, gives a summary of the options, and
contains detailed reference information for each compiler option.
● Data representation describes the available data types, pointers, and structure types.
This chapter also gives information about type and object attributes.
● Extended keywords gives reference information about each of the AVR-specific
keywords that are extensions to the standard C/C++ language.
● Pragma directives gives reference information about the pragma directives.
● Intrinsic functions gives reference information about functions to use for accessing
AVR-specific low-level features.
● The preprocessor gives a brief overview of the preprocessor, including reference
information about the different preprocessor directives, symbols, and other related
information.
● C/C++ standard library functions gives an introduction to the C or C++ library
functions, and summarizes the header files.
● Segment reference gives reference information about the compiler’s use of
segments.
● The stack usage control file describes the syntax and semantics of stack usage
control files.
● Implementation-defined behavior for Standard C describes how the compiler
handles the implementation-defined areas of Standard C.
● Implementation-defined behavior for C89 describes how the compiler handles the
implementation-defined areas of the C language standard C89.
Other documentation
User documentation is available as hypertext PDFs and as a context-sensitive online
help system in HTML format. You can access the documentation from the Information
Center or from the Help menu in the IAR Embedded Workbench IDE. The online help
system is also available via the F1 key.
25
AFE1_AFE2-1:1
Other documentation
USER AND REFERENCE GUIDES

The complete set of IAR Systems development tools is described in a series of guides.
Information about:
● System requirements and information about how to install and register the IAR
Systems products, is available in the booklet Quick Reference (available in the
product box) and the Installation and Licensing Guide.
● Using the IDE for project management and building, is available in the IDE Project
Management and Building Guide for AVR.
● Using the IAR C-SPY® Debugger, is available in the C-SPY® Debugging Guide
for AVR.
● Programming for the IAR C/C++ Compiler for AVR, is available in the IAR C/C++
Compiler User Guide for AVR.
● Using the IAR XLINK Linker, the IAR XAR Library Builder, and the IAR XLIB
Librarian, is available in the IAR Linker and Library Tools Reference Guide.
● Programming for the IAR Assembler for AVR, is available in the IAR Assembler
User Guide for AVR.
● Performing a static analysis using C-STAT and the required checks, is available in
the C-STAT Static Analysis Guide.
● Developing safety-critical applications using the MISRA C guidelines, is available
in the IAR Embedded Workbench® MISRA C:2004 Reference Guide or the IAR
Embedded Workbench® MISRA C:1998 Reference Guide.
● Porting application code and projects created with a previous version of the IAR
Embedded Workbench for AVR, is available in the IAR Embedded Workbench®
Migration Guide.
Note: Additional documentation might be available depending on your product
installation.
THE ONLINE HELP SYSTEM

The context-sensitive online help contains:
● Information about project management, editing, and building in the IDE
● Information about debugging using the IAR C-SPY® Debugger
● Reference information about the menus, windows, and dialog boxes in the IDE
● Compiler reference information
● Keyword reference information for the DLIB library functions. To obtain reference
information for a function, select the function name in the editor window and press
F1. Note that if you select a function name in the editor window and press F1 while

26 for AVR
AFE1_AFE2-1:1
Preface
using the CLIB C standard library, you will get reference information for the DLIB
C/EC++ standard library.
FURTHER READING
These books might be of interest to you when using the IAR Systems development tools:
● Barr, Michael, and Andy Oram, ed. Programming Embedded Systems in C and
C++. O’Reilly & Associates.
● Harbison, Samuel P. and Guy L. Steele (contributor). C: A Reference Manual.
Prentice Hall.
● Labrosse, Jean J. Embedded Systems Building Blocks: Complete and Ready-To-Use
Modules in C. R&D Books.
● Mann, Bernhard. C für Mikrocontroller. Franzis-Verlag. [Written in German.]
● Meyers, Scott. Effective C++: 50 Specific Ways to Improve Your Programs and
Designs. Addison-Wesley.
● Meyers, Scott. More Effective C++. Addison-Wesley.
● Meyers, Scott. Effective STL. Addison-Wesley.
● Sutter, Herb. Exceptional C++: 47 Engineering Puzzles, Programming Problems,
and Solutions. Addison-Wesley.
The web site isocpp.org also has a list of recommended books about C++ programming.
WEB SITES
Recommended web sites:
● The Microchip Technology web site, www.microchip.com, that contains
information and news about the Microchip AVR microcontrollers.
● The IAR Systems web site, www.iar.com, that holds application notes and other
product information.
● The web site of the C standardization working group,
www.open-std.org/jtc1/sc22/wg14.
● The web site of the C++ Standards Committee, www.open-std.org/jtc1/sc22/wg21.
● The C++ programming language web site, isocpp.org.
This web site also has a list of recommended books about C++ programming.
● The C and C++ reference web site, en.cppreference.com.
27
AFE1_AFE2-1:1
Document conventions
When, in the IAR Systems documentation, we refer to the programming language C, the
text also applies to C++, unless otherwise stated.
When referring to a directory in your product installation, for example avr\doc, the full
path to the location is assumed, for example c:\Program Files\IAR
Systems\Embedded Workbench N.n\avr\doc, where the initial digit of the version
number reflects the initial digit of the version number of the IAR Embedded Workbench
shared components.
TYPOGRAPHIC CONVENTIONS
The IAR Systems documentation set uses the following typographic conventions:
Style Used for
computer • Source code examples and file paths.
• Text on the command line.
• Binary, hexadecimal, and octal numbers.
parameter A placeholder for an actual value used as a parameter, for example
filename.h where filename represents the name of the file.
[option] An optional part of a directive, where [ and ] are not part of the actual
directive, but any [, ], {, or } are part of the directive syntax.
{option} A mandatory part of a directive, where { and } are not part of the
actual directive, but any [, ], {, or } are part of the directive syntax.
[option] An optional part of a command.
[a|b|c] An optional part of a command with alternatives.
{a|b|c} A mandatory part of a command with alternatives.
bold Names of menus, menu commands, buttons, and dialog boxes that
appear on the screen.
italic • A cross-reference within this guide or to another guide.
• Emphasis.
… An ellipsis indicates that the previous item can be repeated an arbitrary
number of times.
Identifies instructions specific to the IAR Embedded Workbench® IDE
interface.
Identifies instructions specific to the command line interface.
Identifies helpful tips and programming hints.
Table 1: Typographic conventions used in this guide

28 for AVR
AFE1_AFE2-1:1
Preface
Style Used for

Identifies warnings.
Table 1: Typographic conventions used in this guide (Continued)
NAMING CONVENTIONS
The following naming conventions are used for the products and tools from IAR
Systems®, when referred to in the documentation:
Brand name Generic term
IAR Embedded Workbench® for AVR IAR Embedded Workbench®
IAR Embedded Workbench® IDE for AVR the IDE
IAR C-SPY® Debugger for AVR C-SPY, the debugger
IAR C-SPY® Simulator the simulator
IAR C/C++ Compiler™ for AVR the compiler
IAR Assembler™ for AVR the assembler
IAR XLINK Linker™ XLINK, the linker
IAR XAR Library Builder™ the library builder
IAR XLIB Librarian™ the librarian
IAR DLIB Runtime Environment™ the DLIB runtime environment
IAR CLIB Runtime Environment™ the CLIB runtime environment
Table 2: Naming conventions used in this guide
29
AFE1_AFE2-1:1

30 for AVR
AFE1_AFE2-1:1
Part 1. Using the compiler
This part of the IAR C/C++ Compiler User Guide for AVR includes these
chapters:
● Introduction to the IAR build tools
● Developing embedded applications
● Data storage
● Functions
● Linking overview
● Linking your application
● The DLIB runtime environment
● The CLIB runtime environment
● Assembler language interface
● Using C
● Using C++
● Application-related considerations
● Efficient coding for embedded applications.
31
32
Introduction to the IAR
build tools
● The IAR build tools—an overview
● IAR language overview
● Device support
● Special support for embedded systems
The IAR build tools—an overview

In the IAR product installation you can find a set of tools, code examples, and user
documentation, all suitable for developing software for AVR-based embedded
applications. The tools allow you to develop your application in C, C++, or in assembler
language.
IAR Embedded Workbench® is a very powerful Integrated Development Environment
(IDE) that allows you to develop and manage complete embedded application projects.
It provides an easy-to-learn and highly efficient development environment with
maximum code inheritance capabilities, comprehensive and specific target support. IAR
Embedded Workbench promotes a useful working methodology, and thus a significant
reduction of the development time.
For information about the IDE, see the IDE Project Management and Building Guide
for AVR.
The compiler, assembler, and linker can also be run from a command line environment,
if you want to use them as external tools in an already established project environment.
IAR C/C++ COMPILER

The IAR C/C++ Compiler for AVR is a state-of-the-art compiler that offers the standard
features of the C and C++ languages, plus extensions designed to take advantage of the
AVR-specific facilities.
33
AFE1_AFE2-1:1
IAR language overview
IAR ASSEMBLER
The IAR Assembler for AVR is a powerful relocating macro assembler with a versatile
set of directives and expression operators. The assembler features a built-in C language
preprocessor and supports conditional assembly.
The IAR Assembler for AVR uses the same mnemonics and operand syntax as the
Microchip Technology AVR Assembler, which simplifies the migration of existing
code. For more information, see the IAR Assembler User Guide for AVR.
THE IAR XLINK LINKER

The IAR XLINK Linker is a powerful, flexible software tool for use in the development
of embedded controller applications. It is equally well suited for linking small,
single-file, absolute assembler programs as it is for linking large, relocatable input,
multi-module, C/C++, or mixed C/C++ and assembler programs.
To handle libraries, the library tools XAR and XLIB are included.
EXTERNAL TOOLS
For information about how to extend the tool chain in the IDE, see the IDE Project
IAR language overview

The IAR C/C++ Compiler for AVR supports:
● C, the most widely used high-level programming language in the embedded systems
industry. You can build freestanding applications that follow these standards:
● Standard C—also known as C99. Hereafter, this standard is referred to as
Standard C in this guide.
● C89—also known as C94, C90, C89, and ANSI C. This standard is required
when MISRA C is enabled.
● C++, a modern object-oriented programming language with a full-featured library
well suited for modular programming. Any of these standards can be used:
● Embedded C++ (EC++)—a subset of the C++ programming standard, which is
intended for embedded systems programming. It is defined by an industry
consortium, the Embedded C++ Technical committee. See the chapter Using
C++.
● IAR Extended Embedded C++ (EEC++)—EC++ with additional features such
as full template support, multiple inheritance, namespace support, the new cast
operators, as well as the Standard Template Library (STL).

34 for AVR
AFE1_AFE2-1:1
Introduction to the IAR build tools
Each of the supported languages can be used in strict or relaxed mode, or relaxed with
IAR extensions enabled. The strict mode adheres to the standard, whereas the relaxed
mode allows some common deviations from the standard.
For more information about C, see the chapter Using C.
For more information about Embedded C++ and Extended Embedded C++, see the
chapter Using C++.
For information about how the compiler handles the implementation-defined areas of
the languages, see the chapter Implementation-defined behavior for Standard C.
It is also possible to implement parts of the application, or the whole application, in
assembler language. See the IAR Assembler User Guide for AVR.
Device support
To get a smooth start with your product development, the IAR product installation
comes with a wide range of device-specific support.
SUPPORTED AVR DEVICES

The IAR C/C++ Compiler for AVR supports all devices based on the standard
Microchip Technology AVR microcontroller.
PRECONFIGURED SUPPORT FILES

The IAR product installation contains preconfigured files for supporting different
devices. If you need additional files for device support, they can be created using one of
the provided ones as a template.
Header files for I/O

Standard peripheral units are defined in device-specific I/O header files with the
filename extension h.The product package supplies I/O files for many of the devices that
are available at the time of the product release. You can find these files in the avr\inc
directory. Make sure to include the appropriate include file in your application source
files. If you need additional I/O header files, they can be created using one of the
provided ones as a template.
Linker configuration files

The avr\config directory and the avr\src\template directory contain ready-made
linker configuration files for all supported devices. The files have the filename extension
xcl and contain the information required by the linker. For more information about the
35
AFE1_AFE2-1:1
Special support for embedded systems
linker configuration file, see Placing code and data—the linker configuration file, page
74 as well as the IAR Linker and Library Tools Reference Guide.
Device description files

The debugger handles several of the device-specific requirements, such as definitions of
available memory areas, peripheral registers and groups of these, by using device
description files. These files are located in the avr\config directory and they have the
filename extension ddf. The peripheral registers and groups of these can be defined in
separate files (filename extension sfr), which in that case are included in the ddf file.
For more information about these files, see the C-SPY® Debugging Guide for AVR.
EXAMPLES FOR GETTING STARTED

The avr\examples directory contains examples of working applications to give you a
smooth start with your development.

This section briefly describes the extensions provided by the compiler to support
specific features of the AVR microcontroller.
EXTENDED KEYWORDS
The compiler provides a set of keywords that can be used for configuring how the code
is generated. For example, there are keywords for controlling how to access and store
data objects, as well as for controlling how a function should work internally and how
it should be called/returned.
By default, language extensions are enabled in the IDE.
The command line option -e makes the extended keywords available, and reserves them
so that they cannot be used as variable names. See, -e, page 250 for additional
information.
For more information about the extended keywords, see the chapter Extended keywords.
See also, Data storage, page 53 and Functions, page 63.
PRAGMA DIRECTIVES
The pragma directives control the behavior of the compiler, for example how it allocates
memory, whether it allows extended keywords, and whether it issues warning messages.
The pragma directives are always enabled in the compiler. They are consistent with
standard C, and are very useful when you want to make sure that the source code is
portable.

36 for AVR
AFE1_AFE2-1:1
Introduction to the IAR build tools
For more information about the pragma directives, see the chapter Pragma directives.
PREDEFINED SYMBOLS
With the predefined preprocessor symbols, you can inspect your compile-time
environment, for example time of compilation or the build number of the compiler.
For more information about the predefined symbols, see the chapter The preprocessor.
ACCESSING LOW-LEVEL FEATURES

For hardware-related parts of your application, accessing low-level features is essential.
The compiler supports several ways of doing this: intrinsic functions, mixing C and
assembler modules, and inline assembler. For information about the different methods,
see Mixing C and assembler, page 149.
37
AFE1_AFE2-1:1

38 for AVR
AFE1_AFE2-1:1
Developing embedded
applications
● Developing embedded software using IAR build tools
● The build process—an overview
● Application execution—an overview
● Building applications—an overview
● Basic project configuration
Developing embedded software using IAR build tools

Typically, embedded software written for a dedicated microcontroller is designed as an
endless loop waiting for some external events to happen. The software is located in
ROM and executes on reset. You must consider several hardware and software factors
when you write this kind of software. To your help, you have compiler options, extended
keywords, pragma directives, etc.
CPU FEATURES AND CONSTRAINTS

When developing software for the AVR microcontroller, you must consider some CPU
features and constraints. For example, the instruction set, processor configuration, and
memory model.
The compiler supports this by means of compiler options, extended keywords, pragma
directives, etc.
MAPPING OF MEMORY
Embedded systems typically contain various types of memory, such as on-chip RAM,
external DRAM or SRAM, ROM, EEPROM, or flash memory.
As an embedded software developer, you must understand the features of the different
types of memory. For example, on-chip RAM is often faster than other types of
memories, and variables that are accessed often would in time-critical applications
benefit from being placed here. Conversely, some configuration data might be accessed
seldom but must maintain their value after power off, so they should be saved in
EEPROM or flash memory.
39
AFE1_AFE2-1:1
Developing embedded software using IAR build tools
For efficient memory usage, the compiler provides several mechanisms for controlling
placement of functions and data objects in memory. For more information, see
Controlling data and function placement in memory, page 210. The linker places
sections of code and data in memory according to the directives you specify in the linker
configuration file, see Placing code and data—the linker configuration file, page 74.
COMMUNICATION WITH PERIPHERAL UNITS

If external devices are connected to the microcontroller, you might need to initialize and
control the signaling interface, for example by using chip select pins, and detect and
handle external interrupt signals. Typically, this must be initialized and controlled at
runtime. The normal way to do this is to use special function registers (SFR). These are
typically available at dedicated addresses, containing bits that control the chip
configuration.
Standard peripheral units are defined in device-specific I/O header files with the
filename extension h. See Device support, page 35. For an example, see Accessing
special function registers, page 222.
EVENT HANDLING
In embedded systems, using interrupts is a method for handling external events
immediately; for example, detecting that a button was pressed. In general, when an
interrupt occurs in the code, the microcontroller immediately stops executing the code
it runs, and starts executing an interrupt routine instead.
The compiler provides various primitives for managing hardware and software
interrupts, which means that you can write your interrupt routines in C, see Primitives
for interrupts, concurrency, and OS-related programming, page 64.
SYSTEM STARTUP
In all embedded systems, system startup code is executed to initialize the system—both
the hardware and the software system—before the main function of the application is
called.
As an embedded software developer, you must ensure that the startup code is located at
the dedicated memory addresses, or can be accessed using a pointer from the vector
table. This means that startup code and the initial vector table must be placed in
non-volatile memory, such as ROM, EPROM, or flash.
A C/C++ application further needs to initialize all global variables. This initialization is
handled by the linker and the system startup code in conjunction. For more information,
see Application execution—an overview, page 44.

40 for AVR
AFE1_AFE2-1:1
Developing embedded applications
REAL-TIME OPERATING SYSTEMS

In many cases, the embedded application is the only software running in the system.
However, using an RTOS has some advantages.
For example, the timing of high-priority tasks is not affected by other parts of the
program which are executed in lower priority tasks. This typically makes a program
more deterministic and can reduce power consumption by using the CPU efficiently and
putting the CPU in a lower-power state when idle.
Using an RTOS can make your program easier to read and maintain, and in many cases
smaller as well. Application code can be cleanly separated in tasks which are truly
independent of each other. This makes teamwork easier, as the development work can
be easily split into separate tasks which are handled by one developer or a group of
developers.
Finally, using an RTOS reduces the hardware dependence and creates a clean interface
to the application, making it easier to port the program to different target hardware.
See also Managing a multithreaded environment, page 135.
The build process—an overview

This section gives an overview of the build process; how the various build tools—
compiler, assembler, and linker—fit together, going from source code to an executable
image.
To get familiar with the process in practice, you should run one or more of the tutorials
available from the IAR Information Center.
THE TRANSLATION PROCESS

There are two tools in the IDE that translate application source files to intermediary
object files. The IAR C/C++ Compiler and the IAR Assembler. Both produce
relocatable object files in the IAR UBROF format.
Note: The compiler can also be used for translating C source code into assembler source
code. If required, you can modify the assembler source code which then can be
assembled into object code. For more information about the IAR Assembler, see the IAR
Assembler User Guide for AVR.
41
AFE1_AFE2-1:1
The build process—an overview
This illustration shows the translation process:
After the translation, you can choose to pack any number of modules into an archive, or
in other words, a library. The important reason you should use libraries is that each
module in a library is conditionally linked in the application, or in other words, is only
included in the application if the module is used directly or indirectly by a module
supplied as an object file. Optionally, you can create a library; then use the IAR XAR
Library Builder or the IAR XLIB Librarian.
THE LINKING PROCESS

The relocatable modules, in object files and libraries, produced by the IAR compiler and
assembler cannot be executed as is. To become an executable application, they must be
linked.
The IAR XLINK Linker (xlink.exe) is used for building the final application.
Normally, the linker requires the following information as input:
● Several object files and possibly certain libraries
● The standard library containing the runtime environment and the standard language
functions
● A program start label (set by default)
● The linker configuration file that describes placement of code and data in the
memory of the target system
● Information about the output format.
The IAR XLINK Linker produces output according to your specifications. Choose
the output format that suits your purpose. You might want to load the output to a

42 for AVR
AFE1_AFE2-1:1
debugger—which means that you need output with debug information. Alternatively,
you might want to load output to a flash loader or a PROM programmer—in which
case you need output without debug information, such as Intel hex or Motorola S–
records. The option -F can be used for specifying the output format.
This illustration shows the linking process:
Note: The Standard C/C++ library contains support routines for the compiler, and the
implementation of the C/C++ standard library functions.
During the linking, the linker might produce error messages and logging messages on
stdout and stderr. The log messages are useful for understanding why an application
was linked the way it was, for example, why a module was included or a section
removed.
For more information about the procedure performed by the linker, see the IAR Linker
and Library Tools Reference Guide.
AFTER LINKING
The IAR XLINK Linker produces an absolute object file in the output format you
specify. After linking, the produced absolute executable image can be used for:
● Loading into the IAR C-SPY Debugger or any other compatible external debugger
that reads UBROF.
● Programming to a flash/PROM using a flash/PROM programmer.
43
AFE1_AFE2-1:1
Application execution—an overview
This illustration shows the possible uses of the absolute output files:

This section gives an overview of the execution of an embedded application divided into
three phases, the:
● Initialization phase
● Execution phase
● Termination phase.
THE INITIALIZATION PHASE

Initialization is executed when an application is started (the CPU is reset) but before the
main function is entered. The initialization phase can for simplicity be divided into:
● Hardware initialization, which generally at least initializes the stack pointer.

The hardware initialization is typically performed in the system startup code
cstartup.s90 and if required, by an extra low-level routine that you provide. It
might include resetting/starting the rest of the hardware, setting up the CPU, etc, in
preparation for the software C/C++ system initialization.
● Software C/C++ system initialization
Typically, this includes assuring that every global (statically linked) C/C++ symbol
receives its proper initialization value before the main function is called.
● Application initialization
This depends entirely on your application. It can include setting up an RTOS kernel
and starting initial tasks for an RTOS-driven application. For a bare-bone application,

44 for AVR
AFE1_AFE2-1:1
it can include setting up various interrupts, initializing communication, initializing

devices, etc.
For a ROM/flash-based system, constants and functions are already placed in ROM. All
symbols placed in RAM must be initialized before the main function is called. The
linker has already divided the available RAM into different areas for variables, stack,
heap, etc.
The following sequence of illustrations gives a simplified overview of the different
stages of the initialization.
1 When an application is started, the system startup code first performs hardware
initialization, such as initialization of the stack pointer to point at the end of the
predefined stack area
45
AFE1_AFE2-1:1
2 Then, memories that should be zero-initialized are cleared, in other words, filled with
zeros.
Typically, this is data referred to as zero-initialized data; variables declared as, for
example, int i = 0;
3 For initialized data, data declared, for example, like int i = 6; the initializers are
copied from ROM to RAM.

46 for AVR
AFE1_AFE2-1:1
4 Finally, the main function is called.
For more information about each stage, see System startup and termination, page 118.
For more information about initialization of data, see Initialization at system startup,
page 75.
Note: The AVR microcontroller is based on the Harvard architecture—thus code and
data have separate memory spaces and require different access mechanisms. For more
information, see the chapter Data storage.
THE EXECUTION PHASE

The software of an embedded application is typically implemented as a loop which is
either interrupt-driven or uses polling for controlling external interaction or internal
events. For an interrupt-driven system, the interrupts are typically initialized at the
beginning of the main function.
In a system with real-time behavior and where responsiveness is critical, a multi-task
system might be required. This means that your application software should be
complemented with a real-time operating system. In this case, the RTOS and the
different tasks must also be initialized at the beginning of the main function.
THE TERMINATION PHASE

Typically, the execution of an embedded application should never end. If it does, you
must define a proper end behavior.
47
AFE1_AFE2-1:1
Building applications—an overview
To terminate an application in a controlled way, either call one of the Standard C library
functions exit, _Exit, or abort, or return from main. If you return from main, the
exit function is executed, which means that C++ destructors for static and global
variables are called (C++ only) and all open files are closed.
Of course, in case of incorrect program logic, the application might terminate in an
uncontrolled and abnormal way—a system crash.
For more information about this, see System termination, page 121.
Building applications—an overview

In the command line interface, this line compiles the source file myfile.c into the
object file myfile.r90 using the default settings:
iccavr myfile.c
You must also specify some critical options, see Basic project configuration, page 48.
On the command line, this line can be used for starting the linker:
xlink myfile.r90 myfile2.r90 -o a.d90 -f my_configfile.xcl -r
In this example, myfile.r90 and myfile2.r90 are object files, and

my_configfile.xcl is the linker configuration file. The option -o specifies the name
of the output file. The option -r is used for specifying the output format UBROF, which
can be used for debugging in C-SPY®.
Note: By default, the label where the application starts is __program_start. You can
use the -s command line option to change this.
When building a project, the IAR Embedded Workbench IDE can produce extensive
build information in the Build messages window. This information can be useful, for
example, as a base for producing batch files for building on the command line. You can
copy the information and paste it in a text file. To activate extensive build information,
right-click in the Build messages window and select All on the context menu.
Basic project configuration

This section gives an overview of the basic settings for the project setup that are needed
to make the compiler and linker generate the best code for the AVR device you are using.
You can specify the options either from the command line interface or in the IDE.
You need to make settings for:
● Processor configuration
● Memory model

48 for AVR
AFE1_AFE2-1:1
● Size of double floating-point type

● Optimization settings
● Runtime environment, see Setting up the runtime environment, page 103
● Customizing the XLINK configuration, see the chapter Linking your application.
In addition to these settings, other options and settings can fine-tune the result even
further. For information about how to set options and for a list of all available options,
see the chapter Compiler options and the IDE Project Management and Building Guide
for AVR, respectively.
PROCESSOR CONFIGURATION
To make the compiler generate optimum code, you should configure it for the AVR
microcontroller you are using.
The --cpu option versus the -v option

There are two processor options that can be used for configuring the processor support:
--cpu=device and -vn
Your application may only use one processor option at a time, and the same processor
option must be used by all user and library modules to maintain consistency.
Both options set up default behavior—implicit assumptions—but note that the --cpu
option is more precise because it contains more information about the intended target
than the more generic -v option. The --cpu option knows, for example, how much flash
memory is available in the given target.
The --cpu=device option implicitly sets up all internal compiler settings needed to
generate code for the processor variant you are using. These options are implicitly
controlled when you use the --cpu option: --eecr_address, --eeprom_size,
--enhanced_core, --spmcr_address, -v and --64k_flash.
Because these options are automatically set when you use the --cpu option, you cannot
set them explicitly. For information about implicit assumptions when using the -v
option, see Summary of processor configuration for -v, page 50. For more information
about the generated code, see -v, page 275.
Use the --cpu or -v option to specify the AVR device; see the chapter Compiler options
for syntax information.
See the IDE Project Management and Building Guide for AVR for information about
setting project options in the IDE.
49
AFE1_AFE2-1:1
Summary of processor configuration for -v

This table summarizes the memory characteristics for each -v option:
Function Max module
Generic processor Available Max addressable
memory and/or program
option memory models data
attribute size
-v0 (default) Tiny __nearfunc <= 256 bytes <= 8 Kbytes
-v1 Tiny, Small __nearfunc <= 64 Kbytes <= 8 Kbytes
-v2 Tiny __nearfunc <= 256 bytes <= 128 Kbytes
-v3 Tiny, Small __nearfunc <= 64 Kbytes <= 128 Kbytes
-v4 Small, Large, Huge __nearfunc <= 16 Mbytes <= 128 Kbytes
-v5 Tiny, Small __farfunc* <= 64 Kbytes <= 8 Mbytes
-v6 Small, Large, Huge __farfunc* <= 16 Mbytes <= 8 Mbytes
Table 3: Summary of processor configuration
Note:
● *) When using the -v5 or the -v6 option, it is possible, for individual functions, to
override the __farfunc attribute and instead use the __nearfunc attribute
● Pointers with function memory attributes have restrictions in implicit and explicit
casts between pointers and between pointers and integer types. For details about the
restrictions, see Casting, page 289.
● -v2: There are currently no devices that match this processor option, which has
been added to support future devices
● All implicit assumptions for a given -v option are also true for corresponding
--cpu options.
It is important to be aware of the fact that the -v option does not reflect the amount of
used data, but the maximum amount of addressable data. This means that, for example,
if you are using a microcontroller with 16 Mbytes addressable data, but you are not
using more than 256 bytes or 64 Kbytes of data, you must still use either the -v4 or the
-v6 option for 16 Mbytes data.
SIZE OF DOUBLE FLOATING-POINT TYPE

Floating-point values are represented by 32- and 64-bit numbers in standard IEEE 754
format. If you use the compiler option --64bit_doubles, you can make the compiler
use 64-bit doubles. The data type float is always represented using 32 bits.

50 for AVR
AFE1_AFE2-1:1
OPTIMIZATION FOR SPEED AND SIZE

The compiler’s optimizer performs, among other things, dead-code elimination,
constant propagation, inlining, common sub-expression elimination, and precision
reduction. It also performs loop optimizations, such as induction variable elimination.
You can choose between several optimization levels, and for the highest level you can
choose between different optimization goals—size. Most optimizations will make the
application both smaller and faster. However, when this is not the case, the compiler uses
the selected optimization goal to decide how to perform the optimization.
The optimization level and goal can be specified for the entire application, for individual
files, and for individual functions. In addition, some individual optimizations, such as
function inlining, can be disabled.
For information about compiler optimizations and for more information about efficient
coding techniques, see the chapter Efficient coding for embedded applications.
51
AFE1_AFE2-1:1

52 for AVR
AFE1_AFE2-1:1
Data storage
● Introduction
● Memory types
● Memory models
● Storage of auto variables and parameters
● Dynamic memory on the heap
Introduction
The AVR microcontroller is based on the Harvard architecture—thus code and data have
separate memory spaces and require different access mechanisms. Code and different
types of data are located in memory spaces as follows:
● The internal flash space, which is used for code, __flash declared objects, and
initializers
● The data space, which can consist of external ROM, used for constants, and RAM
areas used for the stack, for registers, and for variables
● The EEPROM space, which is used for variables.
DIFFERENT WAYS TO STORE DATA

In a typical application, data can be stored in memory in three different ways:
● Auto variables
All variables that are local to a function, except those declared static, are stored either
in registers or on the stack. These variables can be used as long as the function
executes. When the function returns to its caller, the memory space is no longer valid.
For more information, see Storage of auto variables and parameters, page 60.
● Global variables, module-static variables, and local variables declared static
In this case, the memory is allocated once and for all. The word static in this context
means that the amount of memory allocated for this kind of variables does not change
while the application is running. For more information, see Memory types, page 54.
● Dynamically allocated data.
An application can allocate data on the heap, where the data remains valid until it is
explicitly released back to the system by the application. This type of memory is
53
AFE1_AFE2-1:1
Memory types
useful when the number of objects is not known until the application executes. Note
that there are potential risks connected with using dynamically allocated data in
systems with a limited amount of memory, or systems that are expected to run for a
long time. For more information, see Dynamic memory on the heap, page 61.
Memory types
This section describes the concept of memory types used for accessing data by the
compiler. It also discusses pointers in the presence of multiple memory types. For each
memory type, the capabilities and limitations are discussed.
INTRODUCTION TO MEMORY TYPES

The compiler uses different memory types to access data that is placed in different areas
of the memory. There are different methods for reaching memory areas, and they have
different costs when it comes to code space, execution speed, and register usage. The
access methods range from generic but expensive methods that can access the full
memory space, to cheap methods that can access limited memory areas. Each memory
type corresponds to one memory access method. If you map different memories—or
part of memories—to memory types, the compiler can generate code that can access
data efficiently.
For example, the memory accessed using the near memory access method is called
memory of near type, or simply near memory.
By selecting a memory model, you have selected a default memory type that your
application will use. However, it is possible to specify—for individual variables or
pointers—different memory types. This makes it possible to create an application that
can contain a large amount of data, and at the same time make sure that variables that
are used often are placed in memory that can be efficiently accessed.
USING DATA MEMORY ATTRIBUTES

The compiler provides a set of extended keywords, which can be used as data memory
attributes. These keywords let you override the default memory type for individual data
objects, which means that you can place data objects in other memory areas than the
default memory. This also means that you can fine-tune the access method for each
individual data object, which results in smaller code size.

54 for AVR
AFE1_AFE2-1:1
Data storage
This table summarizes the available memory types and their corresponding keywords:
Memory Pointer Memory Max object

Address range
attribute size space size
__tiny 1 byte Data 0x0-0xFF 127 bytes
__near 2 bytes Data 0x0-0xFFFF 32 Kbytes-1
__far 3 bytes Data 0x0-0xFFFFFF (16-bit pointer 32 Kbytes-1
arithmetics)
__huge 3 bytes Data 0x0-0xFFFFFF 16 Mbytes-1
__tinyflash 1 byte Code 0x0-0xFF 127 bytes
__flash 2 bytes Code 0x0-0xFFFF 32 Kbytes-1
__farflash 3 bytes Code 0x0-0xFFFFFF (16-bit pointer 32 Kbytes-1
arithmetics)
__hugeflash 3 bytes Code 0x0-0xFFFFFF 8 Mbytes-1
__eeprom 1 bytes EEPROM 0x0-0xFF 255 bytes
__eeprom 2 bytes EEPROM 0x0-0xFFFF 64 Kbytes-1
__io N/A I/O space 0x0–0x3F 4 bytes
__io N/A Data 0x60–0xFF 4 bytes
__ext_io N/A Data 0x100–0xFFFF 4 bytes
__generic 2 bytes Data or Code The most significant bit (MSB) 32 Kbytes-1
3 bytes determines whether this pointer 8 Mbytes-1
points to code space (1) or data
space (0). The small generic pointer
is generated for the processor
options -v0 and -v1.
__regvar N/A Data 0x4–0x0F 4 bytes
Table 4: Memory types and their corresponding memory attributes
The keywords are only available if language extensions are enabled in the compiler.
In the IDE, language extensions are enabled by default.
Use the -e compiler option to enable language extensions. See -e, page 250 for
additional information.
For more information about each keyword, see Descriptions of extended keywords, page
300.
55
AFE1_AFE2-1:1
Memory types
Syntax
The keywords follow the same syntax as the type qualifiers const and volatile. The
memory attributes are type attributes and therefore they must be specified both when
variables are defined and in the declaration, see General syntax rules for extended
keywords, page 295.
The following declarations place the variables i and j in EEPROM memory. The
variables k and l will also be placed in EEPROM memory. The position of the keyword
does not have any effect in this case:
__eeprom int i, j;
int __eeprom k, l;
Note that the keyword affects both identifiers. If no memory type is specified, the default
memory type is used.
The #pragma type_attribute directive can also be used for specifying the memory
attributes. The advantage of using pragma directives for specifying keywords is that it
offers you a method to make sure that the source code is portable. Refer to the chapter
Pragma directives for details about how to use the extended keywords together with
pragma directives.
Type definitions
Storage can also be specified using type definitions. These two declarations are
equivalent:
/* Defines via a typedef */
typedef char __far Byte;
typedef Byte *BytePtr;
Byte aByte;
BytePtr aBytePointer;
/* Defines directly */
__far char aByte;
char __far *aBytePointer;
POINTERS AND MEMORY TYPES

Pointers are used for referring to the location of data. In general, a pointer has a type.
For example, a pointer that has the type int * points to an integer.
In the compiler, a pointer also points to some type of memory. The memory type is
specified using a keyword before the asterisk. For example, a pointer that points to an
integer stored in far memory is declared by:
int __far * MyPtr;

56 for AVR
AFE1_AFE2-1:1
Data storage
Note that the location of the pointer variable MyPtr is not affected by the keyword. In
the following example, however, the pointer variable MyPtr2 is placed in near memory.
Like MyPtr, MyPtr2 points to a character in far memory.
char __far * __near MyPtr2;
Whenever possible, pointers should be declared without memory attributes. For

example, the functions in the standard library are all declared without explicit memory
types.
Differences between pointer types

A pointer must contain information needed to specify a memory location of a certain
memory type. This means that the pointer sizes are different for different memory types.
In the IAR C/C++ Compiler for AVR, the size of the near and far pointers are 16 and 24
bits, respectively.
In the compiler, it is illegal to convert pointers between different types without explicit
casts. For more information, see Casting, page 289.
For more information about pointers, see Pointer types, page 287.
STRUCTURES AND MEMORY TYPES

For structures, the entire object is placed in the same memory type. It is not possible to
place individual structure members in different memory types.
In the example below, the variable gamma is a structure placed in eeprom memory.
struct MyStruct
{
int mAlpha;
int mBeta;
};
__eeprom struct MyStruct gamma;
This declaration is incorrect:

struct MyStruct
{
int mAlpha;
__eeprom int mBeta; /* Incorrect declaration */
};
MORE EXAMPLES
The following is a series of examples with descriptions. First, some integer variables are
defined and then pointer variables are introduced. Finally, a function accepting a pointer
57
AFE1_AFE2-1:1
Memory types
to an integer in far memory is declared. The function returns a pointer to an integer in

eeprom memory. It makes no difference whether the memory attribute is placed before
or after the data type.
int MyA; A variable defined in default memory.

int __far MyB; A variable in far memory.
__eeprom int MyC; A variable in eeprom memory.
int * MyD; A pointer stored in default memory. The
pointer points to an integer in default memory.
int __far * MyE; A pointer stored in default memory. The
pointer points to an integer in far memory.
int __far * __eeprom MyF; A pointer stored in eeprom memory pointing to
an integer stored in far memory.
int __eeprom * MyFunction( A declaration of a function that takes a
int __far *); parameter which is a pointer to an integer
stored in far memory. The function returns a
pointer to an integer stored in eeprom memory.
C++ AND MEMORY TYPES

Instances of C++ classes are placed into a memory (just like all other objects) either
implicitly, or explicitly using memory type attributes or other IAR language extensions.
Non-static member variables, like structure fields, are part of the larger object and
cannot be placed individually into specified memories.
In non-static member functions, the non-static member variables of a C++ object can be
referenced via the this pointer, explicitly or implicitly. The this pointer is of the
default data pointer type unless class memory is used, see Using IAR attributes with
Classes, page 177.
Static member variables can be placed individually into a data memory in the same way
as free variables.
All member functions except for constructors and destructors can be placed individually
into a code memory in the same way as free functions.
For more information about C++ classes, see Using IAR attributes with Classes, page
177.

58 for AVR
AFE1_AFE2-1:1
Data storage
Memory models
Technically, the memory model specifies the default memory type attribute and the
default data pointer attribute. This means that the memory model controls the following:
● The default placement of static and global variables, and constant literals
● Dynamically allocated data, for example data allocated with malloc, or, in C++,
the operator new
● The default pointer type
● The placement of the runtime stack. For details about stack placement, see
CSTACK, page 378.
For information about how to specify a memory type for individual objects, see Using
data memory attributes, page 54.
SPECIFYING A MEMORY MODEL

Four memory models are implemented: Tiny, Small, Large, and Huge. These models are
controlled by the --memory_model option. Each model has a default memory type and
a default pointer size. The code size will also be reduced somewhat if the Tiny or Small
memory model is used.
If you do not specify a memory model option, the compiler will use the Tiny memory
model for all processor options, except for -v4 and -v6, where the Small memory
model will be used. For information about processor options, see Summary of processor
configuration for -v, page 50.
Your project can only use one memory model at a time, and the same model must be
used by all user modules and all library modules. However, you can override the default
memory type for individual data objects and pointers by explicitly specifying a memory
attribute, see Using data memory attributes, page 54.
This table summarizes the different memory models:
Default
Memory Default data Supported by
memory Max. stack range
model pointer processor option
attribute
Tiny __tiny __tiny <= 256 bytes in the range 0-FF -v0, -v1, -v2,
-v3, -v5
Small __near __near <= 64 Kbytes in the range 0-FFFF -v1, -v3, -v4,
-v5, -v6
Table 5: Memory model characteristics
59
AFE1_AFE2-1:1
Storage of auto variables and parameters
Default
Memory Default data Supported by
memory Max. stack range
model pointer processor option
attribute
Large __far __far <= 64 Kbytes in the range -v4, -v6
0-FFFFFF, but never cross a
64-Kbyte boundary
Huge __huge __huge <= 64 Kbytes in the range -v4, -v6
0-FFFFFF, but never cross a
64-Kbyte boundary
Table 5: Memory model characteristics (Continued)
See the IDE Project Management and Building Guide for AVR for information about
setting options in the IDE.
Use the --memory_model option to specify the memory model for your project; see
--memory_model, -m, page 258.
Storage of auto variables and parameters

Variables that are defined inside a function—and not declared static—are named auto
variables by the C standard. A few of these variables are placed in processor registers;
the rest are placed on the stack. From a semantic point of view, this is equivalent. The
main differences are that accessing registers is faster, and that less memory is required
compared to when variables are located on the stack.
Auto variables can only live as long as the function executes; when the function returns,
the memory allocated on the stack is released.
THE STACK
The stack can contain:
● Local variables and parameters not stored in registers
● Temporary results of expressions
● The return value of a function (unless it is passed in registers)
● Processor state during interrupts
● Processor registers that should be restored before the function returns (callee-save
registers).
The stack is a fixed block of memory, divided into two parts. The first part contains
allocated memory used by the function that called the current function, and the function
that called it, etc. The second part contains free memory that can be allocated. The
borderline between the two areas is called the top of stack and is represented by the stack

60 for AVR
AFE1_AFE2-1:1
Data storage
pointer, which is a dedicated processor register. Memory is allocated on the stack by

moving the stack pointer.
A function should never refer to the memory in the area of the stack that contains free
memory. The reason is that if an interrupt occurs, the called interrupt function can
allocate, modify, and—of course—deallocate memory on the stack.
See also Stack considerations, page 193 and Setting up stack memory, page 91.
Advantages
The main advantage of the stack is that functions in different parts of the program can
use the same memory space to store their data. Unlike a heap, a stack will never become
fragmented or suffer from memory leaks.
It is possible for a function to call itself either directly or indirectly—a recursive
function—and each invocation can store its own data on the stack.
Potential problems
The way the stack works makes it impossible to store data that is supposed to live after
the function returns. The following function demonstrates a common programming
mistake. It returns a pointer to the variable x, a variable that ceases to exist when the
function returns.
int *MyFunction()
{
int x;
/* Do something here. */
return &x; /* Incorrect */
}
Another problem is the risk of running out of stack space. This will happen when one
function calls another, which in turn calls a third, etc., and the sum of the stack usage of
each function is larger than the size of the stack. The risk is higher if large data objects
are stored on the stack, or when recursive functions are used.
Dynamic memory on the heap

Memory for objects allocated on the heap will live until the objects are explicitly
released. This type of memory storage is very useful for applications where the amount
of data is not known until runtime.
In C, memory is allocated using the standard library function malloc, or one of the
related functions calloc and realloc. The memory is released again using free.
61
AFE1_AFE2-1:1
Dynamic memory on the heap
In C++, a special keyword, new, allocates memory and runs constructors. Memory
allocated with new must be released using the keyword delete.
The compiler supports heaps in more than one memory types:
Used by default in
Memory type Memory attribute Section name
memory model
tiny __tiny TINY_HEAP Tiny
near __near NEAR_HEAP Small
far __far FAR_HEAP Large
huge __huge HUGE_HEAP Huge
Table 6: Heaps supported in memory types
In DLIB, to use a specific heap, add the memory attribute in front of malloc, free,
alloc, and realloc, for example __near_malloc. The default functions will use of
the specific heap variants, depending on project settings such as data model.
For information about how to set up the size for heap memory, see Setting up heap
memory, page 92.
POTENTIAL PROBLEMS
Applications that use heap-allocated data objects must be very carefully designed,
because it is easy to end up in a situation where it is not possible to allocate objects on
the heap.
The heap can become exhausted if your application uses too much memory. It can also
become full if memory that no longer is in use was not released.
For each allocated memory block, a few bytes of data for administrative purposes is
required. For applications that allocate a large number of small blocks, this
administrative overhead can be substantial.
There is also the matter of fragmentation; this means a heap where small sections of free
memory is separated by memory used by allocated objects. It is not possible to allocate
a new object if no piece of free memory is large enough for the object, even though the
sum of the sizes of the free memory exceeds the size of the object.
Unfortunately, fragmentation tends to increase as memory is allocated and released. For
this reason, applications that are designed to run for a long time should try to avoid using
memory allocated on the heap.

62 for AVR
AFE1_AFE2-1:1
Functions
● Function-related extensions
● Function storage
● Primitives for interrupts, concurrency, and OS-related programming
● Inlining functions
Function-related extensions
In addition to supporting Standard C, the compiler provides several extensions for
writing functions in C. Using these, you can:
● Control the storage of functions in memory
● Use primitives for interrupts, concurrency, and OS-related programming
● Control function inlining
● Facilitate function optimization
● Access hardware features.
The compiler uses compiler options, extended keywords, pragma directives, and
intrinsic functions to support this.
For more information about optimizations, see Efficient coding for embedded
applications, page 205. For information about the available intrinsic functions for
accessing hardware operations, see the chapter Intrinsic functions.
Function storage
USING FUNCTION MEMORY ATTRIBUTES
It is possible to override the default placement for individual functions. Use the
appropriate function memory attribute to specify this. These attributes are available:
Memory attribute Address range Pointer size Used in processor option

__nearfunc 0-0x1FFFE (128 Kbytes) 16 bits -v0, -v1, -v2, -v3, -v4
__farfunc 0-0x7FFFFE (8 Mbytes) 24 bits -v5, -v6
Table 7: Function memory attributes
63
AFE1_AFE2-1:1
Primitives for interrupts, concurrency, and OS-related programming
When using the -v5 or the -v6 option, it is possible, for individual functions, to override
the __farfunc attribute and instead use the __nearfunc attribute. The default
memory can be overridden by explicitly specifying a memory attribute in the function
declaration or by using the #pragma type_attribute directive:
#pragma type_attribute=__nearfunc
void MyFunc(int i)
{
...
}
It is possible to call a __nearfunc function from a __farfunc function and vice versa.
Only the size of the function pointer is affected.
It is possible to place functions into named segments using either the @ operator or the
#pragma location directive. For more information, see Controlling data and function
placement in memory, page 210.
Pointers with function memory attributes have restrictions in implicit and explicit casts
between pointers and between pointers and integer types. For information about the
restrictions, see Casting, page 289.
For syntax information and for more information about each attribute, see the chapter
Extended keywords.

The IAR C/C++ Compiler for AVR provides the following primitives related to writing
interrupt functions, concurrent functions, and OS-related functions:
● The extended keywords: __interrupt, __nested, __task, __monitor
● The pragma directive #pragma vector
● The intrinsic functions: __enable_interrupt, __disable_interrupt.
INTERRUPT FUNCTIONS
In embedded systems, using interrupts is a method for handling external events
immediately; for example, detecting that a button was pressed.
Interrupt service routines

In general, when an interrupt occurs in the code, the microcontroller immediately stops
executing the code it runs, and starts executing an interrupt routine instead. It is
important that the environment of the interrupted function is restored after the interrupt
is handled (this includes the values of processor registers and the processor status

64 for AVR
AFE1_AFE2-1:1
Functions
register). This makes it possible to continue the execution of the original code after the
code that handled the interrupt was executed.
The AVR microcontroller supports many interrupt sources. For each interrupt source, an
interrupt routine can be written. Each interrupt routine is associated with a vector
number, which is specified in the AVR microcontroller documentation from the chip
manufacturer. If you want to handle several different interrupts using the same interrupt
routine, you can specify several interrupt vectors.
Interrupt vectors and the interrupt vector table

For the AVR microcontroller, the interrupt vector table always starts at the address 0x0
and is placed in the INTVEC segment. The interrupt vector is the offset into the interrupt
vector table. The interrupt vector table contains pointers to interrupt routines, including
the reset routine. The AT90S80515 device has 13 interrupt vectors and one reset vector.
For this reason, you should specify 14 interrupt vectors, each of two bytes.
If a vector is specified in the definition of an interrupt function, the processor interrupt
vector table is populated. It is also possible to define an interrupt function without a
vector. This is useful if an application is capable of populating or changing the interrupt
vector table at runtime.
Defining an interrupt function—an example

To define an interrupt function, the __interrupt keyword and the #pragma vector
directive can be used. For example:
#pragma vector = 0x14
__interrupt void MyInterruptRoutine(void)
{
/* Do something */
}
Note: An interrupt function must have the return type void, and it cannot specify any
parameters.
Interrupt and C++ member functions

Only static member functions can be interrupt functions.
MONITOR FUNCTIONS
A monitor function causes interrupts to be disabled during execution of the function. At
function entry, the status register is saved and interrupts are disabled. At function exit,
the original status register is restored, and thereby the interrupt status that existed before
the function call is also restored.
65
AFE1_AFE2-1:1
To define a monitor function, you can use the __monitor keyword. For more
information, see __monitor, page 308.
Avoid using the __monitor keyword on large functions, since the interrupt will
otherwise be turned off for too long.
Example of implementing a semaphore in C

In the following example, a binary semaphore—that is, a mutex—is implemented using
one static variable and two monitor functions. A monitor function works like a critical
region, that is no interrupt can occur and the process itself cannot be swapped out. A
semaphore can be locked by one process, and is used for preventing processes from
simultaneously using resources that can only be used by one process at a time, for
example a USART. The __monitor keyword assures that the lock operation is atomic;
in other words it cannot be interrupted.
/* This is the lock-variable. When non-zero, someone owns it. */
static volatile unsigned int sTheLock = 0;
/* Function to test whether the lock is open, and if so take it.

* Returns 1 on success and 0 on failure.
*/
__monitor int TryGetLock(void)

{
if (sTheLock == 0)
{
/* Success, nobody has the lock. */
sTheLock = 1;
return 1;
}
else
{
/* Failure, someone else has the lock. */
return 0;
}
}

66 for AVR
AFE1_AFE2-1:1
Functions
/* Function to unlock the lock.

* It is only callable by one that has the lock.
*/
__monitor void ReleaseLock(void)

{
sTheLock = 0;
}
/* Function to take the lock. It will wait until it gets it. */
void GetLock(void)
{
while (!TryGetLock())
{
/* Normally, a sleep instruction is used here. */
}
}
/* An example of using the semaphore. */
void MyProgram(void)
{
GetLock();
ReleaseLock();
}
Inlining functions
Function inlining means that a function, whose definition is known at compile time, is
integrated into the body of its caller to eliminate the overhead of the function call. This
optimization, which is performed at optimization level High, normally reduces
execution time, but might increase the code size. The resulting code might become more
difficult to debug. Whether the inlining actually occurs is subject to the compiler’s
heuristics.
The compiler heuristically decides which functions to inline. Different heuristics are
used when optimizing for speed, size, or when balancing between size and speed.
Normally, code size does not increase when optimizing for size.
67
AFE1_AFE2-1:1
Inlining functions
C VERSUS C++ SEMANTICS

In C++, all definitions of a specific inline function in separate translation units must be
exactly the same. If the function is not inlined in one or more of the translation units,
then one of the definitions from these translation units will be used as the function
implementation.
In C, you must manually select one translation unit that includes the non-inlined version
of an inline function. You do this by explicitly declaring the function as extern in that
translation unit. If you declare the function as extern in more than one translation unit,
the linker will issue a multiple definition error. In addition, in C, inline functions cannot
refer to static variables or functions.
For example:
// In a header file.
static int sX;
inline void F(void)
{
//static int sY; // Cannot refer to statics.
//sX; // Cannot refer to statics.
}
// In one source file.

// Declare this F as the non-inlined version to use.
extern inline void F();
FEATURES CONTROLLING FUNCTION INLINING

There are several mechanisms for controlling function inlining:
● The inline keyword advises the compiler that the function defined immediately
after the directive should be inlined.
If you compile your function in C or C++ mode, the keyword will be interpreted
according to its definition in Standard C or Standard C++, respectively.
The main difference in semantics is that in Standard C you cannot (in general) simply
supply an inline definition in a header file. You must supply an external definition in
one of the compilation units, by designating the inline definition as being external in
that compilation unit.
● #pragma inline is similar to the inline keyword, but with the difference that the
compiler always uses C++ inline semantics.
By using the #pragma inline directive you can also disable the compiler’s
heuristics to either force inlining or completely disable inlining. For more
information, see inline, page 329.

68 for AVR
AFE1_AFE2-1:1
Functions
● --use_c++_inline forces the compiler to use C++ semantics when compiling a

Standard C source code file.
● --no_inline, #pragma optimize=no_inline, and #pragma inline=never
all disable function inlining. By default, function inlining is enabled at optimization
level High.
The compiler can only inline a function if the definition is known. Normally, this is
restricted to the current translation unit. However, when the --mfc compiler option for
multi-file compilation is used, the compiler can inline definitions from all translation
units in the multi-file compilation unit. For more information, see Multi-file compilation
units, page 214.
For more information about the function inlining optimization, see Function inlining,
page 217.
69
AFE1_AFE2-1:1
Inlining functions

70 for AVR
AFE1_AFE2-1:1
Linking overview
● Linking—an overview
● Segments and memory
● The linking process in detail
● Placing code and data—the linker configuration file
● Initialization at system startup
● Stack usage analysis
Linking—an overview
The IAR XLINK Linker is a powerful, flexible software tool for use in the development
of embedded applications. It is equally well suited for linking small, single-file, absolute
assembler programs as it is for linking large, relocatable, multi-module, C/C++, or
mixed C/C++ and assembler programs.
The linker combines one or more relocatable object files—produced by the IAR
Systems compiler or assembler—with required parts of object libraries to produce an
executable image containing machine code for the microcontroller you are using.
XLINK can generate more than 30 industry-standard loader formats, in addition to the
proprietary format UBROF which is used by the C-SPY debugger.
The linker will automatically load only those library modules that are actually needed
by the application you are linking. Further, the linker eliminates segment parts that are
not required. During linking, the linker performs a full C-level type checking across all
modules.
The linker uses a configuration file where you can specify separate locations for code
and data areas of your target system memory map.
The final output produced by the linker is an absolute, target-executable object file that
can be downloaded to the microcontroller, to C-SPY, or to a compatible hardware
debugging probe. Optionally, the output file can contain debug information depending
on the output format you choose.
To handle libraries, the library tools XAR and XLIB can be used.
71
AFE1_AFE2-1:1
Segments and memory
Segments and memory

In an embedded system, there might be many different types of physical memory. Also,
it is often critical where parts of your code and data are located in the physical memory.
For this reason it is important that the development tools meet these requirements.
WHAT IS A SEGMENT?
A segment is a container for pieces of data or code that should be mapped to a location
in physical memory. Each segment consists of one or more segment parts. Normally,
each function or variable with static storage duration is placed in its own segment part.
A segment part is the smallest linkable unit, which allows the linker to include only
those segment parts that are referred to. A segment can be placed either in RAM or in
ROM. Segments that are placed in RAM generally do not have any content, they only
occupy space.
Note: Here, ROM memory means all types of read-only memory, including flash
memory.
The compiler uses several predefined segments for different purposes. Each segment is
identified by a name that typically describes the contents of the segment, and has a
segment memory type that denotes the type of content. In addition to the predefined
segments, you can also define your own segments.
At compile time, the compiler assigns code and data to the various segments. The IAR
XLINK Linker is responsible for placing the segments in the physical memory range, in
accordance with the rules specified in the linker configuration file. Ready-made linker
configuration files are provided, but, if necessary, they can be modified according to the
requirements of your target system and application. It is important to remember that,
from the linker's point of view, all segments are equal; they are simply named parts of
memory.
Segment memory type

Each segment always has an associated segment memory type. In some cases, an
individual segment has the same name as the segment memory type it belongs to, for
example CODE. Make sure not to confuse the segment name with the segment memory
type in those cases.
By default, the compiler uses these XLINK segment memory types:
Segment memory type Used for
CODE Executable code
XDATA Data placed in EEPROM
Table 8: XLINK segment memory types

72 for AVR
AFE1_AFE2-1:1
Linking overview
Segment memory type Used for

DATA Data placed in RAM
Table 8: XLINK segment memory types (Continued)
XLINK supports more segment memory types than the ones described above. However,
they exist to support other types of microcontrollers.
For more information about individual segments, see the chapter Segment reference.
The linking process in detail

The relocatable modules in object files and libraries, produced by the IAR compiler and
assembler, cannot be executed as is. To make an application executable, the object files
must be linked.
The IAR XLINK Linker is used for the link process. It normally performs the following
procedure (note that some of the steps can be turned off by command line options or by
directives in the linker configuration file):
● Determines which modules to include in the application. Program modules are
always included. Library modules are only included if they provide a definition for
a global symbol that is referenced from an included module. If the object files
containing library modules contain multiple definitions of variables or functions,
only the first definition will be included. This means that the linking order of the
object files is important.
● Determines which segment parts from the included modules to include in the
application. Only those segments that are actually needed by the application are
included. There are several ways to determine of which segment parts that are
needed, for example, the __root object attribute, the #pragma required
directive, and the -g linker option.
● Divides each segment that will be initialized by copying into two segments, one for
the ROM part and one for the RAM part. The RAM part contains the label and the
ROM part the actual bytes. The bytes are conceptually linked as residing in RAM.
● Determines where to place each segment according to the segment placement
directives in the linker configuration file.
● Produces an absolute file that contains the executable image and any debug
information. The contents of each needed segment in the relocatable input files is
calculated using the relocation information supplied in its file and the addresses
determined when placing segments. This process can result in one or more range
errors if some of the requirements for a particular segment are not met, for instance
if placement resulted in the destination address for a PC-relative jump instruction
being out of range for that instruction.
73
AFE1_AFE2-1:1
Placing code and data—the linker configuration file
● Optionally, produces a map file that lists the result of the segment placement, the
address of each global symbol, and finally, a summary of memory usage for each
module.
This illustration shows the linking process:
During the linking, XLINK might produce error messages and optionally a map file. In
the map file you can see the result of the actual linking and is useful for understanding
why an application was linked the way it was, for example, why a segment part was
included. If a segment part is not included although you expect it to be, the reason is
always that the segment part was not referenced to from an included part.
Note: To inspect the actual content of the object files, use XLIB. See the IAR Linker and
Library Tools Reference Guide.
Placing code and data—the linker configuration file

The placement of segments in memory is performed by the IAR XLINK Linker. It uses
a linker configuration file that contains command line options which specify the
locations where the segments can be placed, thereby assuring that your application fits
on the target microcontroller. To use the same source code with different devices, just
rebuild the code with the appropriate linker configuration file.

74 for AVR
AFE1_AFE2-1:1
Linking overview
In particular, the linker configuration file specifies:

● The placement of segments in memory
● The maximum stack size
● The maximum heap size.
The file consists of a sequence of linker commands. This means that the linking process
will be governed by all commands in sequence.
THE CONTENTS OF THE LINKER CONFIGURATION FILE

Among other things, the linker configuration file contains three different types of
XLINK command line options:
● The CPU used:
-ca90
This specifies your target microcontroller.
● Definitions of constants used in the file. These are defined using the XLINK option
-D. Symbols defined using -D can also be accessed from your application.
● The placement directives (the largest part of the linker configuration file). Segments
can be placed using the -Z and -P options. The former will place the segment parts
in the order they are found, while the latter will try to rearrange them to make better
use of the memory. The -P option is useful when the memory where the segment
should be placed is not continuous.
In the linker configuration file, numbers are generally specified in hexadecimal format.
However, neither the prefix 0x nor the suffix h is necessarily used.
Note: The supplied linker configuration file includes comments explaining the contents.
For more information about the linker configuration file and how to customize it, see
Linking considerations, page 87.
See also the IAR Linker and Library Tools Reference Guide.
Initialization at system startup

In Standard C, all static variables—variables that are allocated at a fixed memory
address—must be initialized by the runtime system to a known value at application
startup. Static variables can be divided into these categories:
● Variables that are initialized to a non-zero value
● Variables that are initialized to zero
75
AFE1_AFE2-1:1
● Variables that are located by use of the @ operator or the #pragma location
directive
● Variables that are declared as const and therefore can be stored in ROM
● Variables defined with the __no_init keyword, meaning that they should not be
initialized at all.
STATIC DATA MEMORY SEGMENTS

The compiler generates a specific type of segment for each type of variable
initialization.
The names of the segments consist of two parts—the segment group name and a suffix—
for instance, NEAR_Z. There is a segment group for each memory type, where each
segment in the group holds different categories of declared data. The names of the
segment groups are derived from the memory type and the corresponding keyword, for
example NEAR and __near.
Some of the declared data is placed in non-volatile memory, for example ROM/flash,
and some of the data is placed in RAM. For this reason, it is also important to know the
XLINK segment memory type of each segment. For more information about segment
memory types, see Segment memory type, page 72.
This table summarizes the different suffixes, which XLINK segment memory type they
are, and which category of declared data they denote:
Categories of Segment
Source Segment type Segment name*
declared data content
Zero-initialized int i; Read/write MEMATTR_Z None
data data, zero-init
Zero-initialized int i = 0; Read/write MEMATTR_Z None
data data, zero-init
Initialized data int i = 6; Read/write MEMATTR_I None
(non-zero) data
Initialized data int i = 6; Read-only data MEMATTR_ID Initializer data
(non-zero) for
MEMATTR_I
Constants const int i = 6; Read-only data MEMATTR_C The constant
Non-initialized __no_init int i; Read/write MEMATTR_N None
data data
Data placed in __memattr int i Read-only data MEMATTR_F The constant
flash memory = 6;
Table 9: segments holding initialized data

76 for AVR
AFE1_AFE2-1:1
Linking overview
* The actual segment group name—MEMATTR—depends on the memory where the

variable is placed. See Memory types, page 54.
For more information about each segment, see the chapter Segment reference.
THE INITIALIZATION PROCESS

Initialization of data is handled by the system startup code. If you add more segments,
you must update the system startup code accordingly.
To configure the initialization of variables, you must consider these issues:
● Segments that should be zero-initialized should only be placed in RAM.
● Segments that should be initialized, except for zero-initialized segments:
The system startup code initializes the non-zero variables by copying a block of
ROM to the location of the variables in RAM. This means that the data in the ROM
segment with the suffix ID is copied to the corresponding I segment.
This works when both segments are placed in continuous memory. However, if one
of the segments is divided into smaller pieces, it is very important that:
● The other segment is divided in exactly the same way
● It is legal to read and write the memory that represents the gaps in the sequence.
● Segments that contain constants do not need to be initialized; they should only be
placed in flash/ROM
● Segments holding __no_init declared variables should not be initialized.
● Finally, global C++ object constructors are called.
For more information about and examples of how to configure the initialization, see
Linking considerations, page 87.
Initialization of local aggregates at function invocation

Initialized aggregate auto variables—struct, union, and array variables local to a
function—have the initial values in blocks of memory. As an auto variable is allocated
either in registers or on the stack, the initialization has to take place every time the
function is called. Assume the following example:
void f()
{
struct block b = { 3, 4, 2, 3, 6634, 234 };
...
}
The initializers are copied to the b variable allocated on the stack each time the function
is entered.
77
AFE1_AFE2-1:1
The initializers can either come from the code memory space (flash) or from the data
memory space (optional external ROM). By default, the initializers are located in
segments with the suffix _C and these segments are copied from external ROM to the
stack.
If you use either the -y option or the --initializers_in_flash option, the
aggregate initializers are located in segments with the suffix _F, which are copied from
flash memory to the stack. The advantage of storing these initializers in flash is that
valuable data space is not wasted. The disadvantage is that copying from flash is slower.
Initialization of constant objects

There are different ways of initializing constant objects.
By default, constant objects are placed in segments with the suffix _C, which are located
in the optional external ROM that resides in the data memory space. The reason for this
is that it must be possible for a default pointer—a pointer without explicit memory
attributes—to point to the object, and a default pointer can only point to the data
memory space.
However, if you do not have any external ROM in the data memory space, and for single
chip applications you most likely do not have it, the constant objects have to be placed
in RAM and initialized as any other non-constant variables. To achieve this, use the -y
option, which means the objects are placed in segments with the suffix _ID.
If you want to place an object in flash memory, you can use any of the memory attributes
__tinyflash, __flash, __farflash, or __hugeflash. The object becomes a flash
object, which means you cannot take the address of it and store it in a default pointer.
However, it is possible to store the address in either a __flash pointer or a __generic
pointer, though neither of these are default pointers. Note that if you attempt to take the
address of a constant __flash object and use it as a default pointer object, the compiler
will issue an error. If you make an explicit cast of the object to a default pointer object,
the error message disappears, instead there will be problems at runtime as the cast
cannot copy the object from the flash memory to the data memory.
To access strings located in flash memory, you must use alternative library routines that
expect flash strings. A few such alternative functions are provided—they are declared in
the pgmspace.h header file. They are flash alternatives for some common C library
functions with an extension _P. For your own code, you can always use the __flash
keyword when passing the strings between functions. For reference information about
the alternative functions, see AVR-specific library functions, page 370.

78 for AVR
AFE1_AFE2-1:1
Linking overview
Stack usage analysis

This section describes how to perform a stack usage analysis using the linker.
In the AVR\src directory, you can find an example project that demonstrates stack usage
analysis.
INTRODUCTION TO STACK USAGE ANALYSIS

Under the right circumstances, the linker can accurately calculate the maximum stack
usage for each call graph, starting from the program start, interrupt functions, tasks etc.
(each function that is not called from another function, in other words, the root).
If you enable stack usage analysis, a stack usage chapter will be added to the linker map
file, listing for each call graph root the particular call chain which results in the
maximum stack depth.
The analysis is only accurate if there is accurate stack usage information for each
function in the application.
In general, the compiler will generate this information for each C function, but if there
are indirect calls (calls using function pointers) in your application, you must supply a
list of possible functions that can be called from each calling function.
If you use a stack usage control file, you can also supply stack usage information for
functions in modules that do not have stack usage information.
You can use the check that directive in your stack usage control file to check that the
stack usage calculated by the linker does not exceed the stack space you have allocated.
PERFORMING A STACK USAGE ANALYSIS

1 Enable stack usage analysis:
In the IDE, choose Project>Options>Linker>Advanced>Enable stack usage
analysis
On the command line, use the linker option --enable_stack_usage
See the IAR Linker and Library Tools Reference Guide for information about the linker
option --enable_stack_usage.
2 Enable the linker map file:
In the IDE, choose Project>Options>Linker>List>Generate linker listing
On the command line, use the linker option -l
79
AFE1_AFE2-1:1
3 Link your project. Note that the linker will issue warnings related to stack usage under
certain circumstances, see Situations where warnings are issued, page 84.
4 Review the linker map file, which now contains a stack usage chapter with a summary
of the stack usage for each call graph root. For more information, see Result of an
analysis—the map file contents, page 80.
5 For more details, analyze the call graph log, see Call graph log, page 84.
Note that there are limitations and sources of inaccuracy in the analysis, see Limitations,
page 83.
You might need to specify more information to the linker to get a more representative
result. See Specifying additional stack usage information, page 82
In the IDE, choose Project>Options>Linker>Advanced>Enable stack usage
analysis>Control file
On the command line, use the linker option --stack_usage_control
See the IAR Linker and Library Tools Reference Guide for information about the linker
option --stack_usage_control.
6 To add an automatic check that you have allocated memory enough for the stack, use
the check that directive in your stack usage control file. For example, assuming a
stack segment named MY_STACK, you can write like this:
check that size("MY_STACK") >=maxstack("Program entry",
CSTACK) + totalstack("interrupt", CSTACK) + 100;
or
check that size("MY_STACK") >=maxstack("Program entry",
RSTACK) + totalstack("interrupt", RSTACK) + 100;
When linking, the linker emits an error if the check fails. In this example, an error will
be emitted if the sum of the following exceeds the size of the MY_STACK segment:
● The maximum stack usage in the category Program entry (the main program).
● The sum of each individual maximum stack usage in the category interrupt
(assuming that all interrupt routines need space at the same time).
● A safety margin of 100 bytes (to account for stack usage not visible to the analysis).
See also check that directive, page 392 and Stack considerations, page 193.
RESULT OF AN ANALYSIS—THE MAP FILE CONTENTS

When stack usage analysis is enabled, the linker map file contains a stack usage chapter
with a summary of the stack usage for each call graph root category, and lists the call

80 for AVR
AFE1_AFE2-1:1
Linking overview
chain that results in the maximum stack depth for each call graph root. This is an
example of what the stack usage chapter in the map file might look like:
*****************************************************
* *
* STACK USAGE ANALYSIS *
* *
*****************************************************
Call Graph Root Category Max Use Total Use

------------------------ ------- ---------
interrupt 4 4
Program entry 350 350
Program entry
"__program_start": 0x0000c0f8
Maximum call chain 350 bytes
"__program_start" 0
"main" 4
"WriteObject" 24
"DumpObject" 0
"PrintObject" 8
"fprintf" 4
"_PrintfLarge" 126
"_PutstrLarge" 100
"pad" 14
"_PutcharsLarge" 10
"_FProut" 6
"fputc" 6
"_Fwprep" 6
"fseek" 4
"_Fspos" 14
"fflush" 6
"fflushOne" 6
"__write" 0
"__dwrite" 10
"__DebugBreak" 2
interrupt
"DoStuff()": 0x0000e9ee
Maximum call chain 4 bytes
"DoStuff()" 4
81
AFE1_AFE2-1:1
The summary contains the depth of the deepest call chain in each category as well as the
sum of the depths of the deepest call chains in that category.
Each call graph root belongs to a call graph root category to enable convenient
calculations in check that directives.
SPECIFYING ADDITIONAL STACK USAGE INFORMATION

To specify additional stack usage information you can use either a stack usage control
file (suc) where you specify stack usage control directives or annotate the source code.
You can:
● Specify complete stack usage information (call graph root category, stack usage,
and possible calls) for a function, by using the stack usage control directive
function. Typically, you do this if stack usage information is missing, for example
in an assembler module. In your suc file you can for example write like this:
function MyFunc: (CSTACK 32, RSTACK 4),
calls MyFunc2,
calls MyFunc3, MyFunc4: (CSTACK 16, RSTACK 4);
function [interrupt] MyInterruptHandler: 44;

See also function directive, page 393.
● Exclude certain functions from stack usage analysis, by using the stack usage
control directive exclude. In your suc file you can for example write like this:
exclude MyFunc5, MyFunc6;
See also exclude directive, page 393.
● Specify a list of possible destinations for indirect calls in a function, by using the
stack usage control directive possible calls. Use this for functions which are
known to perform indirect calls and where you know exactly which functions that
might be called in this particular application. In your suc file you can for example
write like this:
possible calls MyFunc7: MyFunc8, MyFunc9;
If the information about which functions that might be called is available at compile
time, consider using the #pragma calls directive instead.
See also possible calls directive, page 394 and calls, page 322.
● Specify that functions are call graph roots, including an optional call graph root
category, by using the stack usage control directive call graph root or the
#pragma call_graph_root directive. In your suc file you can for example write
like this:
call graph root [task]: MyFunc10, MyFunc11;

82 for AVR
AFE1_AFE2-1:1
Linking overview
If your interrupt functions have not already been designated as call graph roots by the
compiler, you must do so manually. You can do this either by using the #pragma
call_graph_root directive in your source code or by specifying a directive in your
suc file, for example:
call graph root [interrupt]: Irq1Handler, Irq2Handler;
See also call graph root directive, page 391 and call_graph_root, page 322.
● Specify a maximum number of iterations through any of the cycles in the recursion
nest of which the function is a member. In your suc file you can for example write
like this:
max recursion depth MyFunc12: 10;
● Selectively suppress the warning about unmentioned functions referenced by a
module for which you have supplied stack usage information in the stack usage
control file. Use the no calls from directive in your suc file, for example like
this:
no calls from [file.r90] to MyFunc13, MyFunc14;
● Instead of specifying stack usage information about assembler modules in a stack
usage control file, you can annotate the assembler source with call frame
information. For more information, see the IAR Assembler User Guide for AVR.
For more information, see the chapter The stack usage control file, page 391.
LIMITATIONS
Apart from missing or incorrect stack usage information, there are also other sources of
inaccuracy in the analysis:
● The linker cannot always identify all functions in object modules that lack stack
usage information. In particular, this might be a problem with object modules
written in assembly language. You can provide stack usage information for such
modules using a stack usage control file, and for assembly language modules you
can also annotate the assembler source code with CFI directives to provide stack
usage information. See the IAR Assembler User Guide for AVR.
● If you use inline assembler to change the frame size or to perform function calls,
this will not be reflected in the analysis.
● Extra space consumed by other sources (the processor, an operating system, etc) is
not accounted for.
● The support for C++ is very incomplete. In particular virtual function calls, dynamic
objects, and various indirect calls from the standard library (like the functions set up
by set_new_handler and register_callback) are not supported at all.
● If you use other forms of function calls, they will not be reflected in the call graph.
83
AFE1_AFE2-1:1
● Using multi-file compilation (--mfc) can interfere with using a stack usage control
file to specify properties of module-local functions in the involved files.
Note that stack usage analysis produces a worst case result. The program might not
actually ever end up in the maximum call chain, by design, or by coincidence.
Stack usage analysis is only a complement to actual measurement. If the result is
important, you need to perform independent validation of the results of the analysis.
SITUATIONS WHERE WARNINGS ARE ISSUED

When stack usage analysis is enabled in the linker, warnings will be generated in the
following circumstances:
● There is a function without stack usage information.
● There is an indirect call site in the application for which a list of possible called
functions has not been supplied.
● There are no known indirect calls, but there is an uncalled function that is not
known to be a call graph root.
● The application contains recursion (a cycle in the call graph) for which no
maximum recursion depth has been supplied, or which is of a form for which the
linker is unable to calculate a reliable estimate of stack usage.
● There are calls to a function declared as a call graph root.
● You have used the stack usage control file to supply stack usage information for
functions in a module that does not have such information, and there are functions
referenced by that module which have not been mentioned as being called in the
stack usage control file.
CALL GRAPH LOG

To help you interpret the results of the stack usage analysis, there is a log output option
that produces a simple text representation of the call graph (--log stack_usage).

84 for AVR
AFE1_AFE2-1:1
Linking overview
Example output:
Program entry:
0 __program_start [350]
0 __data16_memzero [2]
2 - [0]
0 __data16_memcpy [2]
0 memcpy [2]
2 - [0]
2 - [0]
0 main [350]
4 ParseObject [52]
28 GetObject [28]
34 getc [22]
38 _Frprep [18]
44 malloc [12]
44 __data16_malloc [12]
48 __data16_findmem [8]
52 __data16_free [4]
56 - [0]
52 __data16GetMemChunk [2]
54 - [0]
46 - [0]
44 __read [12]
54 __DebugBreak [2]
56 - [0]
36 - [0]
34 CreateObject [18]
40 malloc [12] ***
4 ProcessObject [326]
8 ProcessHigh [76]
34 ProcesMedium [50]
60 ProcessLow [24]
84 - [0]
8 DumpObject [322]
8 PrintObject [322]
16 fprintf [314]
20 _PrintfLarge [310]
10 - [0]
4 WriteObject [346]
28 DumpObject [322] ***
4 DestroyObject [28]
28 free [4]
28 __data16_free [4] ***
30 - [0]
0 exit [38]
0 _exit [38]
4 _Close_all [34]
85
AFE1_AFE2-1:1
8 fclose [30]
14 _Fofree [4]
14 free [4] ***
16 - [0]
14 fflush [24] ***
14 free [4] ***
14 __close [8]
20 __DebugBreak [2] ***
14 remove [8]
8 __write [12] ***
2 __exit [8]
2 - [0]
Each line consists of this information:

● The stack usage at the point of call of the function
● The name of the function, or a single '-' to indicate usage in a function at a point
with no function call (typically in a leaf function)
● The stack usage along the deepest call chain from that point. If no such value could
be calculated, "[---]" is output instead. "***" marks functions that have already
been shown.

86 for AVR
AFE1_AFE2-1:1
Linking your application
● Linking considerations
● Verifying the linked result of code and data placement
Linking considerations
When you set up your project in the IAR Embedded Workbench IDE, a default linker
configuration file is automatically used based on your project settings and you can
simply link your application. For the majority of all projects it is sufficient to configure
the vital parameters that you find in Project>Options>Linker>Config.
When you build from the command line, you can use a ready-made linker command file
provided with your product package.
The config directory contains the information required by XLINK, and are ready to be
used as is. The only change, if any, you will normally have to make to the supplied
configuration file is to customize it so it fits the target system memory map. If, for
example, your application uses additional external RAM, you must also add details
about the external RAM memory area.
To edit a linker configuration file, use the editor in the IDE, or any other suitable editor.
Do not change the original template file. We recommend that you make a copy in the
working directory, and modify the copy instead.
If you find that the default linker configuration file does not meet your requirements, you
might want to consider:
● Placing segments
● Placing data
● Setting up stack memory
● Setting up heap memory
● Placing code
● Keeping modules
● Keeping symbols and segments
● Application startup
● Interaction between XLINK and your application
● Producing other output formats than UBROF
87
AFE1_AFE2-1:1
PLACING SEGMENTS
The placement of segments in memory is performed by the IAR XLINK Linker.
This section describes the most common linker commands and how to customize the
linker configuration file to suit the memory layout of your target system. In
demonstrating the methods, fictitious examples are used.
In demonstrating the methods, fictitious examples are used based on this memory
layout:
● There is 1 Mbyte addressable memory.
● There is ROM memory in the address ranges 0x0000–0x1FFF, 0x3000–0x4FFF,
and 0x10000–0x1FFFF.
● There is RAM memory in the address ranges 0x8000–0xAFFF, 0xD000–0xFFFF,
and 0x20000–0x27FFF.
● There are two addressing modes for data, one for near memory and one for far
memory.
● There is one stack and one heap.
● There are two addressing modes for code, one for nearfunc memory and one for
farfunc memory.
Note: Even though you have a different memory map, for example if you have
additional memory spaces (EEPROM) and additional segments, you can still use the
methods described in the following examples.
The ROM can be used for storing CONST and CODE segment memory types. The RAM
memory can contain segments of DATA type. The main purpose of customizing the linker
configuration file is to verify that your application code and data do not cross the
memory range boundaries, which would lead to application failure.
For the result of each placement directive after linking, inspect the segment map in the
list file (created by using the command line option -x).
General hints for placing segments

When you consider where in memory you should place your segments, it is typically a
good idea to start placing large segments first, then placing small segments.
In addition, you should consider these aspects:
● Start placing the segments that must be placed on a specific address. This is, for
example, often the case with the segment holding the reset vector.
● Then consider placing segments that hold content that requires continuous memory
addresses, for example the segments for the stack and heap.

88 for AVR
AFE1_AFE2-1:1
● When placing code and data segments for different addressing modes, make sure to
place the segments in size order (the smallest memory type first).
Note: Before the linker places any segments in memory, the linker will first place the
absolute segments.
Using the -Z command for sequential placement

Use the -Z command when you must keep a segment in one consecutive chunk, when
you must preserve the order of segment parts in a segment, or, more unlikely, when you
must put segments in a specific order.
The following illustrates how to use the -Z command to place the segment MYSEGMENTA
followed by the segment MYSEGMENTB in CONST memory (that is, ROM) in the memory
range 0x0000-0x1FFF.
-Z(CONST)MYSEGMENTA,MYSEGMENTB=0000-1FFF
To place two segments of different types continuous in the same memory area, do not
specify a range for the second segment. In the following example, the MYSEGMENTA
segment is first located in memory. Then, the rest of the memory range could be used by
MYCODE.
-Z(CONST)MYSEGMENTA=0000-1FFF
-Z(CODE)MYCODE
Two memory ranges can overlap. This allows segments with different placement
requirements to share parts of the memory space; for example:
-Z(CONST)MYSMALLSEGMENT=0000-01FF
-Z(CONST)MYLARGESEGMENT=0000-1FFF
Even though it is not strictly required, make sure to always specify the end of each
memory range. If you do this, the IAR XLINK Linker will alert you if your segments do
not fit in the available memory.
Using the -P command for packed placement

The -P command differs from -Z in that it does not necessarily place the segments (or
segment parts) sequentially. With -P it is possible to put segment parts into holes left by
earlier placements.
The following example illustrates how the XLINK -P option can be used for making
efficient use of the memory area. This command will place the data segment MYDATA in
DATA memory (that is, in RAM) in a fictitious memory range:
-P(DATA)MYDATA=8000-AFFF,D000-FFFF
89
AFE1_AFE2-1:1
If your application has an additional RAM area in the memory range 0x6000-0x67FF,
you can simply add that to the original definition:
-P(DATA)MYDATA=8000-AFFF,D000-FFFF,6000–67FF
The linker can then place some parts of the MYDATA segment in the first range, and some
parts in the second range. If you had used the -Z command instead, the linker would
have to place all segment parts in the same range.
Note: Copy initialization segments—BASENAME_I and BASENAME_ID—and dynamic
initialization segments must be placed using -Z.
Note: The AVR microcontroller is based on the Harvard architecture—thus code and
data have separate memory spaces and require different access mechanisms. For more
information, see the chapter Data storage.
PLACING DATA
Static memory is memory that contains variables that are global or declared static.
To get a clear understanding about how the data segments work, you must be familiar
with the different memory types and the different memory models available in the
compiler. For information about these details, see the chapter Data storage.
Placing static memory data segments

Depending on their memory attribute, static data is placed in specific segments. For
information about the segments used by the compiler, see Static data memory segments,
page 76.
For example, these commands can be used to place the static data segments:
/* First, the segments to be placed in ROM are defined. */
-Z(CONST)NEAR_C=0000-1FFF,3000-4FFF
-Z(CONST)FAR_C=0000-1FFF,3000-4FFF,10000-1FFFF
-Z(CONST)NEAR_ID,FAR_ID=010000-1FFFF
/* Then, the RAM data segments are placed in memory. */

-Z(DATA)NEAR_I,DATA16_Z,NEAR_N=8000-AFFF
-Z(DATA)FAR_I,FAR_Z,FAR_N=20000-27FFF
All the data segments are placed in the area used by on-chip RAM.
Placing located data

A variable that is explicitly placed at an address, for example by using the #pragma
location directive or the @ operator, is placed in for example either the NEAR_AC or
the NEAR_AN segment. The former is used for constant-initialized data, and the latter for
items declared as __no_init. The individual segment part of the segment knows its

90 for AVR
AFE1_AFE2-1:1
location in the memory space, and it does not have to be specified in the linker
configuration file.
Placing user-defined segments

If you create your own segments by using for example the #pragma location directive
or the @ operator, these segments must also be defined in the linker configuration file
using the -Z or -P segment control directives.
SETTING UP STACK MEMORY

In this example, the data segment for holding the stack is called CSTACK. The system
startup code initializes the stack pointer to point to the end of the stack segment.
Allocating a memory area for the stack is performed differently when using the
command line interface, as compared to when using the IDE.
For more information about stack memory, see Stack considerations, page 193.
Stack size allocation in the IDE

Choose Project>Options. In the General Options category, click the System tab.
Add the required stack size in the dedicated text box.
Stack size allocation from the command line

The size of the CSTACK segment is defined in the linker configuration file.
The linker configuration file sets up a constant, representing the size of the stack, at the
beginning of the file. Specify the appropriate size for your application, in this example
1024 bytes:
-D_..X_CSTACK_SIZE=400 /* 1024 bytes of stack size */
Note that the size is written hexadecimally, but not necessarily with the 0x notation.
Placing the stack segment

Further down in the linker configuration file, the actual stack segment is defined in the
memory area available for the stack:
-Z(DATA)CSTACK+_CSTACK_SIZE=8000-AFFF
Note:
● This range does not specify the size of the stack; it specifies the range of the
available memory.
● The = allocates the CSTACK segment at the start of the memory area.
91
AFE1_AFE2-1:1
SETTING UP HEAP MEMORY

The heap contains dynamic data allocated by the C function malloc (or a corresponding
function) or the C++ operator new.
If your application uses dynamic memory allocation, you should be familiar with:
● The linker segments used for the heap, see Dynamic memory on the heap, page 61
● The steps for allocating the heap size, which differs depending on which build
interface you are using
● The steps involved for placing the heap segments in memory.
See also Heap considerations, page 194.
In this example, the data segment for holding the heap is called HEAP.
Heap size allocation in the IDE

Choose Project>Options. In the General Options category, click the Heap
Configuration tab.
Add the required heap size in the dedicated text box.
Heap size allocation from the command line

The size of the HEAP segment is defined in the linker configuration file.
The linker configuration file sets up a constant, representing the size of the heap, at the
beginning of the file. Specify the appropriate size for your application, in this example
1024 bytes:
-D_HEAP_SIZE=400 /* 1024 bytes for heap memory */
Note that the size is written hexadecimally, but not necessarily with the 0x notation.
In many linker configuration files provided with IAR Embedded Workbench, these lines
are prefixed with the comment character // because the IDE controls the heap size
allocation. To make the directive take effect, remove the comment character.
If you use a heap, you should allocate at least 512 bytes for it, to make it work properly.
Placing the heap segment

The actual HEAP segment is allocated in the memory area available for the heap:
-Z(DATA)HEAP+_HEAP_SIZE=8000-AFFF
Note: This range does not specify the size of the heap; it specifies the range of the
available memory.

92 for AVR
AFE1_AFE2-1:1
PLACING CODE
This section contains descriptions of the segments used for storing code and the
interrupt vector table. For information about all segments, see Summary of segments,
page 375.
Normal code
Depending on their memory attribute, functions are placed in various segments, in this
example NEARFUNC and FARFUNC.
Placing the segments is a simple operation in the linker configuration file:
-P(CODE)NEARFUNC=0000-1FFF,3000-4FFF
-P(CODE)FARFUNC=0000-1FFF,3000-4FFF,10000-1FFFF
Here, the -P linker directive is used for allowing XLINK to split up the segments and
pack their contents more efficiently. This is also useful if the memory range is
non-continuous.
For information about segments holding normal code, see the chapter Segment
reference.
Interrupt vectors
The interrupt vector table contains pointers to interrupt routines, including the reset
routine. In this example, the table is placed in the segment INTVEC. The linker directive
would then look like this:
-Z(CONST)INTVEC=0000-00FF
For more information about the interrupt vectors, see Interrupt vectors and the interrupt
vector table, page 65.
C++ dynamic initialization

In C++, all global objects are created before the main function is called. The creation of
objects can involve the execution of a constructor.
The DIFUNCT segment contains a vector of addresses that point to initialization code.
All entries in the vector are called when the system is initialized.
For example:
-Z(CONST)DIFUNCT=0000-1FFF,3000-4FFF
DIFUNCT must be placed using -Z. For more information, see DIFUNCT, page 378.
93
AFE1_AFE2-1:1
Additional compiler-generated segments

The compiler uses a set of internally generated segments, which are used for storing
information that is vital to the operation of your application.
● The SWITCH segment which contains data statements used in the switch library
routines. These tables are encoded in such a way as to use as little space as possible.
● The INITTAB segment contains the segment initialization description blocks that
are used by the __segment_init function which is called by CSTARTUP. This
table consists of a number of SegmentInitBlock_Type objects. This type is
declared in the segment_init.h file which is located in the avr\src\lib
directory.
In the linker configuration file, it can look like this:
-Z(CODE)SWITCH, INITTAB=0–1FFF
KEEPING MODULES
If a module is linked as a program module, it is always kept. That is, it will contribute
to the linked application. However, if a module is linked as a library module, it is
included only if it is symbolically referred to from other parts of the application that
have been included. This is true, even if the library module contains a root symbol. To
assure that such a library module is always included, use -A to make all modules in the
file be treaded as if they were program modules:
-A file.r90
Use -C to makes all modules in the file be treated as if they were library modules:
-C file.r90
KEEPING SYMBOLS AND SEGMENTS

By default, XLINK removes any segments, segment parts, and global symbols that are
not needed by the application. To retain a symbol that does not appear to be needed—or
actually, the segment part it is defined in—you can either use the __root attribute on
the symbol in your C/C++ source code or ROOT in your assembler source code, or use
the XLINK option -g.
For information about included and excluded symbols and segment parts, inspect the
map file (created by using the XLINK option -xm).
For more information about the linking procedure for keeping symbols and sections, see
The linking process in detail, page 73.

94 for AVR
AFE1_AFE2-1:1
APPLICATION STARTUP
By default, the point where the application starts execution is defined by the
__program_start label, which is defined to point at the reset vector. The label is also
communicated via the debugger information to any debugger.
To change the start point of the application to another label, use the XLINK option -s.
INTERACTION BETWEEN XLINK AND YOUR APPLICATION

Use the XLINK option -D to define symbols that can be used for controlling your
application. You can also use symbols to represent the start and end of a continuous
memory area that is defined in the linker configuration file.
To change a reference to one symbol to another symbol, use the XLINK command line
option -e. This is useful, for example, to redirect a reference from a non-implemented
function to a stub function, or to choose one of several different implementations of a
certain function.
For information about the addresses and sizes of all global (statically linked) symbols,
inspect the entry list in the map file (the XLINK option -xm).
PRODUCING OTHER OUTPUT FORMATS THAN UBROF

XLINK can generate more than 30 industry-standard loader formats, in addition to the
proprietary format UBROF which is used by the C-SPY debugger. For a complete list,
see the IAR Linker and Library Tools Reference Guide. To specify a different output
format than the default, use the XLINK option -F. For example:
-F intel-standard
Note that it can be useful to use the XLINK -O option to produce two output files, one
for debugging and one for burning to ROM/flash.
Note also that if you choose to enable debug support using the -r option for certain
low-level I/O functionality for mechanisms like file operations on the host computer etc,
such debug support might have an impact on the performance and responsiveness of
your application. In this case, the debug build will differ from your release build due to
the debug modules included.
Verifying the linked result of code and data placement

The linker has several features that help you to manage code and data placement, for
example, messages at link time and the linker map file.
95
AFE1_AFE2-1:1
SEGMENT TOO LONG ERRORS AND RANGE ERRORS

Code or data that is placed in a relocatable segment will have its absolute address
resolved at link time. Note that it is not known until link time whether all segments will
fit in the reserved memory ranges. If the contents of a segment do not fit in the address
range defined in the linker configuration file, XLINK will issue a segment too long error.
Some instructions do not work unless a certain condition holds after linking, for
example that a branch must be within a certain distance or that an address must be even.
XLINK verifies that the conditions hold when the files are linked. If a condition is not
satisfied, XLINK generates a range error or warning and prints a description of the
error.
For more information about these types of errors, see the IAR Linker and Library Tools
Reference Guide.
LINKER MAP FILE

XLINK can produce an extensive cross-reference listing, which can optionally contain
the following information:
● A segment map which lists all segments in address order
● A module map which lists all segments, local symbols, and entries (public symbols)
for every module in the program. All symbols not included in the output can also be
listed
● A module summary which lists the contribution (in bytes) from each module
● A symbol list which contains every entry (global symbol) in every module.
Use the option Generate linker listing in the IDE, or the option -x on the command
line, and one of their suboptions to generate a linker listing.
Normally, XLINK will not generate an output file if any errors, such as range errors,
occur during the linking process. Use the option Always generate output in the IDE,
or the option -B on the command line, to generate an output file even if a non-fatal error
was encountered.
For more information about the listing options and the linker listing, see the IAR Linker
and Library Tools Reference Guide, and the IDE Project Management and Building
Guide for AVR.
MANAGING MULTIPLE MEMORY SPACES

Output formats that do not support more than one memory space—like MOTOROLA and
INTEL-HEX—might require up to one output file per memory space. This causes no
problems if you are only producing output to one memory space (flash), but if you also

96 for AVR
AFE1_AFE2-1:1
are placing objects in EEPROM or an external ROM in DATA space, the output format
cannot represent this, and the linker issues this error message:
Error[e133]: The output format Format cannot handle multiple
address spaces. Use format variants (-y -O) to specify which
address space is wanted.
To limit the output to flash memory, make a copy of the linker configuration file for the
derivative and memory model you are using, and put it in the project directory. Add this
line to the file:
-y(CODE)
To produce output for the other memory space(s), you must generate one output file per
memory space (because the output format you chose does not support more than one
memory space). Use the XLINK option -O for this purpose.
For each additional output file, you must specify format, XLINK segment type, and file
name. For example:
-Omotorola,(DATA)=external_rom.a90
-Omotorola,(XDATA)=eeprom.a90
Note: As a general rule, an output file is only necessary if you use non-volatile memory.
In other words, output from the data space is only necessary if the data space contains
external ROM.
For more information, see the IAR Linker and Library Tools Reference Guide.
The IAR Postlink utility

You can also use the IAR Postlink utility, delivered with the compiler. This application
takes as input an object file (of the XLINK simple format) and extracts one or more of
its XLINK segment types into one file (which can be in either Intel extended hex
format or Motorola S-record format). For example, it can put all code segments into
one file, and all EEPROM segments into another.
See the postlink.htm document for more information about IAR Postlink.
97
AFE1_AFE2-1:1

98 for AVR
AFE1_AFE2-1:1
The DLIB runtime
environment
● Introduction to the runtime environment
● Setting up the runtime environment
● Additional information on the runtime environment
● Managing a multithreaded environment
DLIB can be used with both the C and the C++ languages. CLIB, on the other
hand, can only be used with the C language. For more information about CLIB,
see the chapter The CLIB runtime environment.
Introduction to the runtime environment

A runtime environment is the environment in which your application executes.
This section contains information about:
● Runtime environment functionality, page 99
● Briefly about input and output (I/O), page 100
● Briefly about C-SPY emulated I/O, page 102
● Briefly about retargeting, page 102
RUNTIME ENVIRONMENT FUNCTIONALITY

The DLIB runtime environment supports Standard C and C++ and consists of:
● The C/C++ standard library, both its interface (provided in the system header files)
and its implementation.
● Startup and exit code.
● Low-level I/O interface for managing input and output (I/O).
● Special compiler support, for instance functions for switch handling or integer
arithmetics.
99
AFE1_AFE2-1:1
● Support for hardware features:

● Direct access to low-level processor operations by means of intrinsic functions,
such as functions for interrupt mask handling
● Peripheral unit registers and interrupt definitions in include files
● Special compiler support for accessing strings in flash memory, see AVR-specific
library functions, page 370.
Runtime environment functions are provided.
The runtime library is delivered both as and (depending on your product package) as
source files. The prebuilt libraries are available in different configurations to meet
various needs, see Runtime library configurations, page 110. You can find the libraries
in the product subdirectories avr\lib and avr\src\lib, respectively.
For more information about the library, see the chapter C/C++ standard library
functions.
BRIEFLY ABOUT INPUT AND OUTPUT (I/O)

Every application must communicate with its environment. The application might for
example display information on an LCD, read a value from a sensor, get the current date
from the operating system, etc. Typically, your application performs I/O via the C/C++
standard library or some third-party library.
There are many functions in the C/C++ standard library that deal with I/O, including
functions for: standard character streams, file system access, time and date,
miscellaneous system actions, and termination and assert. This set of functions is
referred to as the standard I/O interface.
On a desktop computer or a server, the operating system is expected to provide I/O
functionality to the application via the standard I/O interface in the runtime
environment. However, in an embedded system, the runtime library cannot assume that
such functionality is present, or even that there is an operating system at all. Therefore,

100 for AVR
AFE1_AFE2-1:1
The DLIB runtime environment
the low-level part of the standard I/O interface is not completely implemented by
default:
To make the standard I/O interface work, you can:

● Let the C-SPY debugger emulate I/O operations on the host computer, see Briefly
about C-SPY emulated I/O, page 102
● Retarget the standard I/O interface to your target system by providing a suitable
implementation of the interface, see Briefly about retargeting, page 102.
It is possible to mix these two approaches. You can, for example, let debug printouts and
asserts be emulated by the C-SPY debugger, but implement your own file system. The
debug printouts and asserts are useful during debugging, but no longer needed when
running the application stand-alone (not connected to the C-SPY debugger).
101
AFE1_AFE2-1:1
BRIEFLY ABOUT C-SPY EMULATED I/O

C-SPY emulated I/O is a mechanism which lets the runtime environment interact with
the C-SPY debugger to emulate I/O actions on the host computer:
For example, when C-SPY emulated I/O is enabled:

● Standard character streams are directed to the C-SPY Terminal I/O window
● File system operations are performed on the host computer
● Time and date functions return the time and date of the host computer
● Termination and failed asserts break execution and notify the C-SPY debugger.
This behavior can be valuable during the early development of an application, for
example in an application that uses file I/O before any flash file system I/O drivers are
implemented, or if you need to debug constructions in your application that use stdin
and stdout without the actual hardware device for input and output being available.
See Setting up your runtime environment, page 104 and The C-SPY emulated I/O
mechanism, page 116.
BRIEFLY ABOUT RETARGETING

Retargeting is the process where you adapt the runtime environment so that your
application can execute I/O operations on your target system.
The standard I/O interface is large and complex. To make retargeting easier, the DLIB
runtime environment is designed so that it performs all I/O operations through a small
set of simple functions, which is referred to as the DLIB low-level I/O interface. By
default, the functions in the low-level interface lack usable implementations. Some are

102 for AVR
AFE1_AFE2-1:1
unimplemented, others have stub implementations that do not perform anything except
returning error codes.
To retarget the standard I/O interface, all you have to do is to provide implementations
for the functions in the DLIB low-level I/O interface.
For example, if your application calls the functions printf and fputc in the standard
I/O interface, the implementations of those functions both call the low-level function
_ _write to output individual characters. To make them work, you just need to provide
an implementation of the _ _write function—either by implementing it yourself, or by
using a third-party implementation.
For information about how to override library modules with your own implementations,
see Overriding library modules, page 107. See also The DLIB low-level I/O interface,
page 123 for information about the functions that are part of the interface.
Setting up the runtime environment

This section contains these tasks:
● Setting up your runtime environment, page 104
A runtime environment with basic project settings to be used during the initial phase
of development.
● Retargeting—Adapting for your target system, page 106
● Overriding library modules, page 107
103
AFE1_AFE2-1:1
● Customizing and building your own runtime library, page 108

See also:
● Managing a multithreaded environment, page 135 for information about how to
adapt the runtime environment to treat all library objects according to whether they
are global or local to a thread.
SETTING UP YOUR RUNTIME ENVIRONMENT

You can set up the runtime environment based on some basic project settings. It is also
often convenient to let the C-SPY debugger manage things like standard streams, file
I/O, and various other system interactions. This basic runtime environment can be used
for simulation before you have any target hardware.
To set up the runtime environment:
1 Before you build your project, choose Project>Options>General Options to open the
Options dialog box.
2 On the Library Configuration page, verify the following settings:
● Library: choose which library and library configuration to use. Typically, choose
Normal, or Full.
For information about the various library configurations, see Runtime library
configurations, page 110.
CLIB or DLIB—for more information about the libraries, see C/C++ standard
library overview, page 361.
3 On the Library Options page, select Auto for both Printf formatter and Scanf
formatter. This means that the linker will automatically choose the appropriate
formatters based on information from the compiler. For more information about the
available formatters and how to choose one manually, see Formatters for printf, page
113 and Formatters for scanf, page 114, respectively.
4 To enable C-SPY emulated I/O, choose Project>Options>Linker>Output and select
the Format option that matches the level of support you need. See Briefly about C-SPY
emulated I/O, page 102. Choose between:
Linker
Linker option in the
command line Description
IDE
option
Debug information -Fubrof Debug support for C-SPY, but without any support
for C-SPY for C-SPY emulated I/O.
Table 10: Debug information and levels of C-SPY emulated I/O

104 for AVR
AFE1_AFE2-1:1
Linker
Linker option in the
command line Description
IDE
option
With runtime -r The same as -Fubrof but also limited support for
control modules C-SPY emulated I/O handling program abort, exit,
and assertions.
With I/O emulation -rt Full support for C-SPY emulated I/O, which means
modules the same as -r, but also support for I/O handling,
and accessing files on the host computer during
debugging.
Table 10: Debug information and levels of C-SPY emulated I/O (Continued)
Note: The C-SPY Terminal I/O window is not opened automatically; you must open it
manually. For more information about this window, see the C-SPY® Debugging Guide
for AVR.
5 On some systems, terminal output might be slow because the host computer and the
target system must communicate for each character.
For this reason, a replacement for the _ _write function called _ _write_buffered is
included in the runtime library. This module buffers the output and sends it to the
debugger one line at a time, speeding up the output. Note that this function uses about
80 bytes of RAM memory.
To use this feature in the IDE, choose Project>Options>Linker>Output and select the
option Buffered terminal output.
To enable this function on the command line, add this to the linker command line:
-e_ _write_buffered=_ _write
6 Some math functions are available in different versions: default versions, smaller than
the default versions, and larger but more accurate than default versions. Consider
which versions you should use.
For more information, see Math functions, page 116.
7 When you build your project, a suitable prebuilt library and library configuration file
are automatically used based on the project settings you made.
For information about which project settings affect the choice of library file, see
Runtime library configurations, page 110.
You have now set up a runtime environment that can be used while developing your
application source code.
105
AFE1_AFE2-1:1
RETARGETING—ADAPTING FOR YOUR TARGET SYSTEM

Before you can run your application on your target system, you must adapt some parts
of the runtime environment, typically the system initialization and the DLIB low-level
I/O interface functions.
To adapt your runtime environment for your target system:
1 Adapt system initialization.
It is likely that you must adapt the system initialization, for example, your application
might need to initialize interrupt handling, I/O handling, watchdog timers, etc. You do
this by implementing the routine _ _low_level_init, which is executed before the
data segments are initialized. See System startup and termination, page 118 and System
initialization, page 122. Note that you can find device-specific examples on this in the
example projects provided in the product installation; see the Information Center.
2 Adapt the runtime library for your target system. To implement such functions, you
need a good understanding of the DLIB low-level I/O interface, see Briefly about
retargeting, page 102.
Typically, you must implement your own functions if your application uses:
● Standard streams for input and output
If any of these streams are used by your application, for example by the functions
printf and scanf, you must implement your versions of the low-level functions
_ _read and _ _write.
The low-level functions identify I/O streams, such as an open file, with a file handle
that is a unique integer. The I/O streams normally associated with stdin, stdout,
and stderr have the file handles 0, 1, and 2, respectively. When the handle is -1,
all streams should be flushed. Streams are defined in stdio.h.
● File input and output
The library contains a large number of powerful functions for file I/O operations,
such as fopen, fclose, fprintf, fputs, etc. All these functions call a small set of
low-level functions, each designed to accomplish one particular task; for example,
_ _open opens a file, and _ _write outputs characters. Implement your version of
these low-level functions.
● signal and raise
If the default implementation of these functions does not provide the functionality
you need, you can implement your own versions.
● Time and date
To make the time and date functions work, you must implement the functions clock,
_ _time32, _ _time64, and _ _getzone. Whether you use _ _time32 or _ _time64
depends on which interface you use for time_t, see time.h, page 368.

106 for AVR
AFE1_AFE2-1:1
● Assert, see _ReportAssert, page 129.

● Environment interaction
If the default implementation of system or getenv does not provide the
functionality you need, you can implement your own versions.
For more information about the functions, see The DLIB low-level I/O interface, page
123.
The library files that you can override with your own versions are located in the
avr\src\lib directory.
3 When you have implemented your functions of the low-level I/O interface, you must
add your version of these functions to your project. For information about this, see
Overriding library modules, page 107.
Note: If you have implemented a DLIB low-level I/O interface function and added it to
a project that you have built with support for C-SPY emulated I/O, your low-level
function will be used and not the functions provided with C-SPY emulated I/O. For
example, if you implement your own version of _ _write, output to the C-SPY
Terminal I/O window will not be supported. See Briefly about C-SPY emulated I/O,
page 102.
4 Before you can execute your application on your target system, you must rebuild your
project with a Release build configuration. This means that the linker will not include
the C-SPY emulated I/O mechanism and the low-level I/O functions it provides. If your
application calls any of the low-level functions of the standard I/O interface, either
directly or indirectly, and your project does not contain these, the linker will issue an
error for every missing low-level function. Also, note that the NDEBUG symbol is
defined in a Release build configuration, which means asserts will no longer be
generated. For more information, see _ReportAssert, page 129.
OVERRIDING LIBRARY MODULES

To override a library function and replace it with your own
implementation:
1 Use a template source file—a library source file or another template—and place a copy
of it in your project directory.
avr\src\lib directory.
2 Modify the file.

Note: To override the functions in a module, you must provide alternative
implementations for all the needed symbols in the overridden module. Otherwise you
will get error messages about duplicate definitions.
107
AFE1_AFE2-1:1
3 Add the modified file to your project, like any other source file.
Note: If you have implemented a DLIB low-level I/O interface function and added it to
a project that you have built with support for C-SPY emulated I/O, your low-level
function will be used and not the functions provided with C-SPY emulated I/O. For
example, if you implement your own version of _ _write, output to the C-SPY
Terminal I/O window will not be supported. See Briefly about C-SPY emulated I/O,
page 102.
You have now finished the process of overriding the library module with your version.
CUSTOMIZING AND BUILDING YOUR OWN RUNTIME

LIBRARY
If the prebuilt library configurations do not meet your requirements, you can customize
your own library configuration, but that requires that you rebuild the library.
Building a customized library is a complex process. Therefore, consider carefully
whether it is really necessary. You must build your own runtime library when:
● There is no prebuilt library for the required combination of compiler options or
hardware support
● You want to define your own library configuration with support for locale, file
descriptors, multibyte characters, etc. This will include or exclude certain parts of
the DLIB runtime environment.
In those cases, you must:
● Make sure that you have installed the library source code (scr\lib). If not already
installed, you can install it using the IAR License Manager, see the Installation and
Licensing Guide.
● Set up a library project
● Make the required library customizations
● Build your customized runtime library
● Finally, make sure your application project will use the customized runtime library.
To set up a library project:
1 In the IDE, choose Project>Create New Project and use the library project template
which can be used for customizing the runtime environment configuration. There is a
library template for CLIB and a template for DLIB, see Runtime library
configurations, page 110
Note that when you create a new library project from a template, the majority of the files
included in the new project are the original installation files. If you are going to modify

108 for AVR
AFE1_AFE2-1:1
these files, make copies of them first and replace the original files in the project with
these copies.
2 Modify the project options in the created library project to suit your application, see
Basic project configuration, page 48.
To customize the library functionality:
1 The library functionality is determined by a set of configuration symbols. The default
values of these symbols are defined in the file DLib_Defaults.h which you can find
in avr\inc\DLIB\c. This read-only file describes the configuration possibilities.
Note that you should not modify this file.
In addition, your custom library has its own library configuration file
dlavrCustom.h—which you can find in avr\config\template\project—and
which sets up that specific library with the required library configuration. Customize
this file by setting the values of the configuration symbols according to the application
requirements. For example, to enable the ll qualifier in printf format strings, write in
your library configuration file:
#define _DLIB_PRINTF_LONG_LONG 1
For information about configuration symbols that you might want to customize, see:
● Configuration symbols for printf and scanf, page 132
● Configuration symbols for file input and output, page 133
● Locale, page 133
● Strtod, page 135
● Managing a multithreaded environment, page 135
2 When you are finished, build your library project with the appropriate project options.
After you build your library, you must make sure to use it in your application project.
To build IAR Embedded Workbench projects from the command line, use the IAR
Command Line Build Utility (iarbuild.exe). There is also a batch file
(build_libs.bat) provided for building the library from the command line. For
information about the build process and the IAR Command Line Build Utility, see the
IDE Project Management and Building Guide for AVR.
To use the customized runtime library in your application project:
1 In the IDE, choose Project>Options>General Options and click the Library
Configuration tab.
2 From the Library drop-down menu, choose Custom DLIB.
3 In the Library file text box, locate your library file.
109
AFE1_AFE2-1:1
Additional information on the runtime environment
4 In the Configuration file text box, locate your library configuration file.

This section gives additional information on the runtime environment:
● Runtime library configurations, page 110
● Prebuilt runtime libraries, page 111
● Formatters for printf, page 113
● Formatters for scanf, page 114
● The C-SPY emulated I/O mechanism, page 116
● Math functions, page 116
● System startup and termination, page 118
● System initialization, page 122
● The DLIB low-level I/O interface, page 123
● Configuration symbols for printf and scanf, page 132
● Configuration symbols for file input and output, page 133
● Locale, page 133
● Strtod, page 135
RUNTIME LIBRARY CONFIGURATIONS

The runtime library is provided with different library configurations, where each
configuration is suitable for different application requirements.
The runtime library configuration is defined in the library configuration file. It contains
information about what functionality is part of the runtime environment. The less
functionality you need in the runtime environment, the smaller the environment
becomes.
These predefined library configurations are available:
Library configuration Description

Normal DLIB (default) C locale, but no locale interface, no file descriptor support, no
multibyte characters in printf and scanf, and no hexadecimal
floating-point numbers in strtod.
Full DLIB Full locale interface, C locale, file descriptor support, and optionally
multibyte characters in printf and scanf, and hexadecimal
floating-point numbers in strtod.
Table 11: Library configurations

110 for AVR
AFE1_AFE2-1:1
Note: In addition to these predefined library configurations, you can provide your own
configuration, see Customizing and building your own runtime library, page 108
If you do not specify a library configuration explicitly you will get the default
configuration. If you use a prebuilt runtime library, a configuration file that matches the
runtime library file will automatically be used. See Setting up your runtime environment,
page 104.
To override the default library configuration, use one of these methods:
1 Use a prebuilt configuration of your choice—to specify a runtime configuration
explicitly:
Choose Project>Options>General Options>Library Configuration>Library and
change the default setting.
Use the --dlib_config compiler option, see --dlib_config, page 249.
The prebuilt libraries are based on the default configurations, see Runtime library
2 If you have built your own customized library, choose Project>Options>Library
Configuration>Library and choose Custom to use your own configuration. For more
information, see Customizing and building your own runtime library, page 108.
PREBUILT RUNTIME LIBRARIES

The prebuilt runtime libraries are configured for different combinations of these options:
● Type of library
● Processor option (-v)
● Memory model option (--memory_model)
● AVR enhanced core option (--enhanced_core)
● Small flash memory option (--64k_flash)
● 64-bit doubles option (--64bit_doubles)
● Library configuration—Normal or Full.
To choose a library from the command line:

If you build your application from the command line, make the following settings:
● Specify which library file to use on the XLINK command line, like:
dllibname.r90
● If you do not specify a library configuration, the default configuration will be used.
However, you can specify the library configuration explicitly for the compiler:
--dlib_config C:\...\dllibname.h
111
AFE1_AFE2-1:1
Note: All modules in the library have a name that starts with the character ? (question
mark).
You can find the library files in the subdirectory avr\lib\dlib and the library
configuration files in the avr\inc\dlib subdirectory.
Library filename syntax

The names of the libraries are constructed from these elements:
{library} dl for the IAR DLIB runtime environment
{target} is avr
{memModel} Specifies the memory model:
t = Tiny
s = Small
l = Large
h = Huge
{core} is ec_mul or ec_nomul when enhanced core is used

depending on whether the MUL instruction is available or
not, or xmega if the device is from the xmega family. If
neither the enhanced core nor the xmega core is used, this
value is not specified.
{smallFlash} is sf when the small flash memory is available. When small
flash memory is not available, this value is not specified
{64bitDoubles} is 64 when 64-bit doubles are used. When 32-bit doubles
are used, this value is not specified.
{lib_config} Specifies the library configuration:
n = Normal
f = Full
For example, the library dlavr-3s-ec_mul-sf-n.r90 is configured for the Small

memory model, for the enhanced core with the MUL instruction available, for the small
flash memory, and for the normal library configuration.
Note: The library configuration file has the same base name as the library.

112 for AVR
AFE1_AFE2-1:1
These are some examples of how to decode a library name:

Generic
Memory Small 64-bit Library
Library processor Core
model flash doubles configuration
option
dlavr-3s-ec_mul-sf-n.r90 -v3 Small X X -- Normal
dlavr-3s-ec_mul-64-f.r90 -v3 Small X -- X Full
Table 12: Prebuilt libraries
FORMATTERS FOR PRINTF

The printf function uses a formatter called _Printf. The full version is quite large,
and provides facilities not required in many embedded applications. To reduce the
memory consumption, three smaller, alternative versions are also provided. Note that
the wprintf variants are not affected.
This table summarizes the capabilities of the different formatters:
Small/ Large/ Full/

Formatting capabilities Tiny
SmallNoMb† LargeNoMb† FullNoMb†
Basic specifiers c, d, i, o, p, s, u, X, x, and % Yes Yes Yes Yes
Multibyte support No Yes/No Yes/No Yes/No
Floating-point specifiers a, and A No No No Yes
Floating-point specifiers e, E, f, F, g, and G No No Yes Yes
Conversion specifier n No No Yes Yes
Format flag +, -, #, 0, and space No Yes Yes Yes
Length modifiers h, l, L, s, t, and Z No Yes Yes Yes
Field width and precision, including * No Yes Yes Yes
long long support No No Yes Yes
Table 13: Formatters for printf
† NoMb means without multibytes.

Note: It is possible to optimize these functions even further, but that requires that you
rebuild the library. See Configuration symbols for printf and scanf, page 132.
The compiler can automatically detect which formatting capabilities are needed in a
direct call to printf, if the formatting string is a string literal. This information is
passed to the linker, which combines the information from all modules to select a
suitable formatter for the application. However, if the formatting string is a variable, or
if the call is indirect through a function pointer, the compiler cannot perform the
113
AFE1_AFE2-1:1
analysis, forcing the linker to select the Full formatter. In this case, you might want to
override the automatically selected printf formatter.
To override the automatically selected printf formatter in the IDE:
1 Choose Project>Options>General Options to open the Options dialog box.
2 On the Library Options page, select the appropriate formatter.
To override the automatically selected printf formatter from the command
line:
1 Add one of these lines in the linker configuration file you are using:
-e_PrintfFull=_Printf
-e_PrintfFullNoMb=_Printf
-e_PrintfLarge=_Printf
-e_PrintfLargeNoMb=_Printf
_e_PrintfSmall=_Printf
-e_PrintfSmallNoMb=_Printf
-e_PrintfTiny=_Printf
-e_PrintfTinyNoMb=_Printf
To override the default AVR-specific printf_P formatter, type any of the following
lines in your linker command file:
-e_small_write_P=_formatted_write_P
-e_medium_write_P=_formatted_write_P
Note: In the IDE, you can use Project>Options>Linker>Extra Options to override

the default formatter for the AVR-specific library routines.
See also printf_P, page 371.
FORMATTERS FOR SCANF

In a similar way to the printf function, scanf uses a common formatter, called
_Scanf. The full version is quite large, and provides facilities that are not required in
many embedded applications. To reduce the memory consumption, two smaller,
alternative versions are also provided. Note that the wscanf versions are not affected.
This table summarizes the capabilities of the different formatters:
Small/ Large/ Full/

Formatting capabilities
Basic specifiers c, d, i, o, p, s, u, X, x, and % Yes Yes Yes
Multibyte support Yes/No Yes/No Yes/No
Floating-point specifiers a, and A No No Yes
Table 14: Formatters for scanf

114 for AVR
AFE1_AFE2-1:1
Small/ Large/ Full/

Formatting capabilities
Floating-point specifiers e, E, f, F, g, and G No No Yes
Conversion specifier n No No Yes
Scan set [ and ] No Yes Yes
Assignment suppressing * No Yes Yes
long long support No No Yes
Table 14: Formatters for scanf (Continued)
† NoMb means without multibytes.

Note: It is possible to optimize these functions even further, but that requires that you
rebuild the library. See Configuration symbols for printf and scanf, page 132.
The compiler can automatically detect which formatting capabilities are needed in a
direct call to scanf, if the formatting string is a string literal. This information is passed
to the linker, which combines the information from all modules to select a suitable
formatter for the application. However, if the formatting string is a variable, or if the call
is indirect through a function pointer, the compiler cannot perform the analysis, forcing
the linker to select the full formatter. In this case, you might want to override the
automatically selected scanf formatter.
To manually specify the scanf formatter in the IDE:
1 Choose Project>Options>General Options to open the Options dialog box.
2 On the Library Options page, select the appropriate formatter.
To manually specify the scanf formatter from the command line:
1 Add one of these lines in the linker configuration file you are using:
-e_ScanfFull=_Scanf
-e_ScanfFullNoMb=_Scanf
-e_ScanfLarge=_Scanf
-e_ScanfLargeNoMb=_Scanf
_e_ScanfSmall=_Scanf
_e_ScanfSmallNoMb=_Scanf
To override the default AVR-specific scanf_P formatter, type the following line in your
linker command file:
-e_medium_read_P=_formatted_read_P
Note: In the IDE, you can use Project>Options>Linker>Extra Options to override

the default formatter for the AVR-specific library routines.
See also scanf_P, page 372.
115
AFE1_AFE2-1:1
THE C-SPY EMULATED I/O MECHANISM

The C-SPY emulated I/O mechanism works as follows:
1 The debugger will detect the presence of the function _ _DebugBreak, which will
be part of the application if you linked it with the linker option for C-SPY emulated
I/O.
2 In this case, the debugger will automatically set a breakpoint at the _ _DebugBreak
function.
3 When your application calls a function in the DLIB low-level I/O interface, for
example, open, the _ _DebugBreak function is called, which will cause the
application to stop at the breakpoint and perform the necessary services.
4 The execution will then resume.
See also Briefly about C-SPY emulated I/O, page 102.
MATH FUNCTIONS
Some C/C++ standard library math functions are available in different versions:
● The default versions
● Smaller versions (but less accurate)
● More accurate versions (but larger).
Smaller versions
The functions cos, exp, log, log2, log10, _ _iar_Log (a help function for log, log2,
and log10), pow, sin, tan, and _ _iar_Sin (a help function for sin and cos) exist in
additional, smaller versions in the library. They are about 20% smaller and about 20%
faster than the default versions. The functions handle INF and NaN values. The
drawbacks are that they almost always lose some precision and they do not have the
same input range as the default versions.
The names of the functions are constructed like:
_ _iar_xxx_small<f|l>
where f is used for float variants, l is used for long double variants, and no suffix
is used for double variants.

116 for AVR
AFE1_AFE2-1:1
To specify individual math functions from the command line:

1 Redirect the default function names to these names when linking, using these options:
-e__iar_sin_small=sin
-e__iar_cos_small=cos
-e__iar_tan_small=tan
-e__iar_log_small=log
-e__iar_log2_small=log2
-e__iar_log10_small=log10
-e__iar_exp_small=exp
-e__iar_pow_small=pow
-e__iar_Sin_small=__iar_Sin
-e__iar_Log_small=__iar_Log
-e__iar_sin_smallf=sinf
-e__iar_cos_smallf=cosf
-e__iar_tan_smallf=tanf
-e__iar_log_smallf=logf
-e__iar_log2_smallf=log2f
-e__iar_log10_smallf=log10f
-e__iar_exp_smallf=expf
-e__iar_pow_smallf=powf
-e__iar_Sin_smallf=__iar_Sinf
-e__iar_Log_smallf=__iar_Logf
-e__iar_sin_smalll=sinl
-e__iar_cos_smalll=cosl
-e__iar_tan_smalll=tanl
-e__iar_log_smalll=logl
-e__iar_log2_smalll=log2l
-e__iar_log10_smalll=log10l
-e__iar_exp_smalll=expl
-e__iar_pow_smalll=powl
-e__iar_Sin_smalll=__iar_Sinl
-e__iar_Log_smalll=__iar_Logl
Note that if you want to redirect any of the functions sin, cos, or _ _iar_Sin, you must
redirect all three functions.
Note that if you want to redirect any of the functions log, log2, log10, or _ _iar_Log,
you must redirect all four functions.
More accurate versions

The functions cos, pow, sin, and tan, and the help functions _ _iar_Sin and
_ _iar_Pow exist in versions in the library that are more exact and can handle larger
117
AFE1_AFE2-1:1
argument ranges. The drawback is that they are larger and slower than the default
versions.
The names of the functions are constructed like:
_ _iar_xxx_accurate<f|l>
where f is used for float variants, l is used for long double variants, and no suffix
is used for double variants.
To specify individual math functions from the command line:
1 Redirect the default function names to these names when linking, using these options:
-e__iar_sin_accurate=sin
-e__iar_cos_accurate=cos
-e__iar_tan_accurate=tan
-e__iar_pow_accurate=pow
-e__iar_Sin_accurate=__iar_Sin
-e__iar_Pow_accurate=__iar_Pow
-e__iar_sin_accuratef=sinf
-e__iar_cos_accuratef=cosf
-e__iar_tan_accuratef=tanf
-e__iar_pow_accuratef=powf
-e__iar_Sin_accuratef=__iar_Sinf
-e__iar_Pow_accuratef=__iar_Powf
-e__iar_sin_accuratel=sinl
-e__iar_cos_accuratel=cosl
-e__iar_tan_accuratel=tanl
-e__iar_pow_accuratel=powl
-e__iar_Sin_accuratel=__iar_Sinl
-e__iar_Pow_accuratel=__iar_Powl
Note that if you want to redirect any of the functions sin, cos, or _ _iar_Sin, you must
redirect all three functions.
Note that if you want to redirect any of the functions pow or _ _iar_Pow, you must
redirect both functions.
SYSTEM STARTUP AND TERMINATION

This section describes the runtime environment actions performed during startup and
termination of your application.
The code for handling startup and termination is located in the source files
cstartup.s90, _exit.s90, and low_level_init.c located in the avr\src\lib
directory.

118 for AVR
AFE1_AFE2-1:1
For information about how to customize the system startup code, see System
initialization, page 122.
System startup
During system startup, an initialization sequence is executed before the main function
is entered. This sequence performs initializations required for the target hardware and
the C/C++ environment.
For the hardware initialization, it looks like this:
● When the CPU is reset it will start executing at the program entry label
_ _program_start in the system startup code.
● The external data and address buses are enabled if needed.
● The stack pointers are initialized to the end of CSTACK and RSTACK, respectively.
● The function _ _low_level_init is called if you defined it, giving the application
a chance to perform early initializations.
119
AFE1_AFE2-1:1
For the C/C++ initialization, it looks like this:
● Static and global variables are initialized except for _ _no_init, _ _tinyflash,
_ _flash, _ _farflash, _ _hugeflash, and _ _eeprom declared variables. That is,
zero-initialized variables are cleared and the values of other initialized variables are
copied from ROM to RAM memory. This step is skipped if _ _low_level_init
returns zero. For more information, see Initialization at system startup, page 75.
● Static C++ objects are constructed
● The main function is called, which starts the application.

120 for AVR
AFE1_AFE2-1:1
System termination
This illustration shows the different ways an embedded application can terminate in a
controlled way:
An application can terminate normally in two different ways:

● Return from the main function
● Call the exit function.
Because the C standard states that the two methods should be equivalent, the system
startup code calls the exit function if main returns. The parameter passed to the exit
function is the return value of main.
The default exit function is written in C. It calls a small assembler function _exit that
will:
● Call functions registered to be executed when the application ends. This includes
C++ destructors for static and global variables, and functions registered with the
standard function atexit.
● Close all open files
● Call _ _exit
● When _ _exit is reached, stop the system.
An application can also exit by calling the abort or the _Exit function. The abort
function just calls _ _exit to halt the system, and does not perform any type of cleanup.
The _Exit function is equivalent to the abort function, except for the fact that _Exit
takes an argument for passing exit status information.
If you want your application to do anything extra at exit, for example resetting the
system (and if using atexit is not sufficient), you can write your own implementation
of the _ _exit(int) function.
121
AFE1_AFE2-1:1
avr\src\lib directory. See Overriding library modules, page 107.
C-SPY debugging support for system termination

If you have enabled C-SPY emulated I/O during linking, the normal _ _exit and abort
functions are replaced with special ones. C-SPY will then recognize when those
functions are called and can take appropriate actions to emulate program termination.
For more information, see Briefly about C-SPY emulated I/O, page 102.
SYSTEM INITIALIZATION
It is likely that you need to adapt the system initialization. For example, your application
might need to initialize memory-mapped special function registers (SFRs), or omit the
default initialization of data segments performed by the system startup code.
You can do this by implementing your own version of the routine _ _low_level_init,
which is called from the file cstartup before the data segments are initialized.
Modifying the file cstartup.s90 directly should be avoided.
The code for handling system startup is located in the source files cstartup.s90 and
low_level_init.c, located in the avr\src\lib directory.
Note: Normally, you do not need to customize either of the files cstartup.s90 or
_exit.s90.
Note: Regardless of whether you implement your own version of _ _low_level_init

or the file cstartup.s90, you do not have to rebuild the library.
Customizing _ _low_level_init
You can customize the routine _ _low_level_init, which is called from the system
startup code before the data segments are initialized.
The value returned by _ _low_level_init determines whether or not data segments
should be initialized by the system startup code. If the function returns 0, the data
segments will not be initialized.
Note: The file intrinsics.h must be included by low_level_init.c to assure
correct behavior of the _ _low_level_init routine.
Modifying the cstartup file

As noted earlier, you should not modify the file cstartup.s90 if implementing your
own version of _ _low_level_init is enough for your needs. However, if you do need
to modify the file cstartup.s90, we recommend that you follow the general procedure
for creating a modified copy of the file and adding it to your project, see Overriding
library modules, page 107.

122 for AVR
AFE1_AFE2-1:1
Note that you must make sure that the linker uses the start label used in your version of
cstartup.s90. For information about how to change the start label used by the linker,
read about the -s option in the IAR Linker and Library Tools Reference Guide.
THE DLIB LOW-LEVEL I/O INTERFACE

The runtime library uses a set of low-level functions—which are referred to as the DLIB
low-level I/O interface—to communicate with the target system. Most of the low-level
functions have no implementation.
For more information about this, see Briefly about input and output (I/O), page 100.
This table lists the functions in the DLIB low-level I/O interface:
Function in DLIB low-level I/O interface Defined in source file

abort abort.c
clock clock.c
__close close.c
__exit xxexit.c
getenv getenv.c
__getzone getzone.c
__lseek lseek.c
__open open.c
raise raise.c
__read read.c
remove remove.c
rename rename.c
_ReportAssert xreportassert.c
signal signal.c
system system.c
__time32, __time64 time.c
__write write.c
Table 15: DLIB low-level I/O interface functions
Note: You should not use the low-level functions prefixed with _ _ directly in your
application. Instead you should use the standard library functions that use these
functions. For example, to write to stdout, you should use standard library functions
like printf or puts, which in turn calls the low-level function _ _write. If you have
forgot to implement a low-level function and your application will call that function via
123
AFE1_AFE2-1:1
a standard library function, the linker will issues an error when you link in release build
configuration.
Note: If you implement your own variants of the functions in this interface, your
variants will be used even though you have enabled C-SPY emulated I/O, see Briefly
about C-SPY emulated I/O, page 102.
abort
Source file avr\src\lib\abort.c
Description Standard C library function that aborts execution.
C-SPY debug action Notifies that the application has called abort.
Default implementation Calls __exit(EXIT_FAILURE).
See also Briefly about retargeting, page 102

System termination, page 121.
clock
Source file avr\src\lib\clock.c
Description Standard C library function that accesses the processor time.
C-SPY debug action Returns the clock on the host computer.
Default implementation Returns -1 to indicate that processor time is not available.
See also Briefly about retargeting, page 102.
__close
Source file avr\src\lib\close.c
Description Low-level function that closes a file.
C-SPY debug action Closes the associated host file on the host computer.
Default implementation None.

124 for AVR
AFE1_AFE2-1:1
__exit
Source file avr\src\lib\xxexit.c
Description Low-level function that halts execution.
C-SPY debug action Notifies that the end of the application was reached.
Default implementation Loops forever.
See also Briefly about retargeting, page 102

System termination, page 121.
getenv
Source file avr\src\lib\getenv.c
avr\src\lib\environ.c
C-SPY debug action Accesses the host environment.
Default implementation The getenv function in the library searches the string pointed to by the global variable
_ _environ, for the key that was passed as argument. If the key is found, the value of it
is returned, otherwise 0 (zero) is returned. By default, the string is empty.
To create or edit keys in the string, you must create a sequence of null-terminated strings
where each string has the format:
key=value\0
End the string with an extra null character (if you use a C string, this is added
automatically). Assign the created sequence of strings to the _ _environ variable.
For example:
const char MyEnv[] = "Key=Value\0Key2=Value2\0";
_ _environ = MyEnv;
If you need a more sophisticated environment variable handling, you should implement
your own getenv, and possibly putenv function.
Note: The putenv function is not required by the standard, and the library does not
provide an implementation of it.
125
AFE1_AFE2-1:1
__getzone
Source file avr\src\lib\getzone.c
Description Low-level function that returns the current time zone.
C-SPY debug action Not applicable.
Default implementation Returns ":".
__lseek
Source file avr\src\lib\lseek.c
Description Low-level function for changing the location of the next access in an open file.
C-SPY debug action Searches in the associated host file on the host computer.
__open
Source file avr\src\lib\open.c
Description Low-level function that opens a file.
C-SPY debug action Opens a file on the host computer.

126 for AVR
AFE1_AFE2-1:1
raise
Source file avr\src\lib\raise.c
Description Standard C library function that raises a signal.
Default implementation Calls the signal handler for the raised signal, or terminates with call to
_ _exit(EXIT_FAILURE).
__read
Source file avr\src\lib\read.c
Description Low-level function that reads characters from stdin and from files.
C-SPY debug action Directs stdin to the Terminal I/O window. All other files will read the associated host
file.
127
AFE1_AFE2-1:1
Example The code in this example uses memory-mapped I/O to read from a keyboard, whose port
is assumed to be located at 40:
#include <stddef.h>
__no_init volatile unsigned char kbIO @ 40;
size_t __read(int handle,

unsigned char *buf,
size_t bufSize)
{
size_t nChars = 0;
/* Check for stdin

(only necessary if FILE descriptors are enabled) */
if (handle != 0)
{
return -1;
}
for (/*Empty*/; bufSize > 0; --bufSize)

{
unsigned char c = kbIO;
if (c == 0)
break;
*buf++ = c;
++nChars;
}
return nChars;
}
For information about the handles associated with the streams, see Retargeting—
Adapting for your target system, page 106.
For information about the @ operator, see Controlling data and function placement in
memory, page 210.
remove
Source file avr\src\lib\remove.c
Description Standard C library function that removes a file.

128 for AVR
AFE1_AFE2-1:1
C-SPY debug action Writes a message to the Debug Log window and returns -1.
Default implementation Returns 0 to indicate success, but without removing a file.
rename
Source file avr\src\lib\rename.c
Description Standard C library function that renames a file.
C-SPY debug action None.
Default implementation Returns -1 to indicate failure.
_ReportAssert
Source file avr\src\lib\xreportassert.c
Description Low-level function that handles a failed assert.
C-SPY debug action Notifies the C-SPY debugger about the failed assert.
Default implementation Failed asserts are reported by the function _ _ReportAssert. By default, it prints an
error message and calls abort. If this is not the behavior you require, you can
implement your own version of the function.
The assert macro is defined in the header file assert.h. To turn off assertions, define
the symbol NDEBUG.
In the IDE, the symbol NDEBUG is by default defined in a Release project and not defined
in a Debug project. If you build from the command line, you must explicitly define the
symbol according to your needs. See NDEBUG, page 359.
signal
Source file avr\src\lib\signal.c
129
AFE1_AFE2-1:1
Description Standard C library function that changes signal handlers.
Default implementation As specified by Standard C. You might want to modify this behavior if the environment
supports some kind of asynchronous signals.
system
Source file avr\src\lib\system.c
Description Standard C library function that executes commands.
C-SPY debug action Notifies the C-SPY debugger that system has been called and then returns -1.
Default implementation If you need to use the system function, you must implement it yourself. The system
function available in the library returns 0 if a null pointer is passed to it to indicate that
there is no command processor; otherwise it returns -1 to indicate failure. If this is not
the functionality that you require, you can implement your own version. This does not
require that you rebuild the library.
__time32, __time64
Source file avr\src\lib\time.c
avr\src\lib\time64.c
Description Low-level functions that return the current calendar time.
C-SPY debug action Returns the time on the host computer.
Default implementation Returns -1 to indicate that calendar time is not available.
__write
Source file avr\src\lib\write.c

130 for AVR
AFE1_AFE2-1:1
Description Low-level function that writes to stdout, stderr, or a file.
C-SPY debug action Directs stdout and stderr to the Terminal I/O window. All other files will write to
the associated host file.
Example The code in this example uses memory-mapped I/O to write to an LCD display, whose
port is assumed to be located at address 40:
#include <stddef.h>
__no_init volatile unsigned char lcdIO @ 40;
size_t __write(int handle,

const unsigned char *buf,
size_t bufSize)
{
size_t nChars = 0;
/* Check for the command to flush all handles */

if (handle == -1)
{
return 0;
}
/* Check for stdout and stderr

(only necessary if FILE descriptors are enabled.) */
if (handle != 1 && handle != 2)
{
return -1;
}
for (/* Empty */; bufSize > 0; --bufSize)

{
lcdIO = *buf;
++buf;
++nChars;
}
return nChars;
}
For information about the handles associated with the streams, see Retargeting—
Adapting for your target system, page 106.
131
AFE1_AFE2-1:1
CONFIGURATION SYMBOLS FOR PRINTF AND SCANF

If the provided formatters do not meet your requirements (Formatters for printf, page
113 and Formatters for scanf, page 114), you can customize the Full formatters. Note
that this means you must rebuild the runtime library.
The default behavior of the printf and scanf formatters are defined by configuration
symbols in the file DLib_Defaults.h.
These configuration symbols determine what capabilities the function printf should
have:
Printf configuration symbols Includes support for

_DLIB_PRINTF_MULTIBYTE Multibyte characters
_DLIB_PRINTF_LONG_LONG Long long (ll qualifier)
_DLIB_PRINTF_SPECIFIER_FLOAT Floating-point numbers
_DLIB_PRINTF_SPECIFIER_A Hexadecimal floating-point numbers
_DLIB_PRINTF_SPECIFIER_N Output count (%n)
_DLIB_PRINTF_QUALIFIERS Qualifiers h, l, L, v, t, and z
_DLIB_PRINTF_FLAGS Flags -, +, #, and 0
_DLIB_PRINTF_WIDTH_AND_PRECISION Width and precision
_DLIB_PRINTF_CHAR_BY_CHAR Output char by char or buffered
Table 16: Descriptions of printf configuration symbols
When you build a library, these configurations determine what capabilities the function
scanf should have:
Scanf configuration symbols Includes support for

_DLIB_SCANF_MULTIBYTE Multibyte characters
_DLIB_SCANF_LONG_LONG Long long (ll qualifier)
_DLIB_SCANF_SPECIFIER_FLOAT Floating-point numbers
_DLIB_SCANF_SPECIFIER_N Output count (%n)
_DLIB_SCANF_QUALIFIERS Qualifiers h, j, l, t, z, and L
_DLIB_SCANF_SCANSET Scanset ([*])
_DLIB_SCANF_WIDTH Width
_DLIB_SCANF_ASSIGNMENT_SUPPRESSING Assignment suppressing ([*])
Table 17: Descriptions of scanf configuration symbols

132 for AVR
AFE1_AFE2-1:1
To customize the formatting capabilities, you must;

1 Define the configuration symbols in the library configuration file according to your
application requirements.
2 Rebuild the library, see Customizing and building your own runtime library, page 108.
CONFIGURATION SYMBOLS FOR FILE INPUT AND OUTPUT

File I/O is only supported by libraries with the Full library configuration, see Runtime
library configurations, page 110, or in a customized library when the configuration
symbol _ _DLIB_FILE_DESCRIPTOR is defined. If this symbol is not defined, functions
taking a FILE * argument cannot be used.
To customize your library and rebuild it, see Customizing and building your own
runtime library, page 108.
LOCALE
Locale is a part of the C language that allows language- and country-specific settings for
several areas, such as currency symbols, date and time, and multibyte character
encoding.
Depending on which library configuration you are using, you get different level of locale
support. However, the more locale support, the larger your code will get. It is therefore
necessary to consider what level of support your application needs. See Runtime library
The DLIB runtime library can be used in two main modes:
● With locale interface, which makes it possible to switch between different locales
during runtime
● Without locale interface, where one selected locale is hardwired into the
application.
Locale support in prebuilt libraries

The level of locale support in the prebuilt libraries depends on the library configuration.
● All prebuilt libraries support the C locale only
● All libraries with full library configuration have support for the locale interface. For
prebuilt libraries with locale interface, it is by default only supported to switch
multibyte character encoding at runtime.
● Libraries with normal library configuration do not have support for the locale
interface.
If your application requires a different locale support, you must rebuild the library.
133
AFE1_AFE2-1:1
Customizing the locale support and rebuilding the library

If you decide to rebuild the library, you can choose between these locales:
● The C locale
● The POSIX locale
● A wide range of European locales.
Locale configuration symbols

The configuration symbol _DLIB_FULL_LOCALE_SUPPORT, which is defined in the
library configuration file, determines whether a library has support for a locale interface
or not. The locale configuration symbols _LOCALE_USE_LANG_REGION and
_ENCODING_USE_ENCODING define all the supported locales and encodings:
#define _DLIB_FULL_LOCALE_SUPPORT 1
#define _LOCALE_USE_C /* C locale */
#define _LOCALE_USE_EN_US /* American English */
#define _LOCALE_USE_EN_GB /* British English */
#define _LOCALE_USE_SV_SE /* Swedish in Sweden */
See DLib_Defaults.h for a list of supported locale and encoding settings.

If you want to customize the locale support, you simply define the locale configuration
symbols required by your application. For more information, see Customizing and
building your own runtime library, page 108.
Note: If you use multibyte characters in your C or assembler source code, make sure
that you select the correct locale symbol (the local host locale).
Building a library without support for locale interface

The locale interface is not included if the configuration symbol
_DLIB_FULL_LOCALE_SUPPORT is set to 0 (zero). This means that a hardwired locale
is used—by default the Standard C locale—but you can choose one of the supported
locale configuration symbols. The setlocale function is not available and can
therefore not be used for changing locales at runtime.
Building a library with support for locale interface

Support for the locale interface is obtained if the configuration symbol
_DLIB_FULL_LOCALE_SUPPORT is set to 1. By default, the Standard C locale is used,
but you can define as many configuration symbols as required. Because the setlocale
function will be available in your application, it will be possible to switch locales at
runtime.

134 for AVR
AFE1_AFE2-1:1
Changing locales at runtime

The standard library function setlocale is used for selecting the appropriate portion
of the application’s locale when the application is running.
The setlocale function takes two arguments. The first one is a locale category that is
constructed after the pattern LC_CATEGORY. The second argument is a string that
describes the locale. It can either be a string previously returned by setlocale, or it
can be a string constructed after the pattern:
lang_REGION
or
lang_REGION.encoding
The lang part specifies the language code, and the REGION part specifies a region
qualifier, and encoding specifies the multibyte character encoding that should be used.
The lang_REGION part matches the _LOCALE_USE_LANG_REGION preprocessor
symbols that can be specified in the library configuration file.
Example
This example sets the locale configuration symbols to Swedish to be used in Finland and
UTF8 multibyte character encoding:
setlocale (LC_ALL, "sv_FI.Utf8");
STRTOD
The function strtod does not accept hexadecimal floating-point strings in libraries
with the normal library configuration. To make strtod accept hexadecimal
floating-point strings, you must:
1 Enable the configuration symbol _DLIB_STRTOD_HEX_FLOAT in the library
configuration file.
2 Rebuild the library, see Customizing and building your own runtime library, page 108.
Managing a multithreaded environment

In a multithreaded environment, the standard library must treat all library objects
according to whether they are global or local to a thread. If an object is a true global
object, any updates of its state must be guarded by a locking mechanism to make sure
that only one thread can update it at any given time. If an object is local to a thread, the
static variables containing the object state must reside in a variable area local to that
thread. This area is commonly named thread-local storage (TLS).
135
AFE1_AFE2-1:1
The low-level implementations of locks and TLS are system-specific, and is not
included in the DLIB runtime environment. If you are using an RTOS, check if it
provides some or all of the required functions. Otherwise, you must provide your own.
MULTITHREAD SUPPORT IN THE DLIB RUNTIME

ENVIRONMENT
The DLIB runtime environment uses two kinds of locks—system locks and file stream
locks. The file stream locks are used as guards when the state of a file stream is updated,
and are only needed in the Full library configuration. The following objects are guarded
with system locks:
● The heap (in other words when malloc, new, free, delete, realloc, or calloc
is used).
● The file system (only available in the Full library configuration), but not the file
streams themselves. The file system is updated when a stream is opened or closed,
in other words when fopen, fclose, fdopen, fflush, or freopen is used.
● The signal system (in other words when signal is used).
● The temporary file system (in other words when tmpnam is used).
● Dynamically initialized function local objects with static storage duration.
These library objects use TLS:
Library objects using TLS When these functions are used

Error functions errno, strerror
Locale functions localeconv, setlocale
Time functions asctime, localtime, gmtime, mktime
Multibyte functions mbrlen, mbrtowc, mbsrtowc, mbtowc, wcrtomb,
wcsrtomb, wctomb
Rand functions rand, srand
Miscellaneous functions atexit, strtok
Table 18: Library objects using TLS
Note: If you are using printf/scanf (or any variants) with formatters, each individual
formatter will be guarded, but the complete printf/scanf invocation will not be
guarded.
If one of the C++ variants is used together with the DLIB runtime environment with
multithread support, the compiler option --guard_calls must be used to make sure
that function-static variables with dynamic initializers are not initialized simultaneously
by several threads.

136 for AVR
AFE1_AFE2-1:1
ENABLING MULTITHREAD SUPPORT

There are no prebuilt runtime libraries with support for multithreading, which means
you must first customize the library by adding the line #define
_DLIB_THREAD_SUPPORT 3 in the library configuration file and rebuild your library.
See also Customizing and building your own runtime library, page 108.
To configure multithread support for use with threaded applications:
1 To enable multithread support:
On the command line, use the linker option --threaded_lib and use your own
customized library.
In the IDE, choose Project>Options>Linker>Extra Options and specify the linker
option --threaded_lib. If you are using C++, you must also choose
Project>Options>C/C++ Compiler>Extra Options and specify the compiler option
--guard_calls to make sure that function-static variables with dynamic initializers
are not initialized simultaneously by several threads.
2 To complement the built-in multithread support in the runtime library, you must also:
● Implement code for the library’s system locks interface, see System locks interface,
page 138.
● If file streams are used, implement code for the library’s file stream locks interface
or redirect the interface to the system locks interface (using the linker option -e),
see File streams locks interface, page 138.
● Implement code that handles thread creation, thread destruction, and TLS access
methods for the library, see TLS handling, page 138.
See also Lock usage, page 138 for information about used file streams locks and system
locks.
You can find the required declaration of functions and definitions of macros in the
DLib_Threads.h file, which is included by yvals.h.
3 Build your project.

Note: If you are using a third-party RTOS, check their guidelines for how to enable
multithread support with IAR Systems tools.
137
AFE1_AFE2-1:1
System locks interface

This interface must be fully implemented for system locks to work:
typedef void *__iar_Rmtx; /* Lock info object */
void __iar_system_Mtxinit(__iar_Rmtx *); /* Initialize a system

lock */
void __iar_system_Mtxdst(__iar_Rmtx *);/*Destroy a system lock */
void __iar_system_Mtxlock(__iar_Rmtx *); /* Lock a system lock */
void __iar_system_Mtxunlock(__iar_Rmtx *); /* Unlock a system
lock */
The lock and unlock implementation must survive nested calls.
File streams locks interface

This interface is only needed for the Full library configuration. If file streams are used,
they can either be fully implemented or they can be redirected to the system locks
interface. This interface must be implemented for file streams locks to work:
typedef void *__iar_Rmtx; /* Lock info object */
void __iar_file_Mtxinit(__iar_Rmtx *);/*Initialize a file lock */

void __iar_file_Mtxdst(__iar_Rmtx *); /* Destroy a file lock */
void __iar_file_Mtxlock(__iar_Rmtx *); /* Lock a file lock */
void __iar_file_Mtxunlock(__iar_Rmtx *); /* Unlock a file lock */
The lock and unlock implementation must survive nested calls.
Lock usage
The number of locks that the DLIB runtime environment assumes exist are:
● _FOPEN_MAX—the maximum number of file stream locks. These locks are only
used in the Full library configuration, in other words only if both the macro symbols
_DLIB_FILE_DESCRIPTOR and _FILE_OP_LOCKS are true.
● _MAX_LOCK—the maximum number of system locks.
Note that even if the application uses fewer locks, the runtime environment will initialize
and destroy all the locks above.
For information about the initialization and destruction code, see xsyslock.c.
TLS handling
The DLIB runtime environment supports TLS memory areas for two types of threads:
the main thread (the main function including the system startup and exit code) and
secondary threads.

138 for AVR
AFE1_AFE2-1:1
The main thread’s TLS memory area:

● Is automatically created and initialized by your application’s startup sequence
● Is automatically destructed by the application’s destruct sequence
● Is located in the segment
● Exists also for non-threaded applications.
Each secondary thread’s TLS memory area:
● Must be manually created and initialized
● Must be manually destructed
● Is located in a manually allocated memory area.
If you need the runtime library to support secondary threads, you must override the
function:
void *_ _iar_dlib_perthread_access(void *symbp);
The parameter is the address to the TLS variable to be accessed—in the main thread’s
TLS area—and it should return the address to the symbol in the current TLS area.
Two interfaces can be used for creating and destroying secondary threads. You can use
the following interface that allocates a memory area on the heap and initializes it. At
deallocation, it destroys the objects in the area and then frees the memory.
void *__iar_dlib_perthread_allocate(void);
void __iar_dlib_perthread_deallocate(void *);
Alternatively, if the application handles the TLS allocation, you can use this interface
for initializing and destroying the objects in the memory area:
void __iar_dlib_perthread_initialize(void *);
void __iar_dlib_perthread_destroy(void *);
These macros can be helpful when you implement an interface for creating and
destroying secondary threads:
Macro Description
_ _IAR_DLIB_PERTHREAD_SIZE The size needed for the TLS memory area.
_ _IAR_DLIB_PERTHREAD_INIT_SIZE The initializer size for the TLS memory area.
You should initialize the rest of the TLS
memory area, up to
__IAR_DLIB_PERTHREAD_SIZE to zero.
_ _IAR_DLIB_PERTHREAD_SYMBOL_ The offset to the symbol in the TLS memory
OFFSET(symbolptr) area.
Table 19: Macros for implementing TLS allocation
139
AFE1_AFE2-1:1
Note that the size needed for TLS variables depends on which resources in the DLIB
runtime environment your application uses.
This is an example of how you can handle threads:
#include <yvals.h>
/* A thread's TLS pointer */

void _DLIB_TLS_MEMORY *TLSp;
/* Are we in a secondary thread? */

int InSecondaryThread = 0;
/* Allocate a thread-local TLS memory

area and store a pointer to it in TLSp. */
void AllocateTLS()
{
TLSp = __iar_dlib_perthread_allocate();
}
/* Deallocate the thread-local TLS memory area. */

void DeallocateTLS()
{
__iar_dlib_perthread_deallocate(TLSp);
}
/* Access an object in the

thread-local TLS memory area. */
void _DLIB_TLS_MEMORY *__iar_dlib_perthread_access(
void _DLIB_TLS_MEMORY *symbp)
{
char _DLIB_TLS_MEMORY *p = 0;
if (InSecondaryThread)
p = (char _DLIB_TLS_MEMORY *) TLSp;
else
p = (char _DLIB_TLS_MEMORY *)
__segment_begin("__DLIB_PERTHREAD");
p += __IAR_DLIB_PERTHREAD_SYMBOL_OFFSET(symbp);
return (void _DLIB_TLS_MEMORY *) p;
}
The TLSp variable is unique for each thread, and must be exchanged by the RTOS or
manually whenever a thread switch occurs.

140 for AVR
AFE1_AFE2-1:1
The CLIB runtime
environment
● Using a prebuilt runtime library
● Input and output
● System startup and termination
● Overriding default library modules
● Customizing system initialization
● C-SPY emulated I/O
Note that the CLIB runtime environment does not support any C99
functionality. For example, complex numbers and variable length arrays are not
supported. Neither does CLIB support C++.
The legacy CLIB runtime environment is provided for backward compatibility

and should not be used for new application projects.
For information about migrating from CLIB to DLIB, see the IAR Embedded
Workbench® Migration Guide.
Using a prebuilt runtime library

The prebuilt runtime libraries are configured for different combinations of these
features:
● Type of library
● Processor option (-v)
● Memory model option (--memory_model)
● AVR enhanced core option (--enhanced_core)
● Small flash memory option (--64k_flash)
● 64-bit doubles option (--64bit_doubles).
141
AFE1_AFE2-1:1
Input and output
CHOOSING A RUNTIME LIBRARY

The IDE includes the correct runtime library based on the options you select. See the
IDE Project Management and Building Guide for AVR for more information.
Specify which runtime library file to use on the XLINK command line, for instance:
cllibname.r90
The CLIB runtime environment includes the C standard library. The linker will include
only those routines that are required—directly or indirectly—by your application. For
more information about the runtime libraries, see the chapter C/C++ standard library
functions.
RUNTIME LIBRARY FILENAME SYNTAX

The runtime library names are constructed in this way:
{type}{cpu}{memModel}-{core}-{smallFlash}-{64BitDoubles}.r90
where
● {type} cl for the CLIB runtime library
● {cpu} is a value from 0 to 6, matching the -v option
● {memModel} is one of t, s, l, or h for the Tiny, Small, Large, or Huge memory
model, respectively
● {smallFlash} is sf when the small flash memory is available. When small flash
memory is not available, this value is not specified
● {64BitDoubles} is 64 when 64-bit doubles are used. When 32-bit doubles are
used, this value is not specified.
These are some examples of how to decode a library name:
Generic
Memory 64-bit
Library file processor Core Small flash
model doubles
option
cl0t.r90 -v0 Tiny -- -- --
cl1s-64.r90 -v1 Small -- -- X
cl61-ec_mul-64.r90 -v6 Large X -- X
Table 20: Runtime libraries
Input and output

You can customize:
● The functions related to character-based I/O

142 for AVR
AFE1_AFE2-1:1
The CLIB runtime environment
● The formatters used by printf/sprintf and scanf/sscanf.
CHARACTER-BASED I/O
The functions putchar and getchar are the fundamental C functions for
character-based I/O. For any character-based I/O to be available, you must provide
definitions for these two functions, using whatever facilities the hardware environment
provides.
The creation of new I/O routines is based on these files:
● putchar.c, which serves as the low-level part of functions such as printf
● getchar.c, which serves as the low-level part of functions such as scanf.
The code example below shows how memory-mapped I/O could be used to write to a
memory-mapped I/O device:
__no_init volatile unsigned char devIO @ 40;
int putchar(int outChar)

{
devIO = outChar;
return outChar;
}
The exact address is a design decision. For example, it can depend on the selected
processor variant.
For information about how to include your own modified version of putchar and
getchar in your project build process, see Overriding library modules, page 107.
FORMATTERS USED BY PRINTF AND SPRINTF

The printf and sprintf functions use a common formatter, called
_formatted_write. There are three variants of the formatter:
_large_write
_medium_write
_small_write
By default, the linker automatically uses the most appropriate formatter for your
application.
_large_write
The _large_write formatter supports the C89 printf format directives.
143
AFE1_AFE2-1:1
Input and output
_medium_write
The _medium_write formatter has the same format directives as _large_write,
except that floating-point numbers are not supported. Any attempt to use a %f, %g, %G,
%e, or %E specifier will produce a runtime error:
FLOATS? wrong formatter installed!
_medium_write is considerably smaller than the large version.
_small_write
The _small_write formatter works in the same way as _medium_write, except that
it supports only the %%, %d, %o, %c, %s, and %x specifiers for integer objects, and does
not support field width or precision arguments. The size of _small_write is 10–15%
that of _medium_write.
Specifying the printf formatter in the IDE

1 Choose Project>Options and select the General Options category. Click the Library
Options tab.
2 Choose the appropriate Printf formatter option, which can be either Auto, Small,
Medium, or Large.
Specifying the printf formatter from the command line

To explicitly specify and override the formatter used by default, add one of the following
lines to the linker configuration file:
-e_small_write=_formatted_write
-e_medium_write=_formatted_write
-e_large_write=_formatted_write
Customizing printf
For many embedded applications, sprintf is not required, and even printf with
_small_write provides more facilities than are justified, considering the amount of
memory it consumes. Alternatively, a custom output routine might be required to
support particular formatting needs or non-standard output devices.
For such applications, a much reduced version of the printf function (without
sprintf) is supplied in source form in the file intwri.c. This file can be modified to
meet your requirements, and the compiled module inserted into the library in place of
the original file; see Overriding library modules, page 107.

144 for AVR
AFE1_AFE2-1:1
FORMATTERS USED BY SCANF AND SSCANF

As with the printf and sprintf functions, scanf and sscanf use a common
formatter, called _formatted_read. There are two variants of the formatter:
_large_read
_medium_read
By default, the linker automatically uses the most appropriate formatter for your
application.
_large_read
The _large_read formatter supports the C89 scanf format directives.
_medium_read
The _medium_read formatter has the same format directives as the large version,
except that floating-point numbers are not supported. _medium_read is considerably
smaller than the large version.
Specifying the scanf formatter in the IDE

1 Choose Project>Options and select the General Options category. Click the Library
Options tab.
2 Choose the appropriate Scanf formatter option, which can be either Auto, Medium or
Large.
Specifying the read formatter from the command line

To explicitly specify and override the formatter used by default, add one of the following
lines to the linker configuration file:
-e_medium_read=_formatted_read
-e_large_read=_formatted_read
System startup and termination

This section describes the actions the runtime environment performs during startup and
termination of applications.
The code for handling startup and termination is located in the source files
cstartup.s90, _exit.s90 and low_level_init.c located in the avr\src\lib
directory.
Note: Normally, you do not need to customize either of the files cstartup.s90 or
_exit.s90.
145
AFE1_AFE2-1:1
Overriding default library modules
SYSTEM STARTUP
When an application is initialized, several steps are performed:
● The custom function __low_level_init is called, giving the application a chance
to perform early initializations
● The external data and address buses are enabled if needed
● The stack pointers are initialized to the end of CSTACK and RSTACK, respectively
● Static variables are initialized except for __no_init __tinyflash, __flash,
__farflash, __hugeflash, and __eeprom declared variables; this includes
clearing zero-initialized memory and copying the ROM image of the RAM memory
of the remaining initialized variables
● The main function is called, which starts the application.
Note that the system startup code contains code for more steps than described here. The
other steps are applicable to the DLIB runtime environment.
SYSTEM TERMINATION
An application can terminate normally in two different ways:
● Return from the main function
● Call the exit function.
Because the C standard states that the two methods should be equivalent, the cstartup
code calls the exit function if main returns. The parameter passed to the exit function
is the return value of main. The default exit function is written in assembler.
When the application is built in debug mode, C-SPY stops when it reaches the special
code label ?C_EXIT.
An application can also exit by calling the abort function. The default function just
calls __exit to halt the system, without performing any type of cleanup.
Overriding default library modules

The IAR CLIB Library contains modules which you probably need to override with
your own customized modules, for example for character-based I/O, without rebuilding
the entire library. For information about how to override default library modules, see
Overriding library modules, page 107, in the chapter The DLIB runtime environment.

146 for AVR
AFE1_AFE2-1:1
Customizing system initialization

For information about how to customize system initialization, see System initialization,
page 122.
C-SPY emulated I/O

The low-level I/O interface is used for communication between the application being
debugged and the debugger itself. The interface is simple: C-SPY will place breakpoints
on certain assembler labels in the application. When code located at the special labels is
about to be executed, C-SPY will be notified and can perform an action.
THE DEBUGGER TERMINAL I/O WINDOW

When code at the labels ?C_PUTCHAR and ?C_GETCHAR is executed, data will be sent to
or read from the debugger window.
For the ?C_PUTCHAR routine, one character is taken from the output stream and written.
If everything goes well, the character itself is returned, otherwise -1 is returned.
When the label ?C_GETCHAR is reached, C-SPY returns the next character in the input
field. If no input is given, C-SPY waits until the user types some input and presses the
Return key.
To make the Terminal I/O window available, the application must be linked with the
XLINK option With I/O emulation modules selected. See the IDE Project
TERMINATION
The debugger stops executing when it reaches the special label ?C_EXIT.
147
AFE1_AFE2-1:1
C-SPY emulated I/O

148 for AVR
AFE1_AFE2-1:1
Assembler language
interface
● Mixing C and assembler
● Calling assembler routines from C
● Calling assembler routines from C++
● Calling convention
● Call frame information
Mixing C and assembler

The IAR C/C++ Compiler for AVR provides several ways to access low-level resources:
● Modules written entirely in assembler
● Intrinsic functions (the C alternative)
● Inline assembler.
It might be tempting to use simple inline assembler. However, you should carefully
choose which method to use.
INTRINSIC FUNCTIONS
The compiler provides a few predefined functions that allow direct access to low-level
processor operations without having to use the assembler language. These functions are
known as intrinsic functions. They can be very useful in, for example, time-critical
routines.
An intrinsic function looks like a normal function call, but it is really a built-in function
that the compiler recognizes. The intrinsic functions compile into inline code, either as
a single instruction, or as a short sequence of instructions.
The advantage of an intrinsic function compared to using inline assembler is that the
compiler has all necessary information to interface the sequence properly with register
allocation and variables. The compiler also knows how to optimize functions with such
sequences; something the compiler is unable to do with inline assembler sequences. The
result is that you get the desired sequence properly integrated in your code, and that the
compiler can optimize the result.
149
AFE1_AFE2-1:1
Mixing C and assembler
For more information about the available intrinsic functions, see the chapter Intrinsic
functions.
MIXING C AND ASSEMBLER MODULES

It is possible to write parts of your application in assembler and mix them with your C
or C++ modules.
This gives several benefits compared to using inline assembler:
● The function call mechanism is well-defined
● The code will be easy to read
● The optimizer can work with the C or C++ functions.
This causes some overhead in the form of a function call and return instruction
sequences, and the compiler will regard some registers as scratch registers. However, the
compiler will also assume that all scratch registers are destroyed when calling a function
written in assembler. In many cases, the overhead of the extra instructions is
compensated by the work of the optimizer.
An important advantage is that you will have a well-defined interface between what the
compiler produces and what you write in assembler. When using inline assembler, you
will not have any guarantees that your inline assembler lines do not interfere with the
compiler generated code.
When an application is written partly in assembler language and partly in C or C++, you
are faced with several questions:
● How should the assembler code be written so that it can be called from C?
● Where does the assembler code find its parameters, and how is the return value
passed back to the caller?
● How should assembler code call functions written in C?
● How are global C variables accessed from code written in assembler language?
● Why does not the debugger display the call stack when assembler code is being
debugged?
The first question is discussed in the section Calling assembler routines from C, page
152. The following two are covered in the section Calling convention, page 155.
The answer to the final question is that the call stack can be displayed when you run
assembler code in the debugger. However, the debugger requires information about the
call frame, which must be supplied as annotations in the assembler source file. For more
information, see Call frame information, page 165.

150 for AVR
AFE1_AFE2-1:1
Assembler language interface
The recommended method for mixing C or C++ and assembler modules is described in
Calling assembler routines from C, page 152, and Calling assembler routines from
C++, page 154, respectively.
INLINE ASSEMBLER
Inline assembler can be used for inserting assembler instructions directly into a C or
C++ function.
The asm extended keyword and its alias __asm both insert assembler instructions.
However, when you compile C source code, the asm keyword is not available when the
option --strict is used. The __asm keyword is always available.
Note: Not all assembler directives or operators can be inserted using these keywords.
The syntax is:
asm ("string");
The string can be a valid assembler instruction or a data definition assembler directive,
but not a comment. You can write several consecutive inline assembler instructions, for
example:
asm("label:nop\n"
"jmp label");
where \n (new line) separates each new assembler instruction. Note that you can define
and use local labels in inline assembler instructions.
The following example demonstrates the use of the asm keyword. This example also
shows the risks of using inline assembler.
bool flag;
void Func()
{
while (!flag)
{
asm("IN R0, PIND \n"
"STS flag, R0");
}
}
Note: Because using symbols from inside the inline assembler code is not properly
visible to all parts of the compiler, you must use #pragma required when you
reference an external or module-local symbol only from inline assembler code. If you
do not, you can get an undefined symbol error when compiling. See required, page 334.
151
AFE1_AFE2-1:1
Calling assembler routines from C
Additionally, in this example, the assignment to the global variable flag is not noticed
by the compiler, which means the surrounding code cannot be expected to rely on the
inline assembler statement.
The inline assembler instruction will simply be inserted at the given location in the
program flow. The consequences or side-effects the insertion might have on the
surrounding code are not taken into consideration. If, for example, registers or memory
locations are altered, they might have to be restored within the sequence of inline
assembler instructions for the rest of the code to work properly.
Inline assembler sequences have no well-defined interface with the surrounding code
generated from your C or C++ code. This makes the inline assembler code fragile, and
might also become a maintenance problem if you upgrade the compiler in the future.
There are also several limitations to using inline assembler:
● The compiler’s various optimizations will disregard any effects of the inline
sequences, which will not be optimized at all
● In general, assembler directives will cause errors or have no meaning. Data
definition directives will however work as expected
● Auto variables cannot be accessed.
Inline assembler is therefore often best avoided. If no suitable intrinsic function is
available, we recommend that you use modules written in assembler language instead
of inline assembler, because the function call to an assembler routine normally causes
less performance reduction.
Calling assembler routines from C

An assembler routine that will be called from C must:
● Conform to the calling convention
● Have a PUBLIC entry-point label
● Be declared as external before any call, to allow type checking and optional
promotion of parameters, as in these examples:
extern int foo(void);
or
extern int foo(int i, int j);
One way of fulfilling these requirements is to create skeleton code in C, compile it, and
study the assembler list file.

152 for AVR
AFE1_AFE2-1:1
CREATING SKELETON CODE

The recommended way to create an assembler language routine with the correct
interface is to start with an assembler language source file created by the C compiler.
Note that you must create skeleton code for each function prototype.
The following example shows how to create skeleton code to which you can easily add
the functional body of the routine. The skeleton source code only needs to declare the
variables required and perform simple accesses to them. In this example, the assembler
routine takes an int and a char, and then returns an int:
extern int gInt;
extern char gChar;
int Func(int arg1, char arg2)

{
int locInt = arg1;
gInt = arg1;
gChar = arg2;
return locInt;
}
int main()
{
int locInt = gInt;
gInt = Func(locInt, gChar);
return 0;
}
Note: In this example we use a low optimization level when compiling the code to show
local and global variable access. If a higher level of optimization is used, the required
references to local variables could be removed during the optimization. The actual
function declaration is not changed by the optimization level.
COMPILING THE SKELETON CODE

In the IDE, specify list options on file level. Select the file in the workspace window.
Then choose Project>Options. In the C/C++ Compiler category, select Override
inherited settings. On the List page, deselect Output list file, and instead select the
Output assembler file option and its suboption Include source. Also, be sure to specify
a low level of optimization.
Use these options to compile the skeleton code:
iccavr skeleton.c -lA . -On -e
The -lA option creates an assembler language output file including C or C++ source
lines as assembler comments. The . (period) specifies that the assembler file should be
153
AFE1_AFE2-1:1
Calling assembler routines from C++
named in the same way as the C or C++ module (skeleton), but with the filename
extension s90. The -On option means that no optimization will be used and -e enables
language extensions. In addition, make sure to use relevant compiler options, usually the
same as you use for other C or C++ source files in your project.
The result is the assembler source output file skeleton.s90.
Note: The -lA option creates a list file containing call frame information (CFI)
directives, which can be useful if you intend to study these directives and how they are
used. If you only want to study the calling convention, you can exclude the CFI
directives from the list file.
In the IDE, choose Project>Options>C/C++ Compiler>List and deselect the
suboption Include call frame information.
On the command line, use the option -lB instead of -lA. Note that CFI information
must be included in the source code to make the C-SPY Call Stack window work.
The output file

The output file contains the following important information:
● The calling convention
● The return values
● The global variables
● The function parameters
● How to create space on the stack (auto variables)
● Call frame information (CFI).
The CFI directives describe the call frame information needed by the Call Stack
window in the debugger. For more information, see Call frame information, page 165.
Calling assembler routines from C++

The C calling convention does not apply to C++ functions. Most importantly, a function
name is not sufficient to identify a C++ function. The scope and the type of the function
are also required to guarantee type-safe linkage, and to resolve overloading.
Another difference is that non-static member functions get an extra, hidden argument,
the this pointer.

154 for AVR
AFE1_AFE2-1:1
However, when using C linkage, the calling convention conforms to the C calling
convention. An assembler routine can therefore be called from C++ when declared in
this manner:
extern "C"
{
int MyRoutine(int);
}
In C++, data structures that only use C features are known as PODs (“plain old data
structures”), they use the same memory layout as in C. However, we do not recommend
that you access non-PODs from assembler routines.
The following example shows how to achieve the equivalent to a non-static member
function, which means that the implicit this pointer must be made explicit. It is also
possible to “wrap” the call to the assembler routine in a member function. Use an inline
member function to remove the overhead of the extra call—this assumes that function
inlining is enabled:
class MyClass;
extern "C"
{
void DoIt(MyClass *ptr, int arg);
}
class MyClass
{
public:
inline void DoIt(int arg)
{
::DoIt(this, arg);
}
};
Note: Support for C++ names from assembler code is extremely limited. This means
that:
● Assembler list files resulting from compiling C++ files cannot, in general, be passed
through the assembler.
● It is not possible to refer to or define C++ functions that do not have C linkage in
assembler.
Calling convention
A calling convention is the way a function in a program calls another function. The
compiler handles this automatically, but, if a function is written in assembler language,
155
AFE1_AFE2-1:1
Calling convention
you must know where and how its parameters can be found, how to return to the program
location from where it was called, and how to return the resulting value.
It is also important to know which registers an assembler-level routine must preserve. If
the program preserves too many registers, the program might be ineffective. If it
preserves too few registers, the result would be an incorrect program.
The compiler provides two calling conventions—one old, which is used in version 1.x
of the compiler, and one new which is the default. This section describes the calling
conventions used by the compiler. These items are examined:
● Choosing a calling convention
● Function declarations
● C and C++ linkage
● Preserved versus scratch registers
● Function entrance
● Function exit
● Return address handling
At the end of the section, some examples are shown to describe the calling convention
in practice.
CHOOSING A CALLING CONVENTION

You can choose between these calling conventions:
● The old calling convention offers a simple assembler interface. It is compatible with
the calling convention used in version 1.x of the compiler. Even though this
convention is not used by default, it is recommended for use when mixing C and
assembler code
● The new calling convention is default. It is more efficient than the old calling
convention, but also more complex to understand and subject to change in later
versions of the compiler.
To choose another calling convention than the default, use the --version1_calls
command line option. You can also declare individual functions to use the old calling
convention by using the __version1 function attribute, for example:
extern
__version_1 void doit(int arg);
For details about the --version1_calls option and the __version1 attribute, see
--version1_calls, page 276 and __version_1, page 315, respectively.
In the IDE, choose Use ICCA90 1.x calling convention on the Project>C/C++
Compiler>Code page.

156 for AVR
AFE1_AFE2-1:1
Hints for using the default calling convention

The default calling convention is very complex, and therefore not recommended for use
when calling assembler routines. However, if you intend to use it for your assembler
routines, you should create a list file and see how the compiler assigns the different
parameters to the available registers. For an example, see Creating skeleton code, page
153.
If you intend to use the default calling convention, you should also specify a value to the
runtime model attribute __rt_version using the RTMODEL assembler directive:
RTMODEL "__rt_version", "value"
The parameter value should have the same value as the one used internally by the
compiler. For information about what value to use, see the generated list file. If the
calling convention changes in future compiler versions, the runtime model value used
internally by the compiler will also change. Using this method gives a module
consistency check, because the linker produces errors for mismatches between the
values.
FUNCTION DECLARATIONS
In C, a function must be declared in order for the compiler to know how to call it. A
declaration could look as follows:
int MyFunction(int first, char * second);
This means that the function takes two parameters: an integer and a pointer to a
character. The function returns a value, an integer.
In the general case, this is the only knowledge that the compiler has about a function.
Therefore, it must be able to deduce the calling convention from this information.
USING C LINKAGE IN C++ SOURCE CODE

In C++, a function can have either C or C++ linkage. To call assembler routines from
C++, it is easiest if you make the C++ function have C linkage.
This is an example of a declaration of a function with C linkage:
extern "C"
{
int F(int);
}
157
AFE1_AFE2-1:1
Calling convention
It is often practical to share header files between C and C++. This is an example of a
declaration that declares a function with C linkage in both C and C++:
#ifdef __cplusplus
extern "C"
{
#endif
int F(int);
#ifdef __cplusplus
}
#endif
PRESERVED VERSUS SCRATCH REGISTERS

The general AVR CPU registers are divided into three separate sets, which are described
in this section.
Scratch registers
Any function is permitted to destroy the contents of a scratch register. If a function needs
the register value after a call to another function, it must store it during the call, for
example on the stack.
For both calling conventions, these 14 registers can be used as scratch registers by a
function:
R0-R3, R16-R23, and R30-R31
Preserved registers
Preserved registers, on the other hand, are preserved across function calls. The called
function can use the register for other purposes, but must save the value before using the
register and restore it at the exit of the function.
For both calling conventions, these registers are preserved registers:
R4-R15 and R24-R27
Note that the registers R4-R15 can be locked from the command line and used for global
register variables; see --lock_regs, page 258 and __regvar, page 312.

158 for AVR
AFE1_AFE2-1:1
Special registers
For some registers, you must consider certain prerequisites:
● The stack pointer—register Y—must at all times point to the last element on the
stack. In the eventuality of an interrupt, everything below the point the stack pointer
points to, will be destroyed.
● If using the -v4 or -v6 processor option, the RAMPY register is part of the data stack
pointer.
● If using a processor option which uses any of the registers EIND, RAMPX, or RAMPZ,
these registers are treated as scratch registers.
FUNCTION ENTRANCE
During a function call, the calling function:
● passes the parameters, either in registers or on the stack
● pushes any other parameters on the data stack (CSTACK)
Control is then passed to the called function with the return address being automatically
pushed on the return address stack (RSTACK).
The called function:
● stores any local registers required by the function on the data stack
● allocates space for its auto variables and temporary values
● proceeds to run the function itself.
Register parameters versus stack parameters

Parameters can be passed to a function using one of two basic methods: in registers or
on the stack. It is much more efficient to use registers than to take a detour via memory,
so the calling convention is designed to use registers as much as possible. Only a limited
number of registers can be used for passing parameters; when no more registers are
available, the remaining parameters are passed on the stack.
Hidden parameters
In addition to the parameters visible in a function declaration and definition, there can
be hidden parameters:
● A function returning structures or unions larger than 4 bytes gets an extra hidden
parameter, which is a default pointer—depending on the used memory model—
pointing to the location where the result should be stored. This pointer must be
returned to the callee.
159
AFE1_AFE2-1:1
Calling convention
● For non-static C++ member functions, the this pointer is passed as the first
parameter (but placed after the return structure pointer, if there is one). Note that
static member functions do not have a this pointer.
Register parameters
For both calling conventions, the registers available for passing parameters are:
R16–R23
Parameters are allocated to registers using a first-fit algorithm, using parameter

alignment requirements according to this table:
Parameters Alignment Passed in registers
8-bit values 1 R16, R17, R18, R19, R20, R21, R22, R23
16-bit values 2 R17:R16, R19:R18, R21:R20, R23:R22
24-bit values 4 R18:R17:R16, R22:R21:R20
32-bit values 4 R19:R18:R17:R16, R23:R22:R21:R20
64-bit values 8 R23:R22:R21:R20:R19:R18:R17:R16
Table 21: Registers used for passing parameters
Register assignment using the default calling convention

In the default calling convention, as many parameters as possible are passed in registers.
The remaining parameters are passed on the stack. The compiler may change the order
of the parameters in order to achieve the most efficient register usage.
The algorithm for assigning parameters to registers is quite complex in the default
calling convention. For details, you should create a list file and see how the compiler
assigns the different parameters to the available registers, see Creating skeleton code,
page 153.
Below follows some examples of combinations of register assignment.
A function with the following signature (not C++):
void Func(char __far * a, int b, char c, int d)
would have a allocated to R18:R17:R16, b to R21:R20 (alignment requirement

prevents R20:R19), c to R19 (first fit), and d to R23:R22 (first fit).
Another example:
void bar(char a, int b, long c, char d)
This would result in a being allocated to R16 (first fit), b to R19:R18 (alignment), c to
R23:R22:R21:R20 (first fit), and d to R17 (first fit).

160 for AVR
AFE1_AFE2-1:1
A third example:
void baz(char a, char __far * b, int c, int d)
This would give, a being allocated to R16, b to R22:R21:R20, c to R19:R18, and d to

the stack.
Register assignment using the old calling convention

In the old calling convention, the two left-most parameters are passed in registers if they
are scalar and up to 32 bits in size.
This table shows some of the possible combinations:
Parameters* Parameter 1 Parameter 2
f(b1,b2,…) R16 R20
f(b1,w2,…) R16 R20, R21
f(w1,l1,…) R16, R17 R20, R21, R22, R23
f(l1,b2,…) R16, R17, R18, R19 R20
f(l1,l2,…) R16, R17, R18, R19 R20, R21, R22, R23
Table 22: Passing parameters in registers
* Where b denotes an 8-bit data type, w denotes a 16-bit data type, and l denotes a 32-bit
data type. If the first and/or second parameter is a 3-byte pointer, it will be passed in
R16–R19 or R20–R22 respectively.
161
AFE1_AFE2-1:1
Calling convention
Stack parameters and layout

A function call creates a stack frame as follows:
High
address Stack parameters
Saved registers
Auto variables
Temporary storage
Low
address Y
Note that only registers that are used will be saved.
FUNCTION EXIT
A function can return a value to the function or program that called it, or it can have the
return type void.
The return value of a function, if any, can be scalar (such as integers and pointers),
floating-point, or a structure.
Registers used for returning values

For the old calling convention, the registers available for returning values are R16-R19.
The default calling convention can use the registers R16-R23 for returning values.
Return values Passed in registers
8-bit values R16
16-bit values R17:R16
24-bit values R18:R17:R16
32-bit values R19:R18:R17:R16
64-bit values R23:R22:R21:R20:R19:R18:R17:R16
Table 23: Registers used for returning values

162 for AVR
AFE1_AFE2-1:1
Note that the size of a returned pointer depends on the memory model in use;
appropriate registers are used accordingly.
Stack layout at function exit

Normally, it is the responsibility of the called function to clean the stack. The only
exception is for ellipse functions—functions with a variable argument list such as
printf—for which it is the responsibility of the caller to clean the stack.
Return address handling

A function written in assembler language should, when finished, return to the caller. At
a function call, the return address is automatically stored on the RSTACK (not the
CSTACK).
Typically, a function returns by using the RET instruction.
RESTRICTIONS FOR SPECIAL FUNCTION TYPES
Interrupt functions
Interrupt functions differ from ordinary C functions in that:
● If used, flags and scratch registers are saved
● Calls to interrupt functions are made via interrupt vectors; direct calls are not
allowed
● No arguments can be passed to an interrupt function
● Interrupt functions return by using the RETI function.
For more information about interrupt functions, see Interrupt functions, page 64.
Monitor functions
A monitor function causes interrupts to be disabled during execution of the function. At
function entry, the status register SREG is saved and global interrupts are disabled. At
function exit, the global interrupt enable bit (I) is restored in the SREG register, and
thereby the interrupt status existing before the function call is also restored.
For more information about monitor functions, see Monitor functions, page 65.
EXAMPLES
The following section shows a series of declaration examples and the corresponding
calling conventions. The complexity of the examples increases toward the end.
163
AFE1_AFE2-1:1
Calling convention
Example 1
Assume this function declaration:
int add1(int);
This function takes one parameter in the register R17:R16, and the return value is passed
back to its caller in the register R17:R16.
This assembler routine is compatible with the declaration; it will return a value that is
one number higher than the value of its parameter:
subi R16,FF
sbci R17,FF
ret
Example 2
This example shows how structures are passed on the stack. Assume these declarations:
struct MyStruct
{
long long mA;
long long mB;
};
int MyFunction(struct MyStruct x, int y);
The calling function must reserve 16 bytes on the top of the stack and copy the contents
of the struct to that location. The integer parameter y is passed in the register
R17:R16. The return value is passed back to its caller in the register R17:R16.
Example 3
The function below will return a structure of type struct MyStruct.
struct MyStruct
{
long long mA;
long long mB;
};
int MyFunction(struct MyStruct x, int y);
It is the responsibility of the calling function to allocate a memory location for the return
value and pass a pointer to it as a hidden first parameter. The pointer to the location
where the return value should be stored is passed in the first suitable register/register
pair, which is R16, R17:R16, and R18:R17:R16 for the Tiny, Small, Large, and Huge
memory model, respectively. The parameter x is passed in R19:R18, R19:R18, and
R21:R20 for the Tiny, Small, Large, and Huge memory model, respectively.

164 for AVR
AFE1_AFE2-1:1
Assume that the function instead was declared to return a pointer to the structure:
struct MyStruct *MyFunction(int x);
In this case, the return value is a scalar, so there is no hidden parameter. The parameter
x is passed in R17:R16, and the return value is returned in R16, R17:R16, and
R18:R17:R16 for the Tiny, Small, and Large memory model, respectively.
FUNCTION DIRECTIVES
Note: This type of directive is primarily intended to support static overlay, a feature
which is useful in some smaller microcontrollers. The IAR C/C++ Compiler for AVR
does not use static overlay, because it has no use for it.
The function directives FUNCTION, ARGFRAME, LOCFRAME, and FUNCALL are generated
by the compiler to pass information about functions and function calls to the IAR
XLINK Linker. These directives can be seen if you use the compiler option Assembler
file (-lA) to create an assembler list file.
For more information about the function directives, see the IAR Assembler User Guide
for AVR.
Call frame information

When you debug an application using C-SPY, you can view the call stack, that is, the
chain of functions that called the current function. To make this possible, the compiler
supplies debug information that describes the layout of the call frame, in particular
information about where the return address is stored.
If you want the call stack to be available when debugging a routine written in assembler
language, you must supply equivalent debug information in your assembler source using
the assembler directive CFI. This directive is described in detail in the IAR Assembler
User Guide for AVR.
CFI DIRECTIVES
The CFI directives provide C-SPY with information about the state of the calling
function(s). Most important of this is the return address, and the value of the stack
pointer at the entry of the function or assembler routine. Given this information, C-SPY
can reconstruct the state for the calling function, and thereby unwind the stack.
A full description about the calling convention might require extensive call frame
information. In many cases, a more limited approach will suffice.
165
AFE1_AFE2-1:1
Call frame information
When describing the call frame information, the following three components must be
present:
● A names block describing the available resources to be tracked
● A common block corresponding to the calling convention
● A data block describing the changes that are performed on the call frame. This
typically includes information about when the stack pointer is changed, and when
permanent registers are stored or restored on the stack.
For AVR, the defined resources depend on which processor option you are using. You
can find the resources in the list file section CFI Names.
CREATING ASSEMBLER SOURCE WITH CFI SUPPORT

The recommended way to create an assembler language routine that handles call frame
information correctly is to start with an assembler language source file created by the
compiler.
1 Start with suitable C source code, for example:
int F(int);
int cfiExample(int i)
{
return i + F(i);
}
2 Compile the C source code, and make sure to create a list file that contains call frame
information—the CFI directives.
On the command line, use the option -lA.
In the IDE, choose Project>Options>C/C++ Compiler>List and make sure the

suboption Include call frame information is selected.

166 for AVR
AFE1_AFE2-1:1
Using C
● C language overview
● Extensions overview
● IAR C language extensions
C language overview
The IAR C/C++ Compiler for AVR supports the ISO/IEC 9899:1999 standard
(including up to technical corrigendum No.3), also known as C99. In this guide, this
standard is referred to as Standard C and is the default standard used in the compiler.
This standard is stricter than C89.
In addition, the compiler also supports the ISO 9899:1990 standard (including all
technical corrigenda and addenda), also known as C94, C90, C89, and ANSI C. In this
guide, this standard is referred to as C89. Use the --c89 compiler option to enable this
standard.
The C99 standard is derived from C89, but adds features like these:
● The inline keyword advises the compiler that the function defined immediately
after the keyword should be inlined
● Declarations and statements can be mixed within the same scope
● A declaration in the initialization expression of a for loop
● The bool data type
● The long long data type
● The complex floating-point type
● C++ style comments
● Compound literals
● Incomplete arrays at the end of structs
● Hexadecimal floating-point constants
● Designated initializers in structures and arrays
● The preprocessor operator _Pragma()
● Variadic macros, which are the preprocessor macro equivalents of printf style
functions
● VLA (variable length arrays) must be explicitly enabled with the compiler option
--vla
167
AFE1_AFE2-1:1
Extensions overview
● Inline assembler using the asm or the __asm keyword, see Inline assembler, page
151.
Note: Even though it is a C99 feature, the IAR C/C++ Compiler for AVR does not
support UCNs (universal character names).
Note: CLIB does not support any C99 functionality. For example, complex numbers and
variable length arrays are not supported.
Extensions overview
The compiler offers the features of Standard C and a wide set of extensions, ranging
from features specifically tailored for efficient programming in the embedded industry
to the relaxation of some minor standards issues.
This is an overview of the available extensions:
● IAR C language extensions
For information about available language extensions, see IAR C language extensions,
page 169. For more information about the extended keywords, see the chapter
Extended keywords. For information about C++, the two levels of support for the
language, and C++ language extensions; see the chapter Using C++.
● Pragma directives
The #pragma directive is defined by Standard C and is a mechanism for using
vendor-specific extensions in a controlled way to make sure that the source code is
still portable.
The compiler provides a set of predefined pragma directives, which can be used for
controlling the behavior of the compiler, for example how it allocates memory,
whether it allows extended keywords, and whether it outputs warning messages.
Most pragma directives are preprocessed, which means that macros are substituted
in a pragma directive. The pragma directives are always enabled in the compiler. For
several of them there is also a corresponding C/C++ language extension. For
information about available pragma directives, see the chapter Pragma directives.
● Preprocessor extensions
The preprocessor of the compiler adheres to Standard C. The compiler also makes
several preprocessor-related extensions available to you. For more information, see
the chapter The preprocessor.
● Intrinsic functions
The intrinsic functions provide direct access to low-level processor operations and
can be very useful in, for example, time-critical routines. The intrinsic functions
compile into inline code, either as a single instruction or as a short sequence of
instructions. For more information about using intrinsic functions, see Mixing C and

168 for AVR
AFE1_AFE2-1:1
Using C
assembler, page 149. For information about available functions, see the chapter
Intrinsic functions.
● Library functions
The DLIB runtime environment provides the C and C++ library definitions in the
C/C++ standard library that apply to embedded systems. For more information, see
DLIB runtime environment—implementation details, page 363. For information
about AVR-specific library functions, see AVR-specific library functions, page 370.
Note: Any use of these extensions, except for the pragma directives, makes your source
code inconsistent with Standard C.
ENABLING LANGUAGE EXTENSIONS

You can choose different levels of language conformance by means of project options:
Command line IDE* Description
--strict Strict All IAR C language extensions are disabled;
errors are issued for anything that is not part
of Standard C.
None Standard All extensions to Standard C are enabled, but no
extensions for embedded systems programming.
For information about extensions, see IAR C
language extensions, page 169.
-e Standard with IAR All IAR C language extensions are enabled.
extensions
Table 24: Language extensions
* In the IDE, choose Project>Options>C/C++ Compiler>Language 1>Language

conformance and select the appropriate option. Note that language extensions are
enabled by default.
IAR C language extensions

The compiler provides a wide set of C language extensions. To help you to find the
extensions required by your application, they are grouped like this in this section:
● Extensions for embedded systems programming—extensions specifically tailored
for efficient embedded programming for the specific microcontroller you are using,
typically to meet memory restrictions
● Relaxations to Standard C—that is, the relaxation of some minor Standard C issues
and also some useful but minor syntax extensions, see Relaxations to Standard C,
page 172.
169
AFE1_AFE2-1:1
EXTENSIONS FOR EMBEDDED SYSTEMS PROGRAMMING

The following language extensions are available both in the C and the C++
programming languages and they are well suited for embedded systems programming:
● Memory attributes, type attributes, and object attributes
For information about the related concepts, the general syntax rules, and for
reference information, see the chapter Extended keywords.
● Placement at an absolute address or in a named segment
The @ operator or the directive #pragma location can be used for placing global
and static variables at absolute addresses, or placing a variable or function in a named
segment. For more information about using these features, see Controlling data and
function placement in memory, page 210, and location, page 331.
● Alignment control
Each data type has its own alignment; for more information, see Alignment, page
281. If you want to change the alignment the #pragma data_alignment directive
available. If you want to check the alignment of an object, use the __ALIGNOF__()
operator.
The __ALIGNOF__ operator is used for accessing the alignment of an object. It takes
one of two forms:
● __ALIGNOF__ (type)
● __ALIGNOF__ (expression)
In the second form, the expression is not evaluated.
● Anonymous structs and unions
C++ includes a feature called anonymous unions. The compiler allows a similar
feature for both structs and unions in the C programming language. For more
information, see Anonymous structs and unions, page 208.
● Bitfields and non-standard types
In Standard C, a bitfield must be of the type int or unsigned int. Using IAR C
language extensions, any integer type or enumeration can be used. The advantage is
that the struct will sometimes be smaller. For more information, see Bitfields, page
284.
● static_assert()
The construction static_assert(const-expression,"message"); can be
used in C/C++. The construction will be evaluated at compile time and if
const-expression is false, a message will be issued including the message
string.

170 for AVR
AFE1_AFE2-1:1
Using C
● Parameters in variadic macros

Variadic macros are the preprocessor macro equivalents of printf style functions.
The preprocessor accepts variadic macros with no arguments, which means if no
parameter matches the ... parameter, the comma is then deleted in the ",
##__VA_ARGS__" macro definition. According to Standard C, the ... parameter
must be matched with at least one argument.
Dedicated segment operators

The compiler supports getting the start address, end address, and size for a segment with
these built-in segment operators:
__segment_begin Returns the address of the first byte of the named segment.
__segment_end Returns the address of the first byte after the named segment.
__segment_size Returns the size of the named segment in bytes.
Note: The aliases __segment_begin/__sfb, __segment_end/__sfe, and

__segment_size/__sfs can also be used.
The operators can be used on named segments defined in the linker configuration file.
These operators behave syntactically as if declared like:
void * __segment_begin(char const * segment)
void * __segment_end(char const * segment)
size_t __segment_size(char const * segment)
When you use the @ operator or the #pragma location directive to place a data object
or a function in a user-defined segment in the linker configuration file, the segment
operators can be used for getting the start and end address of the memory range where
the segments were placed.
The named segment must be a string literal and it must have been declared earlier with
the #pragma segment directive. If the segment was declared with a memory attribute
memattr, the type of the __segment_begin operator is a pointer to memattr void.
Otherwise, the type is a default pointer to void. Note that you must enable language
extensions to use these operators.
Example
In this example, the type of the __segment_begin operator is void __near *.
#pragma segment="MYSEGMENT" __near
...
segment_start_address = __segment_begin("MYSEGMENT");
See also segment, page 336, and location, page 331.
171
AFE1_AFE2-1:1
RELAXATIONS TO STANDARD C
This section lists and briefly describes the relaxation of some Standard C issues and also
some useful but minor syntax extensions:
● Arrays of incomplete types
An array can have an incomplete struct, union, or enum type as its element type.
The types must be completed before the array is used (if it is), or by the end of the
compilation unit (if it is not).
● Forward declaration of enum types
The extensions allow you to first declare the name of an enum and later resolve it by
specifying the brace-enclosed list.
● Accepting missing semicolon at the end of a struct or union specifier
A warning—instead of an error—is issued if the semicolon at the end of a struct
or union specifier is missing.
● Null and void
In operations on pointers, a pointer to void is always implicitly converted to another
type if necessary, and a null pointer constant is always implicitly converted to a null
pointer of the right type if necessary. In Standard C, some operators allow this kind
of behavior, while others do not allow it.
● Casting pointers to integers in static initializers
In an initializer, a pointer constant value can be cast to an integral type if the integral
type is large enough to contain it. For more information about casting pointers, see
Casting, page 289.
● Taking the address of a register variable
In Standard C, it is illegal to take the address of a variable specified as a register
variable. The compiler allows this, but a warning is issued.
● long float means double
The type long float is accepted as a synonym for double.
● Repeated typedef declarations
Redeclarations of typedef that occur in the same scope are allowed, but a warning
is issued.
● Mixing pointer types
Assignment and pointer difference is allowed between pointers to types that are
interchangeable but not identical; for example, unsigned char * and char *. This
includes pointers to integral types of the same size. A warning is issued.
Assignment of a string constant to a pointer to any kind of character is allowed, and
no warning is issued.

172 for AVR
AFE1_AFE2-1:1
Using C
● Non-top level const

Assignment of pointers is allowed in cases where the destination type has added type
qualifiers that are not at the top level (for example, int ** to int const **).
Comparing and taking the difference of such pointers is also allowed.
● Non-lvalue arrays
A non-lvalue array expression is converted to a pointer to the first element of the
array when it is used.
● Comments at the end of preprocessor directives
This extension, which makes it legal to place text after preprocessor directives, is
enabled unless the strict Standard C mode is used. The purpose of this language
extension is to support compilation of legacy code; we do not recommend that you
write new code in this fashion.
● An extra comma at the end of enum lists
Placing an extra comma is allowed at the end of an enum list. In strict Standard C
mode, a warning is issued.
● A label preceding a }
In Standard C, a label must be followed by at least one statement. Therefore, it is
illegal to place the label at the end of a block. The compiler allows this, but issues a
warning.
Note that this also applies to the labels of switch statements.
● Empty declarations
An empty declaration (a semicolon by itself) is allowed, but a remark is issued
(provided that remarks are enabled).
● Single-value initialization
Standard C requires that all initializer expressions of static arrays, structs, and unions
are enclosed in braces.
Single-value initializers are allowed to appear without braces, but a warning is
issued. The compiler accepts this expression:
struct str
{
int a;
} x = 10;
173
AFE1_AFE2-1:1
● Declarations in other scopes

External and static declarations in other scopes are visible. In the following example,
the variable y can be used at the end of the function, even though it should only be
visible in the body of the if statement. A warning is issued.
int test(int x)
{
if (x)
{
extern int y;
y = 1;
}
return y;
}
● Expanding function names into strings with the function as context
Use any of the symbols __func__ or __FUNCTION__ inside a function body to
make the symbol expand into a string that contains the name of the current function.
Use the symbol __PRETTY_FUNCTION__ to also include the parameter types and
return type. The result might, for example, look like this if you use the
__PRETTY_FUNCTION__ symbol:
"void func(char)"
These symbols are useful for assertions and other trace utilities and they require that
language extensions are enabled, see -e, page 250.
● Static functions in function and block scopes
Static functions may be declared in function and block scopes. Their declarations are
moved to the file scope.
● Numbers scanned according to the syntax for numbers
Numbers are scanned according to the syntax for numbers rather than the
pp-number syntax. Thus, 0x123e+1 is scanned as three tokens instead of one valid
token. (If the --strict option is used, the pp-number syntax is used instead.)

174 for AVR
AFE1_AFE2-1:1
Using C++
● Overview—EC++ and EEC++
● Enabling support for C++
● EC++ feature descriptions
● EEC++ feature description
● C++ language extensions
Overview—EC++ and EEC++

IAR Systems supports the C++ language. You can choose between the industry-standard
Embedded C++ and Extended Embedded C++. Using C++ describes what you need to
consider when using the C++ language.
Embedded C++ is a proper subset of the C++ programming language which is intended
for embedded systems programming. It was defined by an industry consortium, the
Embedded C++ Technical Committee. Performance and portability are particularly
important in embedded systems development, which was considered when defining the
language. EC++ offers the same object-oriented benefits as C++, but without some
features that can increase code size and execution time in ways that are hard to predict.
EMBEDDED C++
These C++ features are supported:
● Classes, which are user-defined types that incorporate both data structure and
behavior; the essential feature of inheritance allows data structure and behavior to
be shared among classes
● Polymorphism, which means that an operation can behave differently on different
classes, is provided by virtual functions
● Overloading of operators and function names, which allows several operators or
functions with the same name, provided that their argument lists are sufficiently
different
● Type-safe memory management using the operators new and delete
● Inline functions, which are indicated as particularly suitable for inline expansion.
175
AFE1_AFE2-1:1
Overview—EC++ and EEC++
C++ features that are excluded are those that introduce overhead in execution time or
code size that are beyond the control of the programmer. Also excluded are features
added very late before Standard C++ was defined. Embedded C++ thus offers a subset
of C++ which is efficient and fully supported by existing development tools.
Embedded C++ lacks these features of C++:
● Templates
● Multiple and virtual inheritance
● Exception handling
● Runtime type information
● New cast syntax (the operators dynamic_cast, static_cast,
reinterpret_cast, and const_cast)
● Namespaces
● The mutable attribute.
The exclusion of these language features makes the runtime library significantly more
efficient. The Embedded C++ library furthermore differs from the full C++ library in
that:
● The standard template library (STL) is excluded
● Streams, strings, and complex numbers are supported without the use of templates
● Library features which relate to exception handling and runtime type information
(the headers except, stdexcept, and typeinfo) are excluded.
Note: The library is not in the std namespace, because Embedded C++ does not
support namespaces.
EXTENDED EMBEDDED C++

IAR Systems’ Extended EC++ is a slightly larger subset of C++ which adds these
features to the standard EC++:
● Full template support
● Multiple and virtual inheritance
● Namespace support
● The mutable attribute
● The cast operators static_cast, const_cast, and reinterpret_cast.
All these added features conform to the C++ standard.
To support Extended EC++, this product includes a version of the standard template
library (STL), in other words, the C++ standard chapters utilities, containers, iterators,
algorithms, and some numerics. This STL is tailored for use with the Extended EC++

176 for AVR
AFE1_AFE2-1:1
Using C++
language, which means no exceptions and no support for runtime type information
(rtti). Moreover, the library is not in the std namespace.
Note: A module compiled with Extended EC++ enabled is fully link-compatible with a
module compiled without Extended EC++ enabled.
Enabling support for C++

In the compiler, the default language is C.
To compile files written in Embedded C++, use the --ec++ compiler option. See
--ec++, page 250.
To take advantage of Extended Embedded C++ features in your source code, use the
--eec++ compiler option. See --eec++, page 251.
To enable EC++ or EEC++ in the IDE, choose Project>Options>C/C++

Compiler>Language 1 and select the appropriate standard.
EC++ feature descriptions

When you write C++ source code for the IAR C/C++ Compiler for AVR, you must be
aware of some benefits and some possible quirks when mixing C++ features—such as
classes, and class members—with IAR language extensions, such as IAR-specific
attributes.
USING IAR ATTRIBUTES WITH CLASSES

Static data members of C++ classes are treated the same way global variables are, and
can have any applicable IAR type, memory, and object attribute.
Member functions are in general treated the same way free functions are, and can have
any applicable IAR type, memory, and object attributes. Virtual member functions can
only have attributes that are compatible with default function pointers, and constructors
and destructors cannot have any such attributes.
The location operator @ and the #pragma location directive can be used on static data
members and with all member functions.
177
AFE1_AFE2-1:1
Example of using attributes with classes

class MyClass
{
public:
// Locate a static variable in __near memory at address 60
static __near __no_init int mI @ 60;
// Locate a static function in __nearfunc memory

static __nearfunc void F();
// Locate a function in __nearfunc memory

__nearfunc void G();
// Locate a virtual function in __nearfunc memory

virtual __nearfunc void H();
// Locate a virtual function into SPECIAL

virtual void M() const volatile @ "SPECIAL";
};
The this pointer

The this pointer used for referring to a class object or calling a member function of a
class object will by default have the data memory attribute for the default data pointer
type. This means that such a class object can only be defined to reside in memory from
which pointers can be implicitly converted to a default data pointer. This restriction
might also apply to objects residing on a stack, for example temporary objects and auto
objects.
Class memory
To compensate for this limitation, a class can be associated with a class memory type.
The class memory type changes:
● the this pointer type in all member functions, constructors, and destructors into a
pointer to class memory
● the default memory for static storage duration variables—that is, not auto
variables—of the class type, into the specified class memory
● the pointer type used for pointing to objects of the class type, into a pointer to class
memory.

178 for AVR
AFE1_AFE2-1:1
Using C++
Example
class __far C
{
public:
void MyF(); // Has a this pointer of type C __far *
void MyF() const; // Has a this pointer of type
// C __far const *
C(); // Has a this pointer pointing into far
// memory
C(C const &); // Takes a parameter of type C __far
// const & (also true of generated copy
// constructor)
int mI;
};
C Ca; // Resides in far memory instead of the

// default memory
C __near Cb; // Resides in near memory, the 'this'
// pointer still points into far memory
void MyH()
{
C cd; // Resides on the stack
}
C *Cp1; // Creates a pointer to far memory

C __near *Cp2; // Creates a pointer to near memory
Note: To place the C class in huge memory is not allowed, unless using the huge
memory model, because a huge pointer cannot be implicitly converted into a __far
pointer.
Whenever a class type associated with a class memory type, like C, must be declared,
the class memory type must be mentioned as well:
class __far C;
Also note that class types associated with different class memories are not compatible
types.
A built-in operator returns the class memory type associated with a class,
__memory_of(class). For instance, __memory_of(C) returns __far.
179
AFE1_AFE2-1:1
When inheriting, the rule is that it must be possible to convert implicitly a pointer to a
subclass into a pointer to its base class. This means that a subclass can have a more
restrictive class memory than its base class, but not a less restrictive class memory.
class __far D : public C
{ // OK, same class memory
public:
void MyG();
int mJ;
};
class __near E : public C

{ // OK, near memory is inside far
public:
void MyG() // Has a this pointer pointing into near memory
{
MyF(); // Gets a this pointer into far memory
}
int mJ;
};
class F : public C
{ // OK, will be associated with same class memory as C
public:
void MyG();
int mJ;
};
A new expression on the class will allocate memory in the heap associated with the class
memory. A delete expression will naturally deallocate the memory back to the same
heap. To override the default new and delete operator for a class, declare
void *operator new(size_t);
void operator delete(void *);
as member functions, just like in ordinary C++.

For more information about memory types, see Memory types, page 54.
FUNCTION TYPES
A function type with extern "C" linkage is compatible with a function that has C++
linkage.

180 for AVR
AFE1_AFE2-1:1
Using C++
Example
extern "C"
{
typedef void (*FpC)(void); // A C function typedef
}
typedef void (*FpCpp)(void); // A C++ function typedef
FpC F1;
FpCpp F2;
void MyF(FpC);
void MyG()
{
MyF(F1); // Always works
MyF(F2); // FpCpp is compatible with FpC
}
NEW AND DELETE OPERATORS

There are operators for new and delete for each memory that can have a heap, that is,
tiny, near, far, and huge memory.
// Assumes that there is a heap in both __near and __far memory
void __near *operator new __near(__near_size_t);
void __far *operator new __far (__far_size_t);
void operator delete(void __near *);
void operator delete(void __far *);
// And correspondingly for array new and delete operators

void __near *operator new[] __near(__near_size_t);
void __far *operator new[] __far (__far_size_t);
void operator delete[](void __near *);
void operator delete[](void __far *);
Use this syntax if you want to override both global and class-specific operator new
and operator delete for any data memory.
Note that there is a special syntax to name the operator new functions for each
memory, while the naming for the operator delete functions relies on normal
overloading.
181
AFE1_AFE2-1:1
New and delete expressions

A new expression calls the operator new function for the memory of the type given. If
a class, struct, or union type with a class memory is used, the class memory will
determine the operator new function called. For example,
void MyF()
{
// Calls operator new __huge(__huge_size_t)
int __huge *p = new __huge int;
// Calls operator new __near(__near_size_t)

int __near *q = new int __near;
// Calls operator new[] __near(__near_size_t)

int __near *r = new __near int[10];
// Calls operator new __huge(__huge_size_t)

class __huge S
{
};
S *s = new S;
// Calls operator delete(void __huge *)

delete p;
// Calls operator delete(void __near *)
delete s;
int __huge *t = new __far int;

delete t; // Error: Causes a corrupt heap
}
Note that the pointer used in a delete expression must have the correct type, that is, the
same type as that returned by the new expression. If you use a pointer to the wrong
memory, the result might be a corrupt heap.
USING STATIC CLASS OBJECTS IN INTERRUPTS

If interrupt functions use static class objects that need to be constructed (using
constructors) or destroyed (using destructors), your application will not work properly
if the interrupt occurs before the objects are constructed, or, during or after the objects
are destroyed.
To avoid this, make sure that these interrupts are not enabled until the static objects have
been constructed, and are disabled when returning from main or calling exit. For
information about system startup, see System startup and termination, page 118.

182 for AVR
AFE1_AFE2-1:1
Using C++
Function local static class objects are constructed the first time execution passes through
their declaration, and are destroyed when returning from main or when calling exit.
USING NEW HANDLERS

To handle memory exhaustion, you can use the set_new_handler function.
New handlers in Embedded C++

If you do not call set_new_handler, or call it with a NULL new handler, and
operator new fails to allocate enough memory, it will call abort. The nothrow
variant of the new operator will instead return NULL.
If you call set_new_handler with a non-NULL new handler, the provided new
handler will be called by operator new if operator new fails to allocate memory. The
new handler must then make more memory available and return, or abort execution in
some manner. The nothrow variant of operator new will never return NULL in the
presence of a new handler.
TEMPLATES
Extended EC++ supports templates according to the C++ standard, but not the export
keyword. The implementation uses a two-phase lookup which means that the keyword
typename must be inserted wherever needed. Furthermore, at each use of a template,
the definitions of all possible templates must be visible. This means that the definitions
of all templates must be in include files or in the actual source file.
DEBUG SUPPORT IN C-SPY

C-SPY® has built-in display support for the STL containers. The logical structure of
containers is presented in the watch views in a comprehensive way that is easy to
understand and follow.
For more information about this, see the C-SPY® Debugging Guide for AVR.
EEC++ feature description

This section describes features that distinguish Extended EC++ from EC++.
TEMPLATES
The compiler supports templates with the syntax and semantics as defined by Standard
C++. However, note that the STL (standard template library) delivered with the product
is tailored for Extended EC++, see Extended Embedded C++, page 176.
183
AFE1_AFE2-1:1
Templates and data memory attributes

For data memory attributes to work as expected in templates, two elements of the
standard C++ template handling were changed—class template partial specialization
matching and function template parameter deduction.
In Extended Embedded C++, the class template partial specialization matching
algorithm works like this:
When a pointer or reference type is matched against a pointer or reference to a template
parameter type, the template parameter type will be the type pointed to, stripped of any
data memory attributes, if the resulting pointer or reference type is the same.
Example
// We assume that __far is the memory type of the default
// pointer.
template<typename> class Z {};
template<typename T> class Z<T *> {};
Z<int __near *> Zn; // T = int __near

Z<int __far *> Zf; // T = int
Z<int *> Zd; // T = int
Z<int __huge *> Zh; // T = int __huge
In Extended Embedded C++, the function template parameter deduction algorithm

works like this:
When function template matching is performed and an argument is used for the
deduction; if that argument is a pointer to a memory that can be implicitly converted to
a default pointer, do the parameter deduction as if it was a default pointer.
When an argument is matched against a reference, do the deduction as if the argument
and the parameter were both pointers.

184 for AVR
AFE1_AFE2-1:1
Using C++
Example
// pointer.
template<typename T> void fun(T *);
void MyF()
{
fun((int __near *) 0); // T = int. The result is different
// than the analogous situation with
// class template specializations.
fun((int *) 0); // T = int
fun((int __far *) 0); // T = int
fun((int __huge *) 0); // T = int __huge
}
For templates that are matched using this modified algorithm, it is impossible to get
automatic generation of special code for pointers to “small” memory types. For “large”
and “other” memory types (memory that cannot be pointed to by a default pointer) it is
possible. To make it possible to write templates that are fully memory-aware—in the
rare cases where this is useful—use the #pragma basic_template_matching
directive in front of the template function declaration. That template function will then
match without the modifications described above.
Example
// pointer.
#pragma basic_template_matching
void MyF()
{
fun((int __near *) 0); // T = int __near
}
Non-type template parameters

It is allowed to have a reference to a memory type as a template parameter, even if
pointers to that memory type are not allowed.
185
AFE1_AFE2-1:1
Example
extern int __near X;
template<__near int &y>

void Func()
{
y = 17;
}
void Bar()
{
Foo<X>();
}
The containers in the STL, like vector and map, are memory attribute aware. This
means that a container can be declared to reside in a specific memory type which has the
following consequences:
● The container itself will reside in the chosen memory
● Allocations of elements in the container will use a heap for the chosen memory
● All references inside it use pointers to the chosen memory.
Example
#include <vector>
vector<int> D; // D placed in default

memory,
// using the default heap,
// uses default pointers
vector<int __near> __near X; // X placed in near memory,
// heap allocation from
// near, uses pointers to
// near memory
vector<int __huge> __near Y; // Y placed in near memory,
// huge, uses pointers to
// huge memory
Note that this is illegal:

vector<int __near> __huge Z;
Note also that map<key, T>, multimap<key, T>, hash_map<key, T>, and
hash_multimap<key, T> all use the memory of T. This means that the value_type
of these collections will be pair<key, const T> mem where mem is the memory type
of T. Supplying a key with a memory type is not useful.

186 for AVR
AFE1_AFE2-1:1
Using C++
Example
Note that two containers that only differ by the data memory attribute they use cannot
be assigned to each other. Instead, the templated assign member method must be used.
#include <vector>
vector<int __near> X;
vector<int __huge> Y;
void MyF()
{
// The templated assign member method will work
X.assign(Y.begin(), Y.end());
Y.assign(X.begin(), X.end());
}
The standard template library

The containers in the STL, like vector and map, are memory attribute aware. This
means that a container can be declared to reside in a specific memory type which has the
following consequences:
● The container itself will reside in the chosen memory
● Allocations of elements in the container will use a heap for the chosen memory
● All references inside it use pointers to the chosen memory.
Example
#include <vector>
vector<int> D; // D placed in default

memory,
// using the default heap,
// uses default pointers
vector<int __near> __near X; // X placed in near memory,
// near, uses pointers to
// near memory
vector<int __far> __near Y; // Y placed in near memory,
// far, uses pointers to
// far memory
Note that this is illegal:

vector<int __near> __far Z;
187
AFE1_AFE2-1:1
Note also that map<key, T>, multimap<key, T>, hash_map<key, T>, and
hash_multimap<key, T> all use the memory of T. This means that the value_type
of these collections will be pair<key, const T> mem where mem is the memory type
of T. Supplying a key with a memory type is not useful.
Example
Note that two containers that only differ by the data memory attribute they use cannot
be assigned to each other. Instead, the templated assign member method must be used.
#include <vector>
vector<int __near> X;
vector<int __far> Y;
void MyF()
{
// The templated assign member method will work
X.assign(Y.begin(), Y.end());
Y.assign(X.begin(), X.end());
}
VARIANTS OF CAST OPERATORS

In Extended EC++ these additional variants of C++ cast operators can be used:
const_cast<to>(from)
static_cast<to>(from)
reinterpret_cast<to>(from)
MUTABLE
The mutable attribute is supported in Extended EC++. A mutable symbol can be
changed even though the whole class object is const.
NAMESPACE
The namespace feature is only supported in Extended EC++. This means that you can
use namespaces to partition your code. Note, however, that the library itself is not placed
in the std namespace.
THE STD NAMESPACE

The std namespace is not used in either standard EC++ or in Extended EC++. If you
have code that refers to symbols in the std namespace, simply define std as nothing;
for example:
#define std

188 for AVR
AFE1_AFE2-1:1
Using C++
You must make sure that identifiers in your application do not interfere with identifiers
in the runtime library.
C++ language extensions

When you use the compiler in any C++ mode and enable IAR language extensions, the
following C++ language extensions are available in the compiler:
● In a friend declaration of a class, the class keyword can be omitted, for example:
class B;
class A
{
friend B; //Possible when using IAR language
//extensions
friend class B; //According to the standard
};
● Constants of a scalar type can be defined within classes, for example:
class A
{
const int mSize = 10; //Possible when using IAR language
//extensions
int mArr[mSize];
};
According to the standard, initialized static data members should be used instead.
● In the declaration of a class member, a qualified name can be used, for example:
struct A
{
int A::F(); // Possible when using IAR language extensions
int G(); // According to the standard
};
● It is permitted to use an implicit type conversion between a pointer to a function
with C linkage (extern "C") and a pointer to a function with C++ linkage
(extern "C++"), for example:
extern "C" void F(); // Function with C linkage
void (*PF)() // PF points to a function with C++ linkage
= &F; // Implicit conversion of function pointer.
According to the standard, the pointer must be explicitly converted.
189
AFE1_AFE2-1:1
● If the second or third operands in a construction that contains the ? operator are
string literals or wide string literals (which in C++ are constants), the operands can
be implicitly converted to char * or wchar_t *, for example:
bool X;
char *P1 = X ? "abc" : "def"; //Possible when using IAR

//language extensions
char const *P2 = X ? "abc" : "def";//According to the standard
● Default arguments can be specified for function parameters not only in the top-level
function declaration, which is according to the standard, but also in typedef
declarations, in pointer-to-function function declarations, and in pointer-to-member
function declarations.
● In a function that contains a non-static local variable and a class that contains a
non-evaluated expression (for example a sizeof expression), the expression can
reference the non-static local variable. However, a warning is issued.
● An anonymous union can be introduced into a containing class by a typedef name.
It is not necessary to first declare the union. For example:
typedef union
{
int i,j;
} U; // U identifies a reusable anonymous union.
class A
{
public:
U; // OK -- references to A::i and A::j are allowed.
};
In addition, this extension also permits anonymous classes and anonymous structs,
as long as they have no C++ features (for example, no static data members or member
functions, and no non-public members) and have no nested types other than other
anonymous classes, structs, or unions. For example:
struct A
{
struct
{
int i,j;
}; // OK -- references to A::i and A::j are allowed.
};

190 for AVR
AFE1_AFE2-1:1
Using C++
● The friend class syntax allows nonclass types as well as class types expressed
through a typedef without an elaborated type name. For example:
typedef struct S ST;
class C
{
public:
friend S; // Okay (requires S to be in scope)
friend ST; // Okay (same as "friend S;")
// friend S const; // Error, cv-qualifiers cannot
// appear directly
};
● It is allowed to specify an array with no size or size 0 as the last member of a struct.
For example:
typedef struct
{
int i;
char ir[0]; // Zero-length array
};
typedef struct
{
int i;
char ir[]; // Zero-length array
};
Note: If you use any of these constructions without first enabling language extensions,
errors are issued.
191
AFE1_AFE2-1:1

192 for AVR
AFE1_AFE2-1:1
Application-related
considerations
● Stack considerations
● Heap considerations
● Interaction between the tools and your application
● Checksum calculation for verifying image integrity
Stack considerations
To make your application use stack memory efficiently, there are some considerations
to be made.
AVAILABLE STACKS
The compiler uses the internal data stack, CSTACK, for a variety of user program
operations. In addition, there is a specific stack for return addresses, RSTACK.
The return address stack

The return address stack is used for storing the return address when a CALL, RCALL,
ICALL, or EICALL instruction is executed. Each call will use two or three bytes of return
address stack. An interrupt will also place a return address on this stack.
The data segment used for holding the return address stack is called RSTACK.
To determine the size of the return address stack, see Setting up stack memory, page 91.
Notice however that if the cross-call optimization has been used (-z9 without
--no_cross_call), the value can be off by as much as a factor of six depending on
how many times the cross-call optimizer has been run (--cross_call_passes). Each
cross-call pass adds one level of calls, for example, two cross-call passes might result in
a tripled stack usage.
If external SRAM is available, it is possible to place the stack there. However, the
external memory is slower than the internal memory so moving the stacks to external
memory will normally decrease the system performance; see --enable_external_bus,
page 252.
193
AFE1_AFE2-1:1
Heap considerations
Allocating a memory area for the stack is done differently using the command line
interface compared to when using the IDE.
STACK SIZE CONSIDERATIONS

The required stack size depends heavily on the application’s behavior. If the given stack
size is too large, RAM will be wasted. If the given stack size is too small, one of two
things can happen, depending on where in memory you located your stack:
● Variable storage will be overwritten, leading to undefined behavior
● The stack will fall outside of the memory area, leading to an abnormal termination
of your application.
Both alternatives are likely to result in application failure. Because the second
alternative is easier to detect, you should consider placing your stack so that it grows
toward the end of the memory.
For more information about the stack size, see Setting up stack memory, page 91, and
Saving stack space and RAM memory, page 219.
Heap considerations
The heap contains dynamic data allocated by use of the C function malloc (or a
corresponding function) or the C++ operator new.
If your application uses dynamic memory allocation, you should be familiar with:
● Linker segments used for the heap
● Allocating the heap size, see Setting up heap memory, page 92.
HEAP SEGMENTS IN DLIB

To access a heap in a specific memory, use the appropriate memory attribute as a prefix
to the standard functions malloc, free, calloc, and realloc, for example:
__near_malloc
If you use any of the standard functions without a prefix, the function will be mapped to
the default memory type near.
Each heap will reside in a segment with the name _HEAP prefixed by a memory attribute,
for example NEAR_HEAP.
For information about available heaps, see Dynamic memory on the heap, page 61.

194 for AVR
AFE1_AFE2-1:1
Application-related considerations
HEAP SEGMENTS IN CLIB

The memory allocated to the heap is placed in the segment HEAP, which is only included
in the application if dynamic memory allocation is actually used.
HEAP SIZE AND STANDARD I/O

If you excluded FILE descriptors from the DLIB runtime environment, as in the Normal
configuration, there are no input and output buffers at all. Otherwise, as in the Full
configuration, be aware that the size of the input and output buffers is set to 512 bytes
in the stdio library header file. If the heap is too small, I/O will not be buffered, which
is considerably slower than when I/O is buffered. If you execute the application using
the simulator driver of the IAR C-SPY® Debugger, you are not likely to notice the speed
penalty, but it is quite noticeable when the application runs on an AVR microcontroller.
If you use the standard I/O library, you should set the heap size to a value which
accommodates the needs of the standard I/O buffer.
Interaction between the tools and your application

The linking process and the application can interact symbolically in four ways:
● Creating a symbol by using the linker command line option -D. The linker will
create a public absolute constant symbol that the application can use as a label, as a
size, as setup for a debugger, etc.
● Using the compiler operators __segment_begin, __segment_end, or
__segment_size, or the assembler operators SFB, SFE, or SIZEOF on a named
segment. These operators provide access to the start address, end address, and size
of a contiguous sequence of segments with the same name
● The command line option -s informs the linker about the start label of the
application. It is used by the linker as a root symbol and to inform the debugger
where to start execution.
The following lines illustrate how to use -D to create a symbol. If you need to use this
mechanism, add these options to your command line like this:
-Dmy_symbol=A
-D_..X_HEAP_SIZE=100
The linker configuration file can look like this:

-Z(DATA)MyHeap+MY_HEAP_SIZE=20000–2FFFF
195
AFE1_AFE2-1:1
Checksum calculation for verifying image integrity
Add these lines to your application source code:

#include <stdlib.h>
/* Use symbol defined by an XLINK option to dynamically allocate

an array of elements with specified size. The value takes the
form of a label.
*/
extern int NrOfElements;
typedef char Elements;

Elements *GetElementArray()
{
return malloc(sizeof(Elements) * (long) &NrOfElements);
}
/* Use a symbol defined by an XLINK option, a symbol that in the

* configuration file was made available to the application.
*/
extern char HeapSize;
/* Declare the section that contains the heap. */

#pragma segment = "MYHEAP"
char *MyHeap()
{
/* First get start of statically allocated section, */
char *p = __segment_begin("MYHEAP");
/* ...then we zero it, using the imported size. */

for (int i = 0; i < (int) &HeapSize; ++i)
{
p[i] = 0;
}
return p;
}

This section contains information about checksum calculation:
● Briefly about checksum calculation, page 197
● Calculating and verifying a checksum, page 198
● Troubleshooting checksum calculation, page 203

196 for AVR
AFE1_AFE2-1:1
BRIEFLY ABOUT CHECKSUM CALCULATION

You can use a checksum to verify that the image is the same at runtime as when the
image’s original checksum was generated. In other words, to verify that the image has
not been corrupted.
This works as follows:
● You need an initial checksum.
You can either use the linker to generate an initial checksum or you might have a
third-party checksum available.
● You must generate a second checksum during runtime.
You can either add specific code to your application source code for calculating a
checksum during runtime or you can use some dedicated hardware on your device
for calculating a checksum during runtime.
● You must add specific code to your application source code for comparing the two
checksums and take an appropriate action if they differ.
If the two checksums have been calculated in the same way, and if there are no errors
in the image, the checksums should be identical. If not, you should first suspect that
the two checksums were not generated in the same way.
No matter which solutions you use for generating the two checksum, you must make
sure that both checksums are calculated in the exact same way. If you use the linker for
the initial checksum and use a software-based calculation during runtime, you have full
control of the generation for both checksums. However, if you are using a third-party
checksum for the initial checksum or some hardware support for the checksum
calculation during runtime, there might be additional requirements that you must
consider.
For the two checksums, there are some choices that you must always consider and there
are some choices to make only if there are additional requirements. Still, all of the details
must be the same for both checksums.
Details always to consider:
● Checksum range
The memory range (or ranges) that you want to verify by means of checksums.
Typically, you might want to calculate a checksum for all ROM memory. However,
you might want to calculate a checksum only for specific ranges. Remember that:
● It is OK to have several ranges for one checksum.
● Typically, the checksum must be calculated from the lowest to the highest
address for every memory range.
● Each memory range must be verified in the same order as defined (for example,
0x100–0x1FF,0x400–0x4FF is not the same as 0x400–0x4FF,0x100–0x1FF).
197
AFE1_AFE2-1:1
● If several checksums are used, you should place them in sections with unique
names and use unique symbol names.
● A checksum should never be calculated on a memory range that contains a
checksum or a software breakpoint.
● Algorithm and size of checksum
You should consider which algorithm is most suitable in your case. There are two
basic choices, Sum (a simple arithmetic algorithm) or CRC (which is the most
commonly used algorithm). For CRC there are different sizes to choose for the
checksum, 2 or 4 bytes where the predefined polynomials are wide enough to suit the
size, for more error detecting power. The predefined polynomials work well for most,
but possibly not for all data sets. If not, you can specify your own polynomial. If you
just want a decent error detecting mechanism, use the predefined CRC algorithm for
your checksum size, typically CRC16 or CRC32.
Note that for an n-bit polynomial, the n:th bit is always considered to be set. For a
16-bit polynomial (for example, CRC16) this means that 0x11021 is the same as
0x1021.
For more information about selecting an appropriate polynomial for data sets with
non-uniform distribution, see for example section 3.5.3 in Tannenbaum, A.S.,
Computer Networks, Prentice Hall 1981, ISBN: 0131646990.
● Fill
Every byte in the checksum range must have a well-defined value before the
checksum can be calculated. Typically, bytes with unknown values are pad bytes that
have been added for alignment. This means that you must specify which fill pattern
to be used during calculation, typically 0xFF or 0x00.
● Initial value
The checksum must always have an explicit initial value.
In addition to these mandatory details, there might be other details to consider.
Typically, this might happen when you have a third-party checksum, you want the
checksum be compliant with the Rocksoft™ checksum model, or when you use
hardware support for generating a checksum during runtime. The linker provides
support for also controlling alignment, complement, bit order, and checksum unit size.
CALCULATING AND VERIFYING A CHECKSUM

In this example procedure, a checksum is calculated for ROM memory from 0x8002 up
to 0x8FFF and the 2-byte calculated checksum is placed at 0x8000.
1 The CHECKSUM segment will only be included in your application if the segment
appears to be needed. If the checksum is not needed by the application itself, use the
linker option -g__checksum to force the segment to be included.

198 for AVR
AFE1_AFE2-1:1
2 When configuring the linker to calculate a checksum, there are some basic choices to
make:
● Checksum algorithm
Choose which checksum algorithm you want to use. In this example, the CRC16
algorithm is used.
● Memory range
Using the IDE, the checksum will by default be calculated for all placement
directives (specified in the linker configuration file) for ROM-based memory. From
the command line, you can specify any ranges.
● Fill pattern
Specify a fill pattern—typically 0xFF or 0x00—for bytes with unknown values. The
fill pattern will be used in all checksum ranges.
For more information, see Briefly about checksum calculation, page 197.
To run the linker from the IDE, choose Project>Options>Linker>Checksum and
make your settings, for example:
In the simplest case, you can ignore (or leave with default settings) these options:
Complement, Bit order, and Checksum unit size.
To make the linker create a checksum from the command line, use the -J linker option,
for example like this:
-J2,crc16,,__checksum,CHECKSUM,1=0x8002–0x8FFF
The checksum will be created when you build your project and will be placed in the
automatically generated segment CHECKSUM. If you are using your own linker
configuration file or if you explicitly want to control the placement of the CHECKSUM
segment, you must update your linker configuration file with placement information
199
AFE1_AFE2-1:1
accordingly. In that case, make sure to place the segment so that it is not part of the
application’s checksum calculation.
3 You can specify several ranges instead of only one range.
If you are using the IDE, perform these steps:
● Choose Project>Options>Linker>Checksum and make sure to deselect Fill
unused code memory.
● Choose Project>Options>Linker>Extra Options and specify the ranges, for
example like this:
-h(CODE)0-3FF,8002-8FFF
-J2,crc16,,,1=0-3FF,8002-8FFF
If you are using the command line, use the -J option and specify the ranges. for example
like this:
-h(CODE)0-3FF,8002-8FFF
-J2,crc16,,,1=0-3FF,8002-8FFF

200 for AVR
AFE1_AFE2-1:1
4 Add a function for checksum calculation to your source code. Make sure that the
function uses the same algorithm and settings as for the checksum calculated by the
linker. For example, a slow variant of the crc16 algorithm but with small memory
footprint (in contrast to the fast variant that uses more memory):
unsigned short SmallCrc16(uint16_t
sum,
unsigned char *p,
unsigned int len)
{
while (len--)
{
int i;
unsigned char byte = *(p++);
for (i = 0; i < 8; ++i)

{
unsigned long oSum = sum;
sum <<= 1;
if (byte & 0x80)
sum |= 1;
if (oSum & 0x8000)
sum ^= 0x1021;
byte <<= 1;
}
}
return sum;
}
You can find the source code for this checksum algorithm in the avr\src\linker
directory of your product installation.
201
AFE1_AFE2-1:1
5 Make sure that your application also contains a call to the function that calculates the
checksum, compares the two checksums, and takes appropriate action if the checksum
values do not match. This code gives an example of how the checksum can be
calculated for your application and to be compared with the linker generated
checksum:
/* The calculated checksum */
/* Linker generated symbols */

extern unsigned short const __checksum;
extern int __checksum_begin;
extern int __checksum_end;
void TestChecksum()
{
unsigned short calc = 0;
unsigned char zeros[2] = {0, 0};
/* Run the checksum algorithm */

calc = SmallCrc16(0,
(unsigned char *) &__checksum_begin,
((unsigned char *) &__checksum_end -
((unsigned char *) &__checksum_begin)+1));
/* Fill the end of the byte sequence with zeros. */

calc = SmallCrc16(calc, zeros, 2);
/* Test the checksum */

if (calc != __checksum)
{
printf("Incorrect checksum!\n");
abort(); /* Failure */
}
/* Checksum is correct */
}
Note: Make sure to define the symbols __checksum_begin and __checksum_end in

your linker configuration file.
6 Build your application project and download it.
During the build, the linker creates a checksum and places it in the specified symbol
__checksum in the segment CHECKSUM.
7 Choose Download and Debug to start the C-SPY debugger.

202 for AVR
AFE1_AFE2-1:1
During execution, the checksum calculated by the linker and the checksum calculated
by your application should be identical.
TROUBLESHOOTING CHECKSUM CALCULATION

If the two checksums do not match, there are several possible causes. These are some
troubleshooting hints:
● If possible, start with a small example when trying to get the checksums to match.
● Verify that the exact same memory range or ranges are used in both checksum
calculations.
To help you do this, The linker produces useful information in the map file about the
exact addresses that were used and the order in which they were accessed.
● Make sure that all checksum symbols are excluded from all checksum calculations.
In the map file, notice the checksum symbol and its size, and for information about
its placement, check the module map or the entry list. Compare that placement with
the checksum range.
● Verify that the checksum calculations use the same polynomial.
● Verify that the bits in the bytes are processed in the same order in both checksum
calculations, from the least to the most significant bit or the other way around. You
control this with the Bit order option (or from the command line, the -m parameter
of the --checksum option).
● If you are using the small variant of CRC, check whether you need to feed
additional bytes into the algorithm.
The number of zeros to add at the end of the byte sequence must match the size of
the checksum, in other words, one zero for a 1-byte checksum, two zeros for a 2-byte
checksum, and four zeros for a 4-byte checksum.
● Any breakpoints in flash memory change the content of the flash. This means that
the checksum which is calculated by your application will no longer match the
initial checksum calculated by the linker. To make the two checksums match again,
you must disable all your breakpoints in flash and any breakpoints set in flash by
C-SPY internally. The stack plugin and the debugger option Run to both require
C-SPY to set breakpoints. Read more about possible breakpoint consumers in the
C-SPY® Debugging Guide for AVR.
203
AFE1_AFE2-1:1

204 for AVR
AFE1_AFE2-1:1
Efficient coding for
embedded applications
● Selecting data types
● Controlling data and function placement in memory
● Controlling compiler optimizations
● Facilitating good code generation
Selecting data types

For efficient treatment of data, you should consider the data types used and the most
efficient placement of the variables. This section provides useful information for
efficient treatment of data:
● Locating strings in ROM, RAM or flash
● Using efficient data types
● Floating-point types
● Memory model and memory attributes for data
● Using the best pointer type
● Anonymous structs and unions.
LOCATING STRINGS IN ROM, RAM OR FLASH

With the IAR C/C++ Compiler for AVR there are three possible locations for storing
strings:
● In external ROM in the data memory space
● In internal RAM in the data memory space
● In flash in the code memory space.
To read more about this, see Placing code and data—the linker configuration file, page
74.
205
AFE1_AFE2-1:1
Locating strings in flash

This can be done on individual strings or for the whole application/file using the option
--string_literals_in_flash. Examples on how to locate individual strings into
flash:
__flash char str1[] = "abcdef";
__flash char str2[] = "ghi";
__flash char __flash * pVar[] = { str1, str2 };
String literals cannot be put in flash automatically, but you can use a local static variable
instead:
#include <pgmspace.h>
void f (int i)
{
static __flash char sl[] = "%d cookies\n";
printf_P(sl, i);
}
This does not result in more code compared to allowing string literals in flash.
To use flash strings, you must use alternative library routines that expect flash strings.
A few such alternative functions are provided— they are declared in the pgmspace.h
header file. They are flash alternatives for some common C library functions with an
extension _P. For your own code, you can always use the __flash keyword when
passing the strings between functions.
For reference information about the alternative functions, see AVR-specific library
functions, page 370.
USING EFFICIENT DATA TYPES

The data types you use should be considered carefully, because this can have a large
impact on code size and code speed.
● Use small and unsigned data types, (unsigned char and unsigned short)
unless your application really requires signed values.
● Bitfields with sizes other than 1 bit should be avoided because they will result in
inefficient code compared to bit operations.
● When using arrays, it is more efficient if the type of the index expression matches
the index type of the memory of the array. For __near this is int.
● Using floating-point types on a microprocessor without a math co-processor is very
inefficient, both in terms of code size and execution speed.
● Declaring a pointer parameter to point to const data might open for better
optimizations in the calling function.

206 for AVR
AFE1_AFE2-1:1
Efficient coding for embedded applications
For information about representation of supported data types, pointers, and structures
types, see the chapter Data representation.
FLOATING-POINT TYPES
Using floating-point types on a microprocessor without a math coprocessor is very
inefficient, both in terms of code size and execution speed. Thus, you should consider
replacing code that uses floating-point operations with code that uses integers, because
these are more efficient.
The compiler supports two floating-point formats—32 and 64 bits. The 32-bit
floating-point type float is more efficient in terms of code size and execution speed.
However, the 64-bit format double supports higher precision and larger numbers.
In the compiler, the floating-point type float always uses the 32-bit format. The format
used by the double floating-point type depends on the setting of the
--64bit_doubles compiler option (Use 64-bit doubles).
Unless the application requires the extra precision that 64-bit floating-point numbers
give, we recommend using 32-bit floating-point numbers instead.
By default, a floating-point constant in the source code is treated as being of the type
double. This can cause innocent-looking expressions to be evaluated in double
precision. In the example below a is converted from a float to a double, the double
constant 1.0 is added and the result is converted back to a float:
double Test(float a)
{
return a + 1.0;
}
To treat a floating-point constant as a float rather than as a double, add the suffix f
to it, for example:
double Test(float a)
{
return a + 1.0f;
}
For more information about floating-point types, see Basic data types—floating-point
types, page 285.
MEMORY MODEL AND MEMORY ATTRIBUTES FOR DATA

For many applications it is sufficient to use the memory model feature to specify the
default memory for the data objects. However, for individual objects it might be
necessary to specify other memory attributes in certain cases, for example:
207
AFE1_AFE2-1:1
● An application where some global variables are accessed from a large number of
locations. In this case they can be declared to be placed in memory with a smaller
pointer type
● An application where all data, with the exception of one large chunk of data, fits
into the region of one of the smaller memory types
● Data that must be placed at a specific memory location.
The IAR C/C++ Compiler for AVR provides memory attributes for placing data objects
in the different memory spaces, and for the DATA and CODE space (flash) there are
different memory attributes for placing data objects in different memory types, see
Using data memory attributes, page 54.
Efficient usage of memory type attributes can significantly reduce the application size.
For details about the memory types, see Memory types, page 54.
USING THE BEST POINTER TYPE

The generic pointers, pointers declared __generic, can point to all memory spaces,
which makes them simple and also tempting to use. However, they carry a cost in that
special code is needed before each pointer access to check which memory a pointer
points to and taking the appropriate actions. Use the smallest pointer type that you can,
and avoid any generic pointers unless necessary.
ANONYMOUS STRUCTS AND UNIONS

When a structure or union is declared without a name, it becomes anonymous. The effect
is that its members will only be seen in the surrounding scope.
Anonymous structures are part of the C++ language; however, they are not part of the C
standard. In the IAR C/C++ Compiler for AVR they can be used in C if language
extensions are enabled.
Use the -e compiler option to enable language extensions. See -e, page 250, for
additional information.

208 for AVR
AFE1_AFE2-1:1
Example
In this example, the members in the anonymous union can be accessed, in function F,
without explicitly specifying the union name:
struct S
{
char mTag;
union
{
long mL;
float mF;
};
} St;
void F(void)
{
St.mL = 5;
}
The member names must be unique in the surrounding scope. Having an anonymous
struct or union at file scope, as a global, external, or static variable is also allowed.
This could for instance be used for declaring I/O registers, as in this example:
__no_init volatile
union
{
unsigned char IOPORT;
struct
{
unsigned char way: 1;
unsigned char out: 1;
};
} @ 40;
/* The variables are used here. */

void Test(void)
{
IOPORT = 0;
way = 1;
out = 1;
}
This declares an I/O register byte IOPORT at address 40. The I/O register has 2 bits
declared, way and out. Note that both the inner structure and the outer union are
anonymous.
209
AFE1_AFE2-1:1
Controlling data and function placement in memory
Anonymous structures and unions are implemented in terms of objects named after the
first field, with a prefix _A_ to place the name in the implementation part of the
namespace. In this example, the anonymous union will be implemented through an
object named _A_IOPORT.

The compiler provides different mechanisms for controlling placement of functions and
data objects in memory. To use memory efficiently, you should be familiar with these
mechanisms and know which one is best suited for different situations. You can use:
● Memory models
Use the different compiler options for memory models to take advantage of the
different addressing modes available for the microcontroller and thereby also place
data objects in different parts of memory. For more information about memory
models, see Memory models, page 59, and Function storage, page 63, respectively.
● Memory attributes
Using IAR-specific keywords or pragma directives, you can override the default
placement of functions, variables, and constants. For more information, see Using
function memory attributes, page 63 and Using data memory attributes, page 54.
● The @ operator and the #pragma location directive for absolute placement.
Using the @ operator or the #pragma location directive, you can place individual
global and static variables at absolute addresses. For more information, see Data
placement at an absolute location, page 210.
● The @ operator and the #pragma location directive for segment placement.
Using the @ operator or the #pragma location directive, you can place individual
functions, variables, and constants in named segments. The placement of these
segments can then be controlled by linker directives. For more information, see Data
and function placement in segments, page 212.
● Using the --segment option, you can set the default segment for functions,
variables, and constants in a particular module. For more information, see
--segment, page 271.
DATA PLACEMENT AT AN ABSOLUTE LOCATION

The @ operator, alternatively the #pragma location directive, can be used for placing
global and static variables at absolute addresses. The variables must be declared using
one of these combinations of keywords:
● __no_init
● __no_init and const (without initializers)

210 for AVR
AFE1_AFE2-1:1
● const (with initializers)
To place a variable at an absolute address, the argument to the @ operator and the
#pragma location directive should be a literal number, representing the actual
address.
Note: All declarations of __no_init variables placed at an absolute address are
tentative definitions. Tentatively defined variables are only kept in the output from the
compiler if they are needed in the module being compiled. Such variables will be
defined in all modules in which they are used, which will work as long as they are
defined in the same way. The recommendation is to place all such declarations in header
files that are included in all modules that use the variables.
Examples
In this example, a __no_init declared variable is placed at an absolute address. This
is useful for interfacing between multiple processes, applications, etc:
__no_init volatile char alpha @ 0xFF2000;/* OK */
The next example contains two const declared objects. The first one is not initialized,
and the second one is initialized to a specific value. Both objects are placed in ROM.
This is useful for configuration parameters, which are accessible from an external
interface. Note that in the second case, the compiler is not obliged to actually read from
the variable, because the value is known.
#pragma location=0xFF2002
__no_init const int beta; /* OK */
const int gamma @ 0xFF2004 = 3; /* OK */
In the first case, the value is not initialized by the compiler; the value must be set by
other means. The typical use is for configurations where the values are loaded to ROM
separately, or for special function registers that are read-only.
This shows incorrect usage:
int delta @ 0xFF2006; /* Error, neither */
/* "__no_init" nor "const".*/
C++ considerations
In C++, module scoped const variables are static (module local), whereas in C they are
global. This means that each module that declares a certain const variable will contain
a separate variable with this name. If you link an application with several such modules
all containing (via a header file), for instance, the declaration:
volatile const __no_init int x @ 0x100; /* Bad in C++ */
211
AFE1_AFE2-1:1
the linker will report that more than one variable is located at address 0x100.
To avoid this problem and make the process the same in C and C++, you should declare
these variables extern, for example:
/* The extern keyword makes x public. */
extern volatile const __no_init int x @ 0x100;
Note: C++ static member variables can be placed at an absolute address just like any
other static variable.
DATA AND FUNCTION PLACEMENT IN SEGMENTS

The following methods can be used for placing data or functions in named segments
other than default:
● The @ operator, alternatively the #pragma location directive, can be used for
placing individual variables or individual functions in named segments. The named
segment can either be a predefined segment, or a user-defined segment. The
variables must be declared either __no_init or const. If declared const, they
can have initializers.
● The --segment option can be used for placing variables and functions, which are
parts of the whole compilation unit, in named segments.
C++ static member variables can be placed in named segments just like any other static
variable.
If you use your own segments, in addition to the predefined segments, the segments
must also be defined in the linker configuration file using the -Z or the -P segment
control directives.
Note: Take care when explicitly placing a variable or function in a predefined segment
other than the one used by default. This is useful in some situations, but incorrect
placement can result in anything from error messages during compilation and linking to
a malfunctioning application. Carefully consider the circumstances; there might be strict
requirements on the declaration and use of the function or variable.
The location of the segments can be controlled from the linker configuration file.
For more information about segments, see the chapter Segment reference.
Examples of placing variables in named segments

In the following examples, a data object is placed in a user-defined segment. If no
memory attribute is specified, the variable will, like any other variable, be treated as if

212 for AVR
AFE1_AFE2-1:1
it is located in the default memory. Note that you must as always ensure that the segment
is placed in the appropriate memory area when linking.
__no_init int alpha @ "MY_NOINIT"; /* OK */
#pragma location="MY_CONSTANTS"
const int beta = 42; /* OK */
const int gamma @ "MY_CONSTANTS" = 17;/* OK */

int theta @ "MY_ZEROS"; /* OK */
int phi @ "MY_INITED" = 4711; /* OK */
The compiler will warn that segments that contain zero-initialized and initialized data
must be handled manually. To do this, you must use the linker option -Q to separate the
initializers into one separate segment and the symbols to be initialized to a different
segment. You must then write source code that copies the initializer segment to the
initialized segment, and zero-initialized symbols must be cleared before they are used.
As usual, you can use memory attributes to select a memory for the variable. Note that
you must as always ensure that the segment is placed in the appropriate memory area
when linking.
__near __no_init int alpha @ "MY_NEAR_NOINIT";/* Placed in
near*/
This example shows incorrect usage:

int delta @ "MY_ZEROS"; /* Error, neither */
/* "__no_init" nor "const" */
Examples of placing functions in named segments

void f(void) @ "MY_FUNCTIONS";
void g(void) @ "MY_FUNCTIONS"

{
}
#pragma location="MY_FUNCTIONS"
void h(void);
Specify a memory attribute to direct the function to a specific memory, and then modify
the segment placement in the linker configuration file accordingly:
__nearfunc void f(void) @ "MY_NEARFUNC_FUNCTIONS";
213
AFE1_AFE2-1:1
Controlling compiler optimizations

The compiler performs many transformations on your application to generate the best
possible code. Examples of such transformations are storing values in registers instead
of memory, removing superfluous code, reordering computations in a more efficient
order, and replacing arithmetic operations by cheaper operations.
The linker should also be considered an integral part of the compilation system, because
some optimizations are performed by the linker. For instance, all unused functions and
variables are removed and not included in the final output.
SCOPE FOR PERFORMED OPTIMIZATIONS

You can decide whether optimizations should be performed on your whole application
or on individual files. By default, the same types of optimizations are used for an entire
project, but you should consider using different optimization settings for individual files.
For example, put code that must execute very quickly into a separate file and compile it
for minimal execution time, and the rest of the code for minimal code size. This will give
a small program, which is still fast enough where it matters.
You can also exclude individual functions from the performed optimizations. The
#pragma optimize directive allows you to either lower the optimization level, or
specify another type of optimization to be performed. See optimize, page 332, for
information about the pragma directive.
MULTI-FILE COMPILATION UNITS

In addition to applying different optimizations to different source files or even functions,
you can also decide what a compilation unit consists of—one or several source code
files.
By default, a compilation unit consists of one source file, but you can also use multi-file
compilation to make several source files in a compilation unit. The advantage is that
interprocedural optimizations such as inlining, cross call, and cross jump have more
source code to work on. Ideally, the whole application should be compiled as one
compilation unit. However, for large applications this is not practical because of
resource restrictions on the host computer. For more information, see --mfc, page 259.
Note: Only one object file is generated, and thus all symbols will be part of that object
file.
If the whole application is compiled as one compilation unit, it is very useful to make
the compiler also discard unused public functions and variables before the
interprocedural optimizations are performed. Doing this limits the scope of the
optimizations to functions and variables that are actually used. For more information,
see --discard_unused_publics, page 248.

214 for AVR
AFE1_AFE2-1:1
OPTIMIZATION LEVELS
The compiler supports different levels of optimizations. This table lists optimizations
that are typically performed on each level:
Optimization level Description
None (Best debug support) Variables live through their entire scope
Low Same as above but variables only live for as long as they are
needed, not necessarily through their entire scope
Dead code elimination
Redundant label elimination
Redundant branch elimination
Medium Same as above, and:
Live-dead analysis and optimization
Code hoisting
Register content analysis and optimization
Common subexpression elimination
Static clustering
High (Balanced) Same as above, and:
Peephole optimization
Cross jumping
Cross call (when optimizing for size)
Function inlining
Code motion
Type-based alias analysis
Table 25: Compiler optimization levels
Note: Some of the performed optimizations can be individually enabled or disabled. For
more information about these, see Fine-tuning enabled transformations, page 216.
A high level of optimization might result in increased compile time, and will most likely
also make debugging more difficult, because it is less clear how the generated code
relates to the source code. For example, at the low, medium, and high optimization
levels, variables do not live through their entire scope, which means processor registers
used for storing variables can be reused immediately after they were last used. Due to
this, the C-SPY Watch window might not be able to display the value of the variable
throughout its scope, or even occasionally display an incorrect value. At any time, if you
experience difficulties when debugging your code, try lowering the optimization level.
SPEED VERSUS SIZE

At the high optimization level, the compiler balances between size and speed
optimizations. However, it is possible to fine-tune the optimizations explicitly for either
size or speed. They only differ in what thresholds that are used; speed will trade size for
215
AFE1_AFE2-1:1
speed, whereas size will trade speed for size. Note that one optimization sometimes
enables other optimizations to be performed, and an application might in some cases
become smaller even when optimizing for speed rather than size.
If you use the optimization level High speed, the --no_size_constraints compiler
option relaxes the normal restrictions for code size expansion and enables more
aggressive optimizations.
You can choose an optimization goal for each module, or even individual functions,
using command line options and pragma directives (see -O, page 265 and optimize, page
332). For a small embedded application, this makes it possible to achieve acceptable
speed performance while minimizing the code size: Typically, only a few places in the
application need to be fast, such as the most frequently executed inner loops, or the
interrupt handlers.
Rather than compiling the whole application with High (Balanced) optimization, you
can use High (Size) in general, but override this to get High (Speed) optimization only
for those functions where the application needs to be fast.
Because of the unpredictable way in which different optimizations interact, where one
optimization can enable other optimizations, sometimes a function becomes smaller
when compiled with High (Speed) optimization than if High (Size) is used. Also, using
multi-file compilation (see --mfc, page 259) can enable many optimizations to improve
both speed and size performance. It is recommended that you experiment with different
optimization settings so that you can pick the best ones for your project.
FINE-TUNING ENABLED TRANSFORMATIONS

At each optimization level you can disable some of the transformations individually. To
disable a transformation, use either the appropriate option, for instance the command
line option --no_inline, alternatively its equivalent in the IDE Function inlining, or
the #pragma optimize directive. These transformations can be disabled individually:
● Common subexpression elimination
● Function inlining
● Code motion
● Type-based alias analysis
● Static clustering
● Cross call
Common subexpression elimination

Redundant re-evaluation of common subexpressions is by default eliminated at
optimization levels Medium and High. This optimization normally reduces both code
size and execution time. However, the resulting code might be difficult to debug.

216 for AVR
AFE1_AFE2-1:1
Note: This option has no effect at optimization levels None and Low.
For more information about the command line option, see --no_cse, page 261.
Function inlining
Function inlining means that a function, whose definition is known at compile time, is
integrated into the body of its caller to eliminate the overhead of the call. This
optimization normally reduces execution time, but might increase the code size.
For more information, see Inlining functions, page 67.
Code motion
Evaluation of loop-invariant expressions and common subexpressions are moved to
avoid redundant re-evaluation. This optimization, which is performed at optimization
level Medium and above, normally reduces code size and execution time. The resulting
code might however be difficult to debug.
Note: This option has no effect at optimization levels below Medium.
For more information about the command line option, see --no_code_motion, page 261.
Type-based alias analysis

When two or more pointers reference the same memory location, these pointers are said
to be aliases for each other. The existence of aliases makes optimization more difficult
because it is not necessarily known at compile time whether a particular value is being
changed.
Type-based alias analysis optimization assumes that all accesses to an object are
performed using its declared type or as a char type. This assumption lets the compiler
detect whether pointers can reference the same memory location or not.
Type-based alias analysis is performed at optimization level High. For application code
conforming to standard C or C++ application code, this optimization can reduce code
size and execution time. However, non-standard C or C++ code might result in the
compiler producing code that leads to unexpected behavior. Therefore, it is possible to
turn this optimization off.
Note: This option has no effect at optimization levels None, Low, and Medium.
For more information about the command line option, see --no_tbaa, page 263.
217
AFE1_AFE2-1:1
Facilitating good code generation
Example
short F(short *p1, long *p2)
{
*p2 = 0;
*p1 = 1;
return *p2;
}
With type-based alias analysis, it is assumed that a write access to the short pointed to
by p1 cannot affect the long value that p2 points to. Thus, it is known at compile time
that this function returns 0. However, in non-standard-conforming C or C++ code these
pointers could overlap each other by being part of the same union. If you use explicit
casts, you can also force pointers of different pointer types to point to the same memory
location.
Static clustering
When static clustering is enabled, static and global variables that are defined within the
same module are arranged so that variables that are accessed in the same function are
stored close to each other. This makes it possible for the compiler to use the same base
pointer for several accesses.
Note: This option has no effect at optimization levels None and Low.
For more information about the command line option, see --no_clustering, page 260.
Cross call
Common code sequences are extracted to local subroutines. This optimization, which is
performed at optimization level High size, can reduce code size, sometimes
dramatically, on behalf of execution time and stack size. The resulting code might
however be difficult to debug. This optimization cannot be disabled using the #pragma
optimize directive.
If you want to use cross calls at any other optimization level, use --do_cross_call.
For more information about related command line options, see --no_cross_call, page
261, --do_cross_call, page 250, and --cross_call_passes, page 242.

This section contains hints on how to help the compiler generate good code:
● Writing optimization-friendly source code, page 219
● Saving stack space and RAM memory, page 219
● Function prototypes, page 220

218 for AVR
AFE1_AFE2-1:1
● Integer types and bit negation, page 221

● Protecting simultaneously accessed variables, page 221
● Accessing special function registers, page 222
● Non-initialized variables, page 223
WRITING OPTIMIZATION-FRIENDLY SOURCE CODE

The following is a list of programming techniques that will, when followed, enable the
compiler to better optimize the application.
● Local variables—auto variables and parameters—are preferred over static or global
variables. The reason is that the optimizer must assume, for example, that called
functions can modify non-local variables. When the life spans for local variables
end, the previously occupied memory can then be reused. Globally declared
variables will occupy data memory during the whole program execution.
● Avoid taking the address of local variables using the & operator. This is inefficient
for two main reasons. First, the variable must be placed in memory, and thus cannot
be placed in a processor register. This results in larger and slower code. Second, the
optimizer can no longer assume that the local variable is unaffected over function
calls.
● Module-local variables—variables that are declared static—are preferred over
global variables (non-static). Also avoid taking the address of frequently accessed
static variables.
● The compiler is capable of inlining functions, see Function inlining, page 217. To
maximize the effect of the inlining transformation, it is good practice to place the
definitions of small functions called from more than one module in the header file
rather than in the implementation file. Alternatively, you can use multi-file
compilation. For more information, see Multi-file compilation units, page 214.
● Avoid using inline assembler. Instead, try writing the code in C/C++, use intrinsic
functions, or write a separate module in assembler language. For more information,
see Mixing C and assembler, page 149.
SAVING STACK SPACE AND RAM MEMORY

The following is a list of programming techniques that save memory and stack space:
● If stack space is limited, avoid long call chains and recursive functions.
● Avoid using large non-scalar types, such as structures, as parameters or return type.
To save stack space, you should instead pass them as pointers or, in C++, as
references.
219
AFE1_AFE2-1:1
FUNCTION PROTOTYPES
It is possible to declare and define functions using one of two different styles:
● Prototyped
● Kernighan & Ritchie C (K&R C)
Both styles are valid C, however it is strongly recommended to use the prototyped style,
and provide a prototype declaration for each public function in a header that is included
both in the compilation unit defining the function and in all compilation units using it.
The compiler will not perform type checking on parameters passed to functions declared
using K&R style. Using prototype declarations will also result in more efficient code in
some cases, as there is no need for type promotion for these functions.
To make the compiler require that all function definitions use the prototyped style, and
that all public functions have been declared before being defined, use the
Project>Options>C/C++ Compiler>Language 1>Require prototypes compiler
option (--require_prototypes).
Prototyped style
In prototyped function declarations, the type for each parameter must be specified.
int Test(char, int); /* Declaration */
int Test(char ch, int i) /* Definition */

{
return i + ch;
}
Kernighan & Ritchie style

In K&R style—pre-Standard C—it is not possible to declare a function prototyped.
Instead, an empty parameter list is used in the function declaration. Also, the definition
looks different.
For example:
int Test(); /* Declaration */
int Test(ch, i) /* Definition */

char ch;
int i;
{
return i + ch;
}

220 for AVR
AFE1_AFE2-1:1
INTEGER TYPES AND BIT NEGATION

In some situations, the rules for integer types and their conversion lead to possibly
confusing behavior. Things to look out for are assignments or conditionals (test
expressions) involving types with different size, and logical operations, especially bit
negation. Here, types also includes types of constants.
In some cases there might be warnings (for example, for constant conditional or
pointless comparison), in others just a different result than what is expected. Under
certain circumstances the compiler might warn only at higher optimizations, for
example, if the compiler relies on optimizations to identify some instances of constant
conditionals. In this example an 8-bit character, a 16-bit integer, and two’s complement
is assumed:
void F1(unsigned char c1)
{
if (c1 == ~0x80)
;
}
Here, the test is always false. On the right hand side, 0x80 is 0x0080, and ~0x0080
becomes 0xFF7F. On the left hand side, c1 is an 8-bit unsigned character, so it cannot
be larger than 255. It also cannot be negative, which means that the integral promoted
value can never have the topmost 8 bits set.
PROTECTING SIMULTANEOUSLY ACCESSED VARIABLES

Variables that are accessed asynchronously, for example by interrupt routines or by code
executing in separate threads, must be properly marked and have adequate protection.
The only exception to this is a variable that is always read-only.
To mark a variable properly, use the volatile keyword. This informs the compiler,
among other things, that the variable can be changed from other threads. The compiler
will then avoid optimizing on the variable (for example, keeping track of the variable in
registers), will not delay writes to it, and be careful accessing the variable only the
number of times given in the source code.
For sequences of accesses to variables that you do not want to be interrupted, use the
__monitor keyword. This must be done for both write and read sequences, otherwise
you might end up reading a partially updated variable. Accessing a small-sized
volatile variable can be an atomic operation, but you should not rely on it unless you
continuously study the compiler output. It is safer to use the __monitor keyword to
ensure that the sequence is an atomic operation. For more information, see __monitor,
page 308.
For more information about the volatile type qualifier and the rules for accessing
volatile objects, see Declaring objects volatile, page 291.
221
AFE1_AFE2-1:1
Protecting the eeprom write mechanism

A typical example of when it can be necessary to use the __monitor keyword is when
protecting the eeprom write mechanism, which can be used from two threads (for
example, main code and interrupts). Servicing an interrupt during an EEPROM write
sequence can in many cases corrupt the written data.
ACCESSING SPECIAL FUNCTION REGISTERS

Specific header files for several AVR devices are included in the IAR product
installation. The header files are named iodevice.h and define the processor-specific
special function registers (SFRs).
To enable bit definitions in IAR Embedded Workbench, select the option General
Options>System>Enable bit definitions in I/O include files.
Note: Each header file contains one section used by the compiler, and one section used
by the assembler.
SFRs with bitfields are declared in the header file. This example is from iom128.h:
__io union
{
unsigned char PORTE; /* The sfrb as 1 byte */
struct
{
unsigned char PORTE_Bit0:1,
PORTE_Bit1:1,
PORTE_Bit2:1,
PORTE_Bit3:1,
PORTE_Bit4:1,
PORTE_Bit5:1,
PORTE_Bit6:1,
PORTE_Bit7:1;
};
} @ 0x1F;
By including the appropriate include file in your code, it is possible to access either the
whole register or any individual bit (or bitfields) from C code as follows:
/* whole register access */
PORTE = 0x12;
/* Bitfield accesses */
PORTE_Bit0 = 1;
You can also use the header files as templates when you create new header files for other
AVR devices.

222 for AVR
AFE1_AFE2-1:1
NON-INITIALIZED VARIABLES
Normally, the runtime environment will initialize all global and static variables when the
application is started.
The compiler supports the declaration of variables that will not be initialized, using the
__no_init type modifier. They can be specified either as a keyword or using the
#pragma object_attribute directive. The compiler places such variables in a
separate segment, according to the specified memory keyword. See the chapter Linking
overview for more information.
For __no_init, the const keyword implies that an object is read-only, rather than that
the object is stored in read-only memory. It is not possible to give a __no_init object
an initial value.
Variables declared using the __no_init keyword could, for example, be large input
buffers or mapped to special RAM that keeps its content even when the application is
turned off.
For more information, see __no_init, page 311. Note that to use this keyword, language
extensions must be enabled; see -e, page 250. For more information, see also
object_attribute, page 332.
223
AFE1_AFE2-1:1

224 for AVR
AFE1_AFE2-1:1
Part 2. Reference
information
This part of the IAR C/C++ Compiler User Guide for AVR contains these
chapters:
● External interface details
● Compiler options
● Data representation
● Extended keywords
● Pragma directives
● Intrinsic functions
● The preprocessor
● C/C++ standard library functions
● Segment reference
● The stack usage control file
● Implementation-defined behavior for Standard C
● Implementation-defined behavior for C89.
225
226
External interface details
● Invocation syntax
● Include file search procedure
● Compiler output
● Diagnostics
Invocation syntax
You can use the compiler either from the IDE or from the command line. See the IDE
Project Management and Building Guide for AVR for information about using the
compiler from the IDE.
COMPILER INVOCATION SYNTAX

The invocation syntax for the compiler is:
iccavr [options] [sourcefile] [options]
For example, when compiling the source file prog.c, use this command to generate an
object file with debug information:
iccavr prog.c --debug
The source file can be a C or C++ file, typically with the filename extension c or cpp,
respectively. If no filename extension is specified, the file to be compiled must have the
extension c.
Generally, the order of options on the command line, both relative to each other and to
the source filename, is not significant. There is, however, one exception: when you use
the -I option, the directories are searched in the same order as they are specified on the
command line.
If you run the compiler from the command line without any arguments, the compiler
version number and all available options including brief descriptions are directed to
stdout and displayed on the screen.
227
AFE1_AFE2-1:1
Include file search procedure
PASSING OPTIONS
There are three different ways of passing options to the compiler:
● Directly from the command line
Specify the options on the command line after the iccavr command, either before
or after the source filename; see Invocation syntax, page 227.
● Via environment variables
The compiler automatically appends the value of the environment variables to every
command line; see Environment variables, page 228.
● Via a text file, using the -f option; see -f, page 254.
For general guidelines for the option syntax, an options summary, and a detailed
description of each option, see the chapter Compiler options.
ENVIRONMENT VARIABLES
These environment variables can be used with the compiler:
Environment variable Description
C_INCLUDE Specifies directories to search for include files; for example:
C_INCLUDE=c:\program files\iar systems\embedded
workbench 7.n\avr\inc;c:\headers
QCCAVR Specifies command line options; for example: QCCAVR=-lA
asm.lst
Table 26: Compiler environment variables
Include file search procedure

This is a detailed description of the compiler’s #include file search procedure:
● If the name of the #include file is an absolute path specified in angle brackets or
double quotes, that file is opened.
● If the compiler encounters the name of an #include file in angle brackets, such as:
#include <stdio.h>
it searches these directories for the file to include:
1 The directories specified with the -I option, in the order that they were
specified, see -I, page 255.
2 The directories specified using the C_INCLUDE environment variable, if any; see
Environment variables, page 228.
3 The automatically set up library system include directories. See --clib, page 241,
--dlib, page 248, and --dlib_config, page 249.

228 for AVR
AFE1_AFE2-1:1
● If the compiler encounters the name of an #include file in double quotes, for
example:
#include "vars.h"
it searches the directory of the source file in which the #include statement occurs,
and then performs the same sequence as for angle-bracketed filenames.
If there are nested #include files, the compiler starts searching the directory of the file
that was last included, iterating upwards for each included file, searching the source file
directory last. For example:
src.c in directory dir\src
#include "src.h"
...
src.h in directory dir\include
#include "config.h"
...
When dir\exe is the current directory, use this command for compilation:
iccavr ..\src\src.c -I..\include -I..\debugconfig
Then the following directories are searched in the order listed below for the file
config.h, which in this example is located in the dir\debugconfig directory:
dir\include Current file is src.h.

dir\src File including current file (src.c).
dir\include As specified with the first -I option.
dir\debugconfig As specified with the second -I option.
Use angle brackets for standard header files, like stdio.h, and double quotes for files
that are part of your application.
Note: Both \ and / can be used as directory delimiters.
For more information, see Overview of the preprocessor, page 351.
Compiler output
The compiler can produce the following output:
● A linkable object file
The object files produced by the compiler use a proprietary format called UBROF,
which stands for Universal Binary Relocatable Object Format. By default, the object
file has the filename extension r90.
229
AFE1_AFE2-1:1
Compiler output
● Optional list files

Various kinds of list files can be specified using the compiler option -l, see -l, page
256. By default, these files will have the filename extension lst.
● Optional preprocessor output files
A preprocessor output file is produced when you use the --preprocess option; by
default, the file will have the filename extension i.
● Diagnostic messages
Diagnostic messages are directed to the standard error stream and displayed on the
screen, and printed in an optional list file. For more information about diagnostic
messages, see Diagnostics, page 231.
● Error return codes
These codes provide status information to the operating system which can be tested
in a batch file, see Error return codes, page 230.
● Size information
Information about the generated amount of bytes for functions and data for each
memory is directed to the standard output stream and displayed on the screen. Some
of the bytes might be reported as shared.
Shared objects are functions or data objects that are shared between modules. If any
of these occur in more than one module, only one copy is retained. For example, in
some cases inline functions are not inlined, which means that they are marked as
shared, because only one instance of each function will be included in the final
application. This mechanism is sometimes also used for compiler-generated code or
data not directly associated with a particular function or variable, and when only one
instance is required in the final application.
ERROR RETURN CODES

The compiler returns status information to the operating system that can be tested in a
batch file.
These command line error codes are supported:
Code Description
0 Compilation successful, but there might have been warnings.
1 Warnings were produced and the option --warnings_affect_exit_code
was used.
2 Errors occurred.
3 Fatal errors occurred, making the compiler abort.
4 Internal errors occurred, making the compiler abort.
Table 27: Error return codes

230 for AVR
AFE1_AFE2-1:1
Diagnostics
This section describes the format of the diagnostic messages and explains how
diagnostic messages are divided into different levels of severity.
MESSAGE FORMAT
All diagnostic messages are issued as complete, self-explanatory messages. A typical
diagnostic message from the compiler is produced in the form:
filename,linenumber level[tag]: message
with these elements:
filename The name of the source file in which the issue was encountered
linenumber The line number at which the compiler detected the issue
level The level of seriousness of the issue
tag A unique tag that identifies the diagnostic message
message An explanation, possibly several lines long
Diagnostic messages are displayed on the screen, as well as printed in the optional list
file.
Use the option --diagnostics_tables to list all possible compiler diagnostic
messages.
SEVERITY LEVELS
The diagnostic messages are divided into different levels of severity:
Remark
A diagnostic message that is produced when the compiler finds a source code construct
that can possibly lead to erroneous behavior in the generated code. Remarks are by
default not issued, but can be enabled, see --remarks, page 269.
Warning
A diagnostic message that is produced when the compiler finds a potential programming
error or omission which is of concern, but which does not prevent completion of the
compilation. Warnings can be disabled by use of the command line option
--no_warnings, see --no_warnings, page 265.
231
AFE1_AFE2-1:1
Diagnostics
Error
A diagnostic message that is produced when the compiler finds a construct which clearly
violates the C or C++ language rules, such that code cannot be produced. An error will
produce a non-zero exit code.
Fatal error
A diagnostic message that is produced when the compiler finds a condition that not only
prevents code generation, but which makes further processing of the source code
pointless. After the message is issued, compilation terminates. A fatal error will produce
a non-zero exit code.
SETTING THE SEVERITY LEVEL

The diagnostic messages can be suppressed or the severity level can be changed for all
diagnostics messages, except for fatal errors and some of the regular errors.
See the chapter Compiler options, for information about the compiler options that are
available for setting severity levels.
See the chapter Pragma directives, for information about the pragma directives that are
available for setting severity levels.
INTERNAL ERROR
An internal error is a diagnostic message that signals that there was a serious and
unexpected failure due to a fault in the compiler. It is produced using this form:
Internal error: message
where message is an explanatory message. If internal errors occur, they should be

reported to your software distributor or IAR Systems Technical Support. Include enough
information to reproduce the problem, typically:
● The product name
● The version number of the compiler, which can be seen in the header of the list files
generated by the compiler
● Your license number
● The exact internal error message text
● The source file of the application that generated the internal error
● A list of the options that were used when the internal error occurred.

232 for AVR
AFE1_AFE2-1:1
Compiler options
● Options syntax
● Summary of compiler options
● Descriptions of compiler options
Options syntax
Compiler options are parameters you can specify to change the default behavior of the
compiler. You can specify options from the command line—which is described in more
detail in this section—and from within the IDE.
See the online help system for information about the compiler options available in the
IDE and how to set them.
TYPES OF OPTIONS
There are two types of names for command line options, short names and long names.
Some options have both.
● A short option name consists of one character, and it can have parameters. You
specify it with a single dash, for example -e
● A long option name consists of one or several words joined by underscores, and it
can have parameters. You specify it with double dashes, for example
--char_is_signed.
For information about the different methods for passing options, see Passing options,
page 228.
RULES FOR SPECIFYING PARAMETERS

There are some general syntax rules for specifying option parameters. First, the rules
depending on whether the parameter is optional or mandatory, and whether the option
has a short or a long name, are described. Then, the rules for specifying filenames and
directories are listed. Finally, the remaining rules are listed.
Rules for optional parameters

For options with a short name and an optional parameter, any parameter should be
specified without a preceding space, for example:
-O or -Oh
233
AFE1_AFE2-1:1
Options syntax
For options with a long name and an optional parameter, any parameter should be
specified with a preceding equal sign (=), for example:
--misrac2004=n
Rules for mandatory parameters

For options with a short name and a mandatory parameter, the parameter can be
specified either with or without a preceding space, for example:
-I..\src or -I ..\src\
For options with a long name and a mandatory parameter, the parameter can be specified
either with a preceding equal sign (=) or with a preceding space, for example:
--diagnostics_tables=MyDiagnostics.lst
or
--diagnostics_tables MyDiagnostics.lst
Rules for options with both optional and mandatory parameters

For options taking both optional and mandatory parameters, the rules for specifying the
parameters are:
● For short options, optional parameters are specified without a preceding space
● For long options, optional parameters are specified with a preceding equal sign (=)
● For short and long options, mandatory parameters are specified with a preceding
space.
For example, a short option with an optional parameter followed by a mandatory
parameter:
-lA MyList.lst
For example, a long option with an optional parameter followed by a mandatory

parameter:
--preprocess=n PreprocOutput.lst
Rules for specifying a filename or directory as parameters

These rules apply for options taking a filename or directory as parameters:
● Options that take a filename as a parameter can optionally take a file path. The path
can be relative or absolute. For example, to generate a listing to the file List.lst
in the directory ..\listings\:
iccavr prog.c -l ..\listings\List.lst

234 for AVR
AFE1_AFE2-1:1
Compiler options
● For options that take a filename as the destination for output, the parameter can be
specified as a path without a specified filename. The compiler stores the output in
that directory, in a file with an extension according to the option. The filename will
be the same as the name of the compiled source file, unless a different name was
specified with the option -o, in which case that name is used. For example:
iccavr prog.c -l ..\listings\
The produced list file will have the default name ..\listings\prog.lst
● The current directory is specified with a period (.). For example:
iccavr prog.c -l .
● / can be used instead of \ as the directory delimiter.
● By specifying -, input files and output files can be redirected to the standard input
and output stream, respectively. For example:
iccavr prog.c -l -
Additional rules
These rules also apply:
● When an option takes a parameter, the parameter cannot start with a dash (-)
followed by another character. Instead, you can prefix the parameter with two
dashes; this example will create a list file called -r:
iccavr prog.c -l ---r
● For options that accept multiple arguments of the same type, the arguments can be
provided as a comma-separated list (without a space), for example:
--diag_warning=Be0001,Be0002
Alternatively, the option can be repeated for each argument, for example:
--diag_warning=Be0001
--diag_warning=Be0002
Summary of compiler options

This table summarizes the compiler command line options:
Command line option Description
--64bit_doubles Forces the compiler to use 64-bit doubles
--64k_flash Specifies a maximum of 64 Kbytes flash memory
--c89 Specifies the C89 dialect
--char_is_signed Treats char as signed
--char_is_unsigned Treats char as unsigned
Table 28: Compiler options summary
235
AFE1_AFE2-1:1

--clib Uses the system include files for the CLIB library
--cpu Specifies a specific device
--cross_call_passes Cross call optimization
-D Defines preprocessor symbols
--debug Generates debug information
--dependencies Lists file dependencies
--diag_error Treats these as errors
--diag_remark Treats these as remarks
--diag_suppress Suppresses these diagnostics
--diag_warning Treats these as warnings
--diagnostics_tables Lists all diagnostic messages
--disable_all_program_memory_ Disables the LPM/ELPM instructions
load_instructions
--disable_direct_mode Disables direct addressing mode
--disable_library_knowledge Disables the knowledge about the support
routines in the library
--disable_mul Disables the MUL instructions
--disable_spm Disables the SPM instructions
--discard_unused_publics Discards unused public symbols
--dlib Uses the system include files for the DLIB library
--dlib_config Uses the system include files for the DLIB library
and determines which configuration of the library
to use
--do_cross_call Forces the compiler to run the cross call
optimizer.
-e Enables language extensions
--ec++ Specifies Embedded C++
--eec++ Specifies Extended Embedded C++
--eecr_address Defines the EECR address
--eeprom_size Specifies the EEPROM size
--enable_external_bus Adds code to enable external data bus
--enable_multibytes Enables support for multibyte characters in source
files
Table 28: Compiler options summary (Continued)

236 for AVR
AFE1_AFE2-1:1
Compiler options

--enable_restrict Enables the Standard C keyword restrict
--enhanced_core Enables enhanced instruction set
--error_limit Specifies the allowed number of errors before
compilation stops
-f Extends the command line
--force_switch_type Forces the switch type
--guard_calls Enables guards for function static variable
initialization
--header_context Lists all referred source files and header files
-I Specifies include file path
--initializers_in_flash Places aggregate initializers in flash memory
-l Creates a list file
--library_module Creates a library module
--lock_regs Locks registers
-m Specifies a memory model
--macro_positions_in Obtains positions inside macros in diagnostic
_diagnostics messages
--memory_model Specifies a memory model
--mfc Enables multi-file compilation
--misrac Enables error messages specific to MISRA-C:1998.
This option is a synonym of --misrac1998 and
is only available for backwards compatibility.
--misrac1998 Enables error messages specific to MISRA-C:1998.
See the IAR Embedded Workbench® MISRA C:1998
Reference Guide.
--misrac2004 Enables error messages specific to MISRA-C:2004.
See the IAR Embedded Workbench® MISRA C:2004
Reference Guide.
--misrac_verbose IAR Embedded Workbench® MISRA C:1998
Reference Guide or the IAR Embedded Workbench®
MISRA C:2004 Reference Guide.
--module_name Sets the object module name
--no_call_frame_info Disables output of call frame information
--no_clustering Disables static clustering optimizations
237
AFE1_AFE2-1:1

--no_code_motion Disables code motion optimization
--no_cross_call Disables cross-call optimization
--no_cse Disables common subexpression elimination
--no_inline Disables function inlining
--no_path_in_file_macros Removes the path from the return value of the
symbols __FILE__ and __BASE_FILE__
--no_rampd Uses RAMPZ instead of RAMPD
--no_size_constraints Relaxes the normal restrictions for code size
expansion when optimizing for speed.
--no_static_destruction Disables destruction of C++ static variables at
program exit
--no_system_include Disables the automatic search for system include
files
--no_tbaa Disables type-based alias analysis
--no_typedefs_in_diagnostics Disables the use of typedef names in diagnostics
--no_ubrof_messages Excludes messages from UBROF files
--no_unroll Disables loop unrolling
--no_warnings Disables all warnings
--no_wrap_diagnostics Disables wrapping of diagnostic messages
-O Sets the optimization level
-o Sets the object filename. Alias for --output.
--omit_types Excludes type information
--only_stdout Uses standard output only
--output Sets the object filename
--pending_instantiations Sets the maximum number of instantiations of a
given C++ template.
--predef_macros Lists the predefined symbols.
--preinclude Includes an include file before reading the source
file
--preprocess Generates preprocessor output
--public_equ Defines a global named assembler label
-r Generates debug information. Alias for --debug.

238 for AVR
AFE1_AFE2-1:1
Compiler options

--relaxed_fp Relaxes the rules for optimizing floating-point
expressions
--remarks Enables remarks
--require_prototypes Verifies that functions are declared before they are
defined
--root_variables Specifies variables as __root
-s Optimizes for speed.
--segment Changes a segment name
--separate_cluster_for_ Separates initialized and non-initialized variables
initialized_variables
--silent Sets silent operation
--spmcr_address Defines the SPMCR address
--strict Checks for strict compliance with Standard
C/C++
--string_literals_in_flash Puts “string” in the __nearflash or
__farflash segment
--system_include_dir Specifies the path for system include files
--use_c++_inline Uses C++ inline semantics in C99
-v Specifies the processor variant
--version
--version1_calls Uses the ICCA90 calling convention
--version2_calls Uses the V2.10-V4.10B calling convention
--vla Enables C99 VLA support
--warn_about_c_style_casts Makes the compiler warn when C-style casts are
used in C++ source code
--warnings_affect_exit_code Warnings affect exit code
--warnings_are_errors Warnings are treated as errors
--xmcra_address Specifies where XMCRA is located
-y Places constants and literals
-z Optimizes for size
--zero_register Specifies register R15 as the zero register
239
AFE1_AFE2-1:1
Descriptions of compiler options

The following section gives detailed reference information about each compiler option.
Note that if you use the options page Extra Options to specify specific command line
options, the IDE does not perform an instant check for consistency problems like
conflicting options, duplication of options, or use of irrelevant options.
--64bit_doubles
Syntax --64bit_doubles
Description Use this option to force the compiler to use 64-bit doubles instead of 32-bit doubles
which is the default.
See also Floating-point types, page 285.

Project>Options>General Options>Target
--64k_flash
Syntax --64k_flash
Description This option tells the compiler that the intended target processor does not have more than
64 Kbytes program flash memory (small flash), and that the AVR core therefore does
not have the RAMPZ register or the ELPM instruction.
This option can only be used together with the -v2, -v3, and -v4 processor options.
Project>Options>General Options>Target (No RAMPZ register)
--c89
Syntax --c89
Description Use this option to enable the C89 C dialect instead of Standard C.
Note: This option is mandatory when the MISRA C checking is enabled.
See also C language overview, page 167.

240 for AVR
AFE1_AFE2-1:1
Compiler options
Project>Options>C/C++ Compiler>Language 1>C dialect>C89
--char_is_signed
Syntax --char_is_signed
Description By default, the compiler interprets the plain char type as unsigned. Use this option to
make the compiler interpret the plain char type as signed instead. This can be useful
when you, for example, want to maintain compatibility with another compiler.
Note: The runtime library is compiled without the --char_is_signed option and
cannot be used with code that is compiled with this option.
Project>Options>C/C++ Compiler>Language 2>Plain ‘char’ is
--char_is_unsigned
Syntax --char_is_unsigned
Description Use this option to make the compiler interpret the plain char type as unsigned. This is
the default interpretation of the plain char type.
Project>Options>C/C++ Compiler>Language 2>Plain ‘char’ is
--clib
Syntax --clib
Description Use this option to use the system header files for the CLIB library; the compiler will
automatically locate the files and use them when compiling.
Note: The CLIB library is used by default. To use the DLIB library, use the --dlib or
the --dlib_config option instead.
See also --dlib, page 248 and --dlib_config, page 249.

To set related options, choose:
Project>Options>General Options>Library Configuration
241
AFE1_AFE2-1:1
--cpu
Syntax --cpu=processor
Parameters
processor Specifies a specific device
Description The compiler supports different processor variants. Use this option to select a specific
processor variant for which the code will be generated.
Note that to specify the processor, you can use either the --cpu option or the -v option.
The --cpu option is, however, more precise because it contains more information about
the intended target than the more generic -v option.
See also -v, page 275, Processor configuration, page 49.

Project>Options>General Options>Target>Processor configuration
--cross_call_passes
Syntax --cross_call_passes=N
Parameters
N The number of times to run the cross call optimizer, which can be
1–5.
Description Use this option to run the cross-call optimizer for decreasing the RSTACK usage. The
default is to run it until no more improvements can be made.
Note: Use this option if you have a target processor with a hardware stack or a small
internal return stack segment, RSTACK.
See also --no_cross_call, page 261.

Project>Options>C/C++ Compiler>Optimizations

242 for AVR
AFE1_AFE2-1:1
Compiler options
-D
Syntax -D symbol[=value]
Parameters
symbol The name of the preprocessor symbol
value The value of the preprocessor symbol
Description Use this option to define a preprocessor symbol. If no value is specified, 1 is used. This
option can be used one or more times on the command line.
The option -D has the same effect as a #define statement at the top of the source file:
-Dsymbol
is equivalent to:
#define symbol 1
To get the equivalence of:

#define FOO
specify the = sign but nothing after, for example:

-DFOO=
Project>Options>C/C++ Compiler>Preprocessor>Defined symbols
--debug, -r
Syntax --debug
-r
Description Use the --debug or -r option to make the compiler include information in the object
modules required by the IAR C-SPY® Debugger and other symbolic debuggers.
Note: Including debug information will make the object files larger than otherwise.
Project>Options>C/C++ Compiler>Output>Generate debug information
243
AFE1_AFE2-1:1
--dependencies
Syntax --dependencies[=[i|m|n][s]] {filename|directory|+}
Parameters
i (default) Lists only the names of files
m Lists in makefile style (multiple rules)
n Lists in makefile style (one rule)
s Suppresses system files
+ Gives the same output as -o, but with the filename extension d
See also Rules for specifying a filename or directory as parameters, page 234.
Description Use this option to make the compiler list the names of all source and header files opened
for input into a file with the default filename extension i.
Example If --dependencies or --dependencies=i is used, the name of each opened input

file, including the full path, if available, is output on a separate line. For example:
c:\iar\product\include\stdio.h
d:\myproject\include\foo.h
If --dependencies=m is used, the output is in makefile style. For each input file, one
line containing a makefile dependency rule is produced. Each line consists of the name
of the object file, a colon, a space, and the name of an input file. For example:
foo.r90: c:\iar\product\include\stdio.h
foo.r90: d:\myproject\include\foo.h
An example of using --dependencies with a popular make utility, such as gmake

(GNU make):
1 Set up the rule for compiling files to be something like:
%.r90 : %.c
$(ICC) $(ICCFLAGS) $< --dependencies=m $*.d
That is, in addition to producing an object file, the command also produces a
dependency file in makefile style (in this example, using the extension .d).
2 Include all the dependency files in the makefile using, for example:
-include $(sources:.c=.d)
Because of the dash (-) it works the first time, when the .d files do not yet exist.

244 for AVR
AFE1_AFE2-1:1
Compiler options
This option is not available in the IDE.
--diag_error
Syntax --diag_error=tag[,tag,...]
Parameters
tag The number of a diagnostic message, for example the
message number Pe117
Description Use this option to reclassify certain diagnostic messages as errors. An error indicates a
violation of the C or C++ language rules, of such severity that object code will not be
generated. The exit code will be non-zero. This option may be used more than once on
the command line.
Project>Options>C/C++ Compiler>Diagnostics>Treat these as errors
--diag_remark
Syntax --diag_remark=tag[,tag,...]
Parameters
Description Use this option to reclassify certain diagnostic messages as remarks. A remark is the
least severe type of diagnostic message and indicates a source code construction that
may cause strange behavior in the generated code. This option may be used more than
once on the command line.
Note: By default, remarks are not displayed; use the --remarks option to display them.
Project>Options>C/C++ Compiler>Diagnostics>Treat these as remarks
245
AFE1_AFE2-1:1
--diag_suppress
Syntax --diag_suppress=tag[,tag,...]
Parameters
Description Use this option to suppress certain diagnostic messages. These messages will not be
displayed. This option may be used more than once on the command line.
Project>Options>C/C++ Compiler>Diagnostics>Suppress these diagnostics
--diag_warning
Syntax --diag_warning=tag[,tag,...]
Parameters
Description Use this option to reclassify certain diagnostic messages as warnings. A warning
indicates an error or omission that is of concern, but which will not cause the compiler
to stop before compilation is completed. This option may be used more than once on the
command line.
Project>Options>C/C++ Compiler>Diagnostics>Treat these as warnings
--diagnostics_tables
Syntax --diagnostics_tables {filename|directory}
Parameters See Rules for specifying a filename or directory as parameters, page 234.
Description Use this option to list all possible diagnostic messages to a named file. This can be
convenient, for example, if you have used a pragma directive to suppress or change the
severity level of any diagnostic messages, but forgot to document why.
Typically, this option cannot be given together with other options.

246 for AVR
AFE1_AFE2-1:1
Compiler options
To set this option, use Project>Options>C/C++ Compiler>Extra Options.
--disable_all_program_memory_load_instructions
Syntax --disable_all_program_memory_load_instructions
Description Use this option to disable the LPM/ELPM instructions.

To set this option, use: Project>Options>C/C++ Compiler>Extra Options.
--disable_direct_mode
Syntax --disable_direct_mode
Description Use this option to prevent the compiler from generating the direct addressing mode
instructions LDS and STS.
--disable_library_knowledge
Syntax --disable_library_knowledge
Description Use this option to disable the knowledge about the support routines in the library.
--disable_mul
Syntax --disable_mul
Description Use this option to disable the MUL instructions.

247
AFE1_AFE2-1:1
--disable_spm
Syntax --disable_spm
Description Use this option to disable the SPM instruction.

--discard_unused_publics
Syntax --discard_unused_publics
Description Use this option to discard unused public functions and variables when compiling with
the --mfc compiler option.
Note: Do not use this option only on parts of the application, as necessary symbols
might be removed from the generated output. Use the object attribute __root to keep
symbols that are used from outside the compilation unit, for example interrupt handlers.
If the symbol does not have the __root attribute and is defined in the library, the library
definition will be used instead.
See also --mfc, page 259 and Multi-file compilation units, page 214.
Project>Options>C/C++ Compiler>Discard unused publics
--dlib
Syntax --dlib
Description Use this option to use the system header files for the DLIB library; the compiler will
automatically locate the files and use them when compiling.
Note: The DLIB library is used by default: To use the CLIB library, use the --clib
option instead.
See also --dlib_config, page 249, --no_system_include, page 263, --system_include_dir, page
274, and --clib, page 241.

248 for AVR
AFE1_AFE2-1:1
Compiler options
--dlib_config
Syntax --dlib_config filename.h|config
Parameters
filename A DLIB configuration header file, see Rules for specifying a
filename or directory as parameters, page 234.
config The default configuration file for the specified configuration
will be used. Choose between:
none, no configuration will be used
normal, the normal library configuration will be used

(default)
full, the full library configuration will be used.
Description Use this option to specify which library configuration to use, either by specifying an
explicit file or by specifying a library configuration—in which case the default file for
that library configuration will be used. Make sure that you specify a configuration that
corresponds to the library you are using. If you do not specify this option, the default
library configuration file will be used.
All prebuilt runtime libraries are delivered with corresponding configuration files. You
can find the library object files and the library configuration files in the directory
avr\lib. For examples and information about prebuilt runtime libraries, see Prebuilt
runtime libraries, page 111.
If you build your own customized runtime library, you should also create a
corresponding customized library configuration file, which must be specified to the
compiler. For more information, see Customizing and building your own runtime
library, page 108.
Note: This option only applies to the IAR DLIB runtime environment.
249
AFE1_AFE2-1:1
--do_cross_call
Syntax --do_cross_call
Description Use this option to force the compiler to run the cross call optimizer, regardless of the
optimization level. The cross call optimizer is otherwise only run at high size
optimization.
See also -s, page 271.

-e
Syntax -e
Description In the command line version of the compiler, language extensions are disabled by
default. If you use language extensions such as extended keywords and anonymous
structs and unions in your source code, you must use this option to enable them.
Note: The -e option and the --strict option cannot be used at the same time.
See also Enabling language extensions, page 169.

Project>Options>C/C++ Compiler>Language 1>Standard with IAR extensions
Note: By default, this option is selected in the IDE.
--ec++
Syntax --ec++
Description In the compiler, the default language is C. If you use Embedded C++, you must use this
option to set the language the compiler uses to Embedded C++.
Project>Options>C/C++ Compiler>Language 1>C++
and
Project>Options>C/C++ Compiler>Language 1>C++ dialect>Embedded C++

250 for AVR
AFE1_AFE2-1:1
Compiler options
--eec++
Syntax --eec++
Description In the compiler, the default language is C. If you take advantage of Extended Embedded
C++ features like namespaces or the standard template library in your source code, you
must use this option to set the language the compiler uses to Extended Embedded C++.
See also Extended Embedded C++, page 176.

Project>Options>C/C++ Compiler>Language 1>C++
and
Project>Options>C/C++ Compiler>Language 1>C++ dialect>Extended
Embedded C++
--eecr_address
Syntax --eecr_address address
Parameters
address The value of the EECR address. The default is 0x1C.
Description If you use the -v processor option, the --eecr_address option can be used for
modifying the value of the EECR address.
If you use the --cpu processor option, the --eecr_address option is implicitly set,
which means you should not use these options together.
See also -v, page 275 and --cpu, page 242.

--eeprom_size
Syntax --eeprom_size=N
Parameters
N The size of the EEPROM in bytes, 0–65536.
251
AFE1_AFE2-1:1
Description Use this option to enable the __eeprom extended keyword by specifying the size of the
built-in EEPROM.
To use the __eeprom extended keyword, language extensions must be enabled.
See also Function storage, page 63, -e, page 250 and language, page 330.
Project>Options>C/C++ Compiler>Code
--enable_external_bus
Syntax --enable_external_bus
Description Use this option to make the compiler add the special __require statement which
makes XLINK include the code in startup.s90 that enables the external data bus. Use
this option if you intend to place RSTACK in external RAM.
Note: The code in cstartup.s90 that enables the external data bus is preferably
placed in low_level_init instead.
See also The return address stack, page 193.

Project>Options>General Options>System
--enable_multibytes
Syntax --enable_multibytes
Description By default, multibyte characters cannot be used in C or C++ source code. Use this option
to make multibyte characters in the source code be interpreted according to the host
computer’s default setting for multibyte support.
Multibyte characters are allowed in C and C++ style comments, in string literals, and in
character constants. They are transferred untouched to the generated code.
Project>Options>C/C++ Compiler>Language 2>Enable multibyte support

252 for AVR
AFE1_AFE2-1:1
Compiler options
--enable_restrict
Syntax --enable_restrict
Description Enables the Standard C keyword restrict. This option can be useful for improving
analysis precision during optimization.
To set this option, use Project>Options>C/C++ Compiler>Extra options
--enhanced_core
Syntax --enhanced_core
Description Use this option to allow the compiler to generate instructions from the enhanced
instruction set that is available in some AVR devices, for example ATmega161.
The enhanced instruction set consists of these instructions:
MUL
MOVW
MULS
MULSU
FMUL
FMULS
FMULSU
LPM Rd,Z
LPM Rd,Z+
ELPM Rd,Z
ELPM Rd,Z+
SPM
Project>Options>General Options>Target>Enhanced core
--error_limit
Syntax --error_limit=n
Parameters
n The number of errors before the compiler stops the
compilation. n must be a positive integer; 0 indicates no
limit.
253
AFE1_AFE2-1:1
Description Use the --error_limit option to specify the number of errors allowed before the
compiler stops the compilation. By default, 100 errors are allowed.
-f
Syntax -f filename
Description Use this option to make the compiler read command line options from the named file,
with the default filename extension xcl.
In the command file, you format the items exactly as if they were on the command line
itself, except that you may use multiple lines, because the newline character acts just as
a space or tab character.
Both C and C++ style comments are allowed in the file. Double quotes behave in the
same way as in the Microsoft Windows command line environment.
--force_switch_type
Syntax --force_switch_type={0|1|2}
Parameters
0 Library call with switch table
1 Inline code with switch table
2 Inline compare/jump logic
Description Use this option to set the switch type.


254 for AVR
AFE1_AFE2-1:1
Compiler options
--guard_calls
Syntax --guard_calls
Description Use this option to enable guards for function static variable initialization. This option
should be used in a threaded C++ environment.
See also Managing a multithreaded environment, page 135.

--header_context
Syntax --header_context
Description Occasionally, to find the cause of a problem it is necessary to know which header file
that was included from which source line. Use this option to list, for each diagnostic
message, not only the source position of the problem, but also the entire include stack at
that point.
-I
Syntax -I path
Parameters
path The search path for #include files
Description Use this option to specify the search paths for #include files. This option can be used
more than once on the command line.
See also Include file search procedure, page 228.

Project>Options>C/C++ Compiler>Preprocessor>Additional include directories
255
AFE1_AFE2-1:1
--initializers_in_flash
Syntax --initializers_in_flash
Description Use this option to place aggregate initializers in flash memory. These initializers are
otherwise placed in the external segments _C or in the flash segments _F if the compiler
option -y was also specified.
An aggregate initializer—an array or a struct—is constant data that is copied to the stack
dynamically at runtime, in this case every time a function is entered.
The drawback of placing data in flash memory is that it takes more time to copy it; the
advantage is that it does not occupy memory in the data space.
Local variables with aggregate initializers are copied from the segments:
Data Default -y --initializers_in_flash
auto aggregates NEAR_C NEAR_F NEAR_F
Table 29: Accessing variables with aggregate initializers
Example void Func ()

{
char buf[4] = { '1', 'd', 'g', 't' };
...
}
See also -y, page 279 and Initialization of local aggregates at function invocation, page 77.
-l
Syntax -l[a|A|b|B|c|C|D][N][H] {filename|directory}
Parameters
a (default) Assembler list file
A Assembler list file with C or C++ source as comments
b Basic assembler list file. This file has the same contents as a
list file produced with -la, except that no extra
compiler-generated information (runtime model attributes,
call frame information, frame size information) is included *

256 for AVR
AFE1_AFE2-1:1
Compiler options
B Basic assembler list file. This file has the same contents as a
list file produced with -lA, except that no extra compiler
generated information (runtime model attributes, call frame
information, frame size information) is included *
c C or C++ list file
C (default) C or C++ list file with assembler source as comments
D C or C++ list file with assembler source as comments, but
without instruction offsets and hexadecimal byte values
N No diagnostics in file
H Include source lines from header files in output. Without this
option, only source lines from the primary source file are
included
* This makes the list file less useful as input to the assembler, but more useful for reading
by a human.
Description Use this option to generate an assembler or C/C++ listing to a file. Note that this option
can be used one or more times on the command line.
Project>Options>C/C++ Compiler>List
--library_module
Syntax --library_module
Description Use this option to make the compiler generate a library module rather than a program
module. A program module is always included during linking. A library module will
only be included if it is referenced in your program.
Project>Options>C/C++ Compiler>Output>Module type>Library Module
257
AFE1_AFE2-1:1
--lock_regs
Syntax --lock_regs N
Parameters
N The number of registers to lock, 0–12.
Description Use this option to lock registers that are to be used for global register variables. When
you use this option, the registers R15 and downwards will be locked.
To maintain module consistency, make sure to lock the same number of registers in all
modules.
Note: Locking more than nine registers might cause linking to fail. Because the
pre-built libraries delivered with the product do not lock any registers, a library function
might potentially use any of the lockable registers. Any such resource conflict between
locked registers and compiler-used registers will be reported at link time.
If you need to lock any registers in your code, the library must therefore be rebuilt with
the same set of locked registers.
--macro_positions_in_diagnostics
Syntax --macro_positions_in_diagnostics
Description Use this option to obtain position references inside macros in diagnostic messages. This
is useful for detecting incorrect source code constructs in macros.
--memory_model, -m
Syntax --memory_model={tiny|t|small|s|large|l|huge|h}
-m{tiny|t|small|s|large|l<huge|h}
Parameters
tiny, t Specifies the Tiny memory model
small, s Specifies the Small memory model

258 for AVR
AFE1_AFE2-1:1
Compiler options
large, l Specifies the Large memory model

huge, h Specifies the Huge memory model
Description Use this option to specify the memory model, for which the code is to be generated.
By default, the compiler generates code for the Tiny memory model for all processor
options except -v4 and -v6 where the Small memory model is the default.
Use the -m or the --memory_model option if you want to generate code for a different
memory model.
Example To generate code for the Large memory model, give the command:
iccavr filename -ml
or:
iccavr filename --memory_model=large
See also Memory models, page 59.

Project>Options>General Options>Memory model
--mfc
Syntax --mfc
Description Use this option to enable multi-file compilation. This means that the compiler compiles
one or several source files specified on the command line as one unit, which enhances
interprocedural optimizations.
Note: The compiler will generate one object file per input source code file, where the
first object file contains all relevant data and the other ones are empty. If you want only
the first file to be produced, use the -o compiler option and specify a certain output file.
Example iccavr myfile1.c myfile2.c myfile3.c --mfc
See also --discard_unused_publics, page 248, --output, -o, page 267, and Multi-file compilation
units, page 214.
Project>Options>C/C++ Compiler>Multi-file compilation
259
AFE1_AFE2-1:1
--module_name
Syntax --module_name=name
Parameters
name An explicit object module name
Description Normally, the internal name of the object module is the name of the source file, without
a directory name or extension. Use this option to specify an object module name
explicitly.
This option is useful when several modules have the same filename, because the
resulting duplicate module name would normally cause a linker error; for example,
when the source file is a temporary file generated by a preprocessor.
Project>Options>C/C++ Compiler>Output>Object module name
--no_call_frame_info
Syntax --no_call_frame_info
Description Normally, the compiler always generates call frame information in the output, to enable
the debugger to display the call stack even in code from modules with no debug
information. Use this option to disable the generation of call frame information.
See also Call frame information, page 165.

--no_clustering
Syntax --no_clustering
Description Use this option to disable static clustering optimizations.

See also Static clustering, page 218.

Project>Options>C/C++ Compiler>Optimizations>Enable
transformations>Clustering of variables

260 for AVR
AFE1_AFE2-1:1
Compiler options
--no_code_motion
Syntax --no_code_motion
Description Use this option to disable code motion optimizations.

See also Code motion, page 217.

transformations>Code motion
--no_cross_call
Syntax --no_cross_call
Description Use this option to disable the cross-call optimization.

Use this option to disable the cross-call optimization. This is highly recommended if
your target processor has a hardware stack or a small internal return stack segment,
RSTACK, because this option reduces the usage of RSTACK.
This optimization is performed at size optimization, level High. Note that, although the
optimization can drastically reduce the code size, this optimization increases the
execution time.
See also --cross_call_passes, page 242, Cross call, page 218.

transformations>Cross call
--no_cse
Syntax --no_cse
Description Use this option to disable common subexpression elimination.

See also Common subexpression elimination, page 216.

transformations>Common subexpression elimination
261
AFE1_AFE2-1:1
--no_inline
Syntax --no_inline
Description Use this option to disable function inlining.
See also Inlining functions, page 67.

transformations>Function inlining
--no_path_in_file_macros
Syntax --no_path_in_file_macros
Description Use this option to exclude the path from the return value of the predefined preprocessor
symbols __FILE__ and __BASE_FILE__.
See also Description of predefined preprocessor symbols, page 352.

--no_rampd
Syntax --no_rampd
Description Use this option to make the compiler use the RAMPZ register instead of RAMPD. This
option corresponds to the instructions LDS and STS.
Note that this option is only useful on processor variants with more than 64 Kbytes data
(-v4 and -v6).
--no_size_constraints
Syntax --no_size_constraints
Description Use this option to relax the normal restrictions for code size expansion when optimizing
for high speed.

262 for AVR
AFE1_AFE2-1:1
Compiler options
Note: This option has no effect unless used with -Ohs.
See also Speed versus size, page 215.

--no_static_destruction
Syntax --no_static_destruction
Description Normally, the compiler emits code to destroy C++ static variables that require
destruction at program exit. Sometimes, such destruction is not needed.
Use this option to suppress the emission of such code.
See also System termination, page 121.

--no_system_include
Syntax --no_system_include
Description By default, the compiler automatically locates the system include files. Use this option
to disable the automatic search for system include files. In this case, you might need to
set up the search path by using the -I compiler option.
See also --dlib, page 248, --dlib_config, page 249, and --system_include_dir, page 274.
Project>Options>C/C++ Compiler>Preprocessor>Ignore standard include
directories
--no_tbaa
Syntax --no_tbaa
Description Use this option to disable type-based alias analysis.

Note: This option has no effect at optimization levels below High.
See also Type-based alias analysis, page 217.
263
AFE1_AFE2-1:1
transformations>Type-based alias analysis
--no_typedefs_in_diagnostics
Syntax --no_typedefs_in_diagnostics
Description Use this option to disable the use of typedef names in diagnostics. Normally, when a
type is mentioned in a message from the compiler, most commonly in a diagnostic
message of some kind, the typedef names that were used in the original declaration are
used whenever they make the resulting text shorter.
Example typedef int (*MyPtr)(char const *);

MyPtr p = "My text string";
will give an error message like this:

Error[Pe144]: a value of type "char *" cannot be used to
initialize an entity of type "MyPtr"
If the --no_typedefs_in_diagnostics option is used, the error message will be like

this:
Error[Pe144]: a value of type "char *" cannot be used to
initialize an entity of type "int (*)(char const *)"
--no_ubrof_messages
Syntax --no_ubrof_messages
Description Use this option to minimize the size of your application object file by excluding
messages from the UBROF files. The file size can decrease by up to 60%. Note that the
XLINK diagnostic messages will, however, be less useful when you use this option.

264 for AVR
AFE1_AFE2-1:1
Compiler options
--no_unroll
Syntax --no_unroll
Description Use this option to disable loop unrolling.

Note: This option has no effect at optimization levels below High.
--no_warnings
Syntax --no_warnings
Description By default, the compiler issues warning messages. Use this option to disable all warning
messages.
--no_wrap_diagnostics
Syntax --no_wrap_diagnostics
Description By default, long lines in diagnostic messages are broken into several lines to make the
message easier to read. Use this option to disable line wrapping of diagnostic messages.
-O
Syntax -O[n|l|m|h|hs|hz]
Parameters
n None* (Best debug support)
l (default) Low*
m Medium
h High, balanced
hs High, favoring speed
265
AFE1_AFE2-1:1
n None* (Best debug support)
hz High, favoring size
*The most important difference between None and Low is that at None, all non-static
variables will live during their entire scope.
Description Use this option to set the optimization level to be used by the compiler when optimizing
the code. If no optimization option is specified, the optimization level is used by default.
If only -O is used without any parameter, the optimization level High balanced is used.
A low level of optimization makes it relatively easy to follow the program flow in the
debugger, and, conversely, a high level of optimization makes it relatively hard.
See also Controlling compiler optimizations, page 214.

--omit_types
Syntax --omit_types
Description By default, the compiler includes type information about variables and functions in the
object output. Use this option if you do not want the compiler to include this type
information in the output, which is useful when you build a library that should not
contain type information. The object file will then only contain type information that is
a part of a symbol’s name. This means that the linker cannot check symbol references
for type correctness.
--only_stdout
Syntax --only_stdout
Description Use this option to make the compiler use the standard output stream (stdout) also for
messages that are normally directed to the error output stream (stderr).

266 for AVR
AFE1_AFE2-1:1
Compiler options
--output, -o
Syntax --output {filename|directory}
-o {filename|directory}
Description By default, the object code output produced by the compiler is located in a file with the
same name as the source file, but with the extension r90. Use this option to explicitly
specify a different output filename for the object code output.
--pending_instantiations
Syntax --pending_instantiations number
Parameters
number An integer that specifies the limit, where 64 is default. If 0
is used, there is no limit.
Description Use this option to specify the maximum number of instantiations of a given C++
template that is allowed to be in process of being instantiated at a given time. This is
used for detecting recursive instantiations.
Project>Options>C/C++ Compiler>Extra Options
--predef_macros
Syntax --predef_macros {filename|directory}
Description Use this option to list the predefined symbols. When using this option, make sure to also
use the same options as for the rest of your project.
If a filename is specified, the compiler stores the output in that file. If a directory is
specified, the compiler stores the output in that directory, in a file with the predef
filename extension.
Note that this option requires that you specify a source file on the command line.
267
AFE1_AFE2-1:1
--preinclude
Syntax --preinclude includefile
Description Use this option to make the compiler read the specified include file before it starts to
read the source file. This is useful if you want to change something in the source code
for the entire application, for instance if you want to define a new symbol.
Project>Options>C/C++ Compiler>Preprocessor>Preinclude file
--preprocess
Syntax --preprocess[=[c][n][l]] {filename|directory}
Parameters
c Preserve comments
n Preprocess only
l Generate #line directives
Description Use this option to generate preprocessed output to a named file.

Project>Options>C/C++ Compiler>Preprocessor>Preprocessor output to file
--public_equ
Syntax --public_equ symbol[=value]
Parameters
symbol The name of the assembler symbol to be defined
value An optional value of the defined assembler symbol

268 for AVR
AFE1_AFE2-1:1
Compiler options
Description This option is equivalent to defining a label in assembler language using the EQU
directive and exporting it using the PUBLIC directive. This option can be used more than
once on the command line.
--relaxed_fp
Syntax --relaxed_fp
Description Use this option to allow the compiler to relax the language rules and perform more
aggressive optimization of floating-point expressions. This option improves
performance for floating-point expressions that fulfill these conditions:
● The expression consists of both single- and double-precision values
● The double-precision values can be converted to single precision without loss of
accuracy
● The result of the expression is converted to single precision.
Note that performing the calculation in single precision instead of double precision
might cause a loss of accuracy.
Example float F(float a, float b)

{
return a + b * 3.0;
}
The C standard states that 3.0 in this example has the type double and therefore the
whole expression should be evaluated in double precision. However, when the
--relaxed_fp option is used, 3.0 will be converted to float and the whole expression
can be evaluated in float precision.
Project>Options>C/C++ Compiler>Language 2>Floating-point semantics
--remarks
Syntax --remarks
Description The least severe diagnostic messages are called remarks. A remark indicates a source
code construct that may cause strange behavior in the generated code. By default, the
269
AFE1_AFE2-1:1
compiler does not generate remarks. Use this option to make the compiler generate
remarks.
See also Severity levels, page 231.

Project>Options>C/C++ Compiler>Diagnostics>Enable remarks
--require_prototypes
Syntax --require_prototypes
Description Use this option to force the compiler to verify that all functions have proper prototypes.
Using this option means that code containing any of the following will generate an error:
● A function call of a function with no declaration, or with a Kernighan & Ritchie
C declaration
● A function definition of a public function with no previous prototype declaration
● An indirect function call through a function pointer with a type that does not include
a prototype.
Project>Options>C/C++ Compiler>Language 1>Require prototypes
--root_variables
Syntax --root_variables
Description Use this option to apply the __root extended keyword to all global and static variables.
This will make sure that the variables are not removed by the IAR XLINK Linker.
Note: The --root_variables option is always available, even if you do not specify
the compiler option -e, language extensions.
Project>Options>C/C++ Compiler>Code>Force generation of all global and static
variables

270 for AVR
AFE1_AFE2-1:1
Compiler options
-s
Syntax -s[2|3|6|9]
Parameters
2 None* (Best debug support)
3 (default) Low*
6 Medium
9 High (Maximum optimization). This corresponds to the -Oh option.
*The most important difference between -s2 and -s3 is that at level 2, all non-static
Description Use this option to make the compiler optimize the code for maximum execution speed.
If no optimization option is specified, the optimization level 3 is used by default.
Note: The -s and -z options cannot be used at the same time.
Note: This option is available for backward compatibility. Use -O instead.
See also -O, page 265 and -z, page 279

Project>Options>C/C++ Compiler>Optimizations>Speed
--segment
Syntax --segment __memory_attribute=NEWSEGMENTNAME
Description The compiler places functions and data objects into named segments which are referred
to by the IAR XLINK Linker. Use the --segment option to perform one of these
operations:
● To place all functions or data objects declared with the __memory_attribute in
segments with names that begin with NEWSEGMENTNAME.
● To change the name of the segments SWITCH, INITTAB, CSTACK, RSTACK, and
INTVEC to a different name.
This is useful if you want to place your code or data in different address ranges and you
find the @ notation, alternatively the #pragma location directive, insufficient. Note
271
AFE1_AFE2-1:1
that any changes to the segment names require corresponding modifications in the linker
configuration file.
Example This command places the __near int a; defined variable in the MYDATA_Z segment:
--segment __near=MYDATA
This command names the segment that contains interrupt vectors to MYINTS instead of
the default name INTVEC:
--segment intvec=MYINTS
See also Controlling data and function placement in memory, page 210 and Summary of extended
keywords, page 299.
--separate_cluster_for_initialized_variables
Syntax --separate_cluster_for_initialized_variables
Description Use this option to separate initialized and non-initialized variables when using variable
clustering. The option makes the *_ID segments smaller but can generate larger code.
--silent
Syntax --silent
Description By default, the compiler issues introductory messages and a final statistics report. Use
this option to make the compiler operate without sending these messages to the standard
output stream (normally the screen).
This option does not affect the display of error and warning messages.

272 for AVR
AFE1_AFE2-1:1
Compiler options
--spmcr_address
Syntax --spmcr_address address
Parameters
address The SPMCR address, where 0x37 is the default.
Description Use this option to set the SPMCR address.

--strict
Syntax --strict
Description By default, the compiler accepts a relaxed superset of Standard C and C++. Use this
option to ensure that the source code of your application instead conforms to strict
Standard C and C++.
Note: The -e option and the --strict option cannot be used at the same time.
See also Enabling language extensions, page 169.

Project>Options>C/C++ Compiler>Language 1>Language conformance>Strict
--string_literals_in_flash
Syntax --string_literals_in_flash
Description Use this option to put “string” in the __nearflash or __farflash segment depending
on the processor option.
When this option is used, library functions taking a string literal as a parameter will no
longer be type-compatible. Use the *_P library function variants (for example
printf_P).
See also AVR-specific library functions, page 370.

273
AFE1_AFE2-1:1
--system_include_dir
Syntax --system_include_dir path
Parameters
path The path to the system include files, see Rules for specifying
a filename or directory as parameters, page 234.
Description By default, the compiler automatically locates the system include files. Use this option
to explicitly specify a different path to the system include files. This might be useful if
you have not installed IAR Embedded Workbench in the default location.
See also --dlib_config, page 249, and --no_system_include, page 263.

--use_c++_inline
Syntax --use_c++_inline
Description Standard C uses slightly different semantics for the inline keyword than C++ does.
Use this option if you want C++ semantics when you are using C.
See also Inlining functions, page 67

Project>Options>C/C++ Compiler>Language 1>C dialect>C99>C++ inline
semantics

274 for AVR
AFE1_AFE2-1:1
Compiler options
-v
Syntax -v{0|1|2|3|4|5|6}
Parameters
0 (default) The code space is physically limited to 8 Kbytes, and the RCALL/RJMP
instructions are used for reaching the code space. Interrupt vectors are 2
bytes long. The compiler assumes that the index registers X, Y, and Z are
eight bits wide when accessing the built-in SRAM. It also assumes that
it is not possible to attach any external memory to the microcontroller
and that it therefore should not generate any constant segment in data
space (_C segment). Instead the compiler adds an implicit -y command
line option. It will also try to place all aggregate initializers in flash
memory, that is, the implicit --initializers_in_flash option is
also added to the command line. Relative function calls are made.
1 The code space is physically limited to 8 Kbytes, and the RCALL/RJMP
instructions are used for reaching the code space. Interrupt vectors are 2
bytes long. The compiler assumes that it is possible to have external
memory and will therefore generate _C segments. Relative function calls
are made.
2 The code space is physically limited to 128 Kbytes, and, if possible, the
RCALL/RJMP instructions are used for reaching the code space. If that is
not possible, CALL/JMP is used. Function calls through function pointers
use ICALL/IJMP. 2 bytes are used for all function pointers. Interrupt
vectors are 4 bytes long. The compiler assumes that the index registers
X, Y, and Z are eight bits wide when accessing the built-in SRAM. It also
assumes that it is not possible to attach any external memory to the
microcontroller and that it should therefore not generate any constant
segment in data space (_C segment). Instead the compiler adds an
implicit -y command line option. It will also try to place all aggregate
initializers in flash memory, that is, the implicit
--initializers_in_flash option is also added to the command
line.
use ICALL/IJMP. Two bytes are used for all function pointers. Interrupt
vectors are four bytes long. The compiler assumes that it is possible to
have external memory and will therefore generate _C segments.
275
AFE1_AFE2-1:1
use ICALL/IJMP. Two bytes are used for all function pointers. Interrupt
vectors are four bytes long. The compiler assumes that it is possible to
have external memory and will therefore generate _C segments.
5 Allows function calls through farfunc function pointers to cover the
entire 8 Mbyte code space by using EICALL/EIJMP. Three bytes are
used for function pointers. Interrupt vectors are four bytes long. The
compiler assumes that it is possible to have external memory and will
therefore generate _C segments.
6 Allows function calls through farfunc function pointers to cover the
entire 8 Mbyte code space by using EICALL/EIJMP. Three bytes are
used for function pointers. Interrupt vectors are four bytes long.The
compiler assumes that it is possible to have external memory and will
therefore generate _C segments.
Description Use this option to select the processor device for which the code is to be generated.
See also --cpu, page 242 and Processor configuration, page 49.
Project>Options>General Options>Processor configuration
--version
Syntax --version
Description Use this option to make the compiler send version information to the console and then
exit.
--version1_calls
Syntax --version1_calls
Description Use this option to make all functions and function calls use the calling convention of the
A90 IAR Compiler, ICCA90, which is described in Calling convention, page 155.

276 for AVR
AFE1_AFE2-1:1
Compiler options
To change the calling convention of a single function, use the __version_1 function
type attribute. See __version_1, page 315 for detailed information.
Project>Options>C/C++ Compiler>Code>Use ICCA90 1.x calling convention
--version2_calls
Syntax --version2_calls
Description Use this option to select the V2.10-V4.10B calling convention.
See also Calling convention, page 155.

--vla
Syntax --vla
Description Use this option to enable support for C99 variable length arrays. Such arrays are located
on the heap. This option requires Standard C and cannot be used together with the
--c89 compiler option.
Note: --vla should not be used together with the longjmp library function, as that can
lead to memory leakages.
See also C language overview, page 167.

Project>Options>C/C++ Compiler>Language 1>C dialect>Allow VLA
--warn_about_c_style_casts
Syntax --warn_about_c_style_casts
Description Use this option to make the compiler warn when C-style casts are used in C++ source
code.
277
AFE1_AFE2-1:1
--warnings_affect_exit_code
Syntax --warnings_affect_exit_code
Description By default, the exit code is not affected by warnings, because only errors produce a
non-zero exit code. With this option, warnings will also generate a non-zero exit code.
--warnings_are_errors
Syntax --warnings_are_errors
Description Use this option to make the compiler treat all warnings as errors. If the compiler
encounters an error, no object code is generated. Warnings that have been changed into
remarks are not treated as errors.
Note: Any diagnostic messages that have been reclassified as warnings by the option
--diag_warning or the #pragma diag_warning directive will also be treated as
errors when --warnings_are_errors is used.
See also --diag_warning, page 246.

Project>Options>C/C++ Compiler>Diagnostics>Treat all warnings as errors
--xmcra_address
Syntax --xmcra_address address
Parameters
address The address of XMCRA. The default is 0x74.
Description Use this option to specify where XMCRA is located.


278 for AVR
AFE1_AFE2-1:1
Compiler options
-y
Syntax -y
Description Use this option to override the default placement of constants and literals.
Without this option, constants and literals are placed in an external const segment,
segmentbasename_C. With this option, constants and literals will instead be placed in
the initialized segmentbasename_I data segments that are copied from the
segmentbasename_ID segments by the system startup code.
Note that -y is implicit in the tiny memory model.

This option can be combined with the option --initializers_in_flash.
See also --initializers_in_flash, page 256

-z
Syntax -z[2|3|6|9]
Parameters
2 None* (Best debug support)
3 (default) Low*
6 Medium
9 High (Maximum optimization). This corresponds to the -Ohz

option.
*The most important difference between -z2 and -z3 is that at level 2, all non-static
Description Use this option to make the compiler optimize the code for minimum size. If no
optimization option is specified, the optimization level 3 is used by default.
Note: The -s and -z options cannot be used at the same time.
Note: This option is available for backward compatibility. Use -O instead.
279
AFE1_AFE2-1:1
See also -O, page 265 and -s, page 271

Project>Options>C/C++ Compiler>Optimizations>Size
--zero_register
Syntax --zero_register
Description Use this option to make the compiler use register R15 as zero register, which means that
register R15 is assumed to always contain zero.
This option can in some cases reduce the size of the generated code, especially in the
Large memory model. The option might be incompatible with the supplied runtime
libraries. The linker will issue a link-time error if any incompatibilities arise.

280 for AVR
AFE1_AFE2-1:1
Data representation
● Alignment
● Byte order
● Basic data types—integer types
● Basic data types—floating-point types
● Pointer types
● Structure types
● Type qualifiers
● Data types in C++
See the chapter Efficient coding for embedded applications for information about
which data types and pointers provide the most efficient code for your
application.
Alignment
Every C data object has an alignment that controls how the object can be stored in
memory. Should an object have an alignment of, for example, 4, it must be stored on an
address that is divisible by 4.
The reason for the concept of alignment is that some processors have hardware
limitations for how the memory can be accessed.
Assume that a processor can read 4 bytes of memory using one instruction, but only
when the memory read is placed on an address divisible by 4. Then, 4-byte objects, such
as long integers, will have alignment 4.
Another processor might only be able to read 2 bytes at a time; in that environment, the
alignment for a 4-byte long integer might be 2.
All data types must have a size that is a multiple of their alignment. Otherwise, only the
first element of an array would be guaranteed to be placed in accordance with the
alignment requirements. This means that the compiler might add pad bytes at the end of
the structure.
281
AFE1_AFE2-1:1
Byte order
Note that with the #pragma data_alignment directive you can increase the
alignment demands on specific variables.
ALIGNMENT ON THE AVR MICROCONTROLLER

The AVR microcontroller does not have any alignment restrictions.
Byte order
The AVR microcontroller stores data in little-endian byte order.
In the little-endian byte order, the least significant byte is stored at the lowest address in
memory. The most significant byte is stored at the highest address.
In the big-endian byte order, the most significant byte is stored at the lowest address in
memory. The least significant byte is stored at the highest address. If you use the
big-endian byte order, it might be necessary to use the
#pragma bitfields=reversed directive to be compatible with code for other
compilers and I/O register definitions of some devices, see Bitfields, page 284.
Basic data types—integer types

The compiler supports both all Standard C basic data types and some additional types.
INTEGER TYPES—AN OVERVIEW

This table gives the size and range of each integer data type:
Data type Size Range Alignment
bool 8 bits 0 to 1 1
char 8 bits 0 to 255 1
signed char 8 bits -128 to 127 1
unsigned char 8 bits 0 to 255 1
signed short 16 bits -32768 to 32767 1
unsigned short 16 bits 0 to 65535 1
signed int 16 bits -32768 to 32767 1
unsigned int 16 bits 0 to 65535 1
signed long 32 bits -231 to 231-1 1
32
unsigned long 32 bits 0 to 2 -1 1
signed long long 64 bits -263 to 263-1 1
Table 30: Integer types

282 for AVR
AFE1_AFE2-1:1
Data representation
Data type Size Range Alignment

unsigned long long 64 bits 0 to 264-1 1
Table 30: Integer types (Continued)
Signed variables are represented using the two’s complement form.
BOOL
The bool data type is supported by default in the C++ language. If you have enabled
language extensions, the bool type can also be used in C source code if you include the
file stdbool.h. This will also enable the boolean values false and true.
THE LONG LONG TYPE

The long long data type is supported with one restriction:
The CLIB runtime library does not support the long long type.
THE ENUM TYPE

The compiler will use the smallest type required to hold enum constants, preferring
signed rather than unsigned.
When IAR Systems language extensions are enabled, and in C++, the enum constants
and types can also be of the type long, unsigned long, long long, or unsigned
long long .
To make the compiler use a larger type than it would automatically use, define an enum
constant with a large enough value. For example:
/* Disables usage of the char type for enum */
enum Cards{Spade1, Spade2,
DontUseChar=257};
THE CHAR TYPE

The char type is by default unsigned in the compiler, but the --char_is_signed
compiler option allows you to make it signed. Note, however, that the library is compiled
with the char type as unsigned.
THE WCHAR_T TYPE

The wchar_t data type is an integer type whose range of values can represent distinct
codes for all members of the largest extended character set specified among the
supported locals.
283
AFE1_AFE2-1:1
Basic data types—integer types
The wchar_t data type is supported by default in the C++ language. To use the
wchar_t type also in C source code, you must include the file stddef.h from the
runtime library.
Note: The IAR CLIB Library has only rudimentary support for wchar_t.
BITFIELDS
In Standard C, int, signed int, and unsigned int can be used as the base type for
integer bitfields. In standard C++, and in C when language extensions are enabled in the
compiler, any integer or enumeration type can be used as the base type. It is
implementation defined whether a plain integer type (char, short, int, etc) results in
a signed or unsigned bitfield.
In the IAR C/C++ Compiler for AVR, plain integer types are treated as signed.
Bitfields in expressions are treated as int if int can represent all values of the bitfield.
Otherwise, they are treated as the bitfield base type.
Each bitfield is placed into a container of its base type from the least significant bit to
the most significant bit. If the last container is of the same type and has enough bits
available, the bitfield is placed into this container, otherwise a new container is
allocated. This allocation scheme is referred to as the disjoint type bitfield allocation.
If you use the directive #pragma bitfields=reversed, bitfields are placed from the
most significant bit to the least significant bit in each container. See bitfields, page 321.
Assume this example:
struct BitfieldExample
{
uint32_t a : 12;
uint16_t b : 3;
uint16_t c : 7;
uint8_t d;
};
The example in the disjoint types bitfield allocation strategy

To place the first bitfield, a, the compiler allocates a 32-bit container at offset 0 and puts
a into the least significant 12 bits of the container.
To place the second bitfield, b, a new container is allocated at offset 4, because the type
of the bitfield is not the same as that of the previous one. b is placed into the least
significant three bits of this container.
The third bitfield, c, has the same type as b and fits into the same container.

284 for AVR
AFE1_AFE2-1:1
Data representation
The fourth member, d, is allocated into the byte at offset 6. d cannot be placed into the
same container as b and c because it is not a bitfield, it is not of the same type, and it
would not fit.
When using reverse order, each bitfield is instead placed starting from the most
significant bit of its container.
This is the layout of bitfield_example:
Basic data types—floating-point types

In the IAR C/C++ Compiler for AVR, floating-point values are represented in standard
IEEE 754 format. The sizes for the different floating-point types are:
Type Size if double=32 Size if double=64
float 32 bits 32 bits
double 32 bits (default) 64 bits
long double 32 bits 64 bits
Table 31: Floating-point types
Note: The size of double and long double depends on the --64bit_doubles
option, see --64bit_doubles, page 240. The type long double uses the same precision
as double.
285
AFE1_AFE2-1:1
Basic data types—floating-point types
FLOATING-POINT ENVIRONMENT
Exception flags are not supported. The feraiseexcept function does not raise any
exceptions.
32-BIT FLOATING-POINT FORMAT

The representation of a 32-bit floating-point number as an integer is:
The exponent is 8 bits, and the mantissa is 23 bits.

The value of the number is:
(-1)S * 2(Exponent-127) * 1.Mantissa
The range of the number is at least:

±1.18E-38 to ±3.39E+38
The precision of the float operators (+, -, *, and /) is approximately 7 decimal digits.
64-BIT FLOATING-POINT FORMAT

The representation of a 64-bit floating-point number as an integer is:
The exponent is 11 bits, and the mantissa is 52 bits.

The value of the number is:
(-1)S * 2(Exponent-1023) * 1.Mantissa
The range of the number is at least:

±2.23E-308 to ±1.79E+308
The precision of the float operators (+, -, *, and /) is approximately 15 decimal digits.
REPRESENTATION OF SPECIAL FLOATING-POINT NUMBERS

This list describes the representation of special floating-point numbers:
● Zero is represented by zero mantissa and exponent. The sign bit signifies positive or
negative zero.
● Infinity is represented by setting the exponent to the highest value and the mantissa
to zero. The sign bit signifies positive or negative infinity.

286 for AVR
AFE1_AFE2-1:1
Data representation
● Not a number (NaN) is represented by setting the exponent to the highest positive
value and the mantissa to a non-zero value. The value of the sign bit is ignored.
● Subnormal numbers are used for representing values smaller than what can be
represented by normal values. The drawback is that the precision will decrease with
smaller values. The exponent is set to 0 to signify that the number is subnormal,
even though the number is treated as if the exponent was 1. Unlike normal numbers,
subnormal numbers do not have an implicit 1 as the most significant bit (the MSB)
of the mantissa. The value of a subnormal number is:
(-1)S * 2(1-BIAS) * 0.Mantissa
where BIAS is 127 and 1023 for 32-bit and 64-bit floating-point values, respectively.
Note: The IAR CLIB Library does not fully support the special cases of floating-point
numbers, such as infinity, NaN, and subnormal numbers. A library function which gets
one of these special cases of floating-point numbers as an argument might behave
unexpectedly.
Pointer types
The compiler has two basic types of pointers: function pointers and data pointers.
FUNCTION POINTERS
The size of function pointers is always 16 or 24 bits, and they can address the entire
memory. The internal representation of a function pointer is the actual address it refers
to divided by two.
These function pointers are available:
Keyword Address range Pointer size Description
__nearfunc 0–0x1FFFE 16 bits Can be called from any part of the
code memory, but must reside in the
first 128 Kbytes of that space.
__farfunc 0–0x7FFFFE 24 bits No restrictions on code placement.
Table 32: Function pointers
DATA POINTERS
Data pointers have three sizes: 8, 16, or 24 bits. These data pointers are available:
Pointer
Keyword Memory space Index type Address range
size
__tiny 8 bits Data signed char 0x–0xFF
Table 33: Data pointers
287
AFE1_AFE2-1:1
Pointer types
Pointer
Keyword Memory space Index type Address range
size
__near 16 bits Data signed int 0x0–0xFFFF
__far 24 bits Data signed int 0x0–0xFFFFFF
(16-bit arithmetics)
__huge 24 bits Data signed long 0x0–0xFFFFFF
__tinyflash 8 bits Code signed char 0x0–0xFF
__flash 16 bits Code signed int 0x0–0xFFFF
__farflash 24 bits Code signed int 0x0–0xFFFFFF
(16-bit arithmetic)
__hugeflash 24 bits Code signed long 0x0–0xFFFFFF
__eeprom 8 bits EEPROM signed char 0x0–0xFF
__eeprom 16 bits EEPROM signed int 0x0–0xFFFF
__generic 16 bits Data/Code signed int The most significant bit
24 bits signed long (MSB) determines
whether __generic
points to CODE (1) or
DATA (0). The small
generic pointer is
generated for the
processor options -v0
and -v1.
Table 33: Data pointers (Continued)
Segmented data pointer comparison

Note that the result of using relational operators (<, <=, >, >=) on data pointers is only
defined if the pointers point into the same object. For segmented data pointers, only the
offset part of the pointer is compared when using these operators. For example:
void MyFunc(char __far * p, char __far * q)
{
if (p == q) /* Compares the entire pointers. */
...
if (p < q) /* Compares only the low 16 bits of the pointers. */
...
}

288 for AVR
AFE1_AFE2-1:1
Data representation
CASTING
Casts between pointers have these characteristics:
● Casting a value of an integer type to a pointer of a smaller type is performed by
truncation
● Casting a value of an integer type to a pointer of a larger type is performed by zero
extension
● Casting a pointer type to a smaller integer type is performed by truncation
● Casting a pointer type to a larger integer type is performed via casting to the largest
possible pointer that fits in the integer
● Casting a data pointer to a function pointer and vice versa is illegal
● Casting a function pointer to an integer type gives an undefined result
● Casting from a smaller pointer to a larger pointer is performed by zero extension
● Casting from a larger pointer to a smaller pointer gives an undefined result.
size_t
size_t is the unsigned integer type of the result of the sizeof operator. In the IAR
C/C++ Compiler for AVR, the type used for size_t depends on the processor option
used:
Generic processor option Typedef

-v0 and -v1 unsigned int
-v2, -v3, -v4, -v5, and -v6 unsigned long
Table 34: size_t typedef
When using the Large or Huge memory model, the typedef for size_t is unsigned
long int.
Note that some data memory types might be able to accommodate larger, or only
smaller, objects than the memory pointed to by default pointers. In this case, the type of
the result of the sizeof operator could be a larger or smaller unsigned integer type.
There exists a corresponding size_t typedef for each memory type, named after the
memory type. In other words, __near_size_t for __near memory.
ptrdiff_t
ptrdiff_t is the signed integer type of the result of subtracting two pointers. This table
shows the typedef of ptrdiff_t depending on the processor option:
289
AFE1_AFE2-1:1
Structure types
Generic processor option Typedef

-v0 and -v1 unsigned int
-v2, -v3, -v4, -v5, and -v6 unsigned long
Table 35: ptrdif_t typedef
Note that subtracting pointers other than default pointers could result in a smaller or
larger integer type. In each case, this integer type is the signed integer variant of the
corresponding size_t type.
Note: It is sometimes possible to create an object that is so large that the result of
subtracting two pointers in that object is negative. See this example:
char buff[60000]; /* Assuming ptrdiff_t is a 16-bit */
char *p1 = buff; /* signed integer type. */
char *p2 = buff + 60000;
ptrdiff_t diff = p2 - p1; /* Result: -5536 */
intptr_t
intptr_t is a signed integer type large enough to contain a void *. In the IAR C/C++
Compiler for AVR, the type used for intptr_t is long int when using the Large or
Huge memory model and int in all other cases.
uintptr_t
uintptr_t is equivalent to intptr_t, with the exception that it is unsigned.
Structure types
The members of a struct are stored sequentially in the order in which they are
declared: the first member has the lowest memory address.
ALIGNMENT OF STRUCTURE TYPES

The struct and union types inherit the alignment requirements of their members. In
addition, the size of a struct is adjusted to allow arrays of aligned structure objects.
GENERAL LAYOUT
Members of a struct are always allocated in the order specified in the declaration.
Each member is placed in the struct according to the specified alignment (offsets).

290 for AVR
AFE1_AFE2-1:1
Data representation
struct First
{
char c;
short s;
} s;
This diagram shows the layout in memory:
The alignment of the structure is 1 byte, and the size is 3 bytes.
Type qualifiers
According to the C standard, volatile and const are type qualifiers.
DECLARING OBJECTS VOLATILE

By declaring an object volatile, the compiler is informed that the value of the object
can change beyond the compiler’s control. The compiler must also assume that any
accesses can have side effects—thus all accesses to the volatile object must be
preserved.
There are three main reasons for declaring an object volatile:
● Shared access; the object is shared between several tasks in a multitasking
environment
● Trigger access; as for a memory-mapped SFR where the fact that an access occurs
has an effect
● Modified access; where the contents of the object can change in ways not known to
the compiler.
Definition of access to volatile objects

The C standard defines an abstract machine, which governs the behavior of accesses to
volatile declared objects. In general and in accordance to the abstract machine:
● The compiler considers each read and write access to an object declared volatile
as an access
● The unit for the access is either the entire object or, for accesses to an element in a
composite object—such as an array, struct, class, or union—the element. For
example:
291
AFE1_AFE2-1:1
Type qualifiers
char volatile a;
a = 5; /* A write access */
a += 6; /* First a read then a write access */
● An access to a bitfield is treated as an access to the underlying type
● Adding a const qualifier to a volatile object will make write accesses to the
object impossible. However, the object will be placed in RAM as specified by the C
standard.
However, these rules are not detailed enough to handle the hardware-related
requirements. The rules specific to the IAR C/C++ Compiler for AVR are described
below.
Rules for accesses

In the IAR C/C++ Compiler for AVR, accesses to volatile declared objects are
subject to these rules:
● All accesses are preserved
● All accesses are complete, that is, the whole object is accessed
● All accesses are performed in the same order as given in the abstract machine
● All accesses are atomic, that is, they cannot be interrupted.
The compiler adheres to these rules for all 8-bit types.
These object types are treated in a special way:
Type of object Treated as
Global register variables Treated as non-triggering volatiles
__io declared variables located An assignment to a bitfield always generates a write access,
in 0x–0x1F in some cases also a read access. If only one bit is updated—
set or cleared—the sci/cbi instructions are executed.
Table 36: Type of volatile accesses treated in a special way
For all combinations of object types not listed, only the rule that states that all accesses
are preserved applies.
DECLARING OBJECTS VOLATILE AND CONST

If you declare a volatile object const, it will be write-protected but it will still be
stored in RAM memory as the C standard specifies.

292 for AVR
AFE1_AFE2-1:1
Data representation
To store the object in read-only memory instead, but still make it possible to access it as
a const volatile object, follow this example:
/* Header */
extern int const xVar;
#define x (*(int const volatile *) &xVar)
/* Source that uses x */

int DoSomething()
{
return x;
}
/* Source that defines x */

#pragma segment = "FLASH"
int const xVar @ "FLASH" = 6;
The segment FLASH contains the initializers. They must be flashed manually when the
application starts up.
Thereafter, the initializers can be reflashed with other values at any time.
DECLARING OBJECTS CONST

The const type qualifier is used for indicating that a data object, accessed directly or
via a pointer, is non-writable. A pointer to const declared data can point to both
constant and non-constant objects. It is good programming practice to use const
declared pointers whenever possible because this improves the compiler’s possibilities
to optimize the generated code and reduces the risk of application failure due to
erroneously modified data.
Static and global objects declared const are allocated in ROM.
If no ROM is available in the DATA memory space, use the compiler option -y to control
the placement of constants. For more information, see -y, page 279.
In C++, objects that require runtime initialization cannot be placed in ROM.
Data types in C++

In C++, all plain C data types are represented in the same way as described earlier in this
chapter. However, if any Embedded C++ features are used for a type, no assumptions
can be made concerning the data representation. This means, for example, that it is not
supported to write assembler code that accesses class members.
293
AFE1_AFE2-1:1
Data types in C++

294 for AVR
AFE1_AFE2-1:1
Extended keywords
● General syntax rules for extended keywords
● Summary of extended keywords
● Descriptions of extended keywords
For information about the address ranges of the different memory areas, see
the chapter Segment reference.
General syntax rules for extended keywords

The compiler provides a set of attributes that can be used on functions or data objects to
support specific features of the AVR microcontroller. There are two types of attributes—
type attributes and object attributes:
● Type attributes affect the external functionality of the data object or function
● Object attributes affect the internal functionality of the data object or function.
The syntax for the keywords differs slightly depending on whether it is a type attribute
or an object attribute, and whether it is applied to a data object or a function.
For more information about each attribute, see Descriptions of extended keywords, page
300. For information about how to use attributes to modify data, see the chapter Data
storage. For information about how to use attributes to modify functions, see the chapter
Functions.
Note: The extended keywords are only available when language extensions are enabled
in the compiler.
Use the -e compiler option to enable language extensions. See -e, page 250.
TYPE ATTRIBUTES
Type attributes define how a function is called, or how a data object is accessed. This
means that if you use a type attribute, it must be specified both when a function or data
object is defined and when it is declared.
You can either place the type attributes explicitly in your declarations, or use the pragma
directive #pragma type_attribute.
295
AFE1_AFE2-1:1
Type attributes can be further divided into memory type attributes and general type
attributes. Memory type attributes are referred to as simply memory attributes in the rest
of the documentation.
Memory attributes
A memory attribute corresponds to a certain logical or physical memory in the
microcontroller.
Available function memory attributes:
__nearfunc, __farfunc
Available data memory attributes:

__tiny, __near, __far, __huge, __regvar, __eeprom, __tinyflash, __flash,
__farflash, __hugeflash, __generic, __io, and __ext_io.
Data objects, functions, and destinations of pointers or C++ references always have a
memory attribute. If no attribute is explicitly specified in the declaration or by the
pragma directive #pragma type_attribute, an appropriate default attribute is
implicitly used by the compiler. You can specify one memory attribute for each level of
pointer indirection.
General type attributes

Available function type attributes (affect how the function should be called):
__interrupt, __task, __version_1, __version_2, __version_3,
__version_4, __x, __x_z, __z, __z_x
Available data type attributes:

, const, volatile
You can specify as many type attributes as required for each level of pointer indirection.
Syntax for type attributes used on data objects

Type attributes use almost the same syntax rules as the type qualifiers const and
volatile. For example:
__near int i;
int __near j;

296 for AVR
AFE1_AFE2-1:1
Extended keywords
Both i and j are placed in near memory.

Unlike const and volatile, when a type attribute is used before the type specifier in
a derived type, the type attribute applies to the object, or typedef itself, except in
structure member declarations.
int __near * p; /* pointer to integer in near memory */
int * __near p; /* pointer in near memory */
__near int * p; /* pointer in near memory */
In all cases, if a memory attribute is not specified, an appropriate default memory type
is used.
Using a type definition can sometimes make the code clearer:
typedef __near int d16_int;
d16_int *q1;
d16_int is a typedef for integers in near memory. The variable q1 can point to such
integers.
You can also use the #pragma type_attributes directive to specify type attributes
for a declaration. The type attributes specified in the pragma directive are applied to the
data object or typedef being declared.
#pragma type_attribute=__near
int * q2;
The variable q2 is placed in near memory.

For more examples of using memory attributes, see More examples, page 57.
Syntax for type attributes used on functions

The syntax for using type attributes on functions differs slightly from the syntax of type
attributes on data objects. For functions, the attribute must be placed either in front of
the return type, or in parentheses, for example:
__interrupt void my_handler(void);
or
void (__interrupt my_handler)(void);
This declaration of my_handler is equivalent with the previous one:

#pragma type_attribute=__interrupt
void my_handler(void);
297
AFE1_AFE2-1:1
To declare a function pointer, use this syntax:

int (__nearfunc * fp) (double);
After this declaration, the function pointer fp points to near memory.

An easier way of specifying storage is to use type definitions:
typedef __nearfunc void FUNC_TYPE(int);
typedef FUNC_TYPE *FUNC_PTR_TYPE;
FUNC_TYPE func();
FUNC_PTR_TYPE funcptr;
Note that #pragma type_attribute can be used together with a typedef

declaration.
OBJECT ATTRIBUTES
Normally, object attributes affect the internal functionality of functions and data objects,
but not directly how the function is called or how the data is accessed. This means that
an object attribute does not normally need to be present in the declaration of an object.
These object attributes are available:
● Object attributes that can be used for variables:
__no_alloc, __no_alloc16, __no_alloc_str, __no_alloc_str16,
__no_init, __ro_placement, __no_runtime_init
● Object attributes that can be used for functions and variables:
location, @, __root
● Object attributes that can be used for functions:
__intrinsic, __monitor, __nested, __noreturn, vector
You can specify as many object attributes as required for a specific function or data
object.
For more information about location and @, see Controlling data and function
placement in memory, page 210. For more information about vector, see vector, page
339.
Syntax for object attributes

The object attribute must be placed in front of the type. For example, to place myarray
in memory that is not initialized at startup:
__no_init int myarray[10];

298 for AVR
AFE1_AFE2-1:1
Extended keywords
The #pragma object_attribute directive can also be used. This declaration is

equivalent to the previous one:
#pragma object_attribute=__no_init
int myarray[10];
Note: Object attributes cannot be used in combination with the typedef keyword.
Summary of extended keywords

This table summarizes the extended keywords:
Extended keyword Description
__eeprom Controls the storage of data objects in code memory space
__ext_io Controls the storage of data objects in I/O memory space
Supports I/O instructions; used for SFRs
__far Controls the storage of data objects in data memory space
__farflash Controls the storage of data objects in code memory space
__farfunc Controls the storage of functions in code memory space
__flash Controls the storage of data objects in code memory space
__generic Declares a generic pointer
__huge Controls the storage of data objects in data memory space
__hugeflash Controls the storage of data objects in code memory space
__interrupt Specifies interrupt functions
__intrinsic Reserved for compiler internal use only
__io Controls the storage of data objects in I/O memory space
Supports I/O instructions; used for SFRs
__monitor Specifies atomic execution of a function
__near Controls the storage of data objects in data memory space
__nearfunc Controls the storage of functions in code memory space
__nested Implements a nested interrupt
__no_alloc, Makes a constant available in the execution file
__no_alloc16
__no_alloc_str, Makes a string literal available in the execution file
__no_alloc_str16
__no_init Places a data object in non-volatile memory
__no_runtime_init Declares initialized variables that are not initialized at runtime.
Table 37: Extended keywords summary
299
AFE1_AFE2-1:1
Descriptions of extended keywords
Extended keyword Description

__noreturn Informs the compiler that the function will not return
__raw Prevents saving used registers in interrupt functions
__regvar Places a data object in a register
__root Ensures that a function or variable is included in the object code
even if unused
__ro_placement Places const volatile data in read-only memory.
__task Relaxes the rules for preserving registers
__tiny Controls the storage of data objects in data memory space
__tinyflash Controls the storage of data objects in code memory space
__version_1 Specifies the old calling convention; available for backward
compatibility
__version_2 Specifies the V2.10-V4.10B calling convention
__version_4 Specifies the default calling convention
__x Places the first pointer of the parameter list in register X
__x_z Places the first pointer of the parameter list in register X and the
second one in register Z
__z Places the first pointer of the parameter list in register Z
__z_x Places the first pointer of the parameter list in register Z and the
second one in register X
Table 37: Extended keywords summary (Continued)

This section gives detailed information about each extended keyword.
__eeprom
Syntax Follows the generic syntax rules for memory type attributes that can be used on data
objects, see Type attributes, page 295.
Description The __eeprom memory attribute overrides the default storage of variables given by the
selected memory model and places individual initialized and non-initialized variables in
the built-in EEPROM of the AVR microcontroller. These variables can be used like any
other variable and provide a convenient way to access the built-in EEPROM.

300 for AVR
AFE1_AFE2-1:1
Extended keywords
You can also override the supplied support functions to make the __eeprom memory
attribute access an EEPROM or flash memory that is placed externally but not on the
normal data bus, for example on an I2C bus.
To do this, you must write a new EEPROM library routines file and include it in your
project. The new file must use the same interface as the eeprom.s90 file in the
avr\src\lib directory (visible register use, the same entry points and the same
semantics).
Variables declared __eeprom are initialized only when a downloadable linker output
file is downloaded to the system, and not every time the system startup code is executed.
You can also use the __eeprom attribute to create a pointer explicitly pointing to an
object located in the EEPROM memory.
Storage information ● Memory space: EEPROM memory space

● Address range: 0-0xFF (255 bytes) if --eeprom_size <= 256 bytes, and 0–
0xFFFF (64 Kbytes) if --eeprom_size > 256 bytes.
● Maximum object size: 255 bytes if --eeprom_size <= 256 bytes, and 64 Kbytes-1
if --eeprom_size > 256 bytes.
● Pointer size: 1byte if --eeprom_size <= 256 bytes, and 2 bytes if
--eeprom_size > 256 bytes.
Example __eeprom int x;
See also Memory types, page 54 and --eeprom_size, page 251.
__ext_io
Description The __ext_io memory attribute overrides the default storage of variables given by the
selected memory model and allows individual objects to be accessed by use of the
special I/O instructions in the AVR microcontroller. The __ext_io memory attribute
implies that the object is __no_init and volatile.
Your application may access the AVR I/O system by using the memory-mapped internal
special function registers (SFRs). To access the AVR I/O system efficiently, the
__ext_io memory attribute should be included in the code.
Storage information ● Memory space: Data memory space

● Address range: 0x100–0xFFFF (64 Kbytes)
301
AFE1_AFE2-1:1
● Maximum object size: 4 bytes.

● Pointer size: Pointers not allowed.
Example __ext_io int x;
See also Memory types, page 54.
__far
Description The __far memory attribute overrides the default storage of variables given by the
selected memory model and places individual variables and constants in far memory.
You can also use the __far attribute to create a pointer explicitly pointing to an object
located in the far memory.

● Address range: 0–0xFFFFFF
● Maximum object size: 32 Kbytes-1. An object cannot cross a 64-Kbyte boundary.
● Pointer size: 3 bytes. Arithmetics is only performed on the two lower bytes, except
comparison which is always performed on the entire 24-bit address.
Example __far int x;
__farflash
Description The __farflash memory attribute overrides the default storage of variables given by
the selected memory model and places individual variables and constants in flash (code)
memory. You can also use the __farflash attribute to create a pointer explicitly
pointing to an object located in the flash (code) memory.
Note: It is preferable to declare flash objects as const. The __farflash memory
attribute is only available for AVR chips with more than 64 Kbytes of flash memory.
Storage information ● Memory space: Code memory space

302 for AVR
AFE1_AFE2-1:1
Extended keywords
● Address range: 0–0xFFFFFF

● Maximum object size: 32 Kbytes-1. An object cannot cross a 64-Kbyte boundary.
● Pointer size: 3 bytes. Arithmetics is only performed on the two lower bytes, except
comparison which is always performed on the entire 24-bit address.
Example __farflash int x;
__farfunc
Syntax Follows the generic syntax rules for memory type attributes that can be used on
functions, see Type attributes, page 295.
Description The __farfunc memory attribute overrides the default storage of functions given by
the selected code model and places individual functions in code memory. You can also
use the __farfunc attribute to create a pointer explicitly pointing to a function located
in FARCODE memory (0-0xFFFFFF).
The default code pointer for the -v5 and -v6 processor options is __farfunc, and it
only affects the size of the function pointers.
Note that pointers with function memory attributes have restrictions in implicit and
explicit casts when casting between pointers and also when casting between pointers
and integer types.

● Address range: 0–0x7FFFFE (8 Mbytes)
● Pointer size: 3 bytes
Example __farfunc void myfunction(void);
See also Function storage, page 63.
__flash
303
AFE1_AFE2-1:1
Description The __flash memory attribute overrides the default storage of variables given by the
selected memory model and places individual variables and constants in flash (code)
memory.
Note that it is preferable to declare flash objects as constant. The __flash keyword is
available for most AVR processors.
Because the AVR microcontrollers have only a small amount of on-board RAM, this
memory should not be wasted on data that could be stored in flash memory (of which
there is much more). However, because of the architecture of the processor, a default
pointer cannot access the flash memory. The __flash keyword is used to identify that
the constant should be stored in flash memory.
A header file called pgmspace.h is installed in the avr\inc directory, to provide some
standard C library functions for taking strings stored in flash memory as arguments.
You can also use the __flash attribute to create a pointer explicitly pointing to an
object located in the flash (code) memory.
Note: The __flash option cannot be used in a reduced ATtiny core.

● Address range: 0–0xFFFF (64 Kbytes)
● Maximum object size: 32 Kbytes-1.
● Pointer size: 2 bytes.
Example A program defines a couple of strings that are stored in flash memory:
__flash char str1[] = "Message 1";
__flash char str2[] = "Message 2";
The program creates a __flash pointer to point to one of these strings, and assigns it
to str1:
char __flash *msg;
msg=str1;
Using the strcpy_P function declared in pgmspace.h, a string in flash memory can be
copied to another string in RAM as follows:
strcpy_P(dest, msg);

304 for AVR
AFE1_AFE2-1:1
Extended keywords
__generic
Description The __generic pointer type attribute declares a generic pointer that can point to
objects in both code and data space. The size of the generic pointer depends on which
processor option is used.
The most significant bit (MSB) determines whether __generic points to Code
(MSB=1) or Data (MSB=0).
It is not possible to place objects in a generic memory area, only to point to it. When
using generic pointers, make sure that objects that have been declared __far and
__huge are located in the range 0x0–0x7FFFFF. Objects may still be placed in the
entire data address space, but a generic pointer cannot point to objects in the upper half
of the data address space.
The __generic keyword cannot be used with the #pragma type_attribute
directive for a pointer.
Access through a __generic pointer is implemented with inline code. Because this
type of access is slow and generates a lot of code, __generic pointers should be
avoided when possible.
Storage information ● Memory space: Data or code memory space

● Address range: 0–0x7FFF (32 Kbytes) for the processor options -v0 and -v1,
0-07FFFFF (8 Mbytes) for the processor options -v2 to -v6
● Maximum object size: 32 Kbytes-1 if the pointer size is 2 bytes, and 8 Mbytes-1 if
the pointer size is 3 bytes.
● Generic pointer size: 1 + 15 bits when the processor options -v0 and -v1 are used.
1 + 23 bits when the processor options -v2 to -v6 are used.
__huge
Description The __huge memory attribute overrides the default storage of variables given by the
selected memory model and places individual variables and constants in huge memory.
305
AFE1_AFE2-1:1
You can also use the __huge attribute to create a pointer explicitly pointing to an object
located in the huge memory.

● Address range: 0–0xFFFFFF (16 Mbytes)
● Maximum object size: 16 Mbytes-1.
Example __huge int x;
__hugeflash
Description The __hugeflash memory attribute overrides the default storage of variables given by
memory. You can also use the __hugeflash attribute to create a pointer explicitly
Note that it is preferable to declare flash objects as constant. The __hugeflash
memory attribute is only available for AVR microcontrollers with at least 64 Kbytes of
flash memory.

● Address range: 0–0x7FFFFF (8 Mbytes)
● Maximum object size: 8 Mbytes-1.
Example __hugeflash int x;

306 for AVR
AFE1_AFE2-1:1
Extended keywords
__interrupt
Syntax See Syntax for type attributes used on functions, page 297.
Description The __interrupt keyword specifies interrupt functions. To specify one or several
interrupt vectors, use the #pragma vector directive. The range of the interrupt vectors
depends on the device used. It is possible to define an interrupt function without a vector,
but then the compiler will not generate an entry in the interrupt vector table.
An interrupt function must have a void return type and cannot have any parameters.
The header file iodevice.h, where device corresponds to the selected device,
contains predefined names for the existing interrupt vectors.
To make sure that the interrupt handler executes as fast as possible, you should compile
it with -Ohs, or use #pragma optimize=speed if the module is compiled with another
optimization goal.
Example #pragma vector=0x14

__interrupt void my_interrupt_handler(void);
See also Interrupt functions, page 64, vector, page 339, and INTVEC, page 385.
__intrinsic
Description The __intrinsic keyword is reserved for compiler internal use only.
__io
Description The __io memory attribute overrides the default storage of variables given by the
selected memory model and places individual variables and constants in I/O memory.

● Maximum object size: 4 bytes (32 bits).
Example __io int x;
307
AFE1_AFE2-1:1
__monitor
Syntax See Syntax for object attributes, page 298.
Description The __monitor keyword causes interrupts to be disabled during execution of the
function. This allows atomic operations to be performed, such as operations on
semaphores that control access to resources by multiple processes. A function declared
with the __monitor keyword is equivalent to any other function in all other respects.
Example __monitor int get_lock(void);
See also Monitor functions, page 65. For information about related intrinsic functions, see
__disable_interrupt, page 343, __enable_interrupt, page 343, __get_interrupt_state,
page 344, and __set_interrupt_state, page 348, respectively.
__near
Description The __near memory attribute overrides the default storage of variables given by the
selected memory model and places individual variables and constants in near memory.
You can also use the __near attribute to create a pointer explicitly pointing to an object
located in the near memory.

● Maximum object size: 32 Kbytes-1.
Example __near int x;
__nearfunc
Syntax Follows the generic syntax rules for memory type attributes that can be used on
functions, see Type attributes, page 295.

308 for AVR
AFE1_AFE2-1:1
Extended keywords
Description The __nearfunc memory attribute overrides the default storage of functions given by
the selected code model and places individual functions in near memory.
Functions declared __nearfunc can be called from the entire code memory area, but
must reside in the first 128 Kbytes of the code memory.
The default for the -v0 to -v4 processor options is __nearfunc, and it only affects the
size of the function pointers.
Note that pointers with function memory attributes have restrictions in implicit and
explicit casts when casting between pointers and also when casting between pointers
and integer types.
You can also use the __nearfunc attribute to create a pointer explicitly pointing to a
function located in CODE memory (0-0xFFFF).

● Address range: 0–0x1FFFE (128 Kbytes)
● Pointer size: 2 bytes
Example __nearfunc void myfunction(void);
See also Function storage, page 63.
__nested
Syntax Follows the generic syntax rules for type attributes that can be used on functions, see
Type attributes, page 295.
Description Use the __nested keyword to implement a nested interrupt, in other words a new
interrupt can be served inside an interrupt function. A nested interrupt service routine
acts like a normal interrupt service routine except that it sets the interrupt enable bit
before any registers are saved.
Example __nested __interrupt void myInterruptFunction()
309
AFE1_AFE2-1:1
__no_alloc, __no_alloc16
Description Use the __no_alloc or __no_alloc16 object attribute on a constant to make the
constant available in the executable file without occupying any space in the linked
application.
You cannot access the contents of such a constant from your application. You can take
its address, which is an integer offset to the segment of the constant. The type of the
offset is unsigned long when __no_alloc is used, and unsigned short when
__no_alloc16 is used.
Example __no_alloc const struct MyData my_data @ "XXX" = {...};
See also __no_alloc_str, __no_alloc_str16, page 310.
__no_alloc_str, __no_alloc_str16
Syntax __no_alloc_str(string_literal @ segment)
and
__no_alloc_str16(string_literal @ segment)
where
string_literal The string literal that you want to make available in the
executable file.
segment The name of the segment to place the string literal in.
Description Use the __no_alloc_str or __no_alloc_str16 operators to make string literals

available in the executable file without occupying any space in the linked application.
The value of the expression is the offset of the string literal in the segment. For
__no_alloc_str, the type of the offset is unsigned long. For __no_alloc_str16,
the type of the offset is unsigned short.

310 for AVR
AFE1_AFE2-1:1
Extended keywords
Example #define MYSEG "YYY"

#define X(str) __no_alloc_str(str @ MYSEG)
extern void dbg_printf(unsigned long fmt, ...)
#define DBGPRINTF(fmt, ...) dbg_printf(X(fmt), __VA_ARGS__)
void
foo(int i, double d)
{
DBGPRINTF("The value of i is: %d, the value of d is: %f",i,d);
}
Depending on your debugger and the runtime support, this could produce trace output
on the host computer. Note that there is no such runtime support in C-SPY, unless you
use an external plugin module.
See also __no_alloc, __no_alloc16, page 310.
__no_init
Description Use the __no_init keyword to place a data object in non-volatile memory. This means
that the initialization of the variable, for example at system startup, is suppressed.
Example __no_init int myarray[10];
See also Non-initialized variables, page 223.
__no_runtime_init
Syntax Follows the generic syntax rules for object attributes, see Object attributes, page 298.
Description The __no_runtime_init keyword is used for declaring initialized variables for which
the initialization is performed when programming the device. These variables are not
initialized at runtime during system startup.
Note: The __no_runtime_init keyword cannot be used in combination with the
typedef keyword.
Example __no_runtime_init int myarray[10];
311
AFE1_AFE2-1:1
__noreturn
Description The __noreturn keyword can be used on a function to inform the compiler that the
function will not return. If you use this keyword on such functions, the compiler can
optimize more efficiently. Examples of functions that do not return are abort and exit.
Note: At optimization levels medium or high, the __noreturn keyword might cause
incorrect call stack debug information at any point where it can be determined that the
current function cannot return.
Example __noreturn void terminate(void);
__raw
Description This keyword prevents saving used registers in interrupt functions.
Example __raw __interrupt void my_interrupt_function()
__regvar
Syntax Follows the generic syntax rules for type attributes that can be used on data, see Type
attributes, page 295.
Description The __regvar extended keyword is used for declaring that a global or static variable
should be placed permanently in the specified register or registers. The registers R4–R15
can be used for this purpose, provided that they have been locked with the
--lock_regs compiler option.
Note:
● An object declared __regvar cannot have an initial value.
● The__regvar option cannot be used in a reduced ATtiny core.

● Address range: 4–F
● Maximum object size: 4 bytes (32 bits).

312 for AVR
AFE1_AFE2-1:1
Extended keywords
Example __regvar __no_init int counter @ 14;
This will create the 16-bit integer variable counter, which will always be available in
R15:R14. At least two registers must be locked.
See also --memory_model, -m, page 258, --lock_regs, page 258, and Preserved registers, page
158.
__root
Description A function or variable with the __root attribute is kept whether or not it is referenced
from the rest of the application, provided its module is included. Program modules are
always included and library modules are only included if needed.
Example __root int myarray[10];
See also For more information about modules, segments, and the link process, see the IAR Linker
and Library Tools Reference Guide.
__ro_placement
Description The __ro_placement attribute specifies that a data object should be placed in
read-only memory. There are two cases where you might want to use this object
attribute:
● Data objects declared const volatile are by default placed in read-write
memory. Use the __ro_placement object attribute to place the data object in
read-only memory instead.
● In C++, a data object declared const and that needs dynamic initialization is placed
in read-write memory and initialized at system startup. If you use the
__ro_placement object attribute, the compiler will give an error message if the
data object needs dynamic initialization.
You can only use the __ro_placement object attribute on const objects.
In some cases (primarily involving simple constructors), the compiler will be able to
optimize C++ dynamic initialization of a data object into static initialization. In that case
no error message will be issued for the object.
313
AFE1_AFE2-1:1
Example __ro_placement const volatile int x = 10;
__task
Syntax See Syntax for type attributes used on functions, page 297.
Description This keyword allows functions to relax the rules for preserving registers. Typically, the
keyword is used on the start function for a task in an RTOS.
By default, functions save the contents of used preserved registers on the stack upon
entry, and restore them at exit. Functions that are declared __task do not save all
registers, and therefore require less stack space.
Because a function declared __task can corrupt registers that are needed by the calling
function, you should only use __task on functions that do not return or call such a
function from assembler code.
The function main can be declared __task, unless it is explicitly called from the
application. In real-time applications with more than one task, the root function of each
task can be declared __task.
Example __task void my_handler(void);
__tiny
Description The __tiny memory attribute overrides the default storage of variables given by the
selected memory model and places individual variables and constants in tiny memory.
You can also use the __tiny attribute to create a pointer explicitly pointing to an object
located in the tiny memory.

● Address range: 0–0xFF (256 bytes)
● Pointer size: 1 byte.
Example __tiny int x;

314 for AVR
AFE1_AFE2-1:1
Extended keywords
__tinyflash
Description The __tinyflash memory attribute overrides the default storage of variables given by
memory. You can also use the __tinyflash attribute to create a pointer explicitly

● Address range: 0–0xFF (256 bytes)
● Pointer size: 1 byte.
Example __tinyflash int x;
__version_1
Description The __version_1 keyword is available for backward compatibility of the interface for
calling assembler routines from C. It makes a function use the calling convention of the
A90 IAR C Compiler instead of the default calling convention. The __version_1
calling convention is preferred when calling assembler functions from C.
Example __version_1 int func(int arg1, double arg2)
__version_2
Description The __version_2 keyword is available for backward compatibility of the interface for
calling assembler routines from C. The keyword makes a function use the
V2.10-V4.10B calling convention instead of the default calling convention.
315
AFE1_AFE2-1:1
__version_4
Description The __version_4 keyword specifies the default calling convention.
__x
Description The __x keyword places the first pointer of the parameter list in the register X. This
keyword is used by parts of the runtime library.
Note: This keyword can give very efficient code in a local perspective, but should be
used with care because it changes the calling convention and might have a negative
effect on the size of the entire application.
__x_z
Description The __x_z keyword places the first pointer of the parameter list in the register X and the
second one in register Z. This keyword is used by parts of the runtime library.

316 for AVR
AFE1_AFE2-1:1
Extended keywords
__z
Description The __z keyword places the first pointer of the parameter list in the register Z. This
keyword is used by parts of the runtime library.
__z_x
Description The __z_x keyword places the first pointer of the parameter list in the register Z and the
second one in register X. This keyword is used by parts of the runtime library.
317
AFE1_AFE2-1:1

318 for AVR
AFE1_AFE2-1:1
Pragma directives
● Summary of pragma directives
● Descriptions of pragma directives
Summary of pragma directives

The #pragma directive is defined by Standard C and is a mechanism for using
vendor-specific extensions in a controlled way to make sure that the source code is still
portable.
The pragma directives control the behavior of the compiler, for example how it allocates
memory for variables and functions, whether it allows extended keywords, and whether
it outputs warning messages.
The pragma directives are always enabled in the compiler.
This table lists the pragma directives of the compiler that can be used either with the
#pragma preprocessor directive or the _Pragma() preprocessor operator:
Pragma directive Description

basic_template_matching Makes a template function fully memory-attribute
aware.
bitfields Controls the order of bitfield members.
calls Lists possible called functions for indirect calls.
call_graph_root Specifies that the function is a call graph root.
constseg Places constant variables in a named segment.
cstat_disable See the C-STAT Static Analysis Guide.
cstat_enable See the C-STAT Static Analysis Guide.
cstat_restore See the C-STAT Static Analysis Guide.
cstat_suppress See the C-STAT Static Analysis Guide.
data_alignment Gives a variable a higher (more strict) alignment.
dataseg Places variables in a named segment.
default_function_attributes Sets default type and object attributes for
declarations and definitions of functions.
default_variable_attributes Sets default type and object attributes for
declarations and definitions of variables.
Table 38: Pragma directives summary
319
AFE1_AFE2-1:1
Summary of pragma directives

diag_default Changes the severity level of diagnostic messages.
diag_error Changes the severity level of diagnostic messages.
diag_remark Changes the severity level of diagnostic messages.
diag_suppress Suppresses diagnostic messages.
diag_warning Changes the severity level of diagnostic messages.
error Signals an error while parsing.
include_alias Specifies an alias for an include file.
inline Controls inlining of a function.
language Controls the IAR Systems language extensions.
location Specifies the absolute address of a variable, or places
groups of functions or variables in named segments.
message Prints a message.
object_attribute Adds object attributes to the declaration or
definition of a variable or function.
optimize Specifies the type and level of an optimization.
__printf_args Verifies that a function with a printf-style format
string is called with the correct arguments.
public_equ Defines a public assembler label and gives it a value.
required Ensures that a symbol that is needed by another
symbol is included in the linked output.
rtmodel Adds a runtime model attribute to the module.
__scanf_args Verifies that a function with a scanf-style format
string is called with the correct arguments.
section This directive is an alias for #pragma segment.
segment Declares a segment name to be used by intrinsic
functions.
STDC CX_LIMITED_RANGE Specifies whether the compiler can use normal
complex mathematical formulas or not.
STDC FENV_ACCESS Specifies whether your source code accesses the
floating-point environment or not.
STDC FP_CONTRACT Specifies whether the compiler is allowed to contract
floating-point expressions or not.
vector Specifies the vector of an interrupt or trap function.
Table 38: Pragma directives summary (Continued)

320 for AVR
AFE1_AFE2-1:1
Pragma directives

weak Makes a definition a weak definition, or creates a
weak alias for a function or a variable.
type_attribute Adds type attributes to a declaration or to
definitions.
Table 38: Pragma directives summary (Continued)
Note: For portability reasons, see also Recognized pragma directives (6.10.6), page
407.
Descriptions of pragma directives

This section gives detailed information about each pragma directive.
basic_template_matching
Syntax #pragma basic_template_matching
Description Use this pragma directive in front of a template function declaration to make the
function fully memory-attribute aware, in the rare cases where this is useful. That
template function will then match the template without the modifications, see Templates
and data memory attributes, page 184.
Example #pragma basic_template_matching

void MyF()
{
fun((int __near *) 0); // T = int __near
}
bitfields
Syntax #pragma bitfields={reversed|default}
Parameters
reversed Bitfield members are placed from the most significant bit to
the least significant bit.
default Bitfield members are placed from the least significant bit to
the most significant bit.
321
AFE1_AFE2-1:1
Description Use this pragma directive to control the order of bitfield members.
Example #pragma bitfields=reversed

/* Structure that uses reversed bitfields. */
struct S
{
unsigned char error : 1;
unsigned char size : 4;
unsigned short code : 10;
};
#pragma bitfields=default /* Restores to default setting. */
See also Bitfields, page 284.
calls
Syntax #pragma calls=function[, function...]
Parameters
function Any declared function
Description Use this pragma directive to list the functions that can be indirectly called in the
following statement. This information can be used for stack usage analysis in the linker.
Note: For an accurate result, you must list all possible called functions.
Example void Fun1(), Fun2();
void Caller(void (*fp)(void))

{
#pragma calls = Fun1, Fun2
(*fp)();
}
See also Stack usage analysis, page 79
call_graph_root
Syntax #pragma call_graph_root[=category]
Parameters
category A string that identifies an optional call graph root category

322 for AVR
AFE1_AFE2-1:1
Pragma directives
Description Use this pragma directive to specify that, for stack usage analysis purposes, the
immediately following function is a call graph root. You can also specify an optional
category. The compiler will usually automatically assign a call graph root category to
interrupt and task functions. If you use the #pragma call_graph_root directive on
such a function you will override the default category. You can specify any string as a
category.
Example #pragma call_graph_root="interrupt"
See also Stack usage analysis, page 79
constseg
Syntax #pragma constseg=[__memoryattribute ]{SEGMENT_NAME|default}
Parameters
__memoryattribute An optional memory attribute denoting in what memory the
segment will be placed; if not specified, default memory is
used.
SEGMENT_NAME A user-defined segment name; cannot be a segment name
predefined for use by the compiler and linker.
default Uses the default segment for constants.
Description Use this pragma directive to place constant variables in a named segment. The segment
name cannot be a segment name predefined for use by the compiler and linker. The
setting remains active until you turn it off again with the #pragma constseg=default
directive.
Example #pragma constseg=__huge MY_CONSTANTS

const int factorySettings[] = {42, 15, -128, 0};
#pragma constseg=default
data_alignment
Syntax #pragma data_alignment=expression
Parameters
expression A constant which must be a power of two (1, 2, 4, etc.).
323
AFE1_AFE2-1:1
Description Use this pragma directive to give the immediately following variable a higher (more
strict) alignment of the start address than it would otherwise have. This directive can be
used on variables with static and automatic storage duration.
When you use this directive on variables with automatic storage duration, there is an
upper limit on the allowed alignment for each function, determined by the calling
convention used.
Note: Normally, the size of a variable is a multiple of its alignment. The
data_alignment directive only affects the alignment of the variable’s start address,
and not its size, and can thus be used for creating situations where the size is not a
multiple of the alignment.
dataseg
Syntax #pragma dataseg=[__memoryattribute]{SEGMENT_NAME|default}
Parameters
__memoryattribute An optional memory attribute denoting in what memory the
segment will be placed; if not specified, default memory is
used.
SEGMENT_NAME A user-defined segment name; cannot be a segment name
default Uses the default segment.
Description Use this pragma directive to place variables in a named segment. The segment name
cannot be a segment name predefined for use by the compiler and linker. The variable
will not be initialized at startup, and can for this reason not have an initializer, which
means it must be declared __no_init. The setting remains active until you turn it off
again with the #pragma dataseg=default directive.
Example #pragma dataseg=__huge MY_SECTIONSEGMENT

__no_init char myBuffer[1000];
#pragma dataseg=default
default_function_attributes
Syntax #pragma default_function_attributes=[ attribute...]
where attribute can be:

type_attribute
object_attribute

324 for AVR
AFE1_AFE2-1:1
Pragma directives
@ segment_name
Parameters
type_attribute See Type attributes, page 295.
object_attribute See Object attributes, page 298.
@ segment_name See Data and function placement in segments, page 212.
Description Use this pragma directive to set default segment placement, type attributes, and object
attributes for function declarations and definitions. The default settings are only used for
declarations and definitions that do not specify type or object attributes or location in
some other way.
Specifying a default_function_attributes pragma directive with no attributes,
restores the initial state where no such defaults have been applied to function
declarations and definitions.
Example /* Place following functions in segment MYSEG" */

#pragma default_function_attributes = @ "MYSEG"
int fun1(int x) { return x + 1; }
int fun2(int x) { return x - 1;
/* Stop placing functions into MYSEG */
#pragma default_function_attributes =
has the same effect as:

int fun1(int x) @ "MYSEG" { return x + 1; }
int fun2(int x) @ "MYSEG" { return x - 1; }
See also location, page 331

object_attribute, page 332
type_attribute, page 338
default_variable_attributes
Syntax #pragma default_variable_attributes=[ attribute...]
where attribute can be:

type_attribute
object_attribute
@ segment_name
325
AFE1_AFE2-1:1
Parameters
type_attribute See Type attributes, page 295.
object_attributes See Object attributes, page 298.
@ segment_name See Data and function placement in segments, page 212.
Description Use this pragma directive to set default segment placement, type attributes, and object
attributes for declarations and definitions of variables with static storage duration. The
default settings are only used for declarations and definitions that do not specify type or
object attributes or location in some other way.
Specifying a default_variable_attributes pragma directive with no attributes
restores the initial state of no such defaults being applied to variables with static storage
duration.
Example /* Place following variables in segment MYSEG" */

#pragma default_variable_attributes = @ "MYSEG"
int var1 = 42;
int var2 = 17;
/* Stop placing variables into MYSEG */
#pragma default_variable_attributes =
has the same effect as:

int var1 @ "MYSEG" = 42;
int var2 @ "MYSEG" = 17;
See also location, page 331

object_attribute, page 332
type_attribute, page 338
diag_default
Syntax #pragma diag_default=tag[,tag,...]
Parameters
message number Pe177.
Description Use this pragma directive to change the severity level back to the default, or to the
severity level defined on the command line by any of the options --diag_error,
--diag_remark, --diag_suppress, or --diag_warnings, for the diagnostic

326 for AVR
AFE1_AFE2-1:1
Pragma directives
messages specified with the tags. This level remains in effect until changed by another
diagnostic-level pragma directive.
See also Diagnostics, page 231.
diag_error
Syntax #pragma diag_error=tag[,tag,...]
Parameters
Description Use this pragma directive to change the severity level to error for the specified
diagnostics. This level remains in effect until changed by another diagnostic-level
pragma directive.
diag_remark
Syntax #pragma diag_remark=tag[,tag,...]
Parameters
Description Use this pragma directive to change the severity level to remark for the specified
diagnostic messages. This level remains in effect until changed by another
diag_suppress
Syntax #pragma diag_suppress=tag[,tag,...]
Parameters
327
AFE1_AFE2-1:1
Description Use this pragma directive to suppress the specified diagnostic messages. This level
remains in effect until changed by another diagnostic-level pragma directive.
diag_warning
Syntax #pragma diag_warning=tag[,tag,...]
Parameters
Description Use this pragma directive to change the severity level to warning for the specified
diagnostic messages. This level remains in effect until changed by another
error
Syntax #pragma error message
Parameters
message A string that represents the error message.
Description Use this pragma directive to cause an error message when it is parsed. This mechanism
is different from the preprocessor directive #error, because the #pragma error
directive can be included in a preprocessor macro using the _Pragma form of the
directive and only causes an error if the macro is used.
Example #if FOO_AVAILABLE

#define FOO ...
#else
#define FOO _Pragma("error\"Foo is not available\"")
#endif
If FOO_AVAILABLE is zero, an error will be signaled if the FOO macro is used in actual
source code.

328 for AVR
AFE1_AFE2-1:1
Pragma directives
include_alias
Syntax #pragma include_alias ("orig_header" , "subst_header")
#pragma include_alias (<orig_header> , <subst_header>)
Parameters
orig_header The name of a header file for which you want to create an
alias.
subst_header The alias for the original header file.
Description Use this pragma directive to provide an alias for a header file. This is useful for
substituting one header file with another, and for specifying an absolute path to a relative
file.
This pragma directive must appear before the corresponding #include directives and
subst_header must match its corresponding #include directive exactly.
Example #pragma include_alias (<stdio.h> , <C:\MyHeaders\stdio.h>)

#include <stdio.h>
This example will substitute the relative file stdio.h with a counterpart located
according to the specified path.
See also Include file search procedure, page 228.
inline
Syntax #pragma inline[=forced|=never]
Parameters
No parameter Has the same effect as the inline keyword.
forced Disables the compiler’s heuristics and forces inlining.
never Disables the compiler’s heuristics and makes sure that the
function will not be inlined.
Description Use #pragma inline to advise the compiler that the function defined immediately after
the directive should be inlined according to C++ inline semantics.
Specifying #pragma inline=forced will always inline the defined function. If the
compiler fails to inline the function for some reason, for example due to recursion, a
warning message is emitted.
329
AFE1_AFE2-1:1
Inlining is normally performed only on the High optimization level. Specifying

#pragma inline=forced will enable inlining of the function also on the Medium
optimization level.
See also Inlining functions, page 67.
language
Syntax #pragma language={extended|default|save|restore}
Parameters
extended Enables the IAR Systems language extensions from the first
use of the pragma directive and onward.
default From the first use of the pragma directive and onward,
restores the settings for the IAR Systems language
extensions to whatever that was specified by compiler
options.
save|restore Saves and restores, respectively, the IAR Systems language
extensions setting around a piece of source code.
Each use of save must be followed by a matching restore
in the same file without any intervening #include directive.
Description Use this pragma directive to control the use of language extensions.
Example At the top of a file that needs to be compiled with IAR Systems extensions enabled:
#pragma language=extended
/* The rest of the file. */
Around a particular part of the source code that needs to be compiled with IAR Systems
extensions enabled, but where the state before the sequence cannot be assumed to be the
same as that specified by the compiler options in use:
#pragma language=save
#pragma language=extended
/* Part of source code. */
#pragma language=restore
See also -e, page 250 and --strict, page 273.

330 for AVR
AFE1_AFE2-1:1
Pragma directives
location
Syntax #pragma location={address|NAME}
Parameters
address The absolute address of the global or static variable for which
you want an absolute location.
NAME A user-defined segment name; cannot be a segment name
Description Use this pragma directive to specify the location—the absolute address—of the global
or static variable whose declaration follows the pragma directive. The variable must be
declared either __no_init or const. Alternatively, the directive can take a string
specifying a segment for placing either a variable or a function whose declaration
follows the pragma directive. Do not place variables that would normally be in different
segments (for example, variables declared as __no_init and variables declared as
const) in the same named segment.
Example #pragma location=0x2000

__no_init volatile char PORT1; /* PORT1 is located at address
0x2000 */
#pragma segment="FLASH"
#pragma location="FLASH"
__no_init char PORT1; /* PORT1 is located in segment FLASH */
/* A better way is to use a corresponding mechanism */

#define FLASH _Pragma("location=\"FLASH\"")
/* ... */
FLASH __no_init int i; /* i is placed in the FLASH segment */
Note: This still places the variables in DATA memory, and not in CODE.
See also Controlling data and function placement in memory, page 210 and Placing user-defined
segments, page 91.
message
Syntax #pragma message(message)
Parameters
message The message that you want to direct to the standard output
stream.
331
AFE1_AFE2-1:1
Description Use this pragma directive to make the compiler print a message to the standard output
stream when the file is compiled.
Example #ifdef TESTING

#pragma message("Testing")
#endif
object_attribute
Syntax #pragma object_attribute=object_attribute[ object_attribute...]
Parameters For information about object attributes that can be used with this pragma directive, see
Object attributes, page 298.
Description Use this pragma directive to add one or more IAR-specific object attributes to the
declaration or definition of a variable or function. Object attributes affect the actual
variable or function and not its type. When you define a variable or function, the union
of the object attributes from all declarations including the definition, is used.
Example #pragma object_attribute=__no_init

char bar;
is equivalent to:
__no_init char bar;
See also General syntax rules for extended keywords, page 295.
optimize
Syntax #pragma optimize=[goal][level][no_optimization...]
Parameters
goal Choose between:
size, optimizes for size
balanced, optimizes balanced between speed and size
speed, optimizes for speed.
no_size_constraints, optimizes for speed, but relaxes the

normal restrictions for code size expansion.

332 for AVR
AFE1_AFE2-1:1
Pragma directives
level Specifies the level of optimization; choose between none,

low, medium, or high.
no_optimization Disables one or several optimizations; choose between:

no_code_motion, disables code motion
no_crosscall, disables interprocedural cross call
no_crossjump, disables interprocedural cross jump
no_cse, disables common subexpression elimination
no_inline, disables function inlining
no_relays, disables interprocedural relays
no_tbaa, disables type-based alias analysis
no_unroll, disables loop unrolling
Description Use this pragma directive to decrease the optimization level, or to turn off some specific
optimizations. This pragma directive only affects the function that follows immediately
after the directive.
The parameters size, balanced, speed, and no_size_constraints only have
effect on the high optimization level and only one of them can be used as it is not
possible to optimize for speed and size at the same time. It is also not possible to use
preprocessor macros embedded in this pragma directive. Any such macro will not be
expanded by the preprocessor.
Note: If you use the #pragma optimize directive to specify an optimization level that
is higher than the optimization level you specify using a compiler option, the pragma
directive is ignored.
Example #pragma optimize=speed

int SmallAndUsedOften()
{
}
#pragma optimize=size
int BigAndSeldomUsed()
{
}
See also Fine-tuning enabled transformations, page 216.
333
AFE1_AFE2-1:1
__printf_args
Syntax #pragma __printf_args
Description Use this pragma directive on a function with a printf-style format string. For any call to
that function, the compiler verifies that the argument to each conversion specifier (for
example %d) is syntactically correct.
Example #pragma __printf_args

int printf(char const *,...);
void PrintNumbers(unsigned short x)

{
printf("%d", x); /* Compiler checks that x is an integer */
}
public_equ
Syntax #pragma public_equ="symbol",value
Parameters
symbol The name of the assembler symbol to be defined (string).
value The value of the defined assembler symbol (integer constant
expression).
Description Use this pragma directive to define a public assembler label and give it a value.
Example #pragma public_equ="MY_SYMBOL",0x123456
See also --public_equ, page 268.
required
Syntax #pragma required=symbol
Parameters
symbol Any statically linked function or variable.
Description Use this pragma directive to ensure that a symbol which is needed by a second symbol
is included in the linked output. The directive must be placed immediately before the
second symbol.

334 for AVR
AFE1_AFE2-1:1
Pragma directives
Use the directive if the requirement for a symbol is not otherwise visible in the
application, for example if a variable is only referenced indirectly through the segment
it resides in.
Example const char copyright[] = "Copyright by me";
#pragma required=copyright
int main()
{
}
Even if the copyright string is not used by the application, it will still be included by the
linker and available in the output.
See also Inline assembler, page 151
rtmodel
Syntax #pragma rtmodel="key","value"
Parameters
"key" A text string that specifies the runtime model attribute.
"value" A text string that specifies the value of the runtime model
attribute. Using the special value * is equivalent to not
defining the attribute at all.
Description Use this pragma directive to add a runtime model attribute to a module, which can be
used by the linker to check consistency between modules.
This pragma directive is useful for enforcing consistency between modules. All modules
that are linked together and define the same runtime attribute key must have the same
value for the corresponding key, or the special value *. It can, however, be useful to state
explicitly that the module can handle any runtime model.
A module can have several runtime model definitions.
Note: The predefined compiler runtime model attributes start with a double underscore.
To avoid confusion, this style must not be used in the user-defined attributes.
Example #pragma rtmodel="I2C","ENABLED"
The linker will generate an error if a module that contains this definition is linked with
a module that does not have the corresponding runtime model attributes defined.
335
AFE1_AFE2-1:1
__scanf_args
Syntax #pragma __scanf_args
Description Use this pragma directive on a function with a scanf-style format string. For any call to
that function, the compiler verifies that the argument to each conversion specifier (for
example %d) is syntactically correct.
Example #pragma __scanf_args

int scanf(char const *,...);
int GetNumber()
{
int nr;
scanf("%d", &nr); /* Compiler checks that
the argument is a
pointer to an integer */
return nr;
}
segment
Syntax #pragma segment="NAME" [__memoryattribute]
alias
#pragma section="NAME" [__memoryattribute] [align]
Parameters
NAME The name of the segment.
__memoryattribute An optional memory attribute identifying the memory the
segment will be placed in; if not specified, default memory is
used.
align Specifies an alignment for the segment. The value must be a
constant integer expression to the power of two.
Description Use this pragma directive to define a segment name that can be used by the segment
operators __segment_begin, __segment_end, and __segment_size. All segment
declarations for a specific segment must have the same memory type attribute and
alignment.
The align and the__memoryattribute parameters are only relevant when used
together with the segment operators __segment_begin, __segment_end, and

336 for AVR
AFE1_AFE2-1:1
Pragma directives
__segment_size. If you consider using align on an individual variable to achieve a

higher alignment, you must instead use the #pragma data_alignment directive.
If an optional memory attribute is used, the return type of the segment operators
__segment_begin and __segment_end is:
void __memoryattribute *.
Note: To place variables or functions in a specific segment, use the #pragma

location directive or the @ operator.
Example #pragma segment="MYNEAR" __near 4
See also Dedicated segment operators, page 171 and the chapters Linking overview and Linking
your application.
STDC CX_LIMITED_RANGE
Syntax #pragma STDC CX_LIMITED_RANGE {ON|OFF|DEFAULT}
Parameters
ON Normal complex mathematic formulas can be used.
OFF Normal complex mathematic formulas cannot be used.
DEFAULT Sets the default behavior, that is OFF.
Description Use this pragma directive to specify that the compiler can use the normal complex
mathematic formulas for * (multiplication), / (division), and abs.
Note: This directive is required by Standard C. The directive is recognized but has no
effect in the compiler.
STDC FENV_ACCESS
Syntax #pragma STDC FENV_ACCESS {ON|OFF|DEFAULT}
Parameters
ON Source code accesses the floating-point environment. Note
that this argument is not supported by the compiler.
OFF Source code does not access the floating-point environment.
DEFAULT Sets the default behavior, that is OFF.
337
AFE1_AFE2-1:1
Description Use this pragma directive to specify whether your source code accesses the
floating-point environment or not.
Note: This directive is required by Standard C.
STDC FP_CONTRACT
Syntax #pragma STDC FP_CONTRACT {ON|OFF|DEFAULT}
Parameters
ON The compiler is allowed to contract floating-point
expressions.
OFF The compiler is not allowed to contract floating-point
expressions. Note that this argument is not supported by the
compiler.
DEFAULT Sets the default behavior, that is ON.
Description Use this pragma directive to specify whether the compiler is allowed to contract
floating-point expressions or not. This directive is required by Standard C.
Example #pragma STDC FP_CONTRACT=ON
type_attribute
Syntax #pragma type_attribute=type_attr[ type_attr...]
Parameters For information about type attributes that can be used with this pragma directive, see
Description Use this pragma directive to specify IAR-specific type attributes, which are not part of
Standard C. Note however, that a given type attribute might not be applicable to all kind
of objects.
This directive affects the declaration of the identifier, the next variable, or the next
function that follows immediately after the pragma directive.
Example In this example, an int object with the memory attribute __near is defined:
#pragma type_attribute=__near
int x;

338 for AVR
AFE1_AFE2-1:1
Pragma directives
This declaration, which uses extended keywords, is equivalent:

__near int x;
See also The chapter Extended keywords.
vector
Syntax #pragma vector=vector1[, vector2, vector3, ...]
Parameters
vectorN The vector number(s) of an interrupt function.
Description Use this pragma directive to specify the vector(s) of an interrupt or trap function whose
declaration follows the pragma directive. Note that several vectors can be defined for
each function.
Example #pragma vector=0x14

__interrupt void my_handler(void);
weak
Syntax #pragma weak symbol1[=symbol2]
Parameters
symbol1 A function or variable with external linkage.
symbol2 A defined function or variable.
Description This pragma directive can be used in one of two ways:

● To make the definition of a function or variable with external linkage a weak
definition.
● To create a weak alias for another function or variable. You can make more than one
alias for the same function or variable.
Example To make the definition of foo a weak definition, write:

#pragma weak foo
To make NMI_Handler a weak alias for Default_Handler, write:

#pragma weak NMI_Handler=Default_Handler
339
AFE1_AFE2-1:1
If NMI_Handler is not defined elsewhere in the program, all references to

NMI_Handler will refer to Default_Handler.

340 for AVR
AFE1_AFE2-1:1
Intrinsic functions
● Summary of intrinsic functions
● Descriptions of intrinsic functions
Summary of intrinsic functions

To use intrinsic functions in an application, include the header file intrinsics.h.
Note that the intrinsic function names start with double underscores, for example:
__disable_interrupt
This table summarizes the intrinsic functions:

Intrinsic function Description
__delay_cycles Inserts a time delay
__DES_decryption Decrypts according to Digital Encryption
Standard
__DES_encryption Encrypts according to Digital Encryption
Standard
__disable_interrupt Disables interrupts
__enable_interrupt Enables interrupts
__extended_load_program_memory Returns one byte from code memory
__fractional_multiply_signed Generates an FMULS instruction
__fractional_multiply_signed_wit Generates an FMULSU instruction
h_unsigned
__fractional_multiply_unsigned Generates an FMUL instruction
__get_interrupt_state Returns the interrupt state
__indirect_jump_to Generates an IJMP instruction
__insert_opcode Assigns a value to a processor register
__lac Provides access to the LAC instruction
__las Provides access to the LAS instruction
__lat Provides access to the LAT instruction
__load_program_memory Returns one byte from program memory
__multiply_signed Generates a MULS instruction
Table 39: Intrinsic functions summary
341
AFE1_AFE2-1:1
Descriptions of intrinsic functions
Intrinsic function Description

__multiply_signed_with_unsigned Generates a MULSU instruction
__multiply_unsigned Generates a MUL instruction
__no_operation Inserts a instruction
__require Sets a constant literal as required
__restore_interrupt Restores the interrupt flag
__reverse Reverses the byte order of a value
__save_interrupt Saves the state of the interrupt flag
__set_interrupt_state Restores the interrupt state
__sleep Inserts a SLEEP instruction
__swap_nibbles Swaps bit 0-3 with bit 4-7
__watchdog_reset Inserts a watchdog reset instruction
__xch Provides access to the XCH instruction
Table 39: Intrinsic functions summary (Continued)

This section gives reference information about each intrinsic function.
__delay_cycles
Syntax void __delay_cycles(unsigned long int);
Description Makes the compiler generate code that takes the given amount of cycles to perform, that
is, it inserts a time delay that lasts the specified number of cycles.
The specified value must be a constant integer expression and not an expression that is
evaluated at runtime.
__DES_decryption
Syntax unsigned long long __DES_decryption(unsigned long long data,
unsigned long long key);
where:
data The data to decrypt

342 for AVR
AFE1_AFE2-1:1
Intrinsic functions
key The key to decrypt against
Description Decrypts according to the Digital Encryption Standard (DES). __DES_decryption

performs a DES decryption of data against key and returns the decrypted data.
Note: This intrinsic function is available only for cores with support for the DES
instruction.
__DES_encryption
Syntax unsigned long long __DES_encryption(unsigned long long data,
unsigned long long key);
where:
data The data to encrypt

key The key to encrypt against
Description Encrypts according to the Digital Encryption Standard (DES). __DES_encryption

performs a DES encryption of data against key and returns the encrypted data.
This intrinsic function is available only for cores with support for the DES instruction.
__disable_interrupt
Syntax void __disable_interrupt(void);
Description Disables interrupts by inserting the instruction.
__enable_interrupt
Syntax void __enable_interrupt(void);
Description Enables interrupts by inserting the instruction.
__extended_load_program_memory
Syntax unsigned char __extended_load_program_memory(unsigned char
__farflash *);
343
AFE1_AFE2-1:1
Description Returns one byte from code memory.

Use this intrinsic function to access constant data in code memory.
__fractional_multiply_signed
Syntax signed int __fractional_multiply_signed(signed char, signed
char);
Description Generates an FMULS instruction.
__fractional_multiply_signed_with_unsigned
Syntax signed int __fractional_multiply_signed_with_unsigned(signed
char, unsigned char);
Description Generates an FMULSU instruction.
__fractional_multiply_unsigned
Syntax unsigned int __fractional_multiply_unsigned(unsigned char,
unsigned char);
Description Generates an FMUL instruction.
__get_interrupt_state
Syntax __istate_t __get_interrupt_state(void);
Description Returns the global interrupt state. The return value can be used as an argument to the
__set_interrupt_state intrinsic function, which will restore the interrupt state.

344 for AVR
AFE1_AFE2-1:1
Intrinsic functions
Example #include "intrinsics.h"
void CriticalFn()
{
__istate_t s = __get_interrupt_state();
__disable_interrupt();
__set_interrupt_state(s);
}
The advantage of using this sequence of code compared to using

__disable_interrupt and __enable_interrupt is that the code in this example
will not enable any interrupts disabled before the call of __get_interrupt_state.
__indirect_jump_to
Syntax void __indirect_jump_to(unsigned long);
Description Jumps to the address specified by the argument by using the IJMP or EIJMP instruction,
depending on the generic processor option:
IJMP -v0 to -v4
EIJMP -v5 and -v6
__insert_opcode
Syntax void __insert_opcode(unsigned short);
Description Inserts a DW unsigned directive.
__lac
Syntax unsigned char __lac(unsigned char, unsigned char *);
Description Provides access to the LAC (Load And Clear) instruction.
345
AFE1_AFE2-1:1
__las
Syntax unsigned char __las(unsigned char, unsigned char *);
Description Provides access to the LAS (Load And Set) instruction.
__lat
Syntax unsigned char __lat(unsigned char, unsigned char *);
Description Provides access to the LAT (Load And Toggle) instruction.
__load_program_memory
Syntax unsigned char __load_program_memory(unsigned char __flash *);
Description Returns one byte from code memory. The constants must be placed within the first 64
Kbytes of memory.
__multiply_signed
Syntax signed int __multiply_signed(signed char, signed char);
Description Generates a MULS instruction.
__multiply_signed_with_unsigned
Syntax signed int __multiply_signed_with_unsigned(signed char, unsigned
char);
Description Generates a MULSU instruction.
__multiply_unsigned
Syntax unsigned int __multiply_unsigned(unsigned char, unsigned char);
Description Generates a MUL instruction.

346 for AVR
AFE1_AFE2-1:1
Intrinsic functions
__no_operation
Syntax void __no_operation(void);
Description Inserts a NOP instruction.
__require
Syntax void __require(void *);
Description Sets a constant literal as required.

One of the prominent features of the IAR XLINK Linker is its ability to strip away
anything that is not needed. This is a very good feature because it reduces the resulting
code size to a minimum. However, in some situations you may want to be able to
explicitly include a piece of code or a variable even though it is not directly used.
The argument to __require could be a variable, a function name, or an exported
assembler label. It must, however, be a constant literal. The label referred to will be
treated as if it would be used at the location of the __require call.
Example In this example, the copyright message will be included in the generated binary file even
though it is not directly used.
#include <intrinsics.h>
char copyright[] = "Copyright 2011 by XXXX";
void main(void)
{
__require(copyright);
[... the rest of the program ...]
}
__restore_interrupt
Syntax void __restore_interrupt(unsigned char oldState);
Description Restores the interrupt flag to the specified state.

Note: The value of oldState must be the result of a call to the __save_interrupt
intrinsic function.
347
AFE1_AFE2-1:1
__reverse
Syntax unsigned int __reverse(unsigned int);
Description Reverses the byte order of the value given as parameter. Avoid using __reverse in
complex expressions as it might introduce extra register copying.
Example signed int __reverse( signed int);

unsigned long __reverse(unsigned long);
signed long __reverse( signed long);
void __far * __reverse(void __far *); /* Only on -v4 */
/* and -v6 */
void __huge * __reverse(void __huge *); /* Only on -v4 */
/* and -v6 */
void __farflash * __reverse(void __farflash *);
/* Only on -v2 through -v6 with > 64k flash memory */
void __hugeflash * __reverse(void __hugeflash *);
/* Only on -v2 through -v6 with > 64k flash memory */
__save_interrupt
Syntax unsigned char __save_interrupt(void);
Description Saves the state of the interrupt flag in the byte returned. This value can then be used for
restoring the state of the interrupt flag with the __restore_interrupt intrinsic
function.
Example unsigned char oldState;
oldState = __save_interrupt();
__disable_interrupt();
/* Critical section goes here */
__restore_interrupt(oldState);
__set_interrupt_state
Syntax void __set_interrupt_state(__istate_t);
Description Restores the interrupt state to a value previously returned by the

__get_interrupt_state function.
For information about the __istate_t type, see __get_interrupt_state, page 344.

348 for AVR
AFE1_AFE2-1:1
Intrinsic functions
__sleep
Syntax void __sleep(void);
Description Inserts a SLEEP instruction.
__swap_nibbles
Syntax unsigned char __swap_nibbles(unsigned char);
Description Swaps bit 0-3 with bit 4-7 of the parameter and returns the swapped value.
__watchdog_reset
Syntax void __watchdog_reset(void);
Description Inserts a watchdog reset instruction.
__xch
Syntax unsigned char __xch(unsigned char, unsigned char *);
Description Provides access to the XCH (Exchange) instruction.
349
AFE1_AFE2-1:1

350 for AVR
AFE1_AFE2-1:1
The preprocessor
● Overview of the preprocessor
● Description of predefined preprocessor symbols
● Descriptions of miscellaneous preprocessor extensions
Overview of the preprocessor

The preprocessor of the IAR C/C++ Compiler for AVR adheres to Standard C. The
compiler also makes these preprocessor-related features available to you:
● Predefined preprocessor symbols
These symbols allow you to inspect the compile-time environment, for example the
time and date of compilation. For more information, see Description of predefined
preprocessor symbols, page 352.
● User-defined preprocessor symbols defined using a compiler option
In addition to defining your own preprocessor symbols using the #define directive,
you can also use the option -D, see -D, page 243.
● Preprocessor extensions
There are several preprocessor extensions, for example many pragma directives; for
more information, see the chapter Pragma directives. For information about the
corresponding _Pragma operator and the other extensions related to the
preprocessor, see Descriptions of miscellaneous preprocessor extensions, page 359.
● Preprocessor output
Use the option --preprocess to direct preprocessor output to a named file, see
--preprocess, page 268.
To specify a path for an include file, use forward slashes:
#include "mydirectory/myfile"
In source code, use forward slashes:

file = fopen("mydirectory/myfile","rt");
Note that backslashes can also be used. In this case, use one in include file paths and two
in source code strings.
351
AFE1_AFE2-1:1
Description of predefined preprocessor symbols

This section lists and describes the preprocessor symbols.
__BASE_FILE__
Description A string that identifies the name of the base source file (that is, not the header file), being
compiled.
See also __FILE__, page 353, and --no_path_in_file_macros, page 262.
__BUILD_NUMBER__
Description A unique integer that identifies the build number of the compiler currently in use. The
build number does not necessarily increase with a compiler that is released later.
__CORE__
Description An integer that identifies the processor variant in use. The value reflects the setting of
the processor option -vn,
__COUNTER__
Description A macro that expands to a new integer each time it is expanded, starting at zero (0) and
counting up.
__cplusplus
Description An integer which is defined when the compiler runs in any of the C++ modes, otherwise
it is undefined. When defined, its value is 199711L. This symbol can be used with
#ifdef to detect whether the compiler accepts C++ code. It is particularly useful when
creating header files that are to be shared by C and C++ code.
This symbol is required by Standard C.
__CPU__
Description A symbol that identifies the processor variant in use. The value reflects the setting of the
processor option -vn,

352 for AVR
AFE1_AFE2-1:1
The preprocessor
__DATE__
Description A string that identifies the date of compilation, which is returned in the form
"Mmm dd yyyy", for example "Oct 30 2014".
__device__
Description An integer that identifies the processor variant in use. The value reflects the setting of
the --cpu option. device corresponds exactly to the device name, except for FpSLic,
which uses the predefined symbol __AT94Kxx__. For example, the symbol is
__AT90S2313__ when the --cpu=2313 option is used, and __ATmega163__, when
--cpu=m163 is used.
__DOUBLE__
Description An integer that identifies the size of the data type double. The symbol is defined to 32
or 64, depending on the setting of the option --64bit_doubles.
__embedded_cplusplus
Description An integer which is defined to 1 when the compiler runs in any of the C++ modes,
otherwise the symbol is undefined. This symbol can be used with #ifdef to detect
whether the compiler accepts C++ code. It is particularly useful when creating header
files that are to be shared by C and C++ code.
__FILE__
Description A string that identifies the name of the file being compiled, which can be both the base
source file and any included header file.
See also __BASE_FILE__, page 352, and --no_path_in_file_macros, page 262.
353
AFE1_AFE2-1:1
__func__
Description A predefined string identifier that is initialized with the name of the function in which
the symbol is used. This is useful for assertions and other trace utilities. The symbol
requires that language extensions are enabled.
See also -e, page 250 and __PRETTY_FUNCTION__, page 356.
__FUNCTION__
Description A predefined string identifier that is initialized with the name of the function in which
the symbol is used. This is useful for assertions and other trace utilities. The symbol
requires that language extensions are enabled.
See also -e, page 250 and __PRETTY_FUNCTION__, page 356.
__HAS_EEPROM__
Description A symbol that determines whether there is internal EEPROM available or not. When this
symbol is defined, there is internal EEPROM available. When this symbol is not
defined, there is no EEPROM available.
__HAS_EIND__
Description A symbol that determines whether the instruction EIND is available or not. When this
symbol is defined, the instruction EIND is available. When this symbol is not defined,
the EIND instruction is not available.
__HAS_ELPM__
Description A symbol that determines whether the instruction ELPM is available or not. When this
symbol is defined, the instruction ELPM is available. When this symbol is not defined,
the ELPM instruction is not available.

354 for AVR
AFE1_AFE2-1:1
The preprocessor
__HAS_ENHANCED_CORE__
Description A symbol that determines whether the enhanced core is used or not. The symbol reflects
the --enhanced_core option and is defined when the enhanced core is used. When
this symbol is not defined, the enhanced core is not used.
__HAS_FISCR__
Description A symbol that determines whether the instruction FISCR is available or not. When this
symbol is defined, the instruction FISCR is available. When this symbol is not defined,
the FISCR instruction is not available.
__HAS_MUL__
Description A symbol that determines whether the instruction MUL is available or not. When this
symbol is defined, the instruction MUL is available. When this symbol is not defined, the
MUL instruction is not available.
__HAS_RAMPD__
Description A symbol that determines whether the register RAMPD is available or not. When this
symbol is defined, the register RAMPD is available. When this symbol is not defined, the
RAMPD register is not available.
__HAS_RAMPX__
Description A symbol that determines whether the register RAMPX is available or not. When this
symbol is defined, the register RAMPX is available. When this symbol is not defined, the
RAMPX register is not available.
__HAS_RAMPY__
Description A symbol that determines whether the register RAMPY is available or not. When this
symbol is defined, the register RAMPY is available. When this symbol is not defined, the
RAMPY register is not available.
355
AFE1_AFE2-1:1
__HAS_RAMPZ__
Description A symbol that determines whether the register RAMPZ is available or not. When this
symbol is defined, the register RAMPZ is available. When this symbol is not defined, the
RAMPZ register is not available.
__IAR_SYSTEMS_ICC__
Description An integer that identifies the IAR compiler platform. The current value is 9. Note that
the number could be higher in a future version of the product. This symbol can be tested
with #ifdef to detect whether the code was compiled by a compiler from IAR Systems.
__ICCAVR__
Description An integer that is set to 1 when the code is compiled with the IAR C/C++ Compiler for
AVR.
__LINE__
Description An integer that identifies the current source line number of the file being compiled,
which can be both the base source file and any included header file.
__LITTLE_ENDIAN__
Description An integer that reflects the byte order and is defined to 1 (little-endian).
__MEMORY_MODEL__
Description An integer that identifies the memory model in use. The value reflects the setting of the
--memory_model option and is defined to 1 for the Tiny memory model, 2 for the Small
memory model, 3 for the Large memory model, and 4 for the Huge memory model.
__PRETTY_FUNCTION__
Description A predefined string identifier that is initialized with the function name, including
parameter types and return type, of the function in which the symbol is used, for

356 for AVR
AFE1_AFE2-1:1
The preprocessor
example "void func(char)". This symbol is useful for assertions and other trace
utilities. The symbol requires that language extensions are enabled.
See also -e, page 250 and__func__, page 354.
__STDC__
Description An integer that is set to 1, which means the compiler adheres to Standard C. This symbol
can be tested with #ifdef to detect whether the compiler in use adheres to Standard C.*
__STDC_VERSION__
Description An integer that identifies the version of the C standard in use. The symbol expands to
199901L,unless the --c89 compiler option is used in which case the symbol expands
to 199409L. This symbol does not apply in EC++ mode.
__SUBVERSION__
Description An integer that identifies the subversion number of the compiler version number, for
example 3 in 1.2.3.4.
__TID__
Description A symbol that expands to the target identifier which contains these parts:
● t, a target identifier, which is unique for each IAR compiler. For the AVR
microcontroller, the target identifier is 9.0.
● c, the value of the --cpu or -v option.
● m, the value reflects the setting of the --memory_model option; the value is defined
to 1 for Tiny, 2 for Small, 3 for Large, and 4 for Huge.
The __TID__ value is constructed as:
((t << 8) | (c << 4) | m)
You can extract the values as follows:

t = (__TID__ >> 8) & 0x7F; /* target id */
c = (__TID__ >> 4) & 0x0F; /* cpu core */
357
AFE1_AFE2-1:1
m = __TID__ & 0x0F; /* memory model */
To find the value of the target identifier for the current compiler, execute:
printf("%ld", (__TID__ >> 8) & 0x7F)
Note: The use of __TID__ is not recommended. We recommend that you use the
symbols __ICCAVR__ and __CORE__ instead.
__TIME__
Description A string that identifies the time of compilation in the form "hh:mm:ss".
__TIMESTAMP__
Description A string constant that identifies the date and time of the last modification of the current
source file. The format of the string is the same as that used by the asctime standard
function (in other words, "Tue Sep 16 13:03:52 2014").
__TINY_AVR__
Description A symbol that determines whether the reduced tiny AVR core is used or not. When this
symbol is defined, the reduced tiny AVR core is used. When this symbol is not defined,
the reduced tiny AVR core is not used.
__VER__
Description An integer that identifies the version number of the IAR compiler in use. The value of
the number is calculated in this way: (100 * the major version number + the
minor version number). For example, for compiler version 3.34, 3 is the major
version number and 34 is the minor version number. Hence, the value of __VER__ is
334.
__VERSION_1_CALLS__
Description A symbol that expands to 1 if the used calling convention is the old calling convention
used in compiler version 1.x. If zero, the new calling convention is used. For more
information about calling conventions, see Calling convention, page 155.

358 for AVR
AFE1_AFE2-1:1
The preprocessor
__XMEGA_CORE__
Description A symbol that determines whether the xmega core is used or not. When this symbol is
defined, the xmega core is used. When this symbol is not defined, the xmega core is not
used.
__XMEGA_USB__
Description A symbol that determines whether the xmega usb core is used or not. When this symbol
is defined, the xmega usb core is used. When this symbol is not defined, the xmega usb
core is not used.
Descriptions of miscellaneous preprocessor extensions

This section gives reference information about the preprocessor extensions that are
available in addition to the predefined symbols, pragma directives, and Standard C
directives.
NDEBUG
Description This preprocessor symbol determines whether any assert macros you have written in
your application shall be included or not in the built application.
If this symbol is not defined, all assert macros are evaluated. If the symbol is defined,
all assert macros are excluded from the compilation. In other words, if the symbol is:
● defined, the assert code will not be included
● not defined, the assert code will be included
This means that if you write any assert code and build your application, you should
define this symbol to exclude the assert code from the final application.
Note that the assert macro is defined in the assert.h standard include file.
In the IDE, the NDEBUG symbol is automatically defined if you build your application in
the Release build configuration.
See also _ReportAssert, page 129.
359
AFE1_AFE2-1:1
Descriptions of miscellaneous preprocessor extensions
#warning message
Syntax #warning message
where message can be any string.
Description Use this preprocessor directive to produce messages. Typically, this is useful for
assertions and other trace utilities, similar to the way the Standard C #error directive
is used. This directive is not recognized when the --strict compiler option is used.

360 for AVR
AFE1_AFE2-1:1
C/C++ standard library
functions
● C/C++ standard library overview
● DLIB runtime environment—implementation details
● CLIB runtime environment—implementation details
● AVR-specific library functions
For detailed reference information about the library functions, see the online
help system.
C/C++ standard library overview

The compiler comes with two different implementations of the C/C++ standard library:
The IAR DLIB Runtime Environment is a complete implementation of the C/C++
standard library, compliant with Standard C and C++. This library also supports
floating-point numbers in IEEE 754 format and it can be configured to include different
levels of support for locale, file descriptors, multibyte characters, etc.
The IAR CLIB Runtime Environment is a light-weight implementation of the C
standard library, which is not fully compliant with Standard C. Neither does it fully
support floating-point numbers in IEEE 754 format and it does not support C++.
For more information about customization, see the chapter The DLIB runtime
environment and The CLIB runtime environment, respectively.
For detailed information about the library functions, see the online documentation
supplied with the product. There is also keyword reference information for the DLIB
library functions. To obtain reference information for a function, select the function
name in the editor window and press F1.
For more information about library functions, see the chapter Implementation-defined
behavior for Standard C in this guide.
HEADER FILES
Your application program gains access to library definitions through header files, which
it incorporates using the #include directive. The definitions are divided into several
361
AFE1_AFE2-1:1
C/C++ standard library overview
different header files, each covering a particular functional area, letting you include just
those that are required.
It is essential to include the appropriate header file before making any reference to its
definitions. Failure to do so can cause the call to fail during execution, or generate error
or warning messages at compile time or link time.
LIBRARY OBJECT FILES

Most of the library definitions can be used without modification, that is, directly from
the library object files that are supplied with the product. For information about how to
choose a runtime library, see Basic project configuration, page 48. The linker will
include only those routines that are required—directly or indirectly—by your
application.
See also Overriding library modules, page 107 for information about how you can
override library modules with your own versions.
ALTERNATIVE MORE ACCURATE LIBRARY FUNCTIONS

The default implementation of cos, sin, tan, and pow is designed to be fast and small.
As an alternative, there are versions designed to provide better accuracy. They are
named __iar_xxx_accuratef for float variants of the functions and
__iar_xxx_accuratel for long double variants of the functions, and where xxx is
cos, sin, etc.
To use any of these more accurate versions, use the -e linker option.
REENTRANCY
A function that can be simultaneously invoked in the main application and in any
number of interrupts is reentrant. A library function that uses statically allocated data is
therefore not reentrant.
Most parts of the DLIB runtime environment are reentrant, but the following functions
and parts are not reentrant because they need static data:
● Heap functions—malloc, free, realloc, calloc, etc. and the C++ operators
new and delete
● Locale functions—localeconv, setlocale
● Multibyte functions—mblen, mbrlen, mbrtowc, mbsrtowc, mbtowc, wcrtomb,
wcsrtomb, wctomb
● Rand functions—rand, srand
● Time functions—asctime, localtime, gmtime, mktime
● The miscellaneous functions atexit, perror, strerror, strtok

362 for AVR
AFE1_AFE2-1:1
C/C++ standard library functions
● Functions that use files or the heap in some way. This includes scanf, sscanf,
getchar, getwchar, putchar, and putwchar. In addition, if you are using the
options --enable_multibyte and --dlib_config=Full, the printf and
sprintf functions (or any variants) can also use the heap.
For the CLIB library, the qsort function and functions that use files in some way are
non-reentrant. This includes printf, scanf, getchar, and putchar. However, the
functions sprintf and sscanf are reentrant.
Functions that can set errno are not reentrant, because an errno value resulting from
one of these functions can be destroyed by a subsequent use of the function before it is
read. This applies to math and string conversion functions, among others.
Remedies for this are:
● Do not use non-reentrant functions in interrupt service routines
● Guard calls to a non-reentrant function by a mutex, or a secure region, etc.
THE LONGJMP FUNCTION

A longjmp is in effect a jump to a previously defined setjmp. Any variable length
arrays or C++ objects residing on the stack during stack unwinding will not be
destroyed. This can lead to resource leaks or incorrect application behavior.
DLIB runtime environment—implementation details

The DLIB runtime environment provides most of the important C and C++ standard
library definitions that apply to embedded systems. These are of the following types:
● Adherence to a free-standing implementation of Standard C. The library supports
most of the hosted functionality, but you must implement some of its base
functionality. For additional information, see the chapter Implementation-defined
behavior for Standard C in this guide.
● Standard C library definitions, for user programs.
● C++ library definitions, for user programs.
● CSTARTUP, the module containing the start-up code, see the chapter The DLIB
runtime environment in this guide.
● Runtime support libraries; for example low-level floating-point routines.
● Intrinsic functions, allowing low-level use of AVR features. See the chapter
Intrinsic functions for more information.
363
AFE1_AFE2-1:1
In addition, the DLIB runtime environment includes some added C functionality, see
Added C functionality, page 367.
C HEADER FILES
This section lists the C header files specific to the DLIB runtime environment. Header
files may additionally contain target-specific definitions; these are documented in the
chapter Using C.
This table lists the C header files:
Header file Usage
assert.h Enforcing assertions when functions execute
complex.h Computing common complex mathematical functions
ctype.h Classifying characters
errno.h Testing error codes reported by library functions
fenv.h Floating-point exception flags
float.h Testing floating-point type properties
inttypes.h Defining formatters for all types defined in stdint.h
iso646.h Using Amendment 1—iso646.h standard header
limits.h Testing integer type properties
locale.h Adapting to different cultural conventions
math.h Computing common mathematical functions
setjmp.h Executing non-local goto statements
signal.h Controlling various exceptional conditions
stdarg.h Accessing a varying number of arguments
stdbool.h Adds support for the bool data type in C.
stddef.h Defining several useful types and macros
stdint.h Providing integer characteristics
stdio.h Performing input and output
stdlib.h Performing a variety of operations
string.h Manipulating several kinds of strings
tgmath.h Type-generic mathematical functions
time.h Converting between various time and date formats
uchar.h Unicode functionality (IAR extension to Standard C)
wchar.h Support for wide characters
Table 40: Traditional Standard C header files—DLIB

364 for AVR
AFE1_AFE2-1:1
Header file Usage

wctype.h Classifying wide characters
Table 40: Traditional Standard C header files—DLIB (Continued)
C++ HEADER FILES

This section lists the C++ header files:
● The C++ library header files
The header files that constitute the Embedded C++ library.
● The C++ standard template library (STL) header files
The header files that constitute STL for the Extended Embedded C++ library.
● The C++ C header files
The C++ header files that provide the resources from the C library.
The C++ library header files

This table lists the header files that can be used in Embedded C++:
Header file Usage
complex Defining a class that supports complex arithmetic
fstream Defining several I/O stream classes that manipulate external files
iomanip Declaring several I/O stream manipulators that take an argument
ios Defining the class that serves as the base for many I/O streams classes
iosfwd Declaring several I/O stream classes before they are necessarily defined
iostream Declaring the I/O stream objects that manipulate the standard streams
istream Defining the class that performs extractions
new Declaring several functions that allocate and free storage
ostream Defining the class that performs insertions
sstream Defining several I/O stream classes that manipulate string containers
streambuf Defining classes that buffer I/O stream operations
string Defining a class that implements a string container
strstream Defining several I/O stream classes that manipulate in-memory character
sequences
Table 41: C++ header files
365
AFE1_AFE2-1:1
The C++ standard template library (STL) header files

The following table lists the standard template library (STL) header files that can be
used in Extended Embedded C++:
Header file Description
algorithm Defines several common operations on sequences
deque A deque sequence container
functional Defines several function objects
hash_map A map associative container, based on a hash algorithm
hash_set A set associative container, based on a hash algorithm
iterator Defines common iterators, and operations on iterators
list A doubly-linked list sequence container
map A map associative container
memory Defines facilities for managing memory
numeric Performs generalized numeric operations on sequences
queue A queue sequence container
set A set associative container
slist A singly-linked list sequence container
stack A stack sequence container
utility Defines several utility components
vector A vector sequence container
Table 42: Standard template library header files
Using Standard C libraries in C++

The C++ library works in conjunction with some of the header files from the Standard
C library, sometimes with small alterations. The header files come in two forms—new
and traditional—for example, cassert and assert.h.
This table shows the new header files:
Header file Usage
cassert Enforcing assertions when functions execute
cctype Classifying characters
cerrno Testing error codes reported by library functions
cfloat Testing floating-point type properties
cinttypes Defining formatters for all types defined in stdint.h
Table 43: New Standard C header files—DLIB

366 for AVR
AFE1_AFE2-1:1
Header file Usage

climits Testing integer type properties
clocale Adapting to different cultural conventions
cmath Computing common mathematical functions
csetjmp Executing non-local goto statements
csignal Controlling various exceptional conditions
cstdarg Accessing a varying number of arguments
cstdbool Adds support for the bool data type in C.
cstddef Defining several useful types and macros
cstdint Providing integer characteristics
cstdio Performing input and output
cstdlib Performing a variety of operations
cstring Manipulating several kinds of strings
ctime Converting between various time and date formats
cwchar Support for wide characters
cwctype Classifying wide characters
Table 43: New Standard C header files—DLIB (Continued)
LIBRARY FUNCTIONS AS INTRINSIC FUNCTIONS

Certain C library functions will under some circumstances be handled as intrinsic
functions and will generate inline code instead of an ordinary function call, for example
memcpy, memset, and strcat.
ADDED C FUNCTIONALITY
The DLIB runtime environment includes some added C functionality.
The following include files provide these features:
● fenv.h
● stdio.h
● stdlib.h
● string.h
fenv.h
In fenv.h, trap handling support for floating-point numbers is defined with the
functions fegettrapenable and fegettrapdisable. No floating-point status flags
are supported.
367
AFE1_AFE2-1:1
stdio.h
These functions provide additional I/O functionality:
fdopen Opens a file based on a low-level file descriptor.

fileno Gets the low-level file descriptor from the file descriptor
(FILE*).
__gets Corresponds to fgets on stdin.
getw Gets a wchar_t character from stdin.
putw Puts a wchar_t character to stdout.
__ungetchar Corresponds to ungetc on stdout.
__write_array Corresponds to fwrite on stdout.
string.h
These are the additional functions defined in string.h:
strdup Duplicates a string on the heap.

strcasecmp Compares strings case-insensitive.
strncasecmp Compares strings case-insensitive and bounded.
strnlen Bounded string length.
time.h
There are two interfaces for using time_t and the associated functions time, ctime,
difftime, gmtime, localtime, and mktime:
● The 32-bit interface supports years from 1900 up to 2035 and uses a 32-bit integer
for time_t. The type and function have names like __time32_t, __time32, etc.
This variant is mainly available for backwards compatibility.
● The 64-bit interface supports years from -9999 up to 9999 and uses a signed
long long for time_t. The type and function have names like __time64_t,
__time64, etc.
In both interfaces, time_t starts at the year 1970.

The interfaces are defined in the system header file time.h.
An application can use either interface, and even mix them by explicitly using the 32-
or 64-bit variants. By default, the library and the header redirect time_t, time etc. to

368 for AVR
AFE1_AFE2-1:1
the 32-bit variants. However, to explicitly redirect them to their 64-bit variants, define
_DLIB_TIME_USES_64 in front of the inclusion of time.h or ctime.
See also, __time32, __time64, page 130.

clock_t is represented by a 32-bit integer type.
SYMBOLS USED INTERNALLY BY THE LIBRARY

The following symbols are used by the library, which means that they are visible in
library source files, etc:
__assignment_by_bitwise_copy_allowed
This symbol determines properties for class objects.

__constrange()
Determines the allowed range for a parameter to an intrinsic function and that the
parameter must be of type const.
__construction_by_bitwise_copy_allowed
This symbol determines properties for class objects.

__has_constructor, __has_destructor
These symbols determine properties for class objects and they function like the sizeof
operator. The symbols are true when a class, base class, or member (recursively) has a
user-defined constructor or destructor, respectively.
__memory_of
Determines the class memory. A class memory determines which memory a class
object can reside in. This symbol can only occur in class definitions as a class memory.
Note: The symbols are reserved and should only be used by the library.
Use the compiler option --predef_macros to determine the value for any predefined
symbols.
CLIB runtime environment—implementation details

The CLIB runtime environment provides most of the important C standard library
definitions that apply to embedded systems. These are of the following types:
● Standard C library definitions available for user programs. These are documented in
this chapter.
369
AFE1_AFE2-1:1
AVR-specific library functions
● The system startup code; see the chapter The CLIB runtime environment in this
guide.
● Runtime support libraries; for example low-level floating-point routines.
● Intrinsic functions, allowing low-level use of AVR features. See the chapter
Intrinsic functions for more information.
LIBRARY DEFINITIONS SUMMARY

This table lists the C header files specific to the CLIB runtime environment:
Header file Description
assert.h Assertions
ctype.h* Character handling
errno.h Error return values
float.h Limits and sizes of floating-point types
iccbutl.h Low-level routines
limits.h Limits and sizes of integral types
math.h Mathematics
setjmp.h Non-local jumps
stdarg.h Variable arguments
stdbool.h Adds support for the bool data type in C
stddef.h Common definitions including size_t, NULL,
ptrdiff_t, and offsetof
stdio.h Input/output
stdlib.h General utilities
string.h String handling
Table 44: CLIB runtime environment header files
* The functions isxxx, toupper, and tolower declared in the header file ctype.h
evaluate their argument more than once. This is not according to the ISO/ANSI
standard.

This section lists the AVR–specific library functions declared in pgmspace.h that allow
access to strings in flash memory. The _P functions allow access to strings in flash
memory only. The _G functions use the __generic pointer instead, which means that

370 for AVR
AFE1_AFE2-1:1
they allow access to both flash and data memory. Both the _P functions and the _G
functions are available in both the IAR CLIB Library and the IAR DLIB Library.
memcmp_G
Syntax int memcmp_G(const void *s1, const void __generic *s2, size_t
n);
Description Identical to memcmp except that s2 can be located in flash or data memory.
memcpy_G
Syntax void * memcpy_G(void *s1, const void __generic *s2, size_t n);
Description Identical to memcpy except that it copies string s2 in flash or data memory to the
location that s1 points to in data memory.
memcpy_P
Syntax void * memcpy_P(void *s1, PGM_P s2, size_t n);
Description Identical to memcpy except that it copies string s2 in flash memory to the location that
s1 points to in data memory.
printf_P
Syntax int printf_P(PGM_P __format,…);
Description Similar to printf except that the format string is in flash memory, not in data memory.
For information about how to override default formatter, see Formatters for printf, page
113.
puts_G
Syntax int puts_G(const char __generic *s);
Description Identical to puts except that the string to be written can be in flash or data memory.
371
AFE1_AFE2-1:1
puts_P
Syntax int puts_P(PGM_P __s);
Description Identical to puts except that the string to be written is in flash memory, not in data
memory.
scanf_P
Syntax int scanf_P(PGM_P __format,…);
Description Identical to scanf except that the format string is in flash memory, not in data memory.
For information about how to override the default formatter, see Formatters for scanf,
page 114.
sprintf_P
Syntax int sprintf_P(char *__s, PGM_P __format,…);
Description Identical to sprintf except that the format string is in flash memory, not in data
memory.
sscanf_P
Syntax int sscanf_P(const char *__s, PGM_P __format,…);
Description Identical to sscanf except that the format string is in flash memory, not in data memory.
strcat_G
Syntax char *strcat_G(char *s1, const char __generic *s2);
Description Identical to strcat except that string s2 can be in flash or data memory.
strcmp_G
Syntax int strcmp_G(const char *s1, const char __generic *s2);
Description Identical to strcmp except that string s2 can be in flash or data memory.

372 for AVR
AFE1_AFE2-1:1
strcmp_P
Syntax int strcmp_P(const char *s1, PGM_P s2);
Description Identical to strcmp except that string s2 is in flash memory, not in data memory.
strcpy_G
Syntax char *strcpy_G(char *s1, const char __generic *s2);
Description Identical to strcpy except that the string s2 being copied can be in flash or data
memory.
strcpy_P
Syntax char * strcpy_P(char *s1, PGM_P s2);
Description Identical to strcpy except that the string s2 being copied is in flash memory, not in data
memory.
strerror_P
Syntax PGM_P strerror_P(int errnum);
Description Identical to strerror except that the string returned is in flash memory, not in data
memory.
strlen_G
Syntax size_t strlen_G(const char __generic *s);
Description Identical to strlen except that the string being tested can be in flash or data memory.
strlen_P
Syntax size_t strlen_P(PGM_P s);
Description Identical to strlen except that the string being tested is in flash memory, not in data
memory.
373
AFE1_AFE2-1:1
strncat_G
Syntax char *strncat_G(char *s1, const char __generic *s2, size_t n);
Description Identical to strncmp except that the string s2 can be in flash or data memory.
strncmp_G
Syntax int strncmp_G(const char *s1, const char __generic *s2, size_t
n);
Description Identical to strncmp except that the string s2 can be in flash or data memory.
strncmp_P
Syntax int strncmp_P(const char *s1, PGM_P s2, size_t n);
Description Identical to strncmp except that the string s2 is in flash memory, not in data memory.
strncpy_G
Syntax char *strncpy_G(char *s1, const char __generic *s2, size_t n);
Description Identical to strncpy except that the source string s2 can be in flash or data memory.
strncpy_P
Syntax char * strncpy_P(char *s1, PGM_P s2, size_t n);
Description Identical to strncpy except that the source string s2 is in flash memory, not in data
memory.

374 for AVR
AFE1_AFE2-1:1
Segment reference
● Summary of segments
● Descriptions of segments
For more information about placement of segments, see the chapter Linking
your application.
Summary of segments
The table below lists the segments that are available in the compiler:
Segment Description
CHECKSUM Holds the checksum generated by the linker.
CODE Holds the program code.
CSTACK Holds the stack used by C or C++ programs.
DIFUNCT Holds pointers to code, typically C++ constructors, that should be
executed by the system startup code before main is called.
EEPROM_I Holds static and global initialized variables that have the memory
attribute __eeprom.
EEPROM_N Holds __no_initstatic and global variables that have the memory
attribute __eeprom.
FARCODE Holds program code declared __farfunc.
FAR_C Holds constant data that has the memory attribute __far.
FAR_F Holds static and global variables that have the memory attribute
__farflash.
FAR_HEAP Holds the far heap.
FAR_I Holds the static and global initialized variables that have the memory
attribute __far.
FAR_ID Holds initial values for static and global initialized variables in FAR_I.
FAR_N Holds __no_init static and global variables that have the memory
attribute __far.
FAR_Z Holds zero-initialized static and global variables that have the memory
attribute __far.
HEAP Holds the heap used for dynamically allocated data using the CLIB library.
Table 45: Segment summary
375
AFE1_AFE2-1:1
Summary of segments
Segment Description
HUGE_C Holds constant data that has the memory attribute__huge.
HUGE_F Holds static and global __hugeflash variables.
HUGE_HEAP Holds the heap used for dynamically allocated data in huge memory using
the DLIB library.
HUGE_I Holds static and global initialized variables that have the memory
attribute__huge.
HUGE_ID Holds initial values for static and global initialized variables in HUGE_I.
HUGE_N Holds __no_initstatic and global variables that have the memory
attribute__huge.
HUGE_Z Holds zero-initialized static and global variables that have the memory
attribute __huge.
INITTAB Contains compiler-generated table entries that describe the segment
initialization that will be performed at system startup.
INTVEC Contains the reset and interrupt vectors.
NEAR_C Holds constant data that has the memory attribute __tiny and
__near.
NEAR_F Holds static and global __flash variables.
NEAR_HEAP Holds the heap used for dynamically allocated data in near memory using
the DLIB library.
NEAR_I Holds static and global initialized variables that have the memory
attribute __near.
NEAR_ID Holds initial values for static and global initialized variables in NEAR_I.
NEAR_N Holds __no_initstatic and global variables that have the memory
attribute __near.
NEAR_Z Holds zero-initialized static and global variables that have the memory
attribute __near.
RSTACK Holds the internal return stack.
SWITCH Holds switch tables for all functions.
TINY_F Holds static and global __tinyflash variables.
TINY_HEAP Holds the heap used for dynamically allocated data in tiny memory using
the DLIB library.
TINY_I Holds static and global initialized variables that have the memory
attribute __tiny.
TINY_ID Holds initial values for static and global variables in TINY_I.
Table 45: Segment summary (Continued)

376 for AVR
AFE1_AFE2-1:1
Segment reference
Segment Description
TINY_N Holds __no_init static and global variables that have the memory
attribute__tiny.
TINY_Z Holds zero-initialized static and global variables that have the memory
attribute __tiny.
Table 45: Segment summary (Continued)
Descriptions of segments
This section gives reference information about each segment.
The segments are placed in memory by the segment placement linker directives -Z and
-P, for sequential and packed placement, respectively. Some segments cannot use
packed placement, as their contents must be continuous. For information about these
directives, see Using the -Z command for sequential placement, page 89 and Using the
-P command for packed placement, page 89, respectively.
For each segment, the segment memory type is specified, which indicates in which type
of memory the segment should be placed; see Segment memory type, page 72.
For information about how to define segments in the linker configuration file, see
Linking your application, page 87.
For more information about the extended keywords mentioned here, see the chapter
Extended keywords.
CHECKSUM
Description Holds the checksum bytes generated by the linker. This segment also holds the
__checksum symbol. Note that the size of this segment is affected by the linker option
-J.
Segment memory type CODE
Memory placement This segment can be placed anywhere in ROM memory.
Access type Read-only
377
AFE1_AFE2-1:1
CODE
Description Holds __nearfunc program code, except the code for system initialization.
Memory placement Flash. This segment must be placed within the address range 0x0–0x01FFFE.
CSTACK
Description Holds the internal data stack.
Segment memory type DATA
Memory placement Data. The address range depends on the memory model:
In the Tiny memory model, this segment must be placed within the address range
0x0-0xFF.
In the Small memory model, this segment must be placed within the address range
0x0-0xFFFF.
In the Large and Huge memory models, this segment must be placed within the address
range 0x0-0xFFFFFF. In the Large and Huge memory models, the stack can be a
maximum of 64 Kbytes, and CSTACK must not cross a 64-Kbyte boundary.
Access type Read-write
See also The stack, page 60 and RSTACK, page 388.
DIFUNCT
Description Holds the dynamic initialization vector used by C++.
Memory placement This segment must be placed in the first 64 Kbytes of memory.

378 for AVR
AFE1_AFE2-1:1
Segment reference
EEPROM_I
Description Holds __eeprom static and global initialized variables initialized by copying from the
segment EEPROM_ID at application startup.
This segment cannot be placed in memory by using the -P directive for packed
placement, because the contents must be continuous. Instead, when you define this
segment in the linker configuration file, the -Z directive must be used.
This segment is not copied to EEPROM during system startup. Instead it is used for
programming the EEPROM during the download of the code.
Segment memory type XDATA
Memory placement This segment must be placed in EEPROM. Use the command line option
--eeprom_size to set the address range for this segment.
See also --eeprom_size, page 251
EEPROM_N
Description Holds static and global __no_init __eeprom variables.
Segment memory type XDATA
Memory placement This segment must be placed in EEPROM. Use the command line option
--eeprom_size to set the address range for this segment.
See also --eeprom_size, page 251
FARCODE
Description Holds __farfunc program code. The __farfunc memory attribute is available when
using the -v5 and -v6 options, in which case the __farfunc is implicitly used for all
functions.
Memory placement This segment must be placed within the address range 0x0–0x7FFFFE in flash memory.
379
AFE1_AFE2-1:1
FAR_C
Description Holds __far constant data. This can include constant variables, string and aggregate
literals, etc.
Note: This segment is located in external ROM. Systems without external ROM cannot
use this segment.
Memory placement External ROM. This segment must be placed within the address range 0x0–0xFFFFFF.
FAR_F
Description Holds the static and global __farflash variables and aggregate initializers.
Memory placement Flash. This segment must be placed within the address range 0x0–0x7FFFFF.
FAR_HEAP
Description Holds the heap used for dynamically allocated data in far memory, in other words data
allocated by far_malloc and far_free, and in C++, new and delete.
Note: This segment is only used when you use the DLIB library.
Memory placement Data. This segment must be placed within the address range 0x0–0xFFFFFF.
See also Setting up heap memory, page 92 and New and Delete operators, page 181.

380 for AVR
AFE1_AFE2-1:1
Segment reference
FAR_I
Description Holds __far static and global initialized variables initialized by copying from the
segment FAR_ID at application startup.
FAR_ID
Description Holds initial values for __far static and global variables in the FAR_I segment. These
values are copied from FAR_ID to FAR_I at application startup.
FAR_N
Description Holds static and global __no_init __far variables.
381
AFE1_AFE2-1:1
FAR_Z
Description Holds zero-initialized __far static and global variables. The contents of this segment
is declared by the system startup code.
HEAP
Description Holds the heap used for dynamically allocated data, in other words data allocated by
malloc and free.
Note: This segment is only used when you use the CLIB library.
Memory placement Data. The address range depends on the memory model.
In the Tiny memory model, this segment must be placed within the address range
0x0-0xFF.
In the Small memory model, this segment must be placed within the address range
0x0-0xFFFF.
In the Large and Huge memory models, this segment must be placed within the address
range 0x0-0xFFFFFF.
HUGE_C
Description Holds __huge constant data. This can include constant variables, string and aggregate
literals, etc.

382 for AVR
AFE1_AFE2-1:1
Segment reference
use this segment.
Memory placement External ROM. This segment must be placed within the address range 0x0–0xFFFFFF.
HUGE_F
Description Holds the static and global __hugeflash variables and aggregate initializers.
Memory placement Flash. This segment must be placed within the address range 0x0–0xFFFFFF.
HUGE_HEAP
Description Holds the heap used for dynamically allocated data in huge memory, in other words data
allocated by huge_malloc and huge_free, and in C++, new and delete.
HUGE_I
Description Holds __huge static and global initialized variables initialized by copying from the
segment HUGE_ID at application startup.
383
AFE1_AFE2-1:1
When the -y compiler option is used, __huge constant data is located in this segment.
HUGE_ID
Description Holds initial values for __huge static and global variables in the HUGE_I segment.
These values are copied from HUGE_ID to HUGE_I at application startup.
HUGE_N
Description Holds static and global __no_init __huge variables.
HUGE_Z
Description Holds zero-initialized __huge static and global variables. The contents of this segment

384 for AVR
AFE1_AFE2-1:1
Segment reference
INITTAB
Description Holds compiler-generated table entries that describe the segment initialization which
will be performed at system startup.
Memory placement Flash. This segment must be placed within the address range 0x0–0xFFFF (0x7FFFFF
if farflash is enabled).
INTVEC
Description Holds the interrupt vector table generated by the use of the __interrupt extended
keyword in combination with the #pragma vector directive.
Memory placement This segment can be placed anywhere in memory.
NEAR_C
Description Holds __tiny and __near constant data. This can include constant variables, string
and aggregate literals, etc. TINY_I is copied from this segment, because TINY_C is not
a possible segment.
use this segment.
Memory placement External ROM. This segment must be placed within the address range 0x0–0xFFFF.
385
AFE1_AFE2-1:1
NEAR_F
Description Holds the static and global __flash variables and aggregate initializers.
Memory placement Flash. This segment must be placed within the address range 0x0–0xFFFF.
NEAR_HEAP
Description Holds the heap used for dynamically allocated data in near memory, in other words data
allocated by near_malloc and near_free, and in C++, new and delete.
Memory placement Data. This segment must be placed within the address range 0x0–0xFFFF.
NEAR_I
Description Holds __near static and global initialized variables initialized by copying from the
segment NEAR_ID at application startup.
When the -y compiler option is used, NEAR_C data (__near) is located in this segment.

386 for AVR
AFE1_AFE2-1:1
Segment reference
NEAR_ID
Description Holds initial values for __near static and global variables in the NEAR_I segment.
These values are copied from NEAR_ID to NEAR_I at application startup.
NEAR_N
Description Holds static and global __no_init __near variables.
NEAR_Z
Description Holds zero-initialized __near static and global variables. The contents of this segment
387
AFE1_AFE2-1:1
RSTACK
Description Holds the internal return stack.
See also CSTACK, page 378
SWITCH
Description Holds switch tables for all functions.
The SWITCH segment is for compiler internal use only and should always be defined.
The segment allocates, if necessary, jump tables for C/C++ switch statements.
Memory placement Flash. This segment must be placed within the address range 0x0–0xFFFF. If the
__farflash extended keyword and the --enhanced_core option are used, the
segment must be placed within the range 0x0–0x7FFFFF. This segment must not cross
a 64-Kbyte boundary.
TINY_F
Description Holds the static and global __tinyflash variables and aggregate initializers.
Memory placement Flash. This segment must be placed within the address range 0x0–0xFF.

388 for AVR
AFE1_AFE2-1:1
Segment reference
TINY_HEAP
Description Holds the heap used for dynamically allocated data in tiny memory, in other words data
allocated by tiny_malloc and tiny_free, and in C++, new and delete.
Memory placement Data. This segment must be placed within the address range 0x0–0xFF.
TINY_I
Description Holds __tiny static and global initialized variables initialized by copying from the
segment TINY_ID at application startup.
When the -y compiler option is used, NEAR_C data is located in this segment.
TINY_ID
Description Holds initial values for __tiny static and global variables in the TINY_I segment.
These values are copied from TINY_ID to TINY_I at application startup.
389
AFE1_AFE2-1:1
TINY_N
Description Holds static and global __no_init __tiny variables.
TINY_Z
Description Holds zero-initialized __tiny static and global variables. The contents of this segment

390 for AVR
AFE1_AFE2-1:1
The stack usage control file
● Overview
● Stack usage control directives
● Syntactic components
Before you read this chapter, see Stack usage analysis, page 79.
Overview
A stack usage control file consists of a sequence of directives that control stack usage
analysis. You can use C ("/*...*/") and C++ ("//...") comments in these files.
The default filename extension for stack usage control files is suc.
C++ NAMES
You can also use wildcards in function names. "#*" matches any sequence of characters,
and "#?" matches a single character.
Stack usage control directives

This section gives detailed reference information about each stack usage control
directive.
call graph root directive

Syntax call graph root [ category ] : func-spec [, func-spec... ];
Parameters
category See category, page 395
func-spec See func-spec, page 395
Description Specifies that the listed functions are call graph roots. You can optionally specify a call
graph root category. Call graph roots are listed under their category in the Stack Usage
chapter in the linker map file.
391
AFE1_AFE2-1:1
The linker will normally issue a warning for functions needed in the application that are
not call graph roots and which do not appear to be called.
Example call graph root [task]: MyFunc10, MyFunc11;
See also call_graph_root, page 322.
check that directive

Syntax check that expression;
Parameters
expression A boolean expression.
Description You can use the check that directive to compare the results of stack usage analysis
against the sizes of blocks and regions. If the expression evaluates to zero, an error is
emitted.
Three extra operators are available for use only in check that expressions:
maxstack(category, stack) The stack depth of stack in the deepest call chain for
any call graph root function in the category.
totalstack(category, The sum of the stack depths of the deepest call chains
stack) for each call graph root function in the category.
size("SEGMENT") The size of the segment.
Example check that maxstack("Program entry", CSTACK)

+ totalstack("interrupt", RSTACK)
+ 1K
<= size("CSTACK");
In the example, maxstack is the deepest call chain of the program on the CSTACK stack
plus the total stack usage on the RSTACK stack plus 1024.
See also Stack usage analysis, page 79.

392 for AVR
AFE1_AFE2-1:1
exclude directive
Syntax exclude func-spec [, func-spec... ];
Parameters
Description Excludes the specified functions, and call trees originating with them, from stack usage
calculations.
Example exclude MyFunc5, MyFunc6;
function directive
Syntax [ override ] function [ category ] func-spec : stack-size
[ , call-info... ];
Parameters
category See category, page 395
call-info See call-info, page 396
stack-size See stack-size, page 397
Description Specifies what the maximum stack usage is in a function and which other functions that
are called from that function.
Normally, an error is issued if there already is stack usage information for the function,
but if you start with override, the error will be suppressed and the information
supplied in the directive will be used instead of the previous information.
Example function MyOtherFunc: (CSTACK 8, RSTACK 2);

function [interrupt] MyInterruptHandler: (CSTACK 44, RSTACK 4);
max recursion depth directive

Syntax max recursion depth func-spec : size;
Parameters
393
AFE1_AFE2-1:1
size See size, page 397
Description Specifies the maximum number of iterations through any of the cycles in the recursion
nest of which the function is a member.
A recursion nest is a set of cycles in the call graph where each cycle shares at least one
node with another cycle in the nest.
Stack usage analysis will base its result on the max recursion depth multiplied by the
stack usage of the deepest cycle in the nest. If the nest is not entered on a point along
one of the deepest cycles, no stack usage result will be calculated for such calls.
Example max recursion depth MyFunc12: 10;
no calls from directive

Syntax no calls from module-spec to func-spec [, func-spec... ];
Parameters
module-spec See module-spec, page 396
Description When you provide stack usage information for some functions in a module without
stack usage information, the linker warns about functions that are referenced from the
module but not listed as called. This is primarily to help avoid problems with C runtime
routines, calls to which are generated by the compiler, beyond user control.
If there actually is no call to some of these functions, use the no calls from directive
to selectively suppress the warning for the specified functions. You can also disable the
warning entirely (--diag_suppress or
Project>Options>Linker>Diagnostics>Suppress these diagnostics).
Example no calls from [file.r90] to MyFunc13, MyFun14;
possible calls directive

Syntax possible calls calling-func : called-func [ , called-func... ];
Parameters
calling-func See func-spec, page 395
called-func See func-spec, page 395

394 for AVR
AFE1_AFE2-1:1
Description Specifies an exhaustive list of possible destinations for all indirect calls in one function.
Use this for functions which are known to perform indirect calls and where you know
exactly which functions that might be called in this particular application. Consider
using the #pragma calls directive if the information about which functions that might
be called is available when compiling.
Example possible calls MyFunc7: MyFunc8, MyFunc9;
When the function does not perform any calls, the list is empty:
possible calls MyFunc8: ;
See also calls, page 322.
Syntactic components
This section describes the syntactical components that can be used by the stack usage
control directives.
category
Syntax [ name ]
Description A call graph root category. You can use any name you like. Categories are not
case-sensitive.
Example category examples:

[interrupt]
[task]
func-spec
Syntax [ ? ] name [ module-spec ]
Description Specifies the name of a symbol, and for module-local symbols, the name of the module
it is defined in. Normally, if func-spec does not match a symbol in the program, a
warning is emitted. Prefixing with ? suppresses this warning.
Example func-spec examples:

xFun
MyFun [file.r90]
395
AFE1_AFE2-1:1
module-spec
Syntax [name [ (name) ]]
Description Specifies the name of a module, and optionally, in parentheses, the name of the library
it belongs to. To distinguish between modules with the same name, you can specify:
● The complete path of the file ("D:\C1\test\file.o")
● As many path elements as are needed at the end of the path ("test\file.o")
● Some path elements at the start of the path, followed by "...", followed by some
path elements at the end ("D:\...\file.o").
Note that when using multi-file compilation (--mfc), multiple files are compiled into a
single module, named after the first file.
Example module-spec examples:
[file.r90]
[file.r90(lib.a)]
["D:\C1\test\file.r90"]
name
Description A name can be either an identifier or a quoted string.
The first character of an identifier must be either a letter or one of the characters "_",
"$", or ".". The rest of the characters can also be digits.
A quoted string starts and ends with " and can contain any character. Two consecutive
" characters can be used inside a quoted string to represent a single ".
Example name examples:
MyFun
file.r90
"file-1.r90"
call-info
Syntax calls func-spec [ , func-spec... ][ : stack-size ]
Description Specifies one or more called functions, and optionally, the stack size at the calls.

396 for AVR
AFE1_AFE2-1:1
Example call-info examples:
calls MyFunc1 : stack 16

calls MyFunc2, MyFunc3, MyFunc4
stack-size
Syntax (CSTACK size, RSTACK size)
Description Specifies the size of a stack frame.
Example stack-size examples:
(CSTACK 16, RSTACK 4)
size
Description A decimal integer, or 0x followed by a hexadecimal integer. Either alternative can
optionally be followed by a suffix indicating a power of two (K=210, M=220, G=230,
T=240, P=250).
Example size examples:
24
0x18
2048
2K
397
AFE1_AFE2-1:1

398 for AVR
AFE1_AFE2-1:1
Implementation-defined
behavior for Standard C
● Descriptions of implementation-defined behavior
If you are using C89 instead of Standard C, see Implementation-defined behavior

for C89, page 415. For a short overview of the differences between Standard
C and C89, see C language overview, page 167.
The text in this chapter applies to the DLIB runtime environment. Because the
CLIB runtime environment does not follow Standard C, its
implementation-defined behavior is not documented. See also The CLIB runtime
environment, page 141.
Descriptions of implementation-defined behavior

This section follows the same order as the C standard. Each item includes references to
the ISO chapter and section (in parenthesis) that explains the implementation-defined
behavior.
Note: The IAR Systems implementation adheres to a freestanding implementation of
Standard C. This means that parts of a standard library can be excluded in the
implementation.
J.3.1 TRANSLATION
Diagnostics (3.10, 5.1.1.3)

Diagnostics are produced in the form:
where filename is the name of the source file in which the error was encountered,
linenumber is the line number at which the compiler detected the error, level is the
level of seriousness of the message (remark, warning, error, or fatal error), tag is a
unique tag that identifies the message, and message is an explanatory message, possibly
several lines.
399
AFE1_AFE2-1:1
White-space characters (5.1.1.2)

At translation phase three, each non-empty sequence of white-space characters is
retained.
J.3.2 ENVIRONMENT
The character set (5.1.1.2)

The source character set is the same as the physical source file multibyte character set.
By default, the standard ASCII character set is used. However, if you use the
--enable_multibytes compiler option, the host character set is used instead.
Main (5.1.2.1)
The function called at program startup is called main. No prototype is declared for
main, and the only definition supported for main is:
int main(void)
To change this behavior for the DLIB runtime environment, see Customizing
__low_level_init, page 122.
The effect of program termination (5.1.2.1)

Terminating the application returns the execution to the startup code (just after the call
to main).
Alternative ways to define main (5.1.2.2.1)

There is no alternative ways to define the main function.
The argv argument to main (5.1.2.2.1)

The argv argument is not supported.
Streams as interactive devices (5.1.2.3)

The streams stdin, stdout, and stderr are treated as interactive devices.
Signals, their semantics, and the default handling (7.14)

In the DLIB runtime environment, the set of supported signals is the same as in Standard
C. A raised signal will do nothing, unless the signal function is customized to fit the
application.

400 for AVR
AFE1_AFE2-1:1
Implementation-defined behavior for Standard C
Signal values for computational exceptions (7.14.1.1)

In the DLIB runtime environment, there are no implementation-defined values that
correspond to a computational exception.
Signals at system startup (7.14.1.1)

In the DLIB runtime environment, there are no implementation-defined signals that are
executed at system startup.
Environment names (7.20.4.5)

In the DLIB runtime environment, there are no implementation-defined environment
names that are used by the getenv function.
The system function (7.20.4.6)

The system function is not supported.
J.3.3 IDENTIFIERS
Multibyte characters in identifiers (6.4.2)

Additional multibyte characters may not appear in identifiers.
Significant characters in identifiers (5.2.4.1, 6.1.2)

The number of significant initial characters in an identifier with or without external
linkage is guaranteed to be no less than 200.
J.3.4 CHARACTERS
Number of bits in a byte (3.6)

A byte contains 8 bits.
Execution character set member values (5.2.1)

The values of the members of the execution character set are the values of the ASCII
character set, which can be augmented by the values of the extra characters in the host
character set.
Alphabetic escape sequences (5.2.2)

The standard alphabetic escape sequences have the values \a–7, \b–8, \f–12, \n–10,
\r–13, \t–9, and \v–11.
401
AFE1_AFE2-1:1
Characters outside of the basic executive character set (6.2.5)

A character outside of the basic executive character set that is stored in a char is not
transformed.
Plain char (6.2.5, 6.3.1.1)

A plain char is treated as an unsigned char.
Source and execution character sets (6.4.4.4, 5.1.1.2)

The source character set is the set of legal characters that can appear in source files. By
default, the source character set is the standard ASCII character set. However, if you use
the command line option --enable_multibytes, the source character set will be the
host computer’s default character set.
The execution character set is the set of legal characters that can appear in the execution
environment. By default, the execution character set is the standard ASCII character set.
However, if you use the command line option --enable_multibytes, the execution
character set will be the host computer’s default character set. The DLIB runtime
environment needs a multibyte character scanner to support a multibyte execution
character set. See Locale, page 133.
Integer character constants with more than one character (6.4.4.4)

An integer character constant that contains more than one character will be treated as an
integer constant. The value will be calculated by treating the leftmost character as the
most significant character, and the rightmost character as the least significant character,
in an integer constant. A diagnostic message will be issued if the value cannot be
represented in an integer constant.
Wide character constants with more than one character (6.4.4.4)

A wide character constant that contains more than one multibyte character generates a
diagnostic message.
Locale used for wide character constants (6.4.4.4)

By default, the C locale is used. If the --enable_multibytes compiler option is used,
the default host locale is used instead.
Locale used for wide string literals (6.4.5)

By default, the C locale is used. If the --enable_multibytes compiler option is used,
the default host locale is used instead.

402 for AVR
AFE1_AFE2-1:1
Source characters as executive characters (6.4.5)

All source characters can be represented as executive characters.
J.3.5 INTEGERS
Extended integer types (6.2.5)

There are no extended integer types.
Range of integer values (6.2.6.2)

The representation of integer values are in the two's complement form. The most
significant bit holds the sign; 1 for negative, 0 for positive and zero.
For information about the ranges for the different integer types, see Basic data types—
integer types, page 282.
The rank of extended integer types (6.3.1.1)

There are no extended integer types.
Signals when converting to a signed integer type (6.3.1.3)

No signal is raised when an integer is converted to a signed integer type.
Signed bitwise operations (6.5)

Bitwise operations on signed integers work the same way as bitwise operations on
unsigned integers; in other words, the sign-bit will be treated as any other bit, except for
the operator >> which will behave as an arithmetic right shift.
J.3.6 FLOATING POINT
Accuracy of floating-point operations (5.2.4.2.2)

The accuracy of floating-point operations is unknown.
Rounding behaviors (5.2.4.2.2)

There are no non-standard values of FLT_ROUNDS.
Evaluation methods (5.2.4.2.2)

There are no non-standard values of FLT_EVAL_METHOD.
403
AFE1_AFE2-1:1
Converting integer values to floating-point values (6.3.1.4)

When an integral value is converted to a floating-point value that cannot exactly
represent the source value, the round-to-nearest rounding mode is used (FLT_ROUNDS is
defined to 1).
Converting floating-point values to floating-point values (6.3.1.5)

When a floating-point value is converted to a floating-point value that cannot exactly
represent the source value, the round-to-nearest rounding mode is used (FLT_ROUNDS is
defined to 1).
Denoting the value of floating-point constants (6.4.4.2)

The round-to-nearest rounding mode is used (FLT_ROUNDS is defined to 1).
Contraction of floating-point values (6.5)

Floating-point values are contracted. However, there is no loss in precision and because
signaling is not supported, this does not matter.
Default state of FENV_ACCESS (7.6.1)

The default state of the pragma directive FENV_ACCESS is OFF.
Additional floating-point mechanisms (7.6, 7.12)

There are no additional floating-point exceptions, rounding-modes, environments, and
classifications.
Default state of FP_CONTRACT (7.12.2)

The default state of the pragma directive FP_CONTRACT is OFF.
J.3.7 ARRAYS AND POINTERS
Conversion from/to pointers (6.3.2.3)

For information about casting of data pointers and function pointers, see Casting, page
289.
ptrdiff_t (6.5.6)
For information about ptrdiff_t, see ptrdiff_t, page 289.

404 for AVR
AFE1_AFE2-1:1
J.3.8 HINTS
Honoring the register keyword (6.7.1)

User requests for register variables are not honored.
Inlining functions (6.7.4)

User requests for inlining functions increases the chance, but does not make it certain,
that the function will actually be inlined into another function. See Inlining functions,
page 67.
J.3.9 STRUCTURES, UNIONS, ENUMERATIONS, AND

BITFIELDS
Sign of 'plain' bitfields (6.7.2, 6.7.2.1)

For information about how a 'plain' int bitfield is treated, see Bitfields, page 284.
Possible types for bitfields (6.7.2.1)

All integer types can be used as bitfields in the compiler’s extended mode, see -e, page
250.
Bitfields straddling a storage-unit boundary (6.7.2.1)

A bitfield is always placed in one—and one only—storage unit, which means that the
bitfield cannot straddle a storage-unit boundary.
Allocation order of bitfields within a unit (6.7.2.1)

For information about how bitfields are allocated within a storage unit, see Bitfields,
page 284.
Alignment of non-bitfield structure members (6.7.2.1)

The alignment of non-bitfield members of structures is the same as for the member
types, see Alignment, page 281.
Integer type used for representing enumeration types (6.7.2.2)

The chosen integer type for a specific enumeration type depends on the enumeration
constants defined for the enumeration type. The chosen integer type is the smallest
possible.
405
AFE1_AFE2-1:1
J.3.10 QUALIFIERS
Access to volatile objects (6.7.3)

Any reference to an object with volatile qualified type is an access, see Declaring
objects volatile, page 291.
J.3.11 PREPROCESSING DIRECTIVES
Mapping of header names (6.4.7)

Sequences in header names are mapped to source file names verbatim. A backslash '\'
is not treated as an escape sequence. See Overview of the preprocessor, page 351.
Character constants in constant expressions (6.10.1)

A character constant in a constant expression that controls conditional inclusion
matches the value of the same character constant in the execution character set.
The value of a single-character constant (6.10.1)

A single-character constant may only have a negative value if a plain character (char)
is treated as a signed character, see --char_is_signed, page 241.
Including bracketed filenames (6.10.2)

For information about the search algorithm used for file specifications in angle brackets
<>, see Include file search procedure, page 228.
Including quoted filenames (6.10.2)

For information about the search algorithm used for file specifications enclosed in
quotes, see Include file search procedure, page 228.
Preprocessing tokens in #include directives (6.10.2)

Preprocessing tokens in an #include directive are combined in the same way as outside
an #include directive.
Nesting limits for #include directives (6.10.2)

There is no explicit nesting limit for #include processing.
Universal character names (6.10.3.2)

Universal character names (UCN) are not supported.

406 for AVR
AFE1_AFE2-1:1
Recognized pragma directives (6.10.6)

In addition to the pragma directives described in the chapter Pragma directives, the
following directives are recognized and will have an indeterminate effect. If a pragma
directive is listed both in the chapter Pragma directives and here, the information
provided in the chapter Pragma directives overrides the information here.
alignment
baseaddr
building_runtime
can_instantiate
codeseg
cspy_support
define_type_info
do_not_instantiate
early_dynamic_initialization
function
function_effects
hdrstop
important_typedef
instantiate
keep_definition
library_default_requirements
library_provides
library_requirement_override
memory
module_name
no_pch
once
system_include
warnings
407
AFE1_AFE2-1:1
Default __DATE__ and __TIME__ (6.10.8)

The definitions for __TIME__ and __DATE__ are always available.
J.3.12 LIBRARY FUNCTIONS
Additional library facilities (5.1.2.1)

Most of the standard library facilities are supported. Some of them—the ones that need
an operating system—require a low-level implementation in the application. For more
information, see The DLIB runtime environment, page 99.
Diagnostic printed by the assert function (7.2.1.1)

The assert() function prints:
filename:linenr expression -- assertion failed
when the parameter evaluates to zero.
Representation of the floating-point status flags (7.6.2.2)

For information about the floating-point status flags, see fenv.h, page 367.
Feraiseexcept raising floating-point exception (7.6.2.3)

For information about the feraiseexcept function raising floating-point exceptions,
see Floating-point environment, page 286.
Strings passed to the setlocale function (7.11.1.1)

For information about strings passed to the setlocale function, see Locale, page 133.
Types defined for float_t and double_t (7.12)

The FLT_EVAL_METHOD macro can only have the value 0.
Domain errors (7.12.1)

No function generates other domain errors than what the standard requires.
Return values on domain errors (7.12.1)

Mathematic functions return a floating-point NaN (not a number) for domain errors.
Underflow errors (7.12.1)

Mathematic functions set errno to the macro ERANGE (a macro in errno.h) and return
zero for underflow errors.

408 for AVR
AFE1_AFE2-1:1
fmod return value (7.12.10.1)

The fmod function returns a floating-point NaN when the second argument is zero.
The magnitude of remquo (7.12.10.3)

The magnitude is congruent modulo INT_MAX.
signal() (7.14.1.1)
The signal part of the library is not supported.
Note: The default implementation of signal does not perform anything. Use the
template source code to implement application-specific signal handling. See signal,
page 129 and raise, page 127, respectively.
NULL macro (7.17)

The NULL macro is defined to 0.
Terminating newline character (7.19.2)

stdout stream functions recognize either newline or end of file (EOF) as the
terminating character for a line.
Space characters before a newline character (7.19.2)

Space characters written to a stream immediately before a newline character are
preserved.
Null characters appended to data written to binary streams (7.19.2)

No null characters are appended to data written to binary streams.
File position in append mode (7.19.3)

The file position is initially placed at the beginning of the file when it is opened in
append-mode.
Truncation of files (7.19.3)

Whether a write operation on a text stream causes the associated file to be truncated
beyond that point, depends on the application-specific implementation of the low-level
file routines. See Briefly about input and output (I/O), page 100.
File buffering (7.19.3)

An open file can be either block-buffered, line-buffered, or unbuffered.
409
AFE1_AFE2-1:1
A zero-length file (7.19.3)

Whether a zero-length file exists depends on the application-specific implementation of
the low-level file routines.
Legal file names (7.19.3)

The legality of a filename depends on the application-specific implementation of the
low-level file routines.
Number of times a file can be opened (7.19.3)

Whether a file can be opened more than once depends on the application-specific
implementation of the low-level file routines.
Multibyte characters in a file (7.19.3)

The encoding of multibyte characters in a file depends on the application-specific
remove() (7.19.4.1)
The effect of a remove operation on an open file depends on the application-specific
implementation of the low-level file routines. See Briefly about input and output (I/O),
page 100.
rename() (7.19.4.2)
The effect of renaming a file to an already existing filename depends on the
application-specific implementation of the low-level file routines. See Briefly about
input and output (I/O), page 100.
Removal of open temporary files (7.19.4.3)

Whether an open temporary file is removed depends on the application-specific
Mode changing (7.19.5.4)

freopen closes the named stream, then reopens it in the new mode. The streams stdin,
stdout, and stderr can be reopened in any new mode.
Style for printing infinity or NaN (7.19.6.1, 7.24.2.1)

The style used for printing infinity or NaN for a floating-point constant is inf and nan
(INF and NAN for the F conversion specifier), respectively. The n-char-sequence is not
used for nan.

410 for AVR
AFE1_AFE2-1:1
%p in printf() (7.19.6.1, 7.24.2.1)

The argument to a %p conversion specifier, print pointer, to printf() is treated as
having the type void *. The value will be printed as a hexadecimal number, similar to
using the %x conversion specifier.
Reading ranges in scanf (7.19.6.2, 7.24.2.1)

A - (dash) character is always treated as a range symbol.
%p in scanf (7.19.6.2, 7.24.2.2)

The %p conversion specifier, scan pointer, to scanf() reads a hexadecimal number and
converts it into a value with the type void *.
File position errors (7.19.9.1, 7.19.9.3, 7.19.9.4)

On file position errors, the functions fgetpos, ftell, and fsetpos store EFPOS in
errno.
An n-char-sequence after NaN (7.20.1.3, 7.24.4.1.1)

An n-char-sequence after a NaN is read and ignored.
errno value at underflow (7.20.1.3, 7.24.4.1.1)

errno is set to ERANGE if an underflow is encountered.
Zero-sized heap objects (7.20.3)

A request for a zero-sized heap object will return a valid pointer and not a null pointer.
Behavior of abort and exit (7.20.4.1, 7.20.4.4)

A call to abort() or _Exit() will not flush stream buffers, not close open streams, and
not remove temporary files.
Termination status (7.20.4.1, 7.20.4.3, 7.20.4.4)

The termination status will be propagated to __exit() as a parameter. exit() and
_Exit() use the input parameter, whereas abort uses EXIT_FAILURE.
The system function return value (7.20.4.6)

The system function is not supported.
411
AFE1_AFE2-1:1
The time zone (7.23.1)

The local time zone and daylight savings time must be defined by the application. For
more information, see __time32, __time64, page 130.
Range and precision of time (7.23)

For information about range and precision, see time.h, page 368. The application must
supply the actual implementation for the functions time and clock. See __time32,
__time64, page 130 and clock, page 124, respectively.
clock() (7.23.2.1)
The application must supply an implementation of the clock function. See clock, page
124.
%Z replacement string (7.23.3.5, 7.24.5.1)

By default, ":" is used as a replacement for %Z. Your application should implement the
time zone handling. See __time32, __time64, page 130.
Math functions rounding mode (F.9)

The functions in math.h honor the rounding direction mode in FLT-ROUNDS.
J.3.13 ARCHITECTURE
Values and expressions assigned to some macros (5.2.4.2, 7.18.2,

7.18.3)
There are always 8 bits in a byte.
MB_LEN_MAX is at the most 6 bytes depending on the library configuration that is used.
For information about sizes, ranges, etc for all basic types, see Data representation, page
281.
The limit macros for the exact-width, minimum-width, and fastest minimum-width
integer types defined in stdint.h have the same ranges as char, short, int, long,
and long long.
The floating-point constant FLT_ROUNDS has the value 1 (to nearest) and the
floating-point constant FLT_EVAL_METHOD has the value 0 (treat as is).
The number, order, and encoding of bytes (6.2.6.1)

See Data representation, page 281.

412 for AVR
AFE1_AFE2-1:1
The value of the result of the sizeof operator (6.5.3.4)

See Data representation, page 281.
J.4 LOCALE
Members of the source and execution character set (5.2.1)

By default, the compiler accepts all one-byte characters in the host’s default character
set. If the compiler option --enable_multibytes is used, the host multibyte
characters are accepted in comments and string literals as well.
The meaning of the additional character set (5.2.1.2)

Any multibyte characters in the extended source character set is translated verbatim into
the extended execution character set. It is up to your application with the support of the
library configuration to handle the characters correctly.
Shift states for encoding multibyte characters (5.2.1.2)

Using the compiler option --enable_multibytes enables the use of the host’s default
multibyte characters as extended source characters.
Direction of successive printing characters (5.2.2)

The application defines the characteristics of a display device.
The decimal point character (7.1.1)

The default decimal-point character is a '.'. You can redefine it by defining the library
configuration symbol _LOCALE_DECIMAL_POINT.
Printing characters (7.4, 7.25.2)

The set of printing characters is determined by the chosen locale.
Control characters (7.4, 7.25.2)

The set of control characters is determined by the chosen locale.
Characters tested for (7.4.1.2, 7.4.1.3, 7.4.1.7, 7.4.1.9, 7.4.1.10,

7.4.1.11, 7.25.2.1.2, 7.25.5.1.3, 7.25.2.1.7, 7.25.2.1.9, 7.25.2.1.10,
7.25.2.1.11)
The sets of characters tested are determined by the chosen locale.
413
AFE1_AFE2-1:1
The native environment (7.1.1.1)

The native environment is the same as the “C” locale.
Subject sequences for numeric conversion functions (7.20.1,

7.24.4.1)
There are no additional subject sequences that can be accepted by the numeric
conversion functions.
The collation of the execution character set (7.21.4.3, 7.24.4.4.2)

The collation of the execution character set is determined by the chosen locale.
Message returned by strerror (7.21.6.2)

The messages returned by the strerror function depending on the argument is:
Argument Message
EZERO no error
EDOM domain error
ERANGE range error
EFPOS file positioning error
EILSEQ multi-byte encoding error
<0 || >99 unknown error
all others error nnn
Table 46: Message returned by strerror()—DLIB runtime environment

414 for AVR
AFE1_AFE2-1:1
Implementation-defined
behavior for C89
● Descriptions of implementation-defined behavior
If you are using Standard C instead of C89, see Implementation-defined behavior

for Standard C, page 399. For a short overview of the differences between
Standard C and C89, see C language overview, page 167.

The descriptions follow the same order as the ISO appendix. Each item covered includes
references to the ISO chapter and section (in parenthesis) that explains the
implementation-defined behavior.
TRANSLATION
Diagnostics (5.1.1.3)
Diagnostics are produced in the form:
where filename is the name of the source file in which the error was encountered,
linenumber is the line number at which the compiler detected the error, level is the
level of seriousness of the message (remark, warning, error, or fatal error), tag is a
unique tag that identifies the message, and message is an explanatory message, possibly
several lines.
ENVIRONMENT
Arguments to main (5.1.2.2.2.1)

The function called at program startup is called main. No prototype was declared for
main, and the only definition supported for main is:
int main(void)
To change this behavior for the DLIB runtime environment, see Customizing
__low_level_init, page 122. To change this behavior for the CLIB runtime environment,
see Customizing system initialization, page 147.
415
AFE1_AFE2-1:1
Interactive devices (5.1.2.3)

The streams stdin and stdout are treated as interactive devices.
IDENTIFIERS
Significant characters without external linkage (6.1.2)

The number of significant initial characters in an identifier without external linkage is
200.
Significant characters with external linkage (6.1.2)

The number of significant initial characters in an identifier with external linkage is 200.
Case distinctions are significant (6.1.2)

Identifiers with external linkage are treated as case-sensitive.
CHARACTERS
Source and execution character sets (5.2.1)

The source character set is the set of legal characters that can appear in source files. The
default source character set is the standard ASCII character set. However, if you use the
command line option --enable_multibytes, the source character set will be the host
computer’s default character set.
The execution character set is the set of legal characters that can appear in the execution
environment. The default execution character set is the standard ASCII character set.
However, if you use the command line option --enable_multibytes, the execution
character set will be the host computer’s default character set. The DLIB runtime
environment needs a multibyte character scanner to support a multibyte execution
character set. The CLIB runtime environment does not support multibyte characters.
See Locale, page 133.
Bits per character in execution character set (5.2.4.2.1)

The number of bits in a character is represented by the manifest constant CHAR_BIT.
The standard include file limits.h defines CHAR_BIT as 8.
Mapping of characters (6.1.3.4)

The mapping of members of the source character set (in character and string literals) to
members of the execution character set is made in a one-to-one way. In other words, the

416 for AVR
AFE1_AFE2-1:1
Implementation-defined behavior for C89
same representation value is used for each member in the character sets except for the
escape sequences listed in the ISO standard.
Unrepresented character constants (6.1.3.4)

The value of an integer character constant that contains a character or escape sequence
not represented in the basic execution character set or in the extended character set for
a wide character constant generates a diagnostic message, and will be truncated to fit the
execution character set.
Character constant with more than one character (6.1.3.4)

An integer character constant that contains more than one character will be treated as an
integer constant. The value will be calculated by treating the leftmost character as the
most significant character, and the rightmost character as the least significant character,
in an integer constant. A diagnostic message will be issued if the value cannot be
represented in an integer constant.
A wide character constant that contains more than one multibyte character generates a
diagnostic message.
Converting multibyte characters (6.1.3.4)

The only locale supported—that is, the only locale supplied with the IAR C/C++
Compiler—is the ‘C’ locale. If you use the command line option
--enable_multibytes, the DLIB runtime environment will support multibyte
characters if you add a locale with multibyte support or a multibyte character scanner to
the library. The CLIB runtime environment does not support multibyte characters.
See Locale, page 133.
Range of 'plain' char (6.2.1.1)

A ‘plain’ char has the same range as an unsigned char.
INTEGERS
Range of integer values (6.1.2.5)

The representation of integer values are in the two's complement form. The most
significant bit holds the sign; 1 for negative, 0 for positive and zero.
See Basic data types—integer types, page 282, for information about the ranges for the
different integer types.
417
AFE1_AFE2-1:1
Demotion of integers (6.2.1.2)

Converting an integer to a shorter signed integer is made by truncation. If the value
cannot be represented when converting an unsigned integer to a signed integer of equal
length, the bit-pattern remains the same. In other words, a large enough value will be
converted into a negative value.
Signed bitwise operations (6.3)

Bitwise operations on signed integers work the same way as bitwise operations on
unsigned integers; in other words, the sign-bit will be treated as any other bit, except for
the operator >> which will behave as an arithmetic right shift.
Sign of the remainder on integer division (6.3.5)

The sign of the remainder on integer division is the same as the sign of the dividend.
Negative valued signed right shifts (6.3.7)

The result of a right-shift of a negative-valued signed integral type preserves the sign-bit.
For example, shifting 0xFF00 down one step yields 0xFF80.
FLOATING POINT
Representation of floating-point values (6.1.2.5)

The representation and sets of the various floating-point numbers adheres to IEEE 854–
1987. A typical floating-point number is built up of a sign-bit (s), a biased exponent (e),
and a mantissa (m).
See Basic data types—floating-point types, page 285, for information about the ranges
and sizes for the different floating-point types: float and double.
Converting integer values to floating-point values (6.2.1.3)

When an integral number is cast to a floating-point value that cannot exactly represent
the value, the value is rounded (up or down) to the nearest suitable value.
Demoting floating-point values (6.2.1.4)

When a floating-point value is converted to a floating-point value of narrower type that
cannot exactly represent the value, the value is rounded (up or down) to the nearest
suitable value.

418 for AVR
AFE1_AFE2-1:1
ARRAYS AND POINTERS
size_t (6.3.3.4, 7.1.1)

See size_t, page 289, for information about size_t.
Conversion from/to pointers (6.3.4)

See Casting, page 289, for information about casting of data pointers and function
pointers.
ptrdiff_t (6.3.6, 7.1.1)

See ptrdiff_t, page 289, for information about the ptrdiff_t.
REGISTERS
Honoring the register keyword (6.5.1)

User requests for register variables are not honored.
STRUCTURES, UNIONS, ENUMERATIONS, AND BITFIELDS
Improper access to a union (6.3.2.3)

If a union gets its value stored through a member and is then accessed using a member
of a different type, the result is solely dependent on the internal storage of the first
member.
Padding and alignment of structure members (6.5.2.1)

See the section Basic data types—integer types, page 282, for information about the
alignment requirement for data objects.
Sign of 'plain' bitfields (6.5.2.1)

A 'plain' int bitfield is treated as a signed int bitfield. All integer types are allowed
as bitfields.
Allocation order of bitfields within a unit (6.5.2.1)

Bitfields are allocated within an integer from least-significant to most-significant bit.
Can bitfields straddle a storage-unit boundary (6.5.2.1)

Bitfields cannot straddle a storage-unit boundary for the chosen bitfield integer type.
419
AFE1_AFE2-1:1
Integer type chosen to represent enumeration types (6.5.2.2)

The chosen integer type for a specific enumeration type depends on the enumeration
constants defined for the enumeration type. The chosen integer type is the smallest
possible.
QUALIFIERS
Access to volatile objects (6.5.3)

Any reference to an object with volatile qualified type is an access.
DECLARATORS
Maximum numbers of declarators (6.5.4)

The number of declarators is not limited. The number is limited only by the available
memory.
STATEMENTS
Maximum number of case statements (6.6.4.2)

The number of case statements (case values) in a switch statement is not limited. The
number is limited only by the available memory.
PREPROCESSING DIRECTIVES
Character constants and conditional inclusion (6.8.1)

The character set used in the preprocessor directives is the same as the execution
character set. The preprocessor recognizes negative character values if a 'plain' character
is treated as a signed character.
Including bracketed filenames (6.8.2)

For file specifications enclosed in angle brackets, the preprocessor does not search
directories of the parent files. A parent file is the file that contains the #include
directive. Instead, it begins by searching for the file in the directories specified on the
compiler command line.
Including quoted filenames (6.8.2)

For file specifications enclosed in quotes, the preprocessor directory search begins with
the directories of the parent file, then proceeds through the directories of any
grandparent files. Thus, searching begins relative to the directory containing the source

420 for AVR
AFE1_AFE2-1:1
file currently being processed. If there is no grandparent file and the file is not found,
the search continues as if the filename was enclosed in angle brackets.
Character sequences (6.8.2)

Preprocessor directives use the source character set, except for escape sequences. Thus,
to specify a path for an include file, use only one backslash:
#include "mydirectory\myfile"
Within source code, two backslashes are necessary:

file = fopen("mydirectory\\myfile","rt");
Recognized pragma directives (6.8.6)

In addition to the pragma directives described in the chapter Pragma directives, the
following directives are recognized and will have an indeterminate effect. If a pragma
directive is listed both in the chapter Pragma directives and here, the information
provided in the chapter Pragma directives overrides the information here.
alignment
baseaddr
building_runtime
can_instantiate
codeseg
cspy_support
define_type_info
do_not_instantiate
early_dynamic_initialization
function
function_effects
hdrstop
important_typedef
instantiate
keep_definition
library_default_requirements
library_provides
421
AFE1_AFE2-1:1
library_requirement_override
memory
module_name
no_pch
once
system_include
warnings
Default __DATE__ and __TIME__ (6.8.8)

The definitions for __TIME__ and __DATE__ are always available.
LIBRARY FUNCTIONS FOR THE IAR DLIB RUNTIME

ENVIRONMENT
Note that some items in this list only apply when file descriptors are supported by the
library configuration. See the chapter The DLIB runtime environment for more
information about runtime library configurations.
NULL macro (7.1.6)

The NULL macro is defined to 0.
Diagnostic printed by the assert function (7.2)

filename:linenr expression -- assertion failed

NaN (Not a Number) will be returned by the mathematic functions on domain errors.
Underflow of floating-point values sets errno to ERANGE (7.5.1)

The mathematics functions set the integer expression errno to ERANGE (a macro in
errno.h) on underflow range errors.
fmod() functionality (7.5.6.4)

If the second argument to fmod() is zero, the function returns NaN; errno is set to
EDOM.

422 for AVR
AFE1_AFE2-1:1
signal() (7.7.1.1)
Note: The default implementation of signal does not perform anything. Use the
template source code to implement application-specific signal handling. See signal,
page 129 and raise, page 127, respectively.

Blank lines (7.9.2)

Space characters written to the stdout stream immediately before a newline character
are preserved. There is no way to read the line through the stdin stream that was
written through the stdout stream.

No null characters are appended to data written to binary streams.
Files (7.9.3)
Whether the file position indicator of an append-mode stream is initially positioned at
the beginning or the end of the file, depends on the application-specific implementation
of the low-level file routines.
Whether a write operation on a text stream causes the associated file to be truncated
beyond that point, depends on the application-specific implementation of the low-level
file routines. See Briefly about input and output (I/O), page 100.
The characteristics of the file buffering is that the implementation supports files that are
unbuffered, line buffered, or fully buffered.
Whether a zero-length file actually exists depends on the application-specific
Rules for composing valid file names depends on the application-specific
Whether the same file can be simultaneously open multiple times depends on the
application-specific implementation of the low-level file routines.
423
AFE1_AFE2-1:1
remove() (7.9.4.1)
The effect of a remove operation on an open file depends on the application-specific
implementation of the low-level file routines. See Briefly about input and output (I/O),
page 100.
rename() (7.9.4.2)
The effect of renaming a file to an already existing filename depends on the
application-specific implementation of the low-level file routines. See Briefly about
input and output (I/O), page 100.
%p in printf() (7.9.6.1)
having the type void *. The value will be printed as a hexadecimal number, similar to
using the %x conversion specifier.
%p in scanf() (7.9.6.2)
converts it into a value with the type void *.
Reading ranges in scanf() (7.9.6.2)

A - (dash) character is always treated as a range symbol.
File position errors (7.9.9.1, 7.9.9.4)

On file position errors, the functions fgetpos and ftell store EFPOS in errno.
Message generated by perror() (7.9.10.4)

The generated message is:
usersuppliedprefix:errormessage
Allocating zero bytes of memory (7.10.3)

The calloc(), malloc(), and realloc() functions accept zero as an argument.
Memory will be allocated, a valid pointer to that memory is returned, and the memory
block can be modified later by realloc.
Behavior of abort() (7.10.4.1)

The abort() function does not flush stream buffers, and it does not handle files,
because this is an unsupported feature.

424 for AVR
AFE1_AFE2-1:1
Behavior of exit() (7.10.4.3)

The argument passed to the exit function will be the return value returned by the main
function to cstartup.
Environment (7.10.4.4)
The set of available environment names and the method for altering the environment list
is described in getenv, page 125.
system() (7.10.4.5)
How the command processor works depends on how you have implemented the system
function. See system, page 130.
Message returned by strerror() (7.11.6.2)

The messages returned by strerror() depending on the argument is:
Argument Message
EZERO no error
EDOM domain error
ERANGE range error
EFPOS file positioning error
EILSEQ multi-byte encoding error
all others error nnn
Table 47: Message returned by strerror()—DLIB runtime environment

The local time zone and daylight savings time implementation is described in __time32,
__time64, page 130.
clock() (7.12.2.1)
From where the system clock starts counting depends on how you have implemented the
clock function. See clock, page 124.
LIBRARY FUNCTIONS FOR THE CLIB RUNTIME

ENVIRONMENT
NULL macro (7.1.6)

The NULL macro is defined to (void *) 0.
425
AFE1_AFE2-1:1
Diagnostic printed by the assert function (7.2)

Assertion failed: expression, file Filename, line linenumber

HUGE_VAL, the largest representable value in a double floating-point type, will be
returned by the mathematic functions on domain errors.
Underflow of floating-point values sets errno to ERANGE (7.5.1)

The mathematics functions set the integer expression errno to ERANGE (a macro in
errno.h) on underflow range errors.
fmod() functionality (7.5.6.4)

If the second argument to fmod() is zero, the function returns zero (it does not change
the integer expression errno).
signal() (7.7.1.1)

Blank lines (7.9.2)

Space characters written to the stdout stream immediately before a newline character
are preserved. There is no way to read the line through the stdin stream that was
written through the stdout stream.

There are no binary streams implemented.
Files (7.9.3)
There are no other streams than stdin and stdout. This means that a file system is not
implemented.

426 for AVR
AFE1_AFE2-1:1
remove() (7.9.4.1)
implemented.
rename() (7.9.4.2)
implemented.
%p in printf() (7.9.6.1)
having the type 'char *'. The value will be printed as a hexadecimal number, similar
to using the %x conversion specifier.
%p in scanf() (7.9.6.2)
converts it into a value with the type 'void *'.
Reading ranges in scanf() (7.9.6.2)

A - (dash) character is always treated explicitly as a - character.
File position errors (7.9.9.1, 7.9.9.4)

implemented.
Message generated by perror() (7.9.10.4)

perror() is not supported.
Allocating zero bytes of memory (7.10.3)

The calloc(), malloc(), and realloc() functions accept zero as an argument.
Memory will be allocated, a valid pointer to that memory is returned, and the memory
block can be modified later by realloc.
Behavior of abort() (7.10.4.1)

The abort() function does not flush stream buffers, and it does not handle files,
because this is an unsupported feature.
Behavior of exit() (7.10.4.3)

The exit() function does not return.
427
AFE1_AFE2-1:1
Environment (7.10.4.4)
Environments are not supported.
system() (7.10.4.5)
The system() function is not supported.
Message returned by strerror() (7.11.6.2)

The messages returned by strerror() depending on the argument are:
Argument Message
EZERO no error
EDOM domain error
ERANGE range error
all others error No.xx
Table 48: Message returned by strerror()—CLIB runtime environment

The time zone function is not supported.
clock() (7.12.2.1)
The clock() function is not supported.

428 for AVR
AFE1_AFE2-1:1
Index
Index
single-value initialization . . . . . . . . . . . . . . . . . . . . . . . 173
A asm, __asm (language extension) . . . . . . . . . . . . . . . . . . . 151
abort assembler code
implementation-defined behavior . . . . . . . . . . . . . . . . 411 calling from C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
implementation-defined behavior in C89 (CLIB) . . . . 427 calling from C++ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
implementation-defined behavior in C89 (DLIB) . . . . 424 inserting inline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
system termination (DLIB) . . . . . . . . . . . . . . . . . . . . . 121 assembler directives
absolute location for call frame information . . . . . . . . . . . . . . . . . . . . . . 165
data, placing at (@) . . . . . . . . . . . . . . . . . . . . . . . . . . . 210 for static overlay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
language support for . . . . . . . . . . . . . . . . . . . . . . . . . . 170 using in inline assembler code . . . . . . . . . . . . . . . . . . . 152
#pragma location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 assembler instructions
address spaces, managing multiple . . . . . . . . . . . . . . . . . . . 96 inserting inline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
addressing. See memory types, data models, assembler labels
and code models default for application startup . . . . . . . . . . . . . . . . . . . . 48
algorithm (STL header file) . . . . . . . . . . . . . . . . . . . . . . . 366 making public (--public_equ) . . . . . . . . . . . . . . . . . . . . 269
alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281 assembler language interface . . . . . . . . . . . . . . . . . . . . . . 149
forcing stricter (#pragma data_alignment) . . . . . . . . . . 323 calling convention. See assembler code
of an object (__ALIGNOF__) . . . . . . . . . . . . . . . . . . . 170 assembler list file, generating . . . . . . . . . . . . . . . . . . . . . . 256
of data types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282 assembler output file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
alignment (pragma directive) . . . . . . . . . . . . . . . . . . 407, 421 asserts
__ALIGNOF__ (operator) . . . . . . . . . . . . . . . . . . . . . . . . 170 implementation-defined behavior of . . . . . . . . . . . . . . 408
anonymous structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 implementation-defined behavior of in C89, (CLIB) . . 426
anonymous symbols, creating . . . . . . . . . . . . . . . . . . . . . . 167 implementation-defined behavior of in C89, (DLIB) . . 422
ANSI C. See C89 including in application . . . . . . . . . . . . . . . . . . . . . . . . 359
application assert.h (CLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . 370
building, overview of . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 assert.h (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . 364
execution, overview of . . . . . . . . . . . . . . . . . . . . . . . . . . 44 __assignment_by_bitwise_copy_allowed, symbol used
startup and termination (CLIB) . . . . . . . . . . . . . . . . . . 145 in library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
startup and termination (DLIB) . . . . . . . . . . . . . . . . . . 118 @ (operator)
ARGFRAME (assembler directive) . . . . . . . . . . . . . . . . . 165 placing at absolute address . . . . . . . . . . . . . . . . . . . . . . 211
argv (argument), implementation-defined behavior . . . . . 400 placing in segments . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
arrays atomic operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
designated initializers in . . . . . . . . . . . . . . . . . . . . . . . 167 __monitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
hints about index type . . . . . . . . . . . . . . . . . . . . . . . . . 206 attributes
implementation-defined behavior . . . . . . . . . . . . . . . . 404 object . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
implementation-defined behavior in C89 . . . . . . . . . . . 419 type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
incomplete at end of structs . . . . . . . . . . . . . . . . . . . . . 167 auto variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
non-lvalue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 at function entrance . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
of incomplete types . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 programming hints for efficient code . . . . . . . . . . . . . . 219
using in inline assembler code . . . . . . . . . . . . . . . . . . . 152
429
DRX-2
AVR C language, overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
supported devices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 call frame information . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
in assembler list file . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
B in assembler list file (-lA) . . . . . . . . . . . . . . . . . . . . . . 256

call frame information, disabling (--no_call_frame_info) . 260
backtrace information See call frame information call graph root (stack usage control directive) . . . . . . . . . . 391
Barr, Michael . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 call stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
baseaddr (pragma directive) . . . . . . . . . . . . . . . . . . . 407, 421 callee-save registers, stored on stack . . . . . . . . . . . . . . . . . . 60
__BASE_FILE__ (predefined symbol) . . . . . . . . . . . . . . . 352 calling convention
basic_template_matching (pragma directive) . . . . . . . . . . 321 C++, requiring C linkage . . . . . . . . . . . . . . . . . . . . . . . 154
using . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 in compiler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
batch files overriding default (__version_1) . . . . . . . . . . . . . . . . . 315
error return codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230 overriding default (__version_2) . . . . . . . . . . . . . . . . . 315
for building library from command line . . . . . . . . . . . . 109 overriding default (__version_4) . . . . . . . . . . . . . . . . . 316
binary streams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409 calloc (library function) . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
binary streams in C89 (CLIB) . . . . . . . . . . . . . . . . . . . . . . 426 See also heap
binary streams in C89 (DLIB). . . . . . . . . . . . . . . . . . . . . . 423 implementation-defined behavior in C89 (CLIB) . . . . 427
bit negation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 implementation-defined behavior in C89 (DLIB) . . . . 424
bitfields calls (pragma directive). . . . . . . . . . . . . . . . . . . . . . . . . . . 322
data representation of . . . . . . . . . . . . . . . . . . . . . . . . . . 284 call_graph_root (pragma directive) . . . . . . . . . . . . . . . . . . 322
hints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 call-info (in stack usage control file). . . . . . . . . . . . . . . . . 396
implementation-defined behavior . . . . . . . . . . . . . . . . 405 can_instantiate (pragma directive) . . . . . . . . . . . . . . 407, 421
implementation-defined behavior in C89 . . . . . . . . . . . 419 cassert (library header file) . . . . . . . . . . . . . . . . . . . . . . . . 366
non-standard types in . . . . . . . . . . . . . . . . . . . . . . . . . . 170 cast operators
bitfields (pragma directive) . . . . . . . . . . . . . . . . . . . . . . . . 321 in Extended EC++ . . . . . . . . . . . . . . . . . . . . . . . . 176, 188
bits in a byte, implementation-defined behavior . . . . . . . . 401 missing from Embedded C++ . . . . . . . . . . . . . . . . . . . 176
bold style, in this guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 casting
bool (data type) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282 between pointer types . . . . . . . . . . . . . . . . . . . . . . . . . . 57
adding support for in CLIB . . . . . . . . . . . . . . . . . . . . . 370 of pointers and integers . . . . . . . . . . . . . . . . . . . . . . . . 289
adding support for in DLIB . . . . . . . . . . . . . . . . . 364, 367 pointers to integers, language extension . . . . . . . . . . . . 172
building_runtime (pragma directive). . . . . . . . . . . . . 407, 421 category (in stack usage control file) . . . . . . . . . . . . . . . . 395
__BUILD_NUMBER__ (predefined symbol) . . . . . . . . . 352 cctype (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . . 366
byte order (of value), reversing . . . . . . . . . . . . . . . . . . . . . 348 cerrno (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . . 366
CFI (assembler directive) . . . . . . . . . . . . . . . . . . . . . . . . . 165
cfloat (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . . 366
C char (data type) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
changing default representation (--char_is_signed) . . . 241
C and C++ linkage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
C/C++ calling convention. See calling convention changing representation (--char_is_unsigned) . . . . . . . 241
C header files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364 implementation-defined behavior . . . . . . . . . . . . . . . . 402

430 for AVR
DRX-2
Index
signed and unsigned . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 codeseg (pragma directive) . . . . . . . . . . . . . . . . . . . . 407, 421

character set, implementation-defined behavior . . . . . . . . 400 command line options
characters See also compiler options
implementation-defined behavior . . . . . . . . . . . . . . . . 401 part of compiler invocation syntax . . . . . . . . . . . . . . . . 227
implementation-defined behavior in C89 . . . . . . . . . . . 416 passing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
character-based I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 typographic convention . . . . . . . . . . . . . . . . . . . . . . . . . 28
--char_is_signed (compiler option) . . . . . . . . . . . . . . . . . . 241 command prompt icon, in this guide . . . . . . . . . . . . . . . . . . 28
--char_is_unsigned (compiler option) . . . . . . . . . . . . . . . . 241 comments
check that (linker directive) . . . . . . . . . . . . . . . . . . . . . . . 392 after preprocessor directives . . . . . . . . . . . . . . . . . . . . 173
checksum C++ style, using in C code . . . . . . . . . . . . . . . . . . . . . . 167
calculation of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196 common block (call frame information) . . . . . . . . . . . . . . 166
CHECKSUM (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . 377 common subexpr elimination (compiler transformation) . 216
cinttypes (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . 366 disabling (--no_cse) . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
class memory (extended EC++) . . . . . . . . . . . . . . . . . . . . 178 compilation date
class template partial specialization exact time of (__TIME__) . . . . . . . . . . . . . . . . . . . . . . 358
matching (extended EC++) . . . . . . . . . . . . . . . . . . . . . . . . 184 identifying (__DATE__) . . . . . . . . . . . . . . . . . . . . . . . 353
CLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369 compiler
library reference information for . . . . . . . . . . . . . . . . . . 26 environment variables . . . . . . . . . . . . . . . . . . . . . . . . . 228
naming convention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 invocation syntax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
runtime environment . . . . . . . . . . . . . . . . . . . . . . . . . . 141 output from . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
summary of definitions . . . . . . . . . . . . . . . . . . . . . . . . 370 compiler listing, generating (-l). . . . . . . . . . . . . . . . . . . . . 256
--clib (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 compiler object file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
climits (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . . 367 including debug information in (--debug, -r) . . . . . . . . 243
clocale (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . 367 output from compiler . . . . . . . . . . . . . . . . . . . . . . . . . . 229
clock (CLIB library function), compiler optimization levels . . . . . . . . . . . . . . . . . . . . . . . 215
implementation-defined behavior in C89 . . . . . . . . . . . . . 428 compiler options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
clock (DLIB library function), passing to compiler . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
implementation-defined behavior in C89 . . . . . . . . . . . . . 425 reading from file (-f) . . . . . . . . . . . . . . . . . . . . . . . . . . 254
clock (library function) specifying parameters . . . . . . . . . . . . . . . . . . . . . . . . . 235
implementation-defined behavior . . . . . . . . . . . . . . . . . . . 412 summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
clustering (compiler transformation) . . . . . . . . . . . . . . . . . 218 syntax. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
disabling (--no_clustering) . . . . . . . . . . . . . . . . . . . . . . 260 for creating skeleton code . . . . . . . . . . . . . . . . . . . . . . 153
cmath (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . . 367 --initializers_in_flash . . . . . . . . . . . . . . . . . . . . . . . . . . 275
code --warnings_affect_exit_code . . . . . . . . . . . . . . . . . . . . 230
facilitating for good generation of . . . . . . . . . . . . . . . . 218 compiler platform, identifying . . . . . . . . . . . . . . . . . . . . . 356
interruption of execution . . . . . . . . . . . . . . . . . . . . . . . . 64 compiler subversion number . . . . . . . . . . . . . . . . . . . . . . . 357
verifying linked result . . . . . . . . . . . . . . . . . . . . . . . . . . 95 compiler transformations . . . . . . . . . . . . . . . . . . . . . . . . . 214
code motion (compiler transformation) . . . . . . . . . . . . . . . 217 compiler version number . . . . . . . . . . . . . . . . . . . . . . . . . 358
disabling (--no_code_motion) . . . . . . . . . . . . . . . . . . . 261
CODE (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
431
DRX-2
compiling CPU, specifying on command line for compiler . . . . . . . . 242
from the command line . . . . . . . . . . . . . . . . . . . . . . . . . 48 cross call (compiler transformation) . . . . . . . . . . . . . . . . . 218
syntax. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227 --cross_call_passes (compiler option) . . . . . . . . . . . . . . . . 242
complex numbers, supported in Embedded C++ . . . . . . . . 176 csetjmp (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . 367
complex (library header file). . . . . . . . . . . . . . . . . . . . . . . 365 csignal (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . 367
complex.h (library header file) . . . . . . . . . . . . . . . . . . . . . 364 cspy_support (pragma directive) . . . . . . . . . . . . . . . . 407, 421
compound literals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 CSTACK (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
computer style, typographic convention . . . . . . . . . . . . . . . 28 See also stack
configuration cstartup (system startup code)
basic project settings . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 customizing system initialization . . . . . . . . . . . . . . . . . 122
__low_level_init . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 source files for (CLIB) . . . . . . . . . . . . . . . . . . . . . . . . . 145
configuration symbols source files for (DLIB). . . . . . . . . . . . . . . . . . . . . . . . . 118
for file input and output . . . . . . . . . . . . . . . . . . . . . . . . 133 cstat_disable (pragma directive) . . . . . . . . . . . . . . . . . . . . 319
for locale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 cstat_enable (pragma directive) . . . . . . . . . . . . . . . . . . . . 319
for printf and scanf . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 cstat_restore (pragma directive) . . . . . . . . . . . . . . . . . . . . 319
for strtod . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 cstat_suppress (pragma directive) . . . . . . . . . . . . . . . . . . . 319
in library configuration files. . . . . . . . . . . . . . . . . . . . . 109 cstdarg (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . 367
const cstdbool (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . 367
declaring objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293 cstddef (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . 367
non-top level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 cstdio (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . . 367
constants, placing in named segment . . . . . . . . . . . . . . . . 323 cstdlib (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . . 367
__constrange(), symbol used in library . . . . . . . . . . . . . . . 369 cstring (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . . 367
__construction_by_bitwise_copy_allowed, symbol used ctime (DLIB header file) . . . . . . . . . . . . . . . . . . . . . . . . . . 367
in library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369 ctype.h (library header file) . . . . . . . . . . . . . . . . . . . . 364, 370
constseg (pragma directive) . . . . . . . . . . . . . . . . . . . . . . . 323 cwctype.h (library header file) . . . . . . . . . . . . . . . . . . . . . 367
const_cast (cast operator) . . . . . . . . . . . . . . . . . . . . . . . . . 176 ?C_EXIT (assembler label) . . . . . . . . . . . . . . . . . . . . . . . . 147
contents, of this guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 ?C_GETCHAR (assembler label) . . . . . . . . . . . . . . . . . . . 147
control characters, C_INCLUDE (environment variable) . . . . . . . . . . . . . . . . 228
implementation-defined behavior . . . . . . . . . . . . . . . . . . . 413 ?C_PUTCHAR (assembler label) . . . . . . . . . . . . . . . . . . . 147
conventions, used in this guide . . . . . . . . . . . . . . . . . . . . . . 28 C-SPY
copyright notice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 debug support for C++ . . . . . . . . . . . . . . . . . . . . . . . . . 183
cos (library function) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362 interface to system termination . . . . . . . . . . . . . . . . . . 122
cos (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . 117–118 low-level interface (CLIB) . . . . . . . . . . . . . . . . . . . . . . 147
cosf (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . 117–118 C-STAT for static analysis, documentation for . . . . . . . . . . 26
cosl (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . 117–118 C++
__COUNTER__ (predefined symbol). . . . . . . . . . . . . . . . 352 See also Embedded C++ and Extended Embedded C++
__cplusplus (predefined symbol) . . . . . . . . . . . . . . . . . . . 352 absolute location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
__CPU__ (predefined symbol) . . . . . . . . . . . . . . . . . . . . . 352 calling convention . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
--cpu (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 dynamic initialization in . . . . . . . . . . . . . . . . . . . . . . . . 93
CPU, identifying (__CPU__) . . . . . . . . . . . . . . . . . . . . . . 352

432 for AVR
DRX-2
Index
header files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365 __DATE__ (predefined symbol). . . . . . . . . . . . . . . . . . . . 353

language extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 date (library function), configuring support for . . . . . . . . . 106
standard template library (STL) . . . . . . . . . . . . . . . . . . 366 --debug (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . 243
static member variables . . . . . . . . . . . . . . . . . . . . . . . . 212 debug information, including in object file . . . . . . . . . . . . 243
support for . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 decimal point, implementation-defined behavior . . . . . . . 413
C++ header files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365 declarations
C++ names, in assembler code . . . . . . . . . . . . . . . . . . . . . 155 empty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
C++ objects, placing in memory type . . . . . . . . . . . . . . . . . 58 in for loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
C++ terminology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Kernighan & Ritchie . . . . . . . . . . . . . . . . . . . . . . . . . . 220
C++-style comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 of functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
C89 declarations and statements, mixing . . . . . . . . . . . . . . . . . 167
implementation-defined behavior . . . . . . . . . . . . . . . . 415 declarators, implementation-defined behavior in C89 . . . . 420
support for . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 define_type_info (pragma directive) . . . . . . . . . . . . . 407, 421
--c89 (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . 240 __delay_cycles (intrinsic function) . . . . . . . . . . . . . . . . . . 342
C99. See Standard C delete operator (extended EC++) . . . . . . . . . . . . . . . . . . . 181
delete (keyword) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
denormalized numbers. See subnormal numbers
D --dependencies (compiler option) . . . . . . . . . . . . . . . . . . . 244
-D (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 deque (STL header file) . . . . . . . . . . . . . . . . . . . . . . . . . . 366
data destructors and interrupts, using . . . . . . . . . . . . . . . . . . . . 182
alignment of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281 __DES_decryption (intrinsic function) . . . . . . . . . . . . . . . 342
different ways of storing . . . . . . . . . . . . . . . . . . . . . . . . 53 __DES_encryption (intrinsic function) . . . . . . . . . . . . . . . 343
located, declaring extern . . . . . . . . . . . . . . . . . . . . . . . 211 device
placing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210, 271, 324 identifying (__device__) . . . . . . . . . . . . . . . . . . . . . . . 353
at absolute location . . . . . . . . . . . . . . . . . . . . . . . . . 210 __device__ (predefined symbol) . . . . . . . . . . . . . . . . . . . . 353
representation of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281 device description files, preconfigured for C-SPY . . . . . . . 36
storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 device, identifying (__device__) . . . . . . . . . . . . . . . . . . . . 353
verifying linked result . . . . . . . . . . . . . . . . . . . . . . . . . . 95 diagnostic messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
data block (call frame information) . . . . . . . . . . . . . . . . . . 166 classifying as compilation errors . . . . . . . . . . . . . . . . . 245
data bus, enabling external . . . . . . . . . . . . . . . . . . . . . . . . 252 classifying as compilation remarks . . . . . . . . . . . . . . . 245
data memory attributes, using . . . . . . . . . . . . . . . . . . . . . . . 54 classifying as compiler warnings . . . . . . . . . . . . . . . . . 246
data pointers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287 disabling compiler warnings . . . . . . . . . . . . . . . . . . . . 265
data types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282 disabling wrapping of in compiler . . . . . . . . . . . . . . . . 265
avoiding signed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 enabling compiler remarks . . . . . . . . . . . . . . . . . . . . . . 269
floating point . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 listing all used by compiler . . . . . . . . . . . . . . . . . . . . . 246
in C++ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293 suppressing in compiler . . . . . . . . . . . . . . . . . . . . . . . . 246
integer types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282 --diagnostics_tables (compiler option) . . . . . . . . . . . . . . . 246
dataseg (pragma directive) . . . . . . . . . . . . . . . . . . . . . . . . 324 diagnostics, implementation-defined behavior . . . . . . . . . 399
data_alignment (pragma directive) . . . . . . . . . . . . . . . . . . 323 diag_default (pragma directive) . . . . . . . . . . . . . . . . . . . . 326
433
DRX-2
--diag_error (compiler option) . . . . . . . . . . . . . . . . . . . . . 245 domain errors, implementation-defined behavior in C89
diag_error (pragma directive) . . . . . . . . . . . . . . . . . . . . . . 327 (CLIB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426
--diag_remark (compiler option) . . . . . . . . . . . . . . . . . . . . 245 domain errors, implementation-defined behavior in C89
diag_remark (pragma directive) . . . . . . . . . . . . . . . . . . . . 327 (DLIB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
--diag_suppress (compiler option) . . . . . . . . . . . . . . . . . . 246 double (data type) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
diag_suppress (pragma directive) . . . . . . . . . . . . . . . . . . . 327 configuring size of floating-point type . . . . . . . . . . . . . . 50
--diag_warning (compiler option) . . . . . . . . . . . . . . . . . . . 246 --do_cross_call (compiler option) . . . . . . . . . . . . . . . . . . . 250
diag_warning (pragma directive) . . . . . . . . . . . . . . . . . . . 328 do_not_instantiate (pragma directive) . . . . . . . . . . . . 407, 421
DIFUNCT (segment) . . . . . . . . . . . . . . . . . . . . . . . . . 93, 378 DW (directive) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
directives dynamic initialization . . . . . . . . . . . . . . . . . . . . . . . . 118, 145
function for static overlay . . . . . . . . . . . . . . . . . . . . . . 165 and C++ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
pragma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36, 319 dynamic memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
directory, specifying as parameter . . . . . . . . . . . . . . . . . . 234
--disable_all_program_memory_load_instructions (compiler
option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 E
--disable_direct_mode (compiler option) . . . . . . . . . . . . . 247 -e (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
__disable_interrupt (intrinsic function) . . . . . . . . . . . . . . . 343 early_initialization (pragma directive) . . . . . . . . . . . 407, 421
--disable_library_knowledge (compiler option) . . . . . . . . 247 --ec++ (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . 250
--disable_mul (compiler option) . . . . . . . . . . . . . . . . . . . . 247 edition, of this guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
--disable_spm (compiler option) . . . . . . . . . . . . . . . . . . . . 248 --eecr_address (compiler option) . . . . . . . . . . . . . . . . . . . 251
--discard_unused_publics (compiler option) . . . . . . . . . . . 248 --eec++ (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . 251
disclaimer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 __eeprom (extended keyword) . . . . . . . . . . . . . . . . . . . . . 300
DLIB. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363 EEPROM_I (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 EEPROM_N (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . 379
configuring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108, 249 --eeprom_size (compiler option) . . . . . . . . . . . . . . . . . . . . 251
naming convention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 ELPM (instruction) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
reference information. See the online help system . . . . 361 Embedded C++ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
runtime environment . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 differences from C++ . . . . . . . . . . . . . . . . . . . . . . . . . . 176
--dlib (compiler option). . . . . . . . . . . . . . . . . . . . . . . . . . . 248 enabling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
--dlib_config (compiler option). . . . . . . . . . . . . . . . . . . . . 249 language extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
DLib_Defaults.h (library configuration file) . . . . . . . . . . . 109 overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
__DLIB_FILE_DESCRIPTOR (configuration symbol) . . 133 embedded systems, IAR special support for . . . . . . . . . . . . 36
document conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 __embedded_cplusplus (predefined symbol) . . . . . . . . . . 353
documentation --enable_external_bus (compiler option) . . . . . . . . . . . . . 252
contents of this . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 __enable_interrupt (intrinsic function) . . . . . . . . . . . . . . . 343
how to use this . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 --enable_multibytes (compiler option) . . . . . . . . . . . . . . . 252
overview of guides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 --enable_restrict (compiler option) . . . . . . . . . . . . . . . . . . 253
who should read this . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 enabling restrict keyword . . . . . . . . . . . . . . . . . . . . . . . . . 253
domain errors, implementation-defined behavior . . . . . . . 408 endianness. See byte order

434 for AVR
DRX-2
Index
--enhanced_core (compiler option) . . . . . . . . . . . . . . . . . . 253 __exit (library function) . . . . . . . . . . . . . . . . . . . . . . . . . . 121

entry label, program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 exp (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
enumerations expf (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
implementation-defined behavior . . . . . . . . . . . . . . . . 405 expl (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
implementation-defined behavior in C89 . . . . . . . . . . . 419 export keyword, missing from Extended EC++ . . . . . . . . 183
enums extended command line file
data representation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 for compiler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
forward declarations of . . . . . . . . . . . . . . . . . . . . . . . . 172 passing options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
environment Extended Embedded C++ . . . . . . . . . . . . . . . . . . . . . . . . . 176
implementation-defined behavior . . . . . . . . . . . . . . . . 400 enabling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
implementation-defined behavior in C89 . . . . . . . . . . . 415 extended keywords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
runtime (CLIB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 enabling (-e) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
runtime (DLIB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
environment names, implementation-defined behavior . . . 401 summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
environment variables syntax. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
C_INCLUDE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228 object attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
environment (native), type attributes on data objects . . . . . . . . . . . . . . . . . 296
implementation-defined behavior . . . . . . . . . . . . . . . . . . . 414 type attributes on functions . . . . . . . . . . . . . . . . . . . 297
EQU (assembler directive) . . . . . . . . . . . . . . . . . . . . . . . . 269 __root. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
ERANGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408 __extended_load_program_memory (intrinsic function) . 343
ERANGE (C89) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422 extern "C" linkage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
errno value at underflow, external data bus, enabling . . . . . . . . . . . . . . . . . . . . . . . . 252
implementation-defined behavior . . . . . . . . . . . . . . . . . . . 411 external memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275–276
errno.h (library header file) . . . . . . . . . . . . . . . . . . . . 364, 370 __ext_io (extended keyword) . . . . . . . . . . . . . . . . . . . . . . 301
error messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
classifying for compiler . . . . . . . . . . . . . . . . . . . . . . . . 245
error return codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
error (pragma directive) . . . . . . . . . . . . . . . . . . . . . . . . . . 328
F
-f (compiler option). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
errors and warnings, __far (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . . 302
listing all used by the compiler (--diagnostics_tables) . . . 246
FARCODE (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
--error_limit (compiler option) . . . . . . . . . . . . . . . . . . . . . 253
__farflash (extended keyword) . . . . . . . . . . . . . . . . . . . . . 302
escape sequences, implementation-defined behavior . . . . 401
__farfunc (extended keyword) . . . . . . . . . . . . . . . . . . . . . 303
exception handling, missing from Embedded C++ . . . . . . 176
__farfunc (function pointer) . . . . . . . . . . . . . . . . . . . . . . . 287
exception vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
FAR_C (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
exclude (stack usage control directive) . . . . . . . . . . . . . . . 393
FAR_F (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
_Exit (library function) . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
FAR_HEAP (segment) . . . . . . . . . . . . . . . . . . . . . . . 380, 382
exit (library function) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
FAR_I (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
implementation-defined behavior . . . . . . . . . . . . . . . . 411
FAR_ID (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
implementation-defined behavior in C89 . . . . . . . 425, 427
FAR_N (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
_exit (library function) . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
435
DRX-2
FAR_Z (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . 382, 384 floating-point constants
fatal error messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 hexadecimal notation . . . . . . . . . . . . . . . . . . . . . . . . . . 167
fdopen, in stdio.h . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368 hints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
fegettrapdisable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367 floating-point environment, accessing or not . . . . . . . . . . 338
fegettrapenable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367 floating-point expressions
FENV_ACCESS, implementation-defined behavior . . . . 404 contracting or not . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
fenv.h (library header file) . . . . . . . . . . . . . . . . . . . . . . . . 364 floating-point format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
additional C functionality. . . . . . . . . . . . . . . . . . . . . . . 367 hints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206–207
fgetpos (library function), implementation-defined implementation-defined behavior . . . . . . . . . . . . . . . . 403
behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411 implementation-defined behavior in C89 . . . . . . . . . . . 418
fgetpos (library function), implementation-defined special cases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
behavior in C89 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424 32-bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
field width, library support for . . . . . . . . . . . . . . . . . . . . . 144 64-bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
__FILE__ (predefined symbol) . . . . . . . . . . . . . . . . . . . . . 353 floating-point numbers, support for in printf formatters . . 144
file buffering, implementation-defined behavior . . . . . . . . 409 floating-point status flags . . . . . . . . . . . . . . . . . . . . . . . . . 367
file dependencies, tracking . . . . . . . . . . . . . . . . . . . . . . . . 244 floating-point type, configuring size of double . . . . . . . . . . 50
file input and output float.h (library header file) . . . . . . . . . . . . . . . . . . . . 364, 370
configuration symbols for . . . . . . . . . . . . . . . . . . . . . . 133 FLT_EVAL_METHOD, implementation-defined
file paths, specifying for #include files . . . . . . . . . . . . . . . 255 behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403, 408, 412
file position, implementation-defined behavior . . . . . . . . . 409 FLT_ROUNDS, implementation-defined
file streams lock interface . . . . . . . . . . . . . . . . . . . . . . . . . 138 behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403, 412
file systems in C89 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426 fmod (library function),
file (zero-length), implementation-defined behavior . . . . . 410 implementation-defined behavior in C89 . . . . . . . . . 422, 426
filename for loops, declarations in . . . . . . . . . . . . . . . . . . . . . . . . . . 167
extension for device description files . . . . . . . . . . . . . . . 36 --force_switch_type (compiler option) . . . . . . . . . . . . . . . 254
extension for header files . . . . . . . . . . . . . . . . . . . . . . . . 35 formats
of object file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 floating-point values . . . . . . . . . . . . . . . . . . . . . . . . . . 285
search procedure for. . . . . . . . . . . . . . . . . . . . . . . . . . . 228 standard IEEE (floating point) . . . . . . . . . . . . . . . . . . . 285
specifying as parameter . . . . . . . . . . . . . . . . . . . . . . . . 234 _formatted_write (library function) . . . . . . . . . . . . . . . . . 143
filenames (legal), implementation-defined behavior . . . . . 410 FP_CONTRACT, implementation-defined behavior . . . . 404
fileno, in stdio.h . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368 __fractional_multiply_signed (intrinsic function) . . . . . . . 344
files, implementation-defined behavior __fractional_multiply_signed_with_unsigned (intrinsic
handling of temporary . . . . . . . . . . . . . . . . . . . . . . . . . 410 function) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
multibyte characters in . . . . . . . . . . . . . . . . . . . . . . . . . 410 __fractional_multiply_unsigned (intrinsic function) . . . . . 344
opening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410 fragmentation, of heap memory . . . . . . . . . . . . . . . . . . . . . 62
__flash (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . 304 free (library function). See also heap . . . . . . . . . . . . . . . . . 61
flash memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275 fsetpos (library function), implementation-defined
library routines for accessing . . . . . . . . . . . . . . . . . . . . 370 behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
placing aggregate initializers . . . . . . . . . . . . . . . . . . . . 256 fstream (library header file) . . . . . . . . . . . . . . . . . . . . . . . 365
float (data type) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 ftell (library function), implementation-defined behavior . 411

436 for AVR
DRX-2
Index
in C89 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424 return values from . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162

Full DLIB (library configuration) . . . . . . . . . . . . . . . . . . . 110 special function types. . . . . . . . . . . . . . . . . . . . . . . . . . . 64
__func__ (predefined symbol) . . . . . . . . . . . . . . . . . 174, 354 verifying linked result . . . . . . . . . . . . . . . . . . . . . . . . . . 95
FUNCALL (assembler directive) . . . . . . . . . . . . . . . . . . . 165 function_effects (pragma directive) . . . . . . . . . . . . . 407, 421
__FUNCTION__ (predefined symbol) . . . . . . . . . . . 174, 354 function-spec (in stack usage control file). . . . . . . . . . . . . 395
function calls
calling convention . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
eliminating overhead of by inlining . . . . . . . . . . . . . . . . 67 G
stack image after . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 __generic (extended keyword) . . . . . . . . . . . . . . . . . . . . . 305
function declarations, Kernighan & Ritchie . . . . . . . . . . . 220 generic pointers, avoiding . . . . . . . . . . . . . . . . . . . . . . . . . 208
function directives for static overlay . . . . . . . . . . . . . . . . . 165 getchar (library function) . . . . . . . . . . . . . . . . . . . . . . . . . 143
function inlining (compiler transformation) . . . . . . . . . . . 217 getw, in stdio.h . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
disabling (--no_inline) . . . . . . . . . . . . . . . . . . . . . . . . . 262 getzone (library function), configuring support for . . . . . . 106
function pointers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287 __get_interrupt_state (intrinsic function) . . . . . . . . . . . . . 344
function prototypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 global variables
enforcing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270 affected by static clustering . . . . . . . . . . . . . . . . . . . . . 218
function return addresses . . . . . . . . . . . . . . . . . . . . . . . . . 163 handled during system termination . . . . . . . . . . . . . . . 121
function storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 hints for not using . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
function template parameter deduction (extended EC++) . 184 initialized during system startup . . . . . . . . . . . . . . . . . 120
function type information, omitting in object output . . . . . 266 --guard_calls (compiler option). . . . . . . . . . . . . . . . . . . . . 255
FUNCTION (assembler directive) . . . . . . . . . . . . . . . . . . 165 guidelines, reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
function (pragma directive) . . . . . . . . . . . . . . . . . . . . 407, 421
function (stack usage control directive) . . . . . . . . . . . . . . 393
functional (STL header file) . . . . . . . . . . . . . . . . . . . . . . . 366
functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
H
Harbison, Samuel P. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
declaring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157, 220 hardware support in compiler . . . . . . . . . . . . . . . . . . . . . . 100
inlining. . . . . . . . . . . . . . . . . . . . . . . . . 167, 217, 219, 329 hash_map (STL header file) . . . . . . . . . . . . . . . . . . . . . . . 366
interrupt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64–65 hash_set (STL header file) . . . . . . . . . . . . . . . . . . . . . . . . 366
intrinsic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149, 219 __has_constructor, symbol used in library . . . . . . . . . . . . 369
monitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 __has_destructor, symbol used in library . . . . . . . . . . . . . 369
omitting type info . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266 __HAS_EEPROM__ (predefined symbol) . . . . . . . . . . . . 354
parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 __HAS_EIND__ (predefined symbol) . . . . . . . . . . . . . . . 354
placing in memory . . . . . . . . . . . . . . . . . . . . 210, 212, 271 __HAS_ELPM__ (predefined symbol) . . . . . . . . . . . . . . . 354
placing segments for . . . . . . . . . . . . . . . . . . . . . . . . 90, 93 __HAS_ENHANCED_CORE__ (predefined symbol) . . . 355
recursive __HAS_FISCR__ (predefined symbol) . . . . . . . . . . . . . . 355
avoiding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 __HAS_MUL__ (predefined symbol). . . . . . . . . . . . . . . . 355
storing data on stack . . . . . . . . . . . . . . . . . . . . . . . . . 61 __HAS_RAMPD__ (predefined symbol) . . . . . . . . . . . . . 355
reentrancy (DLIB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362 __HAS_RAMPX__ (predefined symbol) . . . . . . . . . . . . . 355
related extensions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 __HAS_RAMPY__ (predefined symbol) . . . . . . . . . . . . . 355
437
DRX-2
__HAS_RAMPZ__ (predefined symbol) . . . . . . . . . . . . . 356 __huge (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . 305
hdrstop (pragma directive) . . . . . . . . . . . . . . . . . . . . 407, 421 __hugeflash (extended keyword) . . . . . . . . . . . . . . . . . . . 306
header files HUGE_C (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364 HUGE_F (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
C++ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365 HUGE_HEAP (segment) . . . . . . . . . . . . . . . . . . . . . . . . . 383
library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361 HUGE_I (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
special function registers . . . . . . . . . . . . . . . . . . . . . . . 222 HUGE_ID (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
STL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366 HUGE_N (segment) . . . . . . . . . . . . . . . . . . . . . . . . . 384, 387
DLib_Defaults.h . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
including stdbool.h for bool . . . . . . . . . . . . . . . . . . . . . 283
including stddef.h for wchar_t . . . . . . . . . . . . . . . . . . . 284 I
header names, implementation-defined behavior . . . . . . . 406 -I (compiler option). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
--header_context (compiler option) . . . . . . . . . . . . . . . . . . 255 IAR Command Line Build Utility. . . . . . . . . . . . . . . . . . . 109
heap IAR Postlink (utility) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
dynamic memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 IAR Systems Technical Support . . . . . . . . . . . . . . . . . . . . 232
segments for . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 iarbuild.exe (utility) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
storing data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 __iar_cos_accurate (library routine) . . . . . . . . . . . . . . . . . 118
VLA allocated on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277 __iar_cos_accuratef (library routine) . . . . . . . . . . . . . . . . 118
heap segments __iar_cos_accuratef (library function) . . . . . . . . . . . . . . . 362
CLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 __iar_cos_accuratel (library routine) . . . . . . . . . . . . . . . . 118
DLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 __iar_cos_accuratel (library function) . . . . . . . . . . . . . . . 362
FAR_F (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380 __iar_cos_small (library routine) . . . . . . . . . . . . . . . . . . . 117
FAR_HEAP (segment) . . . . . . . . . . . . . . . . . . . . 380, 382 __iar_cos_smallf (library routine) . . . . . . . . . . . . . . . . . . . 117
HUGE_F (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . 383 __iar_cos_smalll (library routine) . . . . . . . . . . . . . . . . . . . 117
HUGE_HEAP (segment) . . . . . . . . . . . . . . . . . . . . . . . 383 __IAR_DLIB_PERTHREAD_INIT_SIZE (macro) . . . . . 139
NEAR_F (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . 386 __IAR_DLIB_PERTHREAD_SIZE (macro) . . . . . . . . . . 139
NEAR_HEAP (segment) . . . . . . . . . . . . . . . . . . . . . . . 386 __IAR_DLIB_PERTHREAD_SYMBOL_OFFSET
placing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 (symbolptr) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
TINY_F (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 __iar_exp_small (library routine) . . . . . . . . . . . . . . . . . . . 117
TINY_HEAP (segment). . . . . . . . . . . . . . . . . . . . . . . . 389 __iar_exp_smallf (library routine) . . . . . . . . . . . . . . . . . . 117
heap size __iar_exp_smalll (library routine) . . . . . . . . . . . . . . . . . . 117
and standard I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 __iar_log_small (library routine) . . . . . . . . . . . . . . . . . . . 117
changing default. . . . . . . . . . . . . . . . . . . . . . . . . . . . 91–92 __iar_log_smallf (library routine) . . . . . . . . . . . . . . . . . . . 117
HEAP (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 __iar_log_smalll (library routine) . . . . . . . . . . . . . . . . . . . 117
heap (zero-sized), implementation-defined behavior . . . . 411 __iar_log10_small (library routine) . . . . . . . . . . . . . . . . . 117
hints __iar_log10_smallf (library routine) . . . . . . . . . . . . . . . . . 117
for good code generation . . . . . . . . . . . . . . . . . . . . . . . 218 __iar_log10_smalll (library routine) . . . . . . . . . . . . . . . . . 117
implementation-defined behavior . . . . . . . . . . . . . . . . 405 __iar_Powf (library routine) . . . . . . . . . . . . . . . . . . . . . . . 118
using efficient data types . . . . . . . . . . . . . . . . . . . . . . . 206 __iar_Powl (library routine) . . . . . . . . . . . . . . . . . . . . . . . 118

438 for AVR
DRX-2
Index
__iar_Pow_accurate (library routine) . . . . . . . . . . . . . . . . 118 icons, in this guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

__iar_pow_accurate (library routine) . . . . . . . . . . . . . . . . 118 IDE
__iar_Pow_accuratef (library routine) . . . . . . . . . . . . . . . 118 building a library from . . . . . . . . . . . . . . . . . . . . . . . . . 108
__iar_pow_accuratef (library routine). . . . . . . . . . . . . . . . 118 overview of build tools . . . . . . . . . . . . . . . . . . . . . . . . . 33
__iar_pow_accuratef (library function). . . . . . . . . . . . . . . 362 identifiers, implementation-defined behavior . . . . . . . . . . 401
__iar_Pow_accuratel (library routine). . . . . . . . . . . . . . . . 118 identifiers, implementation-defined behavior in C89 . . . . 416
__iar_pow_accuratel (library routine) . . . . . . . . . . . . . . . . 118 IEEE format, floating-point values . . . . . . . . . . . . . . . . . . 285
__iar_pow_accuratel (library function) . . . . . . . . . . . . . . . 362 important_typedef (pragma directive) . . . . . . . . . . . . 407, 421
__iar_pow_small (library routine) . . . . . . . . . . . . . . . . . . 117 include files
__iar_pow_smallf (library routine) . . . . . . . . . . . . . . . . . . 117 including before source files . . . . . . . . . . . . . . . . . . . . 268
__iar_pow_smalll (library routine) . . . . . . . . . . . . . . . . . . 117 specifying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
__iar_program_start (label). . . . . . . . . . . . . . . . . . . . . . . . 119 include_alias (pragma directive) . . . . . . . . . . . . . . . . . . . . 329
__iar_Sin (library routine) . . . . . . . . . . . . . . . . . . . . . . . . 117 __indirect_jump_to (intrinsic function) . . . . . . . . . . . . . . 345
__iar_Sinf (library routine) . . . . . . . . . . . . . . . . . . . . . . . . 118 infinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
__iar_Sinl (library routine) . . . . . . . . . . . . . . . . . . . . . . . . 118 infinity (style for printing), implementation-defined
__iar_Sin_accurate (library routine) . . . . . . . . . . . . . . . . . 118 behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
__iar_sin_accurate (library routine) . . . . . . . . . . . . . . . . . 118 inheritance, in Embedded C++ . . . . . . . . . . . . . . . . . . . . . 175
__iar_Sin_accuratef (library routine) . . . . . . . . . . . . . . . . 118 initialization
__iar_sin_accuratef (library routine) . . . . . . . . . . . . . . . . . 118 dynamic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118, 145
__iar_sin_accuratef (library function) . . . . . . . . . . . . . . . . 362 single-value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
__iar_Sin_accuratel (library routine) . . . . . . . . . . . . . . . . 118 --initializers_in_flash (compiler option) . . . . . . . . . . . . . . 256
__iar_sin_accuratel (library routine) . . . . . . . . . . . . . . . . . 118 initializers, static . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
__iar_sin_accuratel (library function) . . . . . . . . . . . . . . . . 362 INITTAB (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
__iar_Sin_small (library routine) . . . . . . . . . . . . . . . . . . . 117 inline assembler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
__iar_sin_small (library routine). . . . . . . . . . . . . . . . . . . . 117 avoiding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
__iar_Sin_smallf (library routine) . . . . . . . . . . . . . . . . . . . 117 See also assembler language interface
__iar_sin_smallf (library routine) . . . . . . . . . . . . . . . . . . . 117 inline functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
__iar_Sin_smalll (library routine) . . . . . . . . . . . . . . . . . . . 117 in compiler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
__iar_sin_smalll (library routine) . . . . . . . . . . . . . . . . . . . 117 inline (pragma directive) . . . . . . . . . . . . . . . . . . . . . . . . . . 329
__IAR_SYSTEMS_ICC__ (predefined symbol) . . . . . . . 356 inlining functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
__iar_tan_accurate (library routine) . . . . . . . . . . . . . . . . . 118 implementation-defined behavior . . . . . . . . . . . . . . . . 405
__iar_tan_accuratef (library routine). . . . . . . . . . . . . . . . . 118 __insert_opcode (intrinsic function) . . . . . . . . . . . . . . . . . 345
__iar_tan_accuratef (library function). . . . . . . . . . . . . . . . 362 installation directory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
__iar_tan_accuratel (library routine) . . . . . . . . . . . . . . . . . 118 instantiate (pragma directive) . . . . . . . . . . . . . . . . . . 407, 421
__iar_tan_accuratel (library function) . . . . . . . . . . . . . . . . 362 int (data type) signed and unsigned . . . . . . . . . . . . . . . . . . 282
__iar_tan_small (library routine) . . . . . . . . . . . . . . . . . . . 117 integer types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
__iar_tan_smallf (library routine) . . . . . . . . . . . . . . . . . . . 117 casting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
__iar_tan_smalll (library routine) . . . . . . . . . . . . . . . . . . . 117 implementation-defined behavior . . . . . . . . . . . . . . . . 403
iccbutl.h (library header file) . . . . . . . . . . . . . . . . . . . . . . 370 intptr_t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
ptrdiff_t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
439
DRX-2
size_t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289 iso646.h (library header file) . . . . . . . . . . . . . . . . . . . . . . . 364
uintptr_t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290 istream (library header file). . . . . . . . . . . . . . . . . . . . . . . . 365
integers, implementation-defined behavior in C89 . . . . . . 417 italic style, in this guide . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
integral promotion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 iterator (STL header file) . . . . . . . . . . . . . . . . . . . . . . . . . 366
internal error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 I/O register. See SFR
__interrupt (extended keyword) . . . . . . . . . . . . . . . . . 65, 307 I/O, character-based . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
using in pragma directives . . . . . . . . . . . . . . . . . . . . . . 339
interrupt functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
placement in memory . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 K
interrupt handler. See interrupt service routine keep_definition (pragma directive) . . . . . . . . . . . . . . 407, 421
interrupt service routine . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Kernighan & Ritchie function declarations . . . . . . . . . . . . 220
interrupt state, restoring . . . . . . . . . . . . . . . . . . . . . . . . . . 348 disallowing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
interrupt vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 keywords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
specifying with pragma directive . . . . . . . . . . . . . . . . . 339 extended, overview of . . . . . . . . . . . . . . . . . . . . . . . . . . 36
interrupt vector table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
in linker configuration file . . . . . . . . . . . . . . . . . . . . . . . 93
INITTAB segment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
INTVEC segment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
L
-l (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
start address for . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 for creating skeleton code . . . . . . . . . . . . . . . . . . . . . . 153
interrupts labels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
disabling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308 assembler, making public . . . . . . . . . . . . . . . . . . . . . . . 269
during function execution . . . . . . . . . . . . . . . . . . . . . 65 __iar_program_start . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
processor state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 __program_start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
using with C++ destructors . . . . . . . . . . . . . . . . . . . . . 182 Labrosse, Jean J. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
intptr_t (integer type) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290 __lac (intrinsic function) . . . . . . . . . . . . . . . . . . . . . . . . . . 345
__intrinsic (extended keyword). . . . . . . . . . . . . . . . . . . . . 307 language extensions
intrinsic functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Embedded C++ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 enabling using pragma . . . . . . . . . . . . . . . . . . . . . . . . . 330
summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341 enabling (-e) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
intrinsics.h (header file) . . . . . . . . . . . . . . . . . . . . . . . . . . 341 language overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
inttypes.h (library header file) . . . . . . . . . . . . . . . . . . . . . . 364 language (pragma directive) . . . . . . . . . . . . . . . . . . . . . . . 330
INTVEC (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . 93, 385 _large_write (library function) . . . . . . . . . . . . . . . . . . . . . 143
intwri.c (library source code) . . . . . . . . . . . . . . . . . . . . . . 144 __las (intrinsic function) . . . . . . . . . . . . . . . . . . . . . . . . . . 346
invocation syntax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227 __lat (intrinsic function) . . . . . . . . . . . . . . . . . . . . . . . . . . 346
__io (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . . 307 libraries
iomanip (library header file) . . . . . . . . . . . . . . . . . . . . . . . 365 reason for using . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
ios (library header file) . . . . . . . . . . . . . . . . . . . . . . . . . . . 365 standard template library . . . . . . . . . . . . . . . . . . . . . . . 366
iosfwd (library header file) . . . . . . . . . . . . . . . . . . . . . . . . 365 using a prebuilt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
iostream (library header file). . . . . . . . . . . . . . . . . . . . . . . 365 using a prebuilt (CLIB) . . . . . . . . . . . . . . . . . . . . . . . . 141

440 for AVR
DRX-2
Index
library configuration files in the build process . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

DLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
DLib_Defaults.h . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 process for . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
modifying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 list (STL header file) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
specifying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249 listing, generating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
library documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361 literals, compound. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
library features, missing from Embedded C++ . . . . . . . . . 176 literature, recommended . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
library functions __load_program_memory (intrinsic function) . . . . . . . . . 346
for accessing flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370 local variables, See auto variables
summary, CLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370 locale
summary, DLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364 adding support for in library . . . . . . . . . . . . . . . . . . . . 135
online help for . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 changing at runtime . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
library header files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361 implementation-defined behavior . . . . . . . . . . . . 402, 413
library modules removing support for . . . . . . . . . . . . . . . . . . . . . . . . . . 134
creating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 support for . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
overriding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 locale.h (library header file) . . . . . . . . . . . . . . . . . . . . . . . 364
library object files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362 located data segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
library project, building using a template . . . . . . . . . . . . . 108 located data, declaring extern . . . . . . . . . . . . . . . . . . . . . . 211
library_default_requirements (pragma directive) . . . 407, 421 location (pragma directive) . . . . . . . . . . . . . . . . . . . . 211, 331
--library_module (compiler option) . . . . . . . . . . . . . . . . . 257 LOCFRAME (assembler directive) . . . . . . . . . . . . . . . . . 165
library_provides (pragma directive) . . . . . . . . . . . . . 407, 421 --lock_regs (compiler option) . . . . . . . . . . . . . . . . . . . . . . 258
library_requirement_override (pragma directive) . . . 407, 422 log (library routine). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
lightbulb icon, in this guide. . . . . . . . . . . . . . . . . . . . . . . . . 28 logf (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
limits.h (library header file) . . . . . . . . . . . . . . . . . . . 364, 370 logl (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
__LINE__ (predefined symbol) . . . . . . . . . . . . . . . . . . . . 356 log10 (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
linkage, C and C++ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 log10f (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
linker. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 log10l (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
linker configuration file . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 long double (data type) . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
for placing code and data . . . . . . . . . . . . . . . . . . . . . . . . 74 long float (data type), synonym for double . . . . . . . . . . . . 172
in depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391 long long (data type)
overview of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391 restrictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
using the -P command . . . . . . . . . . . . . . . . . . . . . . . . . . 89 long long (data type) signed and unsigned . . . . . . . . . . . . 282
using the -Z command . . . . . . . . . . . . . . . . . . . . . . . . . . 89 long (data type) signed and unsigned . . . . . . . . . . . . . . . . 282
linker map file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 longjmp, restrictions for using . . . . . . . . . . . . . . . . . . . . . 363
linker options loop unrolling (compiler transformation)
typographic convention . . . . . . . . . . . . . . . . . . . . . . . . . 28 disabling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
linker segment. See segment loop-invariant expressions . . . . . . . . . . . . . . . . . . . . . . . . 217
linking __low_level_init . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
from the command line . . . . . . . . . . . . . . . . . . . . . . . . . 48 customizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
441
DRX-2
initialization phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 allocating in C++ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
low_level_init.c . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118, 145 dynamic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
low-level processor operations . . . . . . . . . . . . . . . . . . . . . 168 external . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275–276
accessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
heap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
M non-initialized . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
RAM, saving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
-m (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258 releasing in C++. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
macros stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
embedded in #pragma optimize . . . . . . . . . . . . . . . . . . 333 saving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
ERANGE (in errno.h) . . . . . . . . . . . . . . . . . . . . . 408, 422 used by global or static variables . . . . . . . . . . . . . . . . . . 53
inclusion of assert . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359 memory consumption, reducing . . . . . . . . . . . . . . . . . . . . 143
NULL, implementation-defined behavior . . . . . . . . . . 409 memory management, type-safe . . . . . . . . . . . . . . . . . . . . 175
in C89 for CLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . 425 memory map
in C89 for DLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . 422 initializing SFRs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
substituted in #pragma directives . . . . . . . . . . . . . . . . . 168 linker configuration for . . . . . . . . . . . . . . . . . . . . . . . . . 87
variadic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 memory models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
--macro_positions_in_diagnostics (compiler option) . . . . 258 identifying (__MEMORY_MODEL__) . . . . . . . . 356–357
main (function) specifying on command line (--memory_model) . . . . . 258
definition (C89) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415 memory placement
implementation-defined behavior . . . . . . . . . . . . . . . . 400 of linker segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
malloc (library function) using pragma directive . . . . . . . . . . . . . . . . . . . . . . . . . . 56
See also heap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 using type definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
implementation-defined behavior in C89 . . . . . . . 424, 427 memory segment. See segment
Mann, Bernhard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 memory types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
map (STL header file) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366 C++ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
map, linker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 hints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
math functions rounding mode, placing variables in . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
implementation-defined behavior . . . . . . . . . . . . . . . . . . . 412 pointers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
math functions (library functions) . . . . . . . . . . . . . . . . . . . 116 specifying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
math.h (library header file) . . . . . . . . . . . . . . . . . . . . 364, 370 structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
max recursion depth (stack usage control directive) . . . . . 393 summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
MB_LEN_MAX, implementation-defined behavior . . . . . 412 memory (pragma directive) . . . . . . . . . . . . . . . . . . . . 407, 422
_medium_write (library function) . . . . . . . . . . . . . . . . . . . 144 memory (STL header file) . . . . . . . . . . . . . . . . . . . . . . . . 366
memcmp_G (library function) . . . . . . . . . . . . . . . . . . . . . 371 __MEMORY_MODEL__ (predefined symbol) . . . . 356–357
memcpy_G (library function) . . . . . . . . . . . . . . . . . . . . . . 371 --memory_model (compiler option) . . . . . . . . . . . . . . . . . 258
memcpy_P (library function) . . . . . . . . . . . . . . . . . . . . . . 371 __memory_of
memory operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
accessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 symbol used in library . . . . . . . . . . . . . . . . . . . . . . . . . 369

442 for AVR
DRX-2
Index
message (pragma directive) . . . . . . . . . . . . . . . . . . . . . . . 331

messages N
disabling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272 name (in stack usage control file) . . . . . . . . . . . . . . . . . . . 396
forcing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 names block (call frame information) . . . . . . . . . . . . . . . . 166
Meyers, Scott . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 namespace support
--mfc (compiler option). . . . . . . . . . . . . . . . . . . . . . . . . . . 259 in Extended EC++ . . . . . . . . . . . . . . . . . . . . . . . . 176, 188
migration, from earlier IAR compilers . . . . . . . . . . . . . . . . 26 missing from Embedded C++ . . . . . . . . . . . . . . . . . . . 176
MISRA C naming conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 NaN
--misrac (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . 237 implementation of . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
--misrac_verbose (compiler option) . . . . . . . . . . . . . . . . . 237 implementation-defined behavior . . . . . . . . . . . . . . . . 410
--misrac1998 (compiler option) . . . . . . . . . . . . . . . . . . . . 237 native environment,
--misrac2004 (compiler option) . . . . . . . . . . . . . . . . . . . . 237 implementation-defined behavior . . . . . . . . . . . . . . . . . . . 414
mode changing, implementation-defined behavior . . . . . . 410 NDEBUG (preprocessor symbol) . . . . . . . . . . . . . . . . . . . 359
module consistency __near (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . 308
rtmodel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335 __nearfunc (extended keyword) . . . . . . . . . . . . . . . . . . . . 309
module map, in linker map file . . . . . . . . . . . . . . . . . . . . . . 96 __nearfunc (function pointer) . . . . . . . . . . . . . . . . . . . . . . 287
module name, specifying (--module_name) . . . . . . . . . . . 260 NEAR_C (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
module summary, in linker map file . . . . . . . . . . . . . . . . . . 96 NEAR_F (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
--module_name (compiler option) . . . . . . . . . . . . . . . . . . 260 NEAR_HEAP (segment) . . . . . . . . . . . . . . . . . . . . . . . . . 386
module_name (pragma directive) . . . . . . . . . . . . . . . 407, 422 NEAR_I (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . 386, 389
module-spec (in stack usage control file) . . . . . . . . . . . . . 396 NEAR_ID (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
__monitor (extended keyword) . . . . . . . . . . . . . . . . . . . . . 308 NEAR_Z (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
monitor functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65, 308 __nested (extended keyword) . . . . . . . . . . . . . . . . . . . . . . 309
multibyte character support . . . . . . . . . . . . . . . . . . . . . . . . 252 new calling convention . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
multibyte characters, implementation-defined new operator (extended EC++) . . . . . . . . . . . . . . . . . . . . . 181
behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401, 413 new (keyword) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
multiple address spaces, output for . . . . . . . . . . . . . . . . . . . 96 new (library header file) . . . . . . . . . . . . . . . . . . . . . . . . . . 365
multiple inheritance no calls from (stack usage control directive) . . . . . . . . . . . 394
in Extended EC++ . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 non-initialized variables, hints for . . . . . . . . . . . . . . . . . . . 223
missing from Embedded C++ . . . . . . . . . . . . . . . . . . . 176 non-scalar parameters, avoiding . . . . . . . . . . . . . . . . . . . . 219
multiple output files, from XLINK . . . . . . . . . . . . . . . . . . . 97 NOP (assembler instruction) . . . . . . . . . . . . . . . . . . . . . . . 347
__multiply_signed (intrinsic function) . . . . . . . . . . . . . . . 346 __noreturn (extended keyword) . . . . . . . . . . . . . . . . . . . . 312
__multiply_signed_with_unsigned (intrinsic function) . . . 346 Normal DLIB (library configuration) . . . . . . . . . . . . . . . . 110
__multiply_unsigned (intrinsic function) . . . . . . . . . . . . . 346 Not a number (NaN) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
multithreaded environment . . . . . . . . . . . . . . . . . . . . . . . . 136 __no_alloc (extended keyword) . . . . . . . . . . . . . . . . . . . . 310
multi-file compilation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 __no_alloc_str (operator) . . . . . . . . . . . . . . . . . . . . . . . . . 310
mutable attribute, in Extended EC++ . . . . . . . . . . . . 176, 188 __no_alloc_str16 (operator) . . . . . . . . . . . . . . . . . . . . . . . 310
__no_alloc16 (extended keyword) . . . . . . . . . . . . . . . . . . 310
443
DRX-2
--no_call_frame_info (compiler option) . . . . . . . . . . . . . . 260 offsetof . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
--no_clustering (compiler option) . . . . . . . . . . . . . . . . . . . 260 old calling convention . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
--no_code_motion (compiler option) . . . . . . . . . . . . . . . . 261 --omit_types (compiler option) . . . . . . . . . . . . . . . . . . . . . 266
--no_cross_call (compiler option) . . . . . . . . . . . . . . . . . . . 261 once (pragma directive) . . . . . . . . . . . . . . . . . . . . . . 407, 422
--no_cse (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . 261 --only_stdout (compiler option) . . . . . . . . . . . . . . . . . . . . 266
__no_init (extended keyword) . . . . . . . . . . . . . . . . . 223, 311 operators
--no_inline (compiler option) . . . . . . . . . . . . . . . . . . . . . . 262 See also @ (operator)
__no_operation (intrinsic function) . . . . . . . . . . . . . . . . . . 347 for cast
--no_path_in_file_macros (compiler option). . . . . . . . . . . 262 in Extended EC++. . . . . . . . . . . . . . . . . . . . . . . . . . 176
no_pch (pragma directive) . . . . . . . . . . . . . . . . . . . . 407, 422 missing from Embedded C++ . . . . . . . . . . . . . . . . . 176
--no_rampd (compiler option) . . . . . . . . . . . . . . . . . . . . . . 262 for segment control . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
__no_runtime_init (extended keyword) . . . . . . . . . . . . . . 311 in inline assembler . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
--no_size_constraints (compiler option) . . . . . . . . . . . . . . 262 new and delete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
--no_static_destruction (compiler option) . . . . . . . . . . . . . 263 precision for 32-bit float . . . . . . . . . . . . . . . . . . . . . . . 286
--no_system_include (compiler option) . . . . . . . . . . . . . . 263 precision for 64-bit float . . . . . . . . . . . . . . . . . . . . . . . 286
--no_tbaa (compiler option) . . . . . . . . . . . . . . . . . . . . . . . 263 sizeof, implementation-defined behavior . . . . . . . . . . . 413
--no_typedefs_in_diagnostics (compiler option) . . . . . . . . 264 variants for cast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
--no_ubrof_messages (compiler option) . . . . . . . . . . . . . . 264 _Pragma (preprocessor) . . . . . . . . . . . . . . . . . . . . . . . . 167
--no_unroll (compiler option) . . . . . . . . . . . . . . . . . . . . . . 265 __ALIGNOF__, for alignment control. . . . . . . . . . . . . 170
--no_warnings (compiler option) . . . . . . . . . . . . . . . . . . . 265 __memory_of. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
--no_wrap_diagnostics (compiler option) . . . . . . . . . . . . . 265 ?, language extensions for . . . . . . . . . . . . . . . . . . . . . . 190
NULL @ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
implementation-defined behavior . . . . . . . . . . . . . . . . 409 optimization
implementation-defined behavior in C89 (CLIB) . . . . 425 clustering, disabling . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
implementation-defined behavior in C89 (DLIB) . . . . 422 code motion, disabling . . . . . . . . . . . . . . . . . . . . . . . . . 261
in library header file (CLIB) . . . . . . . . . . . . . . . . . . . . 370 common sub-expression elimination, disabling . . . . . . 261
pointer constant, relaxation to Standard C . . . . . . . . . . 172 configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
numeric conversion functions, disabling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
implementation-defined behavior . . . . . . . . . . . . . . . . . . . 414 function inlining, disabling (--no_inline) . . . . . . . . . . . 262
numeric (STL header file) . . . . . . . . . . . . . . . . . . . . . . . . . 366 hints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
loop unrolling, disabling . . . . . . . . . . . . . . . . . . . . . . . 265
O size, specifying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279

specifying (-O) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
-O (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265 speed, specifying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
-o (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
object attributes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298 type-based alias analysis, disabling (--tbaa) . . . . . . . . . 263
object filename, specifying (-o) . . . . . . . . . . . . . . . . . . . . 267 using inline assembler code . . . . . . . . . . . . . . . . . . . . . 152
object module name, specifying (--module_name) . . . . . . 260 using pragma directive . . . . . . . . . . . . . . . . . . . . . . . . . 332
object_attribute (pragma directive) . . . . . . . . . . . . . . 223, 332 optimization levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215

444 for AVR
DRX-2
Index
optimize (pragma directive) . . . . . . . . . . . . . . . . . . . . . . . 332 data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287

option parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233 function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
options, compiler. See compiler options implementation-defined behavior . . . . . . . . . . . . . . . . 404
Oram, Andy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 implementation-defined behavior in C89 . . . . . . . . . . . 419
ostream (library header file) . . . . . . . . . . . . . . . . . . . . . . . 365 polymorphism, in Embedded C++ . . . . . . . . . . . . . . . . . . 175
output porting, code containing pragma directives . . . . . . . . . . . . 321
from preprocessor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268 possible calls (stack usage control directive) . . . . . . . . . . 394
multiple files from linker . . . . . . . . . . . . . . . . . . . . . . . . 97 Postlink (utility) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
specifying for linker . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 postlink.htm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
supporting non-standard. . . . . . . . . . . . . . . . . . . . . . . . 144 pow (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . 117–118
--output (compiler option). . . . . . . . . . . . . . . . . . . . . . . . . 267 alternative implementation of . . . . . . . . . . . . . . . . . . . 362
overhead, reducing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217 powf (library routine) . . . . . . . . . . . . . . . . . . . . . . . . 117–118
powl (library routine) . . . . . . . . . . . . . . . . . . . . . . . . 117–118
P pragma directives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
parameters basic_template_matching, using . . . . . . . . . . . . . . . . . 185
function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 for absolute located data . . . . . . . . . . . . . . . . . . . . . . . 211
hidden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 list of all recognized . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
non-scalar, avoiding . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 list of all recognized (C89). . . . . . . . . . . . . . . . . . . . . . 421
register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159–160 type_attribute, using. . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
rules for specifying a file or directory . . . . . . . . . . . . . 234 _Pragma (preprocessor operator) . . . . . . . . . . . . . . . . . . . 167
specifying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 precision arguments, library support for . . . . . . . . . . . . . . 144
stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159, 162 predefined symbols
typographic convention . . . . . . . . . . . . . . . . . . . . . . . . . 28 overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
part number, of this guide . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
--pending_instantiations (compiler option) . . . . . . . . . . . . 267 --predef_macro (compiler option) . . . . . . . . . . . . . . . . . . . 267
permanent registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 --preinclude (compiler option) . . . . . . . . . . . . . . . . . . . . . 268
perror (library function), --preprocess (compiler option) . . . . . . . . . . . . . . . . . . . . . 268
implementation-defined behavior in C89 . . . . . . . . . 424, 427 preprocessor
placement operator (_Pragma) . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
in named segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212 output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
of code and data, introduction to . . . . . . . . . . . . . . . . . . 74 preprocessor directives
plain char, implementation-defined behavior . . . . . . . . . . 402 comments at the end of . . . . . . . . . . . . . . . . . . . . . . . . 173
pointer types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287 implementation-defined behavior . . . . . . . . . . . . . . . . 406
differences between . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 implementation-defined behavior in C89 . . . . . . . . . . . 420
mixing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 #pragma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
using the best . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 preprocessor extensions
pointers __VA_ARGS__ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
casting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57, 289 #warning message . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
445
DRX-2
preprocessor symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
defining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 Q
preserved registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 qualifiers
__PRETTY_FUNCTION__ (predefined symbol). . . . . . . 356 const and volatile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
primitives, for special functions . . . . . . . . . . . . . . . . . . . . . 64 implementation-defined behavior . . . . . . . . . . . . . . . . 406
print formatter, selecting . . . . . . . . . . . . . . . . . . . . . . . . . . 114 implementation-defined behavior in C89 . . . . . . . . . . . 420
printf (library function) . . . . . . . . . . . . . . . . . . . . . . . 113, 143 queue (STL header file) . . . . . . . . . . . . . . . . . . . . . . . . . . 366
choosing formatter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
configuration symbols . . . . . . . . . . . . . . . . . . . . . . . . . 132
customizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
R
-r (compiler option). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
RAM
selecting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
initializers copied from ROM . . . . . . . . . . . . . . . . . . . . 46
__printf_args (pragma directive). . . . . . . . . . . . . . . . . . . . 334
saving memory. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
printf_P (library function) . . . . . . . . . . . . . . . . . . . . . . . . . 371
RAMPZ (register) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
printing characters, implementation-defined behavior . . . 413
range errors, in linker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
processor configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
__raw (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . 312
processor operations
read formatter, selecting . . . . . . . . . . . . . . . . . . . . . . 115, 145
accessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
reading guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
low-level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
reading, recommended . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
program entry label . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
realloc (library function) . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
program termination, implementation-defined behavior . . 400
programming hints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
See also heap
__program_start (label) . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
recursive functions
projects
avoiding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
basic settings for . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
storing data on stack . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
setting up for a library . . . . . . . . . . . . . . . . . . . . . . . . . 108
reentrancy (DLIB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
prototypes, enforcing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
reference information, typographic convention . . . . . . . . . . 28
ptrdiff_t (integer type) . . . . . . . . . . . . . . . . . . . . . . . 289, 370
register keyword, implementation-defined behavior . . . . . 405
PUBLIC (assembler directive) . . . . . . . . . . . . . . . . . . . . . 269
register parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 159–160
publication date, of this guide . . . . . . . . . . . . . . . . . . . . . . . . 2
registered trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
--public_equ (compiler option) . . . . . . . . . . . . . . . . . . . . . 268
registers
public_equ (pragma directive) . . . . . . . . . . . . . . . . . . . . . 334
callee-save, stored on stack . . . . . . . . . . . . . . . . . . . . . . 60
putchar (library function) . . . . . . . . . . . . . . . . . . . . . . . . . 143
for function returns . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
putenv (library function), absent from DLIB . . . . . . . . . . 125
implementation-defined behavior in C89 . . . . . . . . . . . 419
puts_G (library function) . . . . . . . . . . . . . . . . . . . . . . . . . 371
in assembler-level routines. . . . . . . . . . . . . . . . . . . . . . 156
puts_P (library function) . . . . . . . . . . . . . . . . . . . . . . . . . . 372
preserved . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
putw, in stdio.h . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
scratch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
__regvar (extended keyword) . . . . . . . . . . . . . . . . . . . . . . 312

446 for AVR
DRX-2
Index
reinterpret_cast (cast operator) . . . . . . . . . . . . . . . . . . . . . 176 runtime libraries (CLIB)

--relaxed_fp (compiler option) . . . . . . . . . . . . . . . . . . . . . 269 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
remark (diagnostic message) . . . . . . . . . . . . . . . . . . . . . . . 231 filename syntax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
classifying for compiler . . . . . . . . . . . . . . . . . . . . . . . . 245 using prebuilt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
enabling in compiler . . . . . . . . . . . . . . . . . . . . . . . . . . 269 runtime libraries (DLIB)
--remarks (compiler option) . . . . . . . . . . . . . . . . . . . . . . . 269 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
remove (library function) customizing system startup code . . . . . . . . . . . . . . . . . 122
implementation-defined behavior . . . . . . . . . . . . . . . . 410 filename syntax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
implementation-defined behavior in C89 (CLIB) . . . . 427 overriding modules in . . . . . . . . . . . . . . . . . . . . . . . . . 107
implementation-defined behavior in C89 (DLIB) . . . . 424 using prebuilt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
remquo, magnitude of . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409 runtime model definitions . . . . . . . . . . . . . . . . . . . . . . . . . 335
rename (library function) runtime type information, missing from Embedded C++ . 176
implementation-defined behavior in C89 (CLIB) . . . . 427
implementation-defined behavior in C89 (DLIB) . . . . 424 S
__ReportAssert (library function) . . . . . . . . . . . . . . . . . . . 129 -s (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
__require (intrinsic function) . . . . . . . . . . . . . . . . . . . . . . 347 __save_interrupt (intrinsic function) . . . . . . . . . . . . . . . . . 348
required (pragma directive) . . . . . . . . . . . . . . . . . . . . . . . . 334 scanf (library function)
--require_prototypes (compiler option) . . . . . . . . . . . . . . . 270 choosing formatter (CLIB). . . . . . . . . . . . . . . . . . . . . . 145
__restore_interrupt (intrinsic function) . . . . . . . . . . . 347–348 choosing formatter (DLIB) . . . . . . . . . . . . . . . . . . . . . 114
restrict keyword, enabling . . . . . . . . . . . . . . . . . . . . . . . . . 253 configuration symbols . . . . . . . . . . . . . . . . . . . . . . . . . 132
return addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 implementation-defined behavior . . . . . . . . . . . . . . . . 411
return data stack implementation-defined behavior in C89 . . . . . . . . . . . 427
reducing usage of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 implementation-defined behavior in C89 (CLIB) . . . . 427
using cross-call optimizations . . . . . . . . . . . . . . . . . . . 261 implementation-defined behavior in C89 (DLIB) . . . . 424
return values, from functions . . . . . . . . . . . . . . . . . . . . . . 162 __scanf_args (pragma directive) . . . . . . . . . . . . . . . . . . . . 336
__root (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . 313 scanf_P (library function) . . . . . . . . . . . . . . . . . . . . . . . . . 372
--root_variables (compiler option) . . . . . . . . . . . . . . . . . . 270 scratch registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
routines, time-critical . . . . . . . . . . . . . . . . . . . . . . . . 149, 168 section (pragma directive). . . . . . . . . . . . . . . . . . . . . . . . . 336
__ro_placement (extended keyword) . . . . . . . . . . . . . . . . 313 segment group name . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
RSTACK (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 segment map, in linker map file . . . . . . . . . . . . . . . . . . . . . 96
example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 segment memory types, in XLINK . . . . . . . . . . . . . . . . . . . 72
See also stack --segment (compiler option) . . . . . . . . . . . . . . . . . . . . . . . 271
rtmodel (pragma directive) . . . . . . . . . . . . . . . . . . . . . . . . 335 segment (pragma directive) . . . . . . . . . . . . . . . . . . . . . . . . 336
rtti support, missing from STL . . . . . . . . . . . . . . . . . . . . . 177 segments
runtime environment allocation of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
CLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 declaring (#pragma segment) . . . . . . . . . . . . . . . . . . . . 336
DLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 definition of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
setting up (DLIB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 HEAP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
447
DRX-2
located data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 signed long long (data type) . . . . . . . . . . . . . . . . . . . . . . . 282
naming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 signed long (data type) . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
packing in memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 signed short (data type) . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
placing in sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 signed values, avoiding . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
RSTACK --silent (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . 272
reducing usage of . . . . . . . . . . . . . . . . . . . . . . . . . . 242 silent operation, specifying in compiler . . . . . . . . . . . . . . 272
using cross-call optimizations . . . . . . . . . . . . . . . . . 261 sin (library function) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
specifying (--segment) . . . . . . . . . . . . . . . . . . . . . . . . . 271 sin (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . . 117–118
summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 sinf (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . 117–118
too long for address range . . . . . . . . . . . . . . . . . . . . . . . 96 sinl (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . 117–118
too long, in linker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 64-bits (floating-point format) . . . . . . . . . . . . . . . . . . . . . 286
__segment_begin (extended operator) . . . . . . . . . . . . . . . 171 size optimization, specifying. . . . . . . . . . . . . . . . . . . . . . . 279
__segment_end (extended operator) . . . . . . . . . . . . . . . . . 171 size (in stack usage control file) . . . . . . . . . . . . . . . . . . . . 397
__segment_size (extended operator) . . . . . . . . . . . . . . . . . 171 size_t (integer type) . . . . . . . . . . . . . . . . . . . . . . . . . 289, 370
semaphores skeleton code, creating for assembler language interface . 153
C example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 SLEEP (assembler instruction) . . . . . . . . . . . . . . . . . . . . . 349
operations on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308 __sleep (intrinsic function) . . . . . . . . . . . . . . . . . . . . . . . . 349
--separate_cluster_for_initialized_variables (compiler slist (STL header file) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272 _small_write (library function) . . . . . . . . . . . . . . . . . . . . . 144
set (STL header file) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366 source files, list all referred . . . . . . . . . . . . . . . . . . . . . . . . 255
setjmp.h (library header file) . . . . . . . . . . . . . . . . . . . 364, 370 space characters, implementation-defined behavior . . . . . 409
setlocale (library function) . . . . . . . . . . . . . . . . . . . . . . . . 135 special function registers (SFR) . . . . . . . . . . . . . . . . . . . . 222
settings, basic for project configuration . . . . . . . . . . . . . . . 48 special function types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
__set_interrupt_state (intrinsic function) . . . . . . . . . . . . . 348 speed optimization, specifying . . . . . . . . . . . . . . . . . . . . . 271
severity level, of diagnostic messages . . . . . . . . . . . . . . . . 231 --spmcr_address (compiler option) . . . . . . . . . . . . . . . . . . 273
specifying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 sprintf (library function) . . . . . . . . . . . . . . . . . . . . . . 113, 143
SFR choosing formatter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
accessing special function registers . . . . . . . . . . . . . . . 222 customizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
declaring extern special function registers . . . . . . . . . . 211 sprintf_P (library function) . . . . . . . . . . . . . . . . . . . . . . . . 372
shared object . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230 sscanf (library function)
short (data type) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282 choosing formatter (CLIB). . . . . . . . . . . . . . . . . . . . . . 145
signal (library function) choosing formatter (DLIB) . . . . . . . . . . . . . . . . . . . . . 114
implementation-defined behavior . . . . . . . . . . . . . . . . 409 sscanf_P (library function) . . . . . . . . . . . . . . . . . . . . . . . . 372
implementation-defined behavior in C89 . . . . . . . . . . . 423 sstream (library header file) . . . . . . . . . . . . . . . . . . . . . . . 365
signals, implementation-defined behavior. . . . . . . . . . . . . 400 stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
at system startup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401 advantages and problems using . . . . . . . . . . . . . . . . . . . 61
signal.h (library header file) . . . . . . . . . . . . . . . . . . . . . . . 364 cleaning after function return . . . . . . . . . . . . . . . . . . . . 163
signed char (data type) . . . . . . . . . . . . . . . . . . . . . . . 282–283 contents of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
specifying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
signed int (data type). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282

448 for AVR
DRX-2
Index
saving space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 stdarg.h (library header file) . . . . . . . . . . . . . . . . . . . 364, 370

setting up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 stdbool.h (library header file) . . . . . . . . . . . . . . 283, 364, 370
size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 __STDC__ (predefined symbol) . . . . . . . . . . . . . . . . . . . . 357
stack parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159, 162 STDC CX_LIMITED_RANGE (pragma directive) . . . . . 337
stack pointer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 STDC FENV_ACCESS (pragma directive) . . . . . . . . . . . 337
stack pointer register, considerations . . . . . . . . . . . . . . . . 159 STDC FP_CONTRACT (pragma directive) . . . . . . . . . . . 338
stack segments __STDC_VERSION__ (predefined symbol) . . . . . . . . . . 357
CSTACK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378 stddef.h (library header file) . . . . . . . . . . . . . . . 284, 364, 370
placing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 stderr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106, 266
RSTACK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 stdin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
stack (STL header file) . . . . . . . . . . . . . . . . . . . . . . . . . . . 366 implementation-defined behavior in C89 (CLIB) . . . . 426
stack-size (in stack usage control file) . . . . . . . . . . . . . . . 397 implementation-defined behavior in C89 (DLIB) . . . . 423
Standard C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253 stdint.h (library header file) . . . . . . . . . . . . . . . . . . . 364, 367
library compliance with . . . . . . . . . . . . . . . . . . . . . . . . 361 stdio.h (library header file) . . . . . . . . . . . . . . . . . . . . 364, 370
specifying strict usage . . . . . . . . . . . . . . . . . . . . . . . . . 273 stdio.h, additional C functionality . . . . . . . . . . . . . . . . . . . 368
standard error stdlib.h (library header file) . . . . . . . . . . . . . . . . . . . 364, 370
redirecting in compiler . . . . . . . . . . . . . . . . . . . . . . . . . 266 stdout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106, 266
See also diagnostic messages . . . . . . . . . . . . . . . . . . . . 230 implementation-defined behavior . . . . . . . . . . . . . . . . 409
standard output implementation-defined behavior in C89 (CLIB) . . . . 426
specifying in compiler . . . . . . . . . . . . . . . . . . . . . . . . . 266 implementation-defined behavior in C89 (DLIB) . . . . 423
standard template library (STL) Steele, Guy L. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
in C++ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366 STL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
in Extended EC++ . . . . . . . . . . . . . . . . . . . . . . . . 176, 183 strcasecmp, in string.h . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
missing from Embedded C++ . . . . . . . . . . . . . . . . . . . 176 strcat_G (library function) . . . . . . . . . . . . . . . . . . . . . . . . 372
startup code strcmp_G (library function) . . . . . . . . . . . . . . . . . . . . . . . 372
cstartup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 strcmp_P (library function) . . . . . . . . . . . . . . . . . . . . . . . . 373
startup system. See system startup strcpy_G (library function) . . . . . . . . . . . . . . . . . . . . . . . . 373
statements, implementation-defined behavior in C89 . . . . 420 strcpy_P (library function) . . . . . . . . . . . . . . . . . . . . . . . . 373
static analysis strdup, in string.h . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
documentation for . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 streambuf (library header file) . . . . . . . . . . . . . . . . . . . . . 365
static clustering (compiler transformation) . . . . . . . . . . . . 218 streams
static data, in configuration file . . . . . . . . . . . . . . . . . . . . . . 90 implementation-defined behavior . . . . . . . . . . . . . . . . 400
static overlay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 supported in Embedded C++ . . . . . . . . . . . . . . . . . . . . 176
static variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 strerror (library function)
taking the address of . . . . . . . . . . . . . . . . . . . . . . . . . . 219 implementation-defined behavior in C89 (CLIB) . . . . . . . 428
static_assert() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 strerror (library function), implementation-defined
static_cast (cast operator) . . . . . . . . . . . . . . . . . . . . . . . . . 176 behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
status flags for floating-point . . . . . . . . . . . . . . . . . . . . . . 367 strerror (library function),
std namespace, missing from EC++ implementation-defined behavior in C89 (DLIB) . . . . . . . 425
and Extended EC++ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188 strerror_P (library function) . . . . . . . . . . . . . . . . . . . . . . . 373
449
DRX-2
--strict (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . 273 invoking compiler . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
string (library header file) . . . . . . . . . . . . . . . . . . . . . . . . . 365 system function, implementation-defined behavior. . 401, 411
strings, supported in Embedded C++ . . . . . . . . . . . . . . . . 176 system locks interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
--string_literals_in_flash (compiler option). . . . . . . . . . . . 273 system startup
string.h (library header file) . . . . . . . . . . . . . . . . . . . 364, 370 CLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
string.h, additional C functionality . . . . . . . . . . . . . . . . . . 368 customizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
strlen_G (library function) . . . . . . . . . . . . . . . . . . . . . . . . 373 DLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
strlen_P (library function) . . . . . . . . . . . . . . . . . . . . . . . . . 373 initialization phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
strncasecmp, in string.h. . . . . . . . . . . . . . . . . . . . . . . . . . . 368 system termination
strncat_G (library function) . . . . . . . . . . . . . . . . . . . . . . . 374 CLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
strncmp_G (library function) . . . . . . . . . . . . . . . . . . . . . . 374 C-SPY interface to . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
strncmp_P (library function) . . . . . . . . . . . . . . . . . . . . . . . 374 DLIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
strncpy_G (library function) . . . . . . . . . . . . . . . . . . . . . . . 374 system (library function)
strncpy_P (library function) . . . . . . . . . . . . . . . . . . . . . . . 374 implementation-defined behavior in C89 . . . . . . . . . . . 428
strnlen, in string.h . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368 implementation-defined behavior in C89 (DLIB) . . . . 425
strstream (library header file) . . . . . . . . . . . . . . . . . . . . . . 365 system_include (pragma directive) . . . . . . . . . . . . . . 407, 422
strtod (library function), configuring support for . . . . . . . 135 --system_include_dir (compiler option) . . . . . . . . . . . . . . 274
structure types
alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
layout of . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290 T
structures tan (library function) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
anonymous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170, 208 tan (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . . 117–118
implementation-defined behavior . . . . . . . . . . . . . . . . 405 tanf (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . 117–118
implementation-defined behavior in C89 . . . . . . . . . . . 419 tanl (library routine) . . . . . . . . . . . . . . . . . . . . . . . . . 117–118
placing in memory type . . . . . . . . . . . . . . . . . . . . . . . . . 57 __task (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . 314
subnormal numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287 technical support, IAR Systems . . . . . . . . . . . . . . . . . . . . 232
support, technical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 template support
Sutter, Herb. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 in Extended EC++ . . . . . . . . . . . . . . . . . . . . . . . . 176, 183
__swap_nibbles (intrinsic function) . . . . . . . . . . . . . . . . . 349 missing from Embedded C++ . . . . . . . . . . . . . . . . . . . 176
SWITCH (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 Terminal I/O window
symbols making available (CLIB) . . . . . . . . . . . . . . . . . . . . . . . 147
anonymous, creating . . . . . . . . . . . . . . . . . . . . . . . . . . 167 not supported when . . . . . . . . . . . . . . . . . . . . . . . 107–108
including in output . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334 termination of system. See system termination
listing in linker map file . . . . . . . . . . . . . . . . . . . . . . . . . 96 termination status, implementation-defined behavior . . . . 411
overview of predefined . . . . . . . . . . . . . . . . . . . . . . . . . 37 terminology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
preprocessor, defining . . . . . . . . . . . . . . . . . . . . . . . . . 243 tgmath.h (library header file) . . . . . . . . . . . . . . . . . . . . . . 364
syntax 32-bits (floating-point format) . . . . . . . . . . . . . . . . . . . . . 286
command line options . . . . . . . . . . . . . . . . . . . . . . . . . 233 this (pointer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
extended keywords . . . . . . . . . . . . . . . . . . . . 56, 296–298 class memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178

450 for AVR
DRX-2
Index
referring to a class object . . . . . . . . . . . . . . . . . . . . . . . 178 typedefs

threaded environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 excluding from diagnostics . . . . . . . . . . . . . . . . . . . . . 264
thread-local storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 repeated . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
__TIME__ (predefined symbol) . . . . . . . . . . . . . . . . . . . . 358 type_attribute (pragma directive) . . . . . . . . . . . . . . . . 56, 338
time zone (library function) type-based alias analysis (compiler transformation) . . . . . 217
implementation-defined behavior in C89 . . . . . . . . . 425, 428 disabling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
time zone (library function), implementation-defined type-safe memory management . . . . . . . . . . . . . . . . . . . . 175
behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412 typographic conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
__TIMESTAMP__ (predefined symbol) . . . . . . . . . . . . . 358
time-critical routines . . . . . . . . . . . . . . . . . . . . . . . . . 149, 168
time.h (library header file) . . . . . . . . . . . . . . . . . . . . . . . . 364
additional C functionality. . . . . . . . . . . . . . . . . . . . . . . 368
U
UBROF
time32 (library function), configuring support for . . . . . . 106 format of linkable object files . . . . . . . . . . . . . . . . . . . 229
time64 (library function), configuring support for . . . . . . 106 specifying, example of . . . . . . . . . . . . . . . . . . . . . . . . . . 48
__tiny (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . 314 uchar.h (library header file) . . . . . . . . . . . . . . . . . . . . . . . 364
__tinyflash (extended keyword) . . . . . . . . . . . . . . . . . . . . 315 uintptr_t (integer type) . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
__TINY_AVR__ (predefined symbol) . . . . . . . . . . . . . . . 358 underflow errors, implementation-defined behavior . 408–409
TINY_F (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 underflow range errors,
TINY_HEAP (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . 389 implementation-defined behavior in C89 . . . . . . . . . 422, 426
TINY_ID (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389 __ungetchar, in stdio.h . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
TINY_N (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390 unions
TINY_Z (segment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390 anonymous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170, 208
tips, programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218 implementation-defined behavior . . . . . . . . . . . . . . . . 405
TLS handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 implementation-defined behavior in C89 . . . . . . . . . . . 419
tools icon, in this guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 universal character names, implementation-defined
trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
transformations, compiler . . . . . . . . . . . . . . . . . . . . . . . . . 214 unsigned char (data type) . . . . . . . . . . . . . . . . . . . . . 282–283
translation changing to signed char . . . . . . . . . . . . . . . . . . . . . . . . 241
implementation-defined behavior . . . . . . . . . . . . . . . . 399 unsigned int (data type) . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
implementation-defined behavior in C89 . . . . . . . . . . . 415 unsigned long long (data type) . . . . . . . . . . . . . . . . . . . . . 283
trap vectors, specifying with pragma directive . . . . . . . . . 339 unsigned long (data type) . . . . . . . . . . . . . . . . . . . . . . . . . 282
type attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295 unsigned short (data type) . . . . . . . . . . . . . . . . . . . . . . . . . 282
specifying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338 --use_c++_inline (compiler option) . . . . . . . . . . . . . . . . . 274
type definitions, used for specifying memory storage . . . . . 56 utility (STL header file) . . . . . . . . . . . . . . . . . . . . . . . . . . 366
type information, omitting . . . . . . . . . . . . . . . . . . . . . . . . 266
type qualifiers
const and volatile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
V
-v (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
implementation-defined behavior in C89 . . . . . . . . . . . 420 variable type information, omitting in object output . . . . . 266
451
DRX-2
variables
auto . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 W
defined inside a function . . . . . . . . . . . . . . . . . . . . . . . . 60 #warning message (preprocessor extension) . . . . . . . . . . . 360
global warnings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
placement in memory . . . . . . . . . . . . . . . . . . . . . . . . 53 classifying in compiler . . . . . . . . . . . . . . . . . . . . . . . . . 246
hints for choosing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 disabling in compiler . . . . . . . . . . . . . . . . . . . . . . . . . . 265
local. See auto variables exit code in compiler . . . . . . . . . . . . . . . . . . . . . . . . . . 278
non-initialized . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 warnings icon, in this guide . . . . . . . . . . . . . . . . . . . . . . . . 29
omitting type info . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266 warnings (pragma directive) . . . . . . . . . . . . . . . . . . . 407, 422
placing at absolute addresses . . . . . . . . . . . . . . . . . . . . 212 --warnings_affect_exit_code (compiler option) . . . . 230, 278
placing in named segments . . . . . . . . . . . . . . . . . . . . . 212 --warnings_are_errors (compiler option) . . . . . . . . . . . . . 278
static --warn_about_c_style_casts (compiler option) . . . . . . . . . 277
placement in memory . . . . . . . . . . . . . . . . . . . . . . . . 53 watchdog reset instruction . . . . . . . . . . . . . . . . . . . . . . . . 349
taking the address of . . . . . . . . . . . . . . . . . . . . . . . . 219 __watchdog_reset (intrinsic function) . . . . . . . . . . . . . . . . 349
variadic macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 wchar_t (data type), adding support for in C . . . . . . . . . . . 283
vector (pragma directive) . . . . . . . . . . . . . . . . . . . . . . 65, 339 wchar.h (library header file) . . . . . . . . . . . . . . . . . . . 364, 367
vector (STL header file) . . . . . . . . . . . . . . . . . . . . . . . . . . 366 wctype.h (library header file) . . . . . . . . . . . . . . . . . . . . . . 365
version weak (pragma directive) . . . . . . . . . . . . . . . . . . . . . . . . . . 339
compiler subversion number . . . . . . . . . . . . . . . . . . . . 357 web sites, recommended . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
identifying C standard in use (__STDC_VERSION__) 357 white-space characters, implementation-defined behavior 400
of compiler (__VER__) . . . . . . . . . . . . . . . . . . . . . . . . 358 write formatter, selecting . . . . . . . . . . . . . . . . . . . . . 144–145
version number __write_array, in stdio.h . . . . . . . . . . . . . . . . . . . . . . . . . . 368
of this guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 __write_buffered (DLIB library function) . . . . . . . . . . . . 105
__version_1 (extended keyword) . . . . . . . . . . . . . . . . . . . 315
__VERSION_1_CALLS__ (predefined symbol) . . . . . . . 358
X
__x (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . . . 316
--version1_calls (compiler option) . . . . . . . . . . . . . . . . . . 276
__xch (intrinsic function) . . . . . . . . . . . . . . . . . . . . . . . . . 349
--version2_calls (compiler option) . . . . . . . . . . . . . . . . . . 277
XLINK errors
--version (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . 276
range error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
--vla (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
segment too long . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
void, pointers to . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
XLINK options
volatile
-O. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
and const, declaring objects . . . . . . . . . . . . . . . . . . . . . 292
-y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
declaring objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
XLINK segment memory types . . . . . . . . . . . . . . . . . . . . . 72
protecting simultaneously accesses variables . . . . . . . . 221
XLINK. See linker
rules for access. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
--xmcra_address (compiler option) . . . . . . . . . . . . . . . . . . 278
__XMEGA_CORE__ (predefined symbol) . . . . . . . . . . . 359
__XMEGA_USB__ (predefined symbol) . . . . . . . . . . . . . 359

452 for AVR
DRX-2
Index
__x_z (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . 316 __device__ (predefined symbol) . . . . . . . . . . . . . . . . . . . . 353

__disable_interrupt (intrinsic function) . . . . . . . . . . . . . . . 343
Y __DLIB_FILE_DESCRIPTOR (configuration symbol) . . 133

__eeprom (extended keyword) . . . . . . . . . . . . . . . . . . . . . 300
-y (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279 __embedded_cplusplus (predefined symbol) . . . . . . . . . . 353
__enable_interrupt (intrinsic function) . . . . . . . . . . . . . . . 343
__exit (library function) . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Z __extended_load_program_memory (intrinsic function) . 343
__ext_io (extended keyword) . . . . . . . . . . . . . . . . . . . . . . 301
__z (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . . . 317
-z (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279 __far (extended keyword) . . . . . . . . . . . . . . . . . . . . . 288, 302
--zero_register (compiler option) . . . . . . . . . . . . . . . . . . . 280 __farflash (extended keyword) . . . . . . . . . . . . . . . . . . . . . 302
__z_x (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . 317 __farfunc (extended keyword) . . . . . . . . . . . . . . . . . . . . . 303
__farfunc (function pointer) . . . . . . . . . . . . . . . . . . . . . . . 287
__FILE__ (predefined symbol) . . . . . . . . . . . . . . . . . . . . . 353
Symbols __flash (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . 304
__fractional_multiply_signed (intrinsic function) . . . . . . . 344
__fractional_multiply_signed_with_unsigned (intrinsic
_Exit (library function) . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
function) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
_exit (library function) . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
__fractional_multiply_unsigned (intrinsic function) . . . . . 344
_formatted_write (library function) . . . . . . . . . . . . . . . . . 143
__FUNCTION__ (predefined symbol) . . . . . . . . . . . 174, 354
_large_write (library function) . . . . . . . . . . . . . . . . . . . . . 143
__func__ (predefined symbol) . . . . . . . . . . . . . . . . . 174, 354
_medium_write (library function) . . . . . . . . . . . . . . . . . . . 144
__generic (extended keyword) . . . . . . . . . . . . . . . . . . . . . 305
_small_write (library function) . . . . . . . . . . . . . . . . . . . . . 144
__gets, in stdio.h . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
__ALIGNOF__ (operator) . . . . . . . . . . . . . . . . . . . . . . . . 170
__get_interrupt_state (intrinsic function) . . . . . . . . . . . . . 344
__asm (language extension) . . . . . . . . . . . . . . . . . . . . . . . 151
__has_constructor, symbol used in library . . . . . . . . . . . . 369
__assignment_by_bitwise_copy_allowed, symbol used
__has_destructor, symbol used in library . . . . . . . . . . . . . 369
in library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
__HAS_EEPROM__ (predefined symbol) . . . . . . . . . . . . 354
__BASE_FILE__ (predefined symbol) . . . . . . . . . . . . . . . 352
__HAS_EIND__ (predefined symbol) . . . . . . . . . . . . . . . 354
__BUILD_NUMBER__ (predefined symbol) . . . . . . . . . 352
__HAS_ELPM__ (predefined symbol) . . . . . . . . . . . . . . . 354
__constrange(), symbol used in library . . . . . . . . . . . . . . . 369
__HAS_ENHANCED_CORE__ (predefined symbol) . . . 355
__construction_by_bitwise_copy_allowed, symbol used
__HAS_FISCR__ (predefined symbol) . . . . . . . . . . . . . . 355
in library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
__HAS_MUL__ (predefined symbol). . . . . . . . . . . . . . . . 355
__COUNTER__ (predefined symbol). . . . . . . . . . . . . . . . 352
__HAS_RAMPD__ (predefined symbol) . . . . . . . . . . . . . 355
__cplusplus (predefined symbol) . . . . . . . . . . . . . . . . . . . 352
__HAS_RAMPX__ (predefined symbol) . . . . . . . . . . . . . 355
__CPU__ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
__HAS_RAMPY__ (predefined symbol) . . . . . . . . . . . . . 355
__CPU__ (predefined symbol) . . . . . . . . . . . . . . . . . . . . . 352
__HAS_RAMPZ__ (predefined symbol) . . . . . . . . . . . . . 356
__DATE__ (predefined symbol). . . . . . . . . . . . . . . . . . . . 353
__huge (extended keyword) . . . . . . . . . . . . . . . . . . . 288, 305
__delay_cycles (intrinsic function) . . . . . . . . . . . . . . . . . . 342
__hugeflash (extended keyword) . . . . . . . . . . . . . . . . . . . 306
__DES_decryption (intrinsic function) . . . . . . . . . . . . . . . 342
__huge_size_t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
__DES_encryption (intrinsic function) . . . . . . . . . . . . . . . 343
__iar_cos_accurate (library routine) . . . . . . . . . . . . . . . . . 118
453
DRX-2
__iar_cos_accuratef (library routine) . . . . . . . . . . . . . . . . 118 __iar_Sin_small (library routine) . . . . . . . . . . . . . . . . . . . 117
__iar_cos_accuratel (library routine) . . . . . . . . . . . . . . . . 118 __iar_sin_small (library routine). . . . . . . . . . . . . . . . . . . . 117
__iar_cos_small (library routine) . . . . . . . . . . . . . . . . . . . 117 __iar_Sin_smallf (library routine) . . . . . . . . . . . . . . . . . . . 117
__iar_cos_smallf (library routine) . . . . . . . . . . . . . . . . . . . 117 __iar_sin_smallf (library routine) . . . . . . . . . . . . . . . . . . . 117
__iar_cos_smalll (library routine) . . . . . . . . . . . . . . . . . . . 117 __iar_Sin_smalll (library routine) . . . . . . . . . . . . . . . . . . . 117
__IAR_DLIB_PERTHREAD_INIT_SIZE (macro) . . . . . 139 __iar_sin_smalll (library routine) . . . . . . . . . . . . . . . . . . . 117
__IAR_DLIB_PERTHREAD_SIZE (macro) . . . . . . . . . . 139 __IAR_SYSTEMS_ICC__ (predefined symbol) . . . . . . . 356
__IAR_DLIB_PERTHREAD_SYMBOL_OFFSET __iar_tan_accurate (library routine) . . . . . . . . . . . . . . . . . 118
(symbolptr) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 __iar_tan_accuratef (library routine). . . . . . . . . . . . . . . . . 118
__iar_exp_small (library routine) . . . . . . . . . . . . . . . . . . . 117 __iar_tan_accuratel (library routine) . . . . . . . . . . . . . . . . . 118
__iar_exp_smallf (library routine) . . . . . . . . . . . . . . . . . . 117 __iar_tan_small (library routine) . . . . . . . . . . . . . . . . . . . 117
__iar_exp_smalll (library routine) . . . . . . . . . . . . . . . . . . 117 __iar_tan_smallf (library routine) . . . . . . . . . . . . . . . . . . . 117
__iar_log_small (library routine) . . . . . . . . . . . . . . . . . . . 117 __iar_tan_smalll (library routine) . . . . . . . . . . . . . . . . . . . 117
__iar_log_smallf (library routine) . . . . . . . . . . . . . . . . . . . 117 __indirect_jump_to (intrinsic function) . . . . . . . . . . . . . . 345
__iar_log_smalll (library routine) . . . . . . . . . . . . . . . . . . . 117 __insert_opcode (intrinsic function) . . . . . . . . . . . . . . . . . 345
__iar_log10_small (library routine) . . . . . . . . . . . . . . . . . 117 __interrupt (extended keyword) . . . . . . . . . . . . . . . . . 65, 307
__iar_log10_smallf (library routine) . . . . . . . . . . . . . . . . . 117 using in pragma directives . . . . . . . . . . . . . . . . . . . . . . 339
__iar_log10_smalll (library routine) . . . . . . . . . . . . . . . . . 117 __intrinsic (extended keyword). . . . . . . . . . . . . . . . . . . . . 307
__iar_Pow (library routine) . . . . . . . . . . . . . . . . . . . . . . . . 118 __io (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . . 307
__iar_Powf (library routine) . . . . . . . . . . . . . . . . . . . . . . . 118 __lac (intrinsic function) . . . . . . . . . . . . . . . . . . . . . . . . . . 345
__iar_Powl (library routine) . . . . . . . . . . . . . . . . . . . . . . . 118 __las (intrinsic function) . . . . . . . . . . . . . . . . . . . . . . . . . . 346
__iar_Pow_accurate (library routine) . . . . . . . . . . . . . . . . 118 __lat (intrinsic function) . . . . . . . . . . . . . . . . . . . . . . . . . . 346
__iar_pow_accurate (library routine) . . . . . . . . . . . . . . . . 118 __LINE__ (predefined symbol) . . . . . . . . . . . . . . . . . . . . 356
__iar_Pow_accuratef (library routine) . . . . . . . . . . . . . . . 118 __load_program_memory (intrinsic function) . . . . . . . . . 346
__iar_pow_accuratef (library routine). . . . . . . . . . . . . . . . 118 __low_level_init . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
__iar_Pow_accuratel (library routine). . . . . . . . . . . . . . . . 118 initialization phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
__iar_pow_accuratel (library routine) . . . . . . . . . . . . . . . . 118 __low_level_init, customizing . . . . . . . . . . . . . . . . . . . . . 122
__iar_pow_small (library routine) . . . . . . . . . . . . . . . . . . 117 __MEMORY_MODEL__ (predefined symbol) . . . . 356–357
__iar_pow_smallf (library routine) . . . . . . . . . . . . . . . . . . 117 __memory_of
__iar_pow_smalll (library routine) . . . . . . . . . . . . . . . . . . 117 operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
__iar_program_start (label). . . . . . . . . . . . . . . . . . . . . . . . 119 symbol used in library . . . . . . . . . . . . . . . . . . . . . . . . . 369
__iar_Sin (library routine) . . . . . . . . . . . . . . . . . . . . 117–118 __monitor (extended keyword) . . . . . . . . . . . . . . . . . . . . . 308
__iar_Sinf (library routine) . . . . . . . . . . . . . . . . . . . . 117–118 __multiply_signed (intrinsic function) . . . . . . . . . . . . . . . 346
__iar_Sinl (library routine) . . . . . . . . . . . . . . . . . . . . 117–118 __multiply_signed_with_unsigned (intrinsic function) . . . 346
__iar_Sin_accurate (library routine) . . . . . . . . . . . . . . . . . 118 __multiply_unsigned (intrinsic function) . . . . . . . . . . . . . 346
__iar_sin_accurate (library routine) . . . . . . . . . . . . . . . . . 118 __near (extended keyword) . . . . . . . . . . . . . . . . . . . . 288, 308
__iar_Sin_accuratef (library routine) . . . . . . . . . . . . . . . . 118 __nearfunc (extended keyword) . . . . . . . . . . . . . . . . . . . . 309
__iar_sin_accuratef (library routine) . . . . . . . . . . . . . . . . . 118 __nearfunc (function pointer) . . . . . . . . . . . . . . . . . . . . . . 287
__iar_Sin_accuratel (library routine) . . . . . . . . . . . . . . . . 118 __nested (extended keyword) . . . . . . . . . . . . . . . . . . . . . . 309
__iar_sin_accuratel (library routine) . . . . . . . . . . . . . . . . . 118

454 for AVR
DRX-2
Index
__noreturn (extended keyword) . . . . . . . . . . . . . . . . . . . . 312 __version_2 (extended keyword) . . . . . . . . . . . . . . . . . . . 315

__no_alloc (extended keyword) . . . . . . . . . . . . . . . . . . . . 310 __version_4 (extended keyword) . . . . . . . . . . . . . . . . . . . 316
__no_alloc_str (operator) . . . . . . . . . . . . . . . . . . . . . . . . . 310 __watchdog_reset (intrinsic function) . . . . . . . . . . . . . . . . 349
__no_alloc_str16 (operator) . . . . . . . . . . . . . . . . . . . . . . . 310 __write_array, in stdio.h . . . . . . . . . . . . . . . . . . . . . . . . . . 368
__no_alloc16 (extended keyword) . . . . . . . . . . . . . . . . . . 310 __write_buffered (DLIB library function) . . . . . . . . . . . . 105
__no_init (extended keyword) . . . . . . . . . . . . . . . . . 223, 311 __x (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . . . 316
__no_operation (intrinsic function) . . . . . . . . . . . . . . . . . . 347 __xch (intrinsic function) . . . . . . . . . . . . . . . . . . . . . . . . . 349
__no_runtime_init (extended keyword) . . . . . . . . . . . . . . 311 __XMEGA_CORE__ (predefined symbol) . . . . . . . . . . . 359
__PRETTY_FUNCTION__ (predefined symbol). . . . . . . 356 __XMEGA_USB__ (predefined symbol) . . . . . . . . . . . . . 359
__printf_args (pragma directive). . . . . . . . . . . . . . . . . . . . 334 __x_z (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . 316
__program_start (label) . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 __z (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . . . 317
__raw (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . 312 __z_x (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . 317
__regvar (extended keyword) . . . . . . . . . . . . . . . . . . . . . . 312 -D (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
__ReportAssert (library function) . . . . . . . . . . . . . . . . . . . 129 -e (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
__require (intrinsic function) . . . . . . . . . . . . . . . . . . . . . . 347 -f (compiler option). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
__require, adding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252 -I (compiler option). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
__restore_interrupt (intrinsic function) . . . . . . . . . . . 347–348 -l (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
__root (extended keyword) . . . . . . . . . . . . . . . . . . . . 270, 313 for creating skeleton code . . . . . . . . . . . . . . . . . . . . . . 153
__ro_placement (extended keyword) . . . . . . . . . . . . . . . . 313 -m (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
__save_interrupt (intrinsic function) . . . . . . . . . . . . . . . . . 348 -O (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
__scanf_args (pragma directive) . . . . . . . . . . . . . . . . . . . . 336 -o (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
__segment_begin (extended operator) . . . . . . . . . . . . . . . 171 -O (XLINK option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
__segment_end (extended operator) . . . . . . . . . . . . . . . . . 171 -r (compiler option). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
__segment_size (extended operator) . . . . . . . . . . . . . . . . . 171 -s (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
__set_interrupt_state (intrinsic function) . . . . . . . . . . . . . 348 -v (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
__sleep (intrinsic function) . . . . . . . . . . . . . . . . . . . . . . . . 349 -y (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
__STDC_VERSION__ (predefined symbol) . . . . . . . . . . 357 -y (XLINK option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
__STDC__ (predefined symbol) . . . . . . . . . . . . . . . . . . . . 357 -z (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
__swap_nibbles (intrinsic function) . . . . . . . . . . . . . . . . . 349 --char_is_signed (compiler option) . . . . . . . . . . . . . . . . . . 241
__task (extended keyword) . . . . . . . . . . . . . . . . . . . . . . . . 314 --char_is_unsigned (compiler option) . . . . . . . . . . . . . . . . 241
__TIMESTAMP__ (predefined symbol) . . . . . . . . . . . . . 358 --clib (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
__TIME__ (predefined symbol) . . . . . . . . . . . . . . . . . . . . 358 --cpu (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
__tiny (extended keyword) . . . . . . . . . . . . . . . . . . . . 287, 314 --cross_call_passes (compiler option) . . . . . . . . . . . . . . . . 242
__tinyflash (extended keyword) . . . . . . . . . . . . . . . . . . . . 315 --c89 (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
__TINY_AVR__ (predefined symbol) . . . . . . . . . . . . . . . 358 --debug (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . 243
__ungetchar, in stdio.h . . . . . . . . . . . . . . . . . . . . . . . . . . . 368 --dependencies (compiler option) . . . . . . . . . . . . . . . . . . . 244
__VA_ARGS__ (preprocessor extension) . . . . . . . . . . . . 167 --diagnostics_tables (compiler option) . . . . . . . . . . . . . . . 246
__version_1 (extended keyword) . . . . . . . . . . . . . . . . . . . 315 --diag_error (compiler option) . . . . . . . . . . . . . . . . . . . . . 245
__VERSION_1_CALLS__ (predefined symbol) . . . . . . . 358 --diag_remark (compiler option) . . . . . . . . . . . . . . . . . . . . 245
455
DRX-2
--diag_suppress (compiler option) . . . . . . . . . . . . . . . . . . 246 --no_inline (compiler option) . . . . . . . . . . . . . . . . . . . . . . 262
--diag_warning (compiler option) . . . . . . . . . . . . . . . . . . . 246 --no_path_in_file_macros (compiler option). . . . . . . . . . . 262
--disable_all_program_memory_load_instructions (compiler --no_rampd (compiler option) . . . . . . . . . . . . . . . . . . . . . . 262
option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 --no_size_constraints (compiler option) . . . . . . . . . . . . . . 262
--disable_direct_mode (compiler option) . . . . . . . . . . . . . 247 --no_static_destruction (compiler option) . . . . . . . . . . . . . 263
--disable_library_knowledge (compiler option) . . . . . . . . 247 --no_system_include (compiler option) . . . . . . . . . . . . . . 263
--disable_mul (compiler option) . . . . . . . . . . . . . . . . . . . . 247 --no_typedefs_in_diagnostics (compiler option) . . . . . . . . 264
--disable_spm (compiler option) . . . . . . . . . . . . . . . . . . . . 248 --no_ubrof_messages (compiler option) . . . . . . . . . . . . . . 264
--discard_unused_publics (compiler option) . . . . . . . . . . . 248 --no_unroll (compiler option) . . . . . . . . . . . . . . . . . . . . . . 265
--dlib (compiler option). . . . . . . . . . . . . . . . . . . . . . . . . . . 248 --no_warnings (compiler option) . . . . . . . . . . . . . . . . . . . 265
--dlib_config (compiler option). . . . . . . . . . . . . . . . . . . . . 249 --no_wrap_diagnostics (compiler option) . . . . . . . . . . . . . 265
--do_cross_call (compiler option) . . . . . . . . . . . . . . . . . . . 250 --omit_types (compiler option) . . . . . . . . . . . . . . . . . . . . . 266
--ec++ (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . 250 --only_stdout (compiler option) . . . . . . . . . . . . . . . . . . . . 266
--eecr_address (compiler option) . . . . . . . . . . . . . . . . . . . 251 --output (compiler option). . . . . . . . . . . . . . . . . . . . . . . . . 267
--eec++ (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . 251 --pending_instantiations (compiler option) . . . . . . . . . . . . 267
--eeprom_size (compiler option) . . . . . . . . . . . . . . . . . . . . 251 --predef_macro (compiler option) . . . . . . . . . . . . . . . . . . . 267
--enable_external_bus (compiler option) . . . . . . . . . . . . . 252 --preinclude (compiler option) . . . . . . . . . . . . . . . . . . . . . 268
--enable_multibytes (compiler option) . . . . . . . . . . . . . . . 252 --preprocess (compiler option) . . . . . . . . . . . . . . . . . . . . . 268
--enable_restrict (compiler option) . . . . . . . . . . . . . . . . . . 253 --relaxed_fp (compiler option) . . . . . . . . . . . . . . . . . . . . . 269
--enhanced_core (compiler option) . . . . . . . . . . . . . . . . . . 253 --remarks (compiler option) . . . . . . . . . . . . . . . . . . . . . . . 269
--error_limit (compiler option) . . . . . . . . . . . . . . . . . . . . . 253 --require_prototypes (compiler option) . . . . . . . . . . . . . . . 270
--force_switch_type (compiler option) . . . . . . . . . . . . . . . 254 --root_variables (compiler option) . . . . . . . . . . . . . . . . . . 270
--guard_calls (compiler option). . . . . . . . . . . . . . . . . . . . . 255 --segment (compiler option) . . . . . . . . . . . . . . . . . . . . . . . 271
--header_context (compiler option) . . . . . . . . . . . . . . . . . . 255 --separate_cluster_for_initialized_variables (compiler
--initializers_in_flash (compiler option) . . . . . . . . . . 256, 275 option) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
--library_module (compiler option) . . . . . . . . . . . . . . . . . 257 --silent (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . 272
--lock_regs (compiler option) . . . . . . . . . . . . . . . . . . . . . . 258 --spmcr_address (compiler option) . . . . . . . . . . . . . . . . . . 273
--macro_positions_in_diagnostics (compiler option) . . . . 258 --strict (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . 273
--memory_model (compiler option) . . . . . . . . . . . . . . . . . 258 --string_literals_in_flash (compiler option). . . . . . . . . . . . 273
--mfc (compiler option). . . . . . . . . . . . . . . . . . . . . . . . . . . 259 --system_include_dir (compiler option) . . . . . . . . . . . . . . 274
--misrac (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . 237 --use_c++_inline (compiler option) . . . . . . . . . . . . . . . . . 274
--misrac_verbose (compiler option) . . . . . . . . . . . . . . . . . 237 --version (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . 276
--misrac1998 (compiler option) . . . . . . . . . . . . . . . . . . . . 237 --version1_calls (compiler option) . . . . . . . . . . . . . . . . . . 276
--misrac2004 (compiler option) . . . . . . . . . . . . . . . . . . . . 237 --version2_calls (compiler option) . . . . . . . . . . . . . . . . . . 277
--module_name (compiler option) . . . . . . . . . . . . . . . . . . 260 --vla (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
--no_call_frame_info (compiler option) . . . . . . . . . . . . . . 260 --warnings_affect_exit_code (compiler option) . . . . 230, 278
--no_clustering (compiler option) . . . . . . . . . . . . . . . . . . . 260 --warnings_are_errors (compiler option) . . . . . . . . . . . . . 278
--no_code_motion (compiler option) . . . . . . . . . . . . . . . . 261 --warn_about_c_style_casts (compiler option) . . . . . . . . . 277
--no_cross_call (compiler option) . . . . . . . . . . . . . . . . . . . 261 --xmcra_address (compiler option) . . . . . . . . . . . . . . . . . . 278
--no_cse (compiler option) . . . . . . . . . . . . . . . . . . . . . . . . 261 --zero_register (compiler option) . . . . . . . . . . . . . . . . . . . 280

456 for AVR
DRX-2
Index
--64bit_doubles (compiler option) . . . . . . . . . . . . . . . . . . 240

--64k_flash (compiler option) . . . . . . . . . . . . . . . . . . . . . . 240
?C_EXIT (assembler label) . . . . . . . . . . . . . . . . . . . . . . . . 147
?C_GETCHAR (assembler label) . . . . . . . . . . . . . . . . . . . 147
?C_PUTCHAR (assembler label) . . . . . . . . . . . . . . . . . . . 147
@ (operator) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
placing at absolute address . . . . . . . . . . . . . . . . . . . . . . 211
placing in segments . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
#include files, specifying . . . . . . . . . . . . . . . . . . . . . 228, 255
#warning message (preprocessor extension) . . . . . . . . . . . 360
%Z replacement string,
implementation-defined behavior . . . . . . . . . . . . . . . . . . . 412
Numerics
32-bits (floating-point format) . . . . . . . . . . . . . . . . . . . . . 286
--64bit_doubles (compiler option) . . . . . . . . . . . . . . . . . . 240
--64k_flash (compiler option) . . . . . . . . . . . . . . . . . . . . . . 240
64-bits (floating-point format) . . . . . . . . . . . . . . . . . . . . . 286
457
DRX-2

EWAVR CompilerGuide PDF

Uploaded by

Copyright:

Available Formats

EWAVR CompilerGuide PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

EWAVR CompilerGuide PDF

Uploaded by

Copyright:

Available Formats

IAR C/C++ Compiler

for Microchip Technology’s

IAR C/C++ Compiler User Guide

Part 1. Using the compiler ....................................................... 31

Introduction to the IAR build tools ........................................................ 33

Developing embedded applications ........................................................ 39

Data storage ...................................................................................................... 53

Linking overview ............................................................................................. 71

Linking your application ................................................................................ 87

The DLIB runtime environment ............................................................. 99

The CLIB runtime environment ............................................................ 141

Assembler language interface ................................................................. 149

Using C ................................................................................................................ 167

Using C++ .......................................................................................................... 175

Application-related considerations ....................................................... 193

Efficient coding for embedded applications ...................................... 205

Part 2. Reference information ........................................... 225

External interface details ............................................................................ 227

Compiler options ........................................................................................... 233

Data representation ...................................................................................... 281

Extended keywords ...................................................................................... 295

Intrinsic functions ........................................................................................... 341

The preprocessor .......................................................................................... 351

C/C++ standard library functions .......................................................... 361

The stack usage control file ...................................................................... 391

Implementation-defined behavior for Standard C ........................ 399

Implementation-defined behavior for C89 ....................................... 415

Index ..................................................................................................................... 429

IAR C/C++ Compiler User Guide

Part 1. Using the compiler ....................................................... 31

Introduction to the IAR build tools ........................................................ 33

The IAR build tools—an overview ................................................... 33

Developing embedded software using IAR build tools .......... 39

Data storage ...................................................................................................... 53

Function-related extensions .............................................................. 63

Linking—an overview ............................................................................ 71

Linking considerations .......................................................................... 87

Introduction to the runtime environment .................................. 99

IAR C/C++ Compiler User Guide

The C-SPY emulated I/O mechanism .............................................. 116

Using a prebuilt runtime library .................................................... 141

Mixing C and assembler ..................................................................... 149

C language overview ........................................................................... 167

Overview—EC++ and EEC++ .......................................................... 175

Stack considerations ........................................................................... 193

IAR C/C++ Compiler User Guide

Interaction between the tools and your application ............. 195

Selecting data types ............................................................................. 205

Part 2. Reference information ........................................... 225

External interface details ............................................................................ 227

Invocation syntax ................................................................................. 227

Options syntax ....................................................................................... 233

IAR C/C++ Compiler User Guide

--memory_model, -m ....................................................................... 258

Data representation ...................................................................................... 281

Alignment ................................................................................................ 281

General syntax rules for extended keywords ........................... 295

IAR C/C++ Compiler User Guide

Summary of extended keywords ................................................... 299