Avr Libc User Manual 1.7.0
Avr Libc User Manual 1.7.0
1.7.0
Contents
1 AVR Libc 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 General information about this library . . . . . . . . . . . . . . . . . 1
1.3 Supported Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.4 avr-libc License . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2 Toolchain Overview 11
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2 FSF and GNU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.3 GCC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.4 GNU Binutils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.5 avr-libc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.6 Building Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.7 AVRDUDE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.8 GDB / Insight / DDD . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.9 AVaRICE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.10 SimulAVR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.11 Utilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.12 Toolchain Distributions (Distros) . . . . . . . . . . . . . . . . . . . . 15
2.13 Open Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4 Memory Sections 21
4.1 The .text Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.2 The .data Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4.3 The .bss Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
9 Benchmarks 52
9.1 A few of libc functions. . . . . . . . . . . . . . . . . . . . . . . . . . 53
9.2 Math functions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
17 Acknowledgments 126
1 AVR Libc
1.1 Introduction
Note:
If you think you’ve found a bug, or have a suggestion for an improvement, ei-
ther in this documentation or in the library itself, please use the bug tracker at
https://savannah.nongnu.org/bugs/?group=avr-libc to ensure
the issue won’t be forgotten.
In general, it has been the goal to stick as best as possible to established standards
while implementing this library. Commonly, this refers to the C library as described by
the ANSI X3.159-1989 and ISO/IEC 9899:1990 ("ANSI-C") standard, as well as parts
of their successor ISO/IEC 9899:1999 ("C99"). Some additions have been inspired by
other standards like IEEE Std 1003.1-1988 ("POSIX.1"), while other extensions are
purely AVR-specific (like the entire program-space string interface).
Unless otherwise noted, functions of this library are not guaranteed to be reentrant. In
particular, any functions that store local state are known to be non-reentrant, as well
as functions that manipulate IO registers like the EEPROM access routines. If these
functions are used within both standard and interrupt contexts undefined behaviour will
result. See the FAQ for a more detailed discussion.
The following is a list of AVR devices currently supported by the library. Note that
actual support for some newer devices depends on the ability of the compiler/assembler
to support these devices at library compile-time.
megaAVR Devices:
• atmega103
• atmega128
• atmega1280
• atmega1281
• atmega1284p
• atmega16
• atmega161
• atmega162
• atmega163
• atmega164a
• atmega164p
• atmega165
• atmega165a
• atmega165p
• atmega168
• atmega168a
• atmega168p
• atmega16a
• atmega2560
• atmega2561
• atmega32
• atmega323
• atmega324a
• atmega324p
• atmega324pa
• atmega325
• atmega325a
• atmega325p
• atmega3250
• atmega3250a
• atmega3250p
• atmega328
• atmega328p
• atmega48
• atmega48a
• atmega48p
• atmega64
• atmega640
• atmega644
• atmega644a
• atmega644p
• atmega644pa
• atmega645
• atmega645a
• atmega645p
• atmega6450
• atmega6450a
• atmega6450p
• atmega8
• atmega88
• atmega88a
• atmega88p
• atmega88pa
• atmega8515
• atmega8535
tinyAVR Devices:
• attiny4
• attiny5
• attiny10
• attiny11 [1]
• attiny12 [1]
• attiny13
• attiny13a
• attiny15 [1]
• attiny20
• attiny22
• attiny24
• attiny24a
• attiny25
• attiny26
• attiny261
• attiny261a
• attiny28 [1]
• attiny2313
• attiny2313a
• attiny40
• attiny4313
• attiny43u
• attiny44
• attiny44a
• attiny45
• attiny461
• attiny461a
• attiny48
• attiny84
• attiny84a
• attiny85
• attiny861
• attiny861a
• attiny87
• attiny88
• atmega64c1
• atmega64m1
• attiny167
• at90can32
• at90can64
• at90can128
• atmega169
• atmega169a
• atmega169p
• atmega169pa
• atmega329
• atmega329a
• atmega329p
• atmega329pa
• atmega3290
• atmega3290a
• atmega3290p
• atmega649
• atmega649a
• atmega6490
• atmega6490a
• atmega6490p
• atmega649p
• at90pwm1
• at90pwm2
• at90pwm2b
• at90pwm216
• at90pwm3
• at90pwm3b
• at90pwm316
• at90pwm81
• atmega8hva
• atmega16hva
• atmega16hva2
• atmega16hvb
• atmega32hvb
• atmega64hve
• atmega406
• at90usb82
• at90usb162
• at90usb646
• at90usb647
• at90usb1286
• at90usb1287
• atmega8u2
• atmega16u2
• atmega16u4
• atmega32u2
• atmega32u4
• atmega32u6
XMEGA Devices:
• atxmega16a4
• atxmega16d4
• atxmega32a4
• atxmega32d4
• atxmega64a1
• atxmega64a1u
• atxmega64a3
• atxmega64d3
• atxmega128a1
• atxmega128a1u
• atxmega128a3
• atxmega128d3
• atxmega192a3
• atxmega192d3
• atxmega256a3
• atxmega256a3b
• atxmega256d3
Miscellaneous Devices:
• at94K [2]
• at76c711 [3]
• at43usb320
• at43usb355
• at86rf401
• at90scr100
• ata6289
• m3000 [4]
• at90s1200 [1]
• at90s2313
• at90s2323
• at90s2333
• at90s2343
• at90s4414
• at90s4433
• at90s4434
• at90s8515
• at90c8534
• at90s8535
Note:
[1] Assembly only. There is no direct support for these devices to be programmed
in C since they do not have a RAM based stack. Still, it could be possible to
program them in C, see the FAQ for an option.
Note:
[2] The at94K devices are a combination of FPGA and AVR microcontroller.
[TRoth-2002/11/12: Not sure of the level of support for these. More information
would be welcomed.]
Note:
[3] The at76c711 is a USB to fast serial interface bridge chip using an AVR core.
Note:
[4] The m3000 is a motor controller AVR ASIC from Intelligent Motion Systems
(IMS) / Schneider Electric.
avr-libc can be freely used and redistributed, provided the following license conditions
are met.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
POSSIBILITY OF SUCH DAMAGE.
2 Toolchain Overview
2.1 Introduction
Welcome to the open source software development toolset for the Atmel AVR!
There is not a single tool that provides everything needed to develop software for the
AVR. It takes many tools working together. Collectively, the group of tools are called a
toolset, or commonly a toolchain, as the tools are chained together to produce the final
executable application for the AVR microcontroller.
The following sections provide an overview of all of these tools. You may be used
to cross-compilers that provide everything with a GUI front-end, and not know what
goes on "underneath the hood". You may be coming from a desktop or server computer
background and not used to embedded systems. Or you may be just learning about the
most common software development toolchain available on Unix and Linux systems.
Hopefully the following overview will be helpful in putting everything in perspective.
According to its website, "the Free Software Foundation (FSF), established in 1985, is
dedicated to promoting computer users’ rights to use, study, copy, modify, and redis-
tribute computer programs. The FSF promotes the development and use of free soft-
ware, particularly the GNU operating system, used widely in its GNU/Linux variant."
The FSF remains the primary sponsor of the GNU project.
The GNU Project was launched in 1984 to develop a complete Unix-like operating
system which is free software: the GNU system. GNU is a recursive acronym for
»GNU’s Not Unix«; it is pronounced guh-noo, approximately like canoe.
One of the main projects of the GNU system is the GNU Compiler Collection, or GCC,
and its sister project, GNU Binutils. These two open source projects provide a foun-
dation for a software development toolchain. Note that these projects were designed to
originally run on Unix-like systems.
2.3 GCC
GCC stands for GNU Compiler Collection. GCC is highly flexible compiler system. It
has different compiler front-ends for different languages. It has many back-ends that
generate assembly code for many different processors and host operating systems. All
share a common "middle-end", containing the generic parts of the compiler, including
a lot of optimizations.
In GCC, a host system is the system (processor/OS) that the compiler runs on. A
target system is the system that the compiler compiles code for. And, a build system
is the system that the compiler is built (from source code) on. If a compiler has the
same system for host and for target, it is known as a native compiler. If a compiler
has different systems for host and target, it is known as a cross-compiler. (And if all
three, build, host, and target systems are different, it is known as a Canadian cross
compiler, but we won’t discuss that here.) When GCC is built to execute on a host
system such as FreeBSD, Linux, or Windows, and it is built to generate code for the
AVR microcontroller target, then it is a cross compiler, and this version of GCC is
commonly known as "AVR GCC". In documentation, or discussion, AVR GCC is
used when referring to GCC targeting specifically the AVR, or something that is AVR
specific about GCC. The term "GCC" is usually used to refer to something generic
about GCC, or about GCC as a whole.
GCC is different from most other compilers. GCC focuses on translating a high-level
language to the target assembly only. AVR GCC has three available compilers for the
AVR: C language, C++, and Ada. The compiler itself does not assemble or link the
final code.
GCC is also known as a "driver" program, in that it knows about, and drives other
programs seamlessly to create the final output. The assembler, and the linker are part
of another open source project called GNU Binutils. GCC knows how to drive the
GNU assembler (gas) to assemble the output of the compiler. GCC knows how to drive
the GNU linker (ld) to link all of the object modules into a final executable.
The two projects, GCC and Binutils, are very much interrelated and many of the same
volunteers work on both open source projects.
When GCC is built for the AVR target, the actual program names are prefixed with
"avr-". So the actual executable name for AVR GCC is: avr-gcc. The name "avr-gcc"
is used in documentation and discussion when referring to the program itself and not
just the whole AVR GCC system.
See the GCC Web Site and GCC User Manual for more information about GCC.
The name GNU Binutils stands for "Binary Utilities". It contains the GNU assembler
(gas), and the GNU linker (ld), but also contains many other utilities that work with
binary files that are created as part of the software development toolchain.
Again, when these tools are built for the AVR target, the actual program names are
prefixed with "avr-". For example, the assembler program name, for a native assembler
is "as" (even though in documentation the GNU assembler is commonly referred to as
"gas"). But when built for an AVR target, it becomes "avr-as". Below is a list of the
programs that are included in Binutils:
avr-as
The Assembler.
avr-ld
The Linker.
avr-ar
Create, modify, and extract from libraries (archives).
avr-ranlib
Generate index to library (archive) contents.
avr-objcopy
avr-objdump
avr-size
List section sizes and total size.
avr-nm
List symbols from object files.
avr-strings
avr-strip
avr-readelf
Display the contents of ELF format files.
avr-addr2line
Convert addresses to file and line.
avr-c++filt
Filter to demangle encoded C++ symbols.
2.5 avr-libc
GCC and Binutils provides a lot of the tools to develop software, but there is one critical
component that they do not provide: a Standard C Library.
There are different open source projects that provide a Standard C Library depending
upon your system time, whether for a native compiler (GNU Libc), for some other
embedded system (newlib), or for some versions of Linux (uCLibc). The open source
AVR toolchain has its own Standard C Library project: avr-libc.
AVR-Libc provides many of the same functions found in a regular Standard C Library
and many additional library functions that is specific to an AVR. Some of the Standard
C Library functions that are commonly used on a PC environment have limitations or
additional issues that a user needs to be aware of when used on an embedded system.
AVR-Libc also contains the most documentation about the whole AVR toolchain.
Even though GCC, Binutils, and avr-libc are the core projects that are used to build
software for the AVR, there is another piece of software that ties it all together: Make.
GNU Make is a program that makes things, and mainly software. Make interprets and
executes a Makefile that is written for a project. A Makefile contains dependency rules,
showing which output files are dependent upon which input files, and instructions on
how to build output files from input files.
Some distributions of the toolchains, and other AVR tools such as MFile, contain a
Makefile template written for the AVR toolchain and AVR applications that you can
copy and modify for your application.
See the GNU Make User Manual for more information.
2.7 AVRDUDE
After creating your software, you’ll want to program your device. You can do this by
using the program AVRDUDE which can interface with various hardware devices to
program your processor.
AVRDUDE is a very flexible package. All the information about AVR processors
and various hardware programmers is stored in a text database. This database can be
modified by any user to add new hardware or to add an AVR processor if it is not
already listed.
The GNU Debugger (GDB) is a command-line debugger that can be used with the rest
of the AVR toolchain. Insight is GDB plus a GUI written in Tcl/Tk. Both GDB and
Insight are configured for the AVR and the main executables are prefixed with the target
name: avr-gdb, and avr-insight. There is also a "text mode" GUI for GDB: avr-gdbtui.
DDD (Data Display Debugger) is another popular GUI front end to GDB, available on
Unix and Linux systems.
2.9 AVaRICE
AVaRICE is a back-end program to AVR GDB and interfaces to the Atmel JTAG In-
Circuit Emulator (ICE), to provide emulation capabilities.
2.10 SimulAVR
2.11 Utilities
There are also other optional utilities available that may be useful to add to your toolset.
SRecord is a collection of powerful tools for manipulating EPROM load files. It
reads and writes numerous EPROM file formats, and can perform many different ma-
nipulations.
All of the various open source projects that comprise the entire toolchain are normally
distributed as source code. It is left up to the user to build the tool application from its
source code. This can be a very daunting task to any potential user of these tools.
Luckily there are people who help out in this area. Volunteers take the time to build the
application from source code on particular host platforms and sometimes packaging
the tools for convenient installation by the end user. These packages contain the binary
executables of the tools, pre-made and ready to use. These packages are known as
"distributions" of the AVR toolchain, or by a more shortened name, "distros".
AVR toolchain distros are available on FreeBSD, Windows, Mac OS X, and certain
flavors of Linux.
All of these tools, from the original source code in the multitude of projects, to the
various distros, are put together by many, many volunteers. All of these projects could
always use more help from other people who are willing to volunteer some of their time.
There are many different ways to help, for people with varying skill levels, abilities,
and available time.
You can help to answer questions in mailing lists such as the avr-gcc-list, or on forums
at the AVR Freaks website. This helps many people new to the open source AVR tools.
If you think you found a bug in any of the tools, it is always a big help to submit a good
bug report to the proper project. A good bug report always helps other volunteers to
analyze the problem and to get it fixed for future versions of the software.
You can also help to fix bugs in various software projects, or to add desirable new
features.
Volunteers are always welcome! :-)
3.1 Introduction
Many of the devices that are possible targets of avr-libc have a minimal amount of
RAM. The smallest parts supported by the C environment come with 128 bytes of
RAM. This needs to be shared between initialized and uninitialized variables (sections
.data and .bss), the dynamic memory allocator, and the stack that is used for calling
subroutines and storing local (automatic) variables.
Also, unlike larger architectures, there is no hardware-supported memory management
which could help in separating the mentioned RAM regions from being overwritten by
each other.
The standard RAM layout is to place .data variables first, from the beginning of the
internal RAM, followed by .bss. The stack is started from the top of internal RAM,
growing downwards. The so-called "heap" available for the dynamic memory allocator
will be placed beyond the end of .bss. Thus, there’s no risk that dynamic memory will
ever collide with the RAM variables (unless there were bugs in the implementation of
the allocator). There is still a risk that the heap and stack could collide if there are large
requirements for either dynamic memory or stack space. The former can even happen
if the allocations aren’t all that large but dynamic memory allocations get fragmented
over time such that new requests don’t quite fit into the "holes" of previously freed
regions. Large stack space requirements can arise in a C function containing large
and/or numerous local variables or when recursively calling function.
Note:
The pictures shown in this document represent typical situations where the RAM
locations refer to an ATmega128. The memory addresses used are not displayed
in a linear scale.
0x0100
0xFFFF
0x10FF
0x1100
.data .bss
variables variables heap
! stack
SP RAMEND
*(__brkval) (<= *SP − *(__malloc_margin))
*(__malloc_heap_start) == __heap_start
__bss_end
__data_end == __bss_start
__data_start
Obviously, the constraints are much harder to satisfy in the default configuration where
only internal RAM is available. Extreme care must be taken to avoid a stack-heap
collision, both by making sure functions aren’t nesting too deeply, and don’t require
too much stack space for local variables, as well as by being cautious with allocating
too much dynamic memory.
If external RAM is available, it is strongly recommended to move the heap into the ex-
ternal RAM, regardless of whether or not the variables from the .data and .bss sections
are also going to be located there. The stack should always be kept in internal RAM.
Some devices even require this, and in general, internal RAM can be accessed faster
since no extra wait states are required. When using dynamic memory allocation and
stack and heap are separated in distinct memory areas, this is the safest way to avoid a
stack-heap collision.
There are a number of variables that can be tuned to adapt the behavior of malloc()
to the expected requirements and constraints of the application. Any changes to these
tunables should be made before the very first call to malloc(). Note that some library
functions might also use dynamic memory (notably those from the <stdio.h>: Stan-
dard IO facilities), so make sure the changes will be done early enough in the startup
sequence.
The variables __malloc_heap_start and __malloc_heap_end can be used
to restrict the malloc() function to a certain memory region. These variables are stati-
cally initialized to point to __heap_start and __heap_end, respectively, where
__heap_start is filled in by the linker to point just beyond .bss, and __heap_end
is set to 0 which makes malloc() assume the heap is below the stack.
If the heap is going to be moved to external RAM, __malloc_heap_end must be
adjusted accordingly. This can either be done at run-time, by writing directly to this
variable, or it can be done automatically at link-time, by adjusting the value of the
symbol __heap_end.
The following example shows a linker command to relocate the entire .data and .bss
segments, and the heap to location 0x1100 in external RAM. The heap will extend up
to address 0xffff.
Note:
See explanation for offset 0x800000. See the chapter about using gcc for the -Wl
options.
The ld (linker) user manual states that using -Tdata=<x> is equivalent to using
--section-start,.data=<x>. However, you have to use --section-start as above be-
cause the GCC frontend also sets the -Tdata option for all MCU types where the
SRAM doesn’t start at 0x800060. Thus, the linker is being faced with two -Tdata
options. Sarting with binutils 2.16, the linker changed the preference, and picks
the "wrong" option in this situation.
0x0100
0xFFFF
0x10FF
0x1100
on−board RAM external RAM
.data .bss
stack variables variables heap
SP *(__malloc_heap_end) == __heap_end
RAMEND *(__brkval)
*(__malloc_heap_start) == __heap_start
__bss_end
__data_end == __bss_start
__data_start
Figure 2: Internal RAM: stack only, external RAM: variables and heap
If dynamic memory should be placed in external RAM, while keeping the variables in
internal RAM, something like the following could be used. Note that for demonstration
purposes, the assignment of the various regions has not been made adjacent in this
example, so there are "holes" below and above the heap in external RAM that remain
completely unaccessible by regular variables or dynamic memory allocations (shown
in light bisque color in the picture below).
external RAM
0x0100
0xFFFF
0x3FFF
0x10FF
0x1100
0x2000
on−board RAM
.data .bss
variables variables stack heap
SP *(__malloc_heap_end) == __heap_end
RAMEND *(__brkval)
__bss_end *(__malloc_heap_start) == __heap_start
__data_end == __bss_start
__data_start
When allocating memory, first the freelist is walked to see if it could satisfy the request.
If there’s a chunk available on the freelist that will fit the request exactly, it will be
taken, disconnected from the freelist, and returned to the caller. If no exact match could
be found, the closest match that would just satisfy the request will be used. The chunk
will normally be split up into one to be returned to the caller, and another (smaller)
one that will remain on the freelist. In case this chunk was only up to two bytes larger
than the request, the request will simply be altered internally to also account for these
additional bytes since no separate freelist entry could be split off in that case.
If nothing could be found on the freelist, heap extension is attempted. This is where
__malloc_margin will be considered if the heap is operating below the stack, or
where __malloc_heap_end will be verified otherwise.
If the remaining memory is insufficient to satisfy the request, NULL will eventually be
returned to the caller.
When calling free(), a new freelist entry will be prepared. An attempt is then made to
aggregate the new entry with possible adjacent entries, yielding a single larger entry
available for further allocations. That way, the potential for heap fragmentation is
hopefully reduced. When deallocating the topmost chunk of memory, the size of the
heap is reduced.
A call to realloc() first determines whether the operation is about to grow or shrink the
current allocation. When shrinking, the case is easy: the existing chunk is split, and the
tail of the region that is no longer to be used is passed to the standard free() function for
insertion into the freelist. Checks are first made whether the tail chunk is large enough
to hold a chunk of its own at all, otherwise realloc() will simply do nothing, and return
the original region.
When growing the region, it is first checked whether the existing allocation can be ex-
tended in-place. If so, this is done, and the original pointer is returned without copying
any data contents. As a side-effect, this check will also record the size of the largest
chunk on the freelist.
If the region cannot be extended in-place, but the old chunk is at the top of heap, and
the above freelist walk did not reveal a large enough chunk on the freelist to satisfy
the new request, an attempt is made to quickly extend this topmost chunk (and thus
the heap), so no need arises to copy over the existing data. If there’s no more space
available in the heap (same check is done as in malloc()), the entire request will fail.
Otherwise, malloc() will be called with the new request size, the existing data will be
copied over, and free() will be called on the old region.
4 Memory Sections
Remarks:
Need to list all the sections which are available to the avr.
Weak Bindings
The .text section contains the actual machine instructions which make up your program.
This section is further subdivided by the .initN and .finiN sections dicussed below.
Note:
The avr-size program (part of binutils), coming from a Unix background,
doesn’t account for the .data initialization space added to the .text section, so in
order to know how much flash the final program will consume, one needs to add
the values for both, .text and .data (but not .bss), while the amount of pre-allocated
SRAM is the sum of .data and .bss.
This section contains static data which was defined in your code. Things like the fol-
lowing would end up in .data:
struct point pt = { 1, 1 };
It is possible to tell the linker the SRAM address of the beginning of the .data section.
This is accomplished by adding -Wl,-Tdata,addr to the avr-gcc command
used to the link your program. Not that addr must be offset by adding 0x800000
the to real SRAM address so that the linker knows that the address is in the SRAM
memory space. Thus, if you want the .data section to start at 0x1100, pass 0x801100
at the address to the linker. [offset explained]
Note:
When using malloc() in the application (which could even happen inside library
calls), additional adjustments are required.
This sections is a part of the .bss section. What makes the .noinit section special is that
variables which are defined as such:
will not be initialized to zero during startup as would normal .bss data.
Only uninitialized variables can be placed in the .noinit section. Thus, the following
code will cause avr-gcc to issue an error:
It is possible to tell the linker explicitly where to place the .noinit section by adding
-Wl,--section-start=.noinit=0x802000 to the avr-gcc command line
at the linking stage. For example, suppose you wish to place the .noinit section at
SRAM address 0x2000:
Note:
Because of the Harvard architecture of the AVR devices, you must manually add
0x800000 to the address you pass to the linker as the start of the section. Oth-
erwise, the linker thinks you want to put the .noinit section into the .text section
instead of .data/.bss and will complain.
Alternatively, you can write your own linker script to automate this. [FIXME: need an
example or ref to dox for writing linker scripts.]
These sections are used to define the startup code from reset up through the start of
main(). These all are subparts of the .text section.
The purpose of these sections is to allow for more specific placement of code within
your program.
Note:
Sometimes, it is convenient to think of the .initN and .finiN sections as functions,
but in reality they are just symbolic names which tell the linker where to stick a
chunk of code which is not a function. Notice that the examples for asm and C can
not be called as functions and should not be jumped into.
.init0:
Weakly bound to __init(). If user defines __init(), it will be jumped into immedi-
ately after a reset.
.init1:
Unused. User definable.
.init2:
In C programs, weakly bound to initialize the stack, and to clear __zero_reg__
(r1).
.init3:
Unused. User definable.
.init4:
For devices with > 64 KB of ROM, .init4 defines the code which takes care of copying
the contents of .data from the flash to SRAM. For all other devices, this code as well
as the code to zero out the .bss section is loaded from libgcc.a.
.init5:
Unused. User definable.
.init6:
Unused for C programs, but used for constructors in C++ programs.
.init7:
Unused. User definable.
.init8:
Unused. User definable.
.init9:
Jumps into main().
These sections are used to define the exit code executed after return from main() or a
call to exit(). These all are subparts of the .text section.
The .finiN sections are executed in descending order from 9 to 0.
.finit9:
Unused. User definable. This is effectively where _exit() starts.
.fini8:
Unused. User definable.
.fini7:
Unused. User definable.
.fini6:
Unused for C programs, but used for destructors in C++ programs.
.fini5:
Unused. User definable.
.fini4:
Unused. User definable.
.fini3:
Unused. User definable.
.fini2:
Unused. User definable.
.fini1:
Unused. User definable.
.fini0:
Goes into an infinite loop after program termination and completion of any _exit()
code (execution of code in the .fini9 -> .fini1 sections).
Example:
#include <avr/io.h>
.section .init1,"ax",@progbits
ldi r0, 0xff
out _SFR_IO_ADDR(PORTB), r0
out _SFR_IO_ADDR(DDRB), r0
Note:
The ,"ax",@progbits tells the assembler that the section is allocatable ("a"),
executable ("x") and contains data ("@progbits"). For more detailed information
on the .section directive, see the gas user manual.
Example:
#include <avr/io.h>
void
my_init_portb (void)
{
PORTB = 0xff;
DDRB = 0xff;
}
Note:
Section .init3 is used in this example, as this ensures the inernal __zero_reg_-
_ has already been set up. The code generated by the compiler might blindly rely
on __zero_reg__ being really 0.
5.1 Introduction
So you have some constant data and you’re running out of room to store it? Many
AVRs have limited amount of RAM in which to store data, but may have more Flash
space available. The AVR is a Harvard architecture processor, where Flash is used for
the program, RAM is used for data, and they each have separate address spaces. It is
a challenge to get constant data to be stored in the Program Space, and to retrieve that
data to use it in the AVR application.
The problem is exacerbated by the fact that the C Language was not designed for
Harvard architectures, it was designed for Von Neumann architectures where code and
data exist in the same address space. This means that any compiler for a Harvard
architecture processor, like the AVR, has to use other means to operate with separate
address spaces.
Some compilers use non-standard C language keywords, or they extend the standard
syntax in ways that are non-standard. The AVR toolset takes a different approach.
GCC has a special keyword, __attribute__ that is used to attach different at-
tributes to things such as function declarations, variables, and types. This keyword is
followed by an attribute specification in double parentheses. In AVR GCC, there is a
special attribute called progmem. This attribute is use on data declarations, and tells
the compiler to place the data in the Program Memory (Flash).
AVR-Libc provides a simple macro PROGMEM that is defined as the attribute syn-
tax of GCC with the progmem attribute. This macro was created as a convenience
to the end user, as we will see below. The PROGMEM macro is defined in the
<avr/pgmspace.h> system header file.
It is difficult to modify GCC to create new extensions to the C language syntax, so
instead, avr-libc has created macros to retrieve the data from the Program Space. These
macros are also found in the <avr/pgmspace.h> system header file.
Many users bring up the idea of using C’s keyword const as a means of declaring
data to be in Program Space. Doing this would be an abuse of the intended meaning of
the const keyword.
const is used to tell the compiler that the data is to be "read-only". It is used to help
make it easier for the compiler to make certain transformations, or to help the compiler
check for incorrect usage of those variables.
For example, the const keyword is commonly used in many functions as a modifier on
the parameter type. This tells the compiler that the function will only use the parameter
as read-only and will not modify the contents of the parameter variable.
const was intended for uses such as this, not as a means to identify where the data
should be stored. If it were used as a means to define data storage, then it loses its
correct meaning (changes its semantics) in other situations such as in the function pa-
rameter example.
{0x0A,0x0B,0x0C,0x0D,0x0E,0x0F,0x10,0x11,0x12,0x13},
{0x14,0x15,0x16,0x17,0x18,0x19,0x1A,0x1B,0x1C,0x1D},
{0x1E,0x1F,0x20,0x21,0x22,0x23,0x24,0x25,0x26,0x27},
{0x28,0x29,0x2A,0x2B,0x2C,0x2D,0x2E,0x2F,0x30,0x31},
{0x32,0x33,0x34,0x35,0x36,0x37,0x38,0x39,0x3A,0x3B},
{0x3C,0x3D,0x3E,0x3F,0x40,0x41,0x42,0x43,0x44,0x45},
{0x46,0x47,0x48,0x49,0x4A,0x4B,0x4C,0x4D,0x4E,0x4F},
{0x50,0x51,0x52,0x53,0x54,0x55,0x56,0x57,0x58,0x59},
{0x5A,0x5B,0x5C,0x5D,0x5E,0x5F,0x60,0x61,0x62,0x63},
{0x64,0x65,0x66,0x67,0x68,0x69,0x6A,0x6B,0x6C,0x6D}
};
and later in your code you access this data in a function and store a single byte into a
variable like so:
byte = mydata[i][j];
Now you want to store your data in Program Memory. Use the PROGMEM macro found
in <avr/pgmspace.h> and put it after the declaration of the variable, but before
the initializer, like so:
#include <avr/pgmspace.h>
.
.
.
unsigned char mydata[11][10] PROGMEM =
{
{0x00,0x01,0x02,0x03,0x04,0x05,0x06,0x07,0x08,0x09},
{0x0A,0x0B,0x0C,0x0D,0x0E,0x0F,0x10,0x11,0x12,0x13},
{0x14,0x15,0x16,0x17,0x18,0x19,0x1A,0x1B,0x1C,0x1D},
{0x1E,0x1F,0x20,0x21,0x22,0x23,0x24,0x25,0x26,0x27},
{0x28,0x29,0x2A,0x2B,0x2C,0x2D,0x2E,0x2F,0x30,0x31},
{0x32,0x33,0x34,0x35,0x36,0x37,0x38,0x39,0x3A,0x3B},
{0x3C,0x3D,0x3E,0x3F,0x40,0x41,0x42,0x43,0x44,0x45},
{0x46,0x47,0x48,0x49,0x4A,0x4B,0x4C,0x4D,0x4E,0x4F},
{0x50,0x51,0x52,0x53,0x54,0x55,0x56,0x57,0x58,0x59},
{0x5A,0x5B,0x5C,0x5D,0x5E,0x5F,0x60,0x61,0x62,0x63},
{0x64,0x65,0x66,0x67,0x68,0x69,0x6A,0x6B,0x6C,0x6D}
};
That’s it! Now your data is in the Program Space. You can compile, link, and check
the map file to verify that mydata is placed in the correct section.
Now that your data resides in the Program Space, your code to access (read) the data
will no longer work. The code that gets generated will retrieve the data that is located
at the address of the mydata array, plus offsets indexed by the i and j variables.
However, the final address that is calculated where to the retrieve the data points to
the Data Space! Not the Program Space where the data is actually located. It is likely
that you will be retrieving some garbage. The problem is that AVR GCC does not
intrinsically know that the data resides in the Program Space.
The solution is fairly simple. The "rule of thumb" for accessing data stored in the
Program Space is to access the data as you normally would (as if the variable is stored
in Data Space), like so:
byte = mydata[i][j];
byte = &(mydata[i][j]);
then use the appropriate pgm_read_∗ macro, and the address of your data becomes
the parameter to that macro:
byte = pgm_read_byte(&(mydata[i][j]));
The pgm_read_∗ macros take an address that points to the Program Space, and re-
trieves the data that is stored at that address. This is why you take the address of the
offset into the array. This address becomes the parameter to the macro so it can gen-
erate the correct code to retrieve the data from the Program Space. There are different
pgm_read_∗ macros to read different sizes of data at the address given.
Now that you can successfully store and retrieve simple data from Program Space you
want to store and retrive strings from Program Space. And specifically you want to
store and array of strings to Program Space. So you start off with your array, like so:
char *string_table[] =
{
"String 1",
"String 2",
"String 3",
"String 4",
"String 5"
};
and then you add your PROGMEM macro to the end of the declaration:
Right? WRONG!
Unfortunately, with GCC attributes, they affect only the declaration that they are at-
tached to. So in this case, we successfully put the string_table variable, the array
itself, in the Program Space. This DOES NOT put the actual strings themselves into
Program Space. At this point, the strings are still in the Data Space, which is probably
not what you want.
In order to put the strings in Program Space, you have to have explicit declarations for
each string, and put each string in Program Space:
Now this has the effect of putting string_table in Program Space, where
string_table is an array of pointers to characters (strings), where each pointer
is a pointer to the Program Space, where each string is also stored.
The PGM_P type above is also a macro that defined as a pointer to a character in the
Program Space.
Retrieving the strings are a different matter. You probably don’t want to pull the string
out of Program Space, byte by byte, using the pgm_read_byte() macro. There are
other functions declared in the <avr/pgmspace.h> header file that work with strings
that are stored in the Program Space.
For example if you want to copy the string from Program Space to a buffer in RAM
(like an automatic variable inside a function, that is allocated on the stack), you can do
this:
void foo(void)
{
char buffer[10];
5.5 Caveats
The macros and functions used to retrieve data from the Program Space have to gen-
erate some extra code in order to actually load the data from the Program Space. This
incurs some extra overhead in terms of code space (extra opcodes) and execution time.
Usually, both the space and time overhead is minimal compared to the space savings
of putting data in Program Space. But you should be aware of this so you can mini-
mize the number of calls within a single function that gets the same piece of data from
Program Space. It is always instructive to look at the resulting disassembly from the
compiler.
6.1 Introduction
There might be several reasons to write code for AVR microcontrollers using plain
assembler source code. Among them are:
• Code for devices that do not have RAM and are thus not supported by the C
compiler.
Usually, all but the first could probably be done easily using the inline assembler facility
of the compiler.
Although avr-libc is primarily targeted to support programming AVR microcontrollers
using the C (and C++) language, there’s limited support for direct assembler usage as
well. The benefits of it are:
• Use of the C preprocessor and thus the ability to use the same symbolic constants
that are available to C programs, as well as a flexible macro concept that can use
any valid C identifier as a macro (whereas the assembler’s macro concept is
basically targeted to use a macro in place of an assembler instruction).
• Use of the runtime framework like automatically assigning interrupt vectors. For
devices that have RAM, initializing the RAM variables can also be utilized.
For the purpose described in this document, the assembler and linker are usually not
invoked manually, but rather using the C compiler frontend (avr-gcc) that in turn
will call the assembler and linker as required.
This approach has the following advantages:
• The invokation of the C preprocessor will be automatic, and will include the
appropriate options to locate required include files in the filesystem.
• The invokation of the linker will be automatic, and will include the appropri-
ate options to locate additional libraries as well as the application start-up code
(crtXXX.o) and linker script.
Note that the invokation of the C preprocessor will be automatic when the filename
provided for the assembler file ends in .S (the capital letter "s"). This would even apply
to operating systems that use case-insensitive filesystems since the actual decision is
made based on the case of the filename suffix given on the command-line, not based on
the actual filename from the file system.
Alternatively, the language can explicitly be specified using the -x
assembler-with-cpp option.
The following annotated example features a simple 100 kHz square wave generator
using an AT90S1200 clocked with a 10.7 MHz crystal. Pin PD6 will be used for the
square wave output.
inttmp = 19
intsav = 0
; Note [4]:
tmconst= 10700000 / 200000 ; 100 kHz => 200000 edges/s
fuzz= 8 ; # clocks in ISR until TCNT0 is set
.section .text
ioinit:
sbi _SFR_IO_ADDR(DDRD), SQUARE
sei
ret
.end
Note [1]
As in C programs, this includes the central processor-specific file containing the IO port
definitions for the device. Note that not all include files can be included into assembler
sources.
Note [2]
#define work 16
Note [3]
Our bit number for the square wave output. Note that the right-hand side consists of a
CPP macro which will be substituted by its value (6 in this case) before actually being
passed to the assembler.
Note [4]
The assembler uses integer operations in the host-defined integer size (32 bits or longer)
when evaluating expressions. This is in contrast to the C compiler that uses the C type
int by default in order to calculate constant integer expressions.
In order to get a 100 kHz output, we need to toggle the PD6 line 200000 times per
second. Since we use timer 0 without any prescaling options in order to get the de-
sired frequency and accuracy, we already run into serious timing considerations: while
accepting and processing the timer overflow interrupt, the timer already continues to
count. When pre-loading the TCCNT0 register, we therefore have to account for the
number of clock cycles required for interrupt acknowledge and for the instructions to
reload TCCNT0 (4 clock cycles for interrupt acknowledge, 2 cycles for the jump from
the interrupt vector, 2 cycles for the 2 instructions that reload TCCNT0). This is what
the constant fuzz is for.
Note [5]
Note [6]
The main loop is just a single jump back to itself. Square wave generation itself is
completely handled by the timer 0 overflow interrupt service. A sleep instruction
(using idle mode) could be used as well, but probably would not conserve much energy
anyway since the interrupt service is executed quite frequently.
Note [7]
Interrupt functions can get the usual names that are also available to C programs. The
linker will then put them into the appropriate interrupt vector slots. Note that they must
be declared .global in order to be acceptable for this purpose. This will only work if
<avr/io.h> has been included. Note that the assembler or linker have no chance
to check the correct spelling of an interrupt function, so it should be double-checked.
(When analyzing the resulting object file using avr-objdump or avr-nm, a name
like __vector_N should appear, with N being a small integer number.)
Note [8]
As explained in the section about special function registers, the actual IO port address
should be obtained using the macro _SFR_IO_ADDR. (The AT90S1200 does not have
RAM thus the memory-mapped approach to access the IO registers is not available. It
would be slower than using in / out instructions anyway.)
Since the operation to reload TCCNT0 is time-critical, it is even performed before
saving SREG. Obviously, this requires that the instructions involved would not change
any of the flag bits in SREG.
Note [9]
Interrupt routines must not clobber the global CPU state. Thus, it is usually necessary
to save at least the state of the flag bits in SREG. (Note that this serves as an example
here only since actually, all the following instructions would not modify SREG either,
but that’s not commonly the case.)
Also, it must be made sure that registers used inside the interrupt routine do not conflict
with those used outside. In the case of a RAM-less device like the AT90S1200, this can
only be done by agreeing on a set of registers to be used exclusively inside the interrupt
routine; there would not be any other chance to "save" a register anywhere.
If the interrupt routine is to be linked together with C modules, care must be taken
to follow the register usage guidelines imposed by the C compiler. Also, any register
modified inside the interrupt sevice needs to be saved, usually on the stack.
Note [10]
As explained in Interrupts, a global "catch-all" interrupt handler that gets all unassigned
interrupt vectors can be installed using the name __vector_default. This must
be .global, and obviously, should end in a reti instruction. (By default, a jump to
location 0 would be implied instead.)
The available pseudo-ops in the assembler are described in the GNU assembler (gas)
manual. The manual can be found online as part of the current binutils release under
http://sources.redhat.com/binutils/.
As gas comes from a Unix origin, its pseudo-op and overall assembler syntax is slightly
different than the one being used by other assemblers. Numeric constants follow the C
notation (prefix 0x for hexadecimal constants), expressions use a C-like syntax.
Some common pseudo-ops include:
• .global (or .globl) declares a public symbol that is visible to the linker (e. g.
function entry point, global variable)
Note that .org is available in gas as well, but is a fairly pointless pseudo-op in an as-
sembler environment that uses relocatable object files, as it is the linker that determines
the final position of some object in ROM or RAM.
Along with the architecture-independent standard operators, there are some AVR-
specific operators available which are unfortunately not yet described in the official
documentation. The most notable operators are:
Example:
This passes the address of function somefunc as the first parameter to function
something.
The GNU C compiler for Atmel AVR RISC processors offers, to embed assembly
language code into C programs. This cool feature may be used for manually optimizing
time critical parts of the software or to use specific processor instruction, which are not
available in the C language.
Because of a lack of documentation, especially for the AVR version of the compiler, it
may take some time to figure out the implementation details by studying the compiler
and assembler source code. There are also a few sample programs available in the net.
Hopefully this document will help to increase their number.
It’s assumed, that you are familiar with writing AVR assembler programs, because this
is not an AVR assembler programming tutorial. It’s not a C language tutorial either.
Note that this document does not cover file written completely in assembler language,
refer to avr-libc and assembler programs for this.
Copyright (C) 2001-2002 by egnite Software GmbH
Permission is granted to copy and distribute verbatim copies of this manual provided
that the copyright notice and this permission notice are preserved on all copies. Permis-
sion is granted to copy and distribute modified versions of this manual provided that
the entire resulting derived work is distributed under the terms of a permission notice
identical to this one.
This document describes version 3.3 of the compiler. There may be some parts, which
hadn’t been completely understood by the author himself and not all samples had been
tested so far. Because the author is German and not familiar with the English language,
there are definitely some typos and syntax errors in the text. As a programmer the
author knows, that a wrong documentation sometimes might be worse than none. Any-
way, he decided to offer his little knowledge to the public, in the hope to get enough
response to improve this document. Feel free to contact the author via e-mail. For the
latest release check http://www.ethernut.de/.
Herne, 17th of May 2002 Harald Kipp harald.kipp-at-egnite.de
Note:
As of 26th of July 2002, this document has been merged into the
documentation for avr-libc. The latest version is now available at
http://savannah.nongnu.org/projects/avr-libc/.
Each asm statement is devided by colons into (up to) four parts:
2. A list of output operands, separated by commas. Our example uses just one:
"=r" (value)
3. A comma separated list of input operands. Again our example uses one operand
only:
"I" (_SFR_IO_ADDR(PORTD))
You can write assembler instructions in much the same way as you would write assem-
bler programs. However, registers and constants are used in a different way if they refer
to expressions of your C program. The connection between registers and C operands is
specified in the second and third part of the asm instruction, the list of input and output
operands, respectively. The general form is
In the code section, operands are referenced by a percent sign followed by a single digit.
0 refers to the first 1 to the second operand and so forth. From the above example:
0 refers to "=r" (value) and
1 refers to "I" (_SFR_IO_ADDR(PORTD)).
This may still look a little odd now, but the syntax of an operand list will be explained
soon. Let us first examine the part of a compiler listing which may have been generated
from our example:
lds r24,value
/* #APP */
in r24, 12
/* #NOAPP */
sts value,r24
The comments have been added by the compiler to inform the assembler that the in-
cluded code was not generated by the compilation of C statements, but by inline as-
sembler statements. The compiler selected register r24 for storage of the value read
from PORTD. The compiler could have selected any other register, though. It may not
explicitely load or store the value and it may even decide not to include your assembler
code at all. All these decisions are part of the compiler’s optimization strategy. For
example, if you never use the variable value in the remaining part of the C program,
the compiler will most likely remove your code unless you switched off optimization.
To avoid this, you can add the volatile attribute to the asm statement:
Alternatively, operands can be given names. The name is prepended in brackets to the
constraints in the operand list, and references to the named operand use the bracketed
name instead of a number after the % sign. Thus, the above example could also be
written as
The last part of the asm instruction, the clobber list, is mainly used to tell the compiler
about modifications done by the assembler code. This part may be omitted, all other
parts are required, but may be left empty. If your assembler routine won’t use any
input or output operand, two colons must still follow the assembler code string. A
good example is a simple statement to disable interrupts:
asm volatile("cli"::);
You can use the same assembler instruction mnemonics as you’d use with any other
AVR assembler. And you can write as many assembler statements into one code string
as you like and your flash memory is able to hold.
Note:
The available assembler directives vary from one assembler to another.
To make it more readable, you should put each statement on a seperate line:
asm volatile("nop\n\t"
"nop\n\t"
"nop\n\t"
"nop\n\t"
::);
The linefeed and tab characters will make the assembler listing generated by the com-
piler more readable. It may look a bit odd for the first time, but that’s the way the
compiler creates it’s own assembler code.
You may also make use of some special registers.
Symbol Register
__SREG__ Status register at address 0x3F
__SP_H__ Stack pointer high byte at address 0x3E
__SP_L__ Stack pointer low byte at address 0x3D
__tmp_reg__ Register r0, used for temporary storage
__zero_reg__ Register r1, always zero
Register r0 may be freely used by your assembler code and need not be restored at
the end of your code. It’s a good idea to use __tmp_reg__ and __zero_reg__
instead of r0 or r1, just in case a new compiler version changes the register usage
definitions.
Note:
The most up-to-date and detailed information on contraints for the avr can be found
in the gcc manual.
The x register is r27:r26, the y register is r29:r28, and the z register is
r31:r30
The following table shows all AVR assembler mnemonics which require operands, and
the related contraints. Because of the improper constraint definitions in version 3.3,
they aren’t strict enough. There is, for example, no constraint, which restricts integer
constants to the range 0 to 7 for bit set and bit clear operations.
Mnemonic Constraints Mnemonic Constraints
adc r,r add r,r
adiw w,I and r,r
andi d,M asr r
bclr I bld r,I
brbc I,label brbs I,label
bset I bst r,I
cbi I,I cbr d,I
com r cp r,r
cpc r,r cpi d,M
cpse r,r dec r
elpm t,z eor r,r
in r,I inc r
ld r,e ldd r,b
ldi d,M lds r,label
lpm t,z lsl r
lsr r mov r,r
movw r,r mul r,r
neg r or r,r
ori d,M out I,r
pop r push r
rol r ror r
sbc r,r sbci d,M
sbi I,I sbic I,I
sbiw w,I sbr d,M
sbrc r,I sbrs r,I
ser d st e,r
std b,r sts label,r
sub r,r subi d,M
swap r
Constraint characters may be prepended by a single constraint modifier. Contraints
without a modifier specify read-only operands. Modifiers are:
Modifier Specifies
= Write-only operand, usually used for all
output operands.
+ Read-write operand
& Register should be used for output only
Output operands must be write-only and the C expression result must be an lvalue,
which means that the operands must be valid on the left side of assignments. Note,
that the compiler will not check if the operands are of reasonable type for the kind of
This statement will swap the nibbles of an 8-bit variable named value. Constraint "0"
tells the compiler, to use the same input register as for the first operand. Note however,
that this doesn’t automatically imply the reverse case. The compiler may choose the
same registers for input and output, even if not told to do so. This is not a problem in
most cases, but may be fatal if the output operator is modified by the assembler code
before the input operator is used. In the situation where your code depends on different
registers used for input and output operands, you must add the & constraint modifier to
your output operand. The following example demonstrates this problem:
In this example an input value is read from a port and then an output value is written to
the same port. If the compiler would have choosen the same register for input and out-
put, then the output value would have been destroyed on the first assembler instruction.
Fortunately, this example uses the & constraint modifier to instruct the compiler not to
select any register for the output value, which is used for any of the input operands.
Back to swapping. Here is the code to swap high and low byte of a 16-bit value:
First you will notice the usage of register __tmp_reg__, which we listed among
other special registers in the Assembler Code section. You can use this register without
saving its contents. Completely new are those letters A and B in %A0 and %B0. In fact
they refer to two different 8-bit registers, both containing a part of value.
Another example to swap bytes of a 32-bit value:
Instead of listing the same operand as both, input and output operand, it can also be
declared as a read-write operand. This must be applied to an output operand, and the
respective input operand list remains empty:
If operands do not fit into a single register, the compiler will automatically assign
enough registers to hold the entire operand. In the assembler code you use %A0 to refer
to the lowest byte of the first operand, %A1 to the lowest byte of the second operand
and so on. The next byte of the first operand will be %B0, the next byte %C0 and so on.
This also implies, that it is often neccessary to cast the type of an input operand to the
desired size.
A final problem may arise while using pointer register pairs. If you define an input
operand
"e" (ptr)
ld r24,Z
If you write
ld r24, %a0
with a lower case a following the percent sign, then the compiler will create the proper
assembler line.
7.4 Clobbers
As stated previously, the last part of the asm statement, the list of clobbers, may be
omitted, including the colon seperator. However, if you are using registers, which
had not been passed as operands, you need to inform the compiler about this. The
following example will do an atomic increment. It increments an 8-bit value pointed
to by a pointer variable in one go, without being interrupted by an interrupt routine
or another thread in a multithreaded environment. Note, that we must use a pointer,
because the incremented value needs to be stored before interrupts are enabled.
asm volatile(
"cli" "\n\t"
"ld r24, %a0" "\n\t"
"inc r24" "\n\t"
"st %a0, r24" "\n\t"
"sei" "\n\t"
:
: "e" (ptr)
: "r24"
);
cli
ld r24, Z
inc r24
st Z, r24
sei
One easy solution to avoid clobbering register r24 is, to make use of the special tem-
porary register __tmp_reg__ defined by the compiler.
asm volatile(
"cli" "\n\t"
"ld __tmp_reg__, %a0" "\n\t"
"inc __tmp_reg__" "\n\t"
"st %a0, __tmp_reg__" "\n\t"
"sei" "\n\t"
:
: "e" (ptr)
);
The compiler is prepared to reload this register next time it uses it. Another problem
with the above code is, that it should not be called in code sections, where interrupts
are disabled and should be kept disabled, because it will enable interrupts at the end.
We may store the current status, but then we need another register. Again we can solve
this without clobbering a fixed, but let the compiler select it. This could be done with
the help of a local C variable.
{
uint8_t s;
asm volatile(
"in %0, __SREG__" "\n\t"
"cli" "\n\t"
"ld __tmp_reg__, %a1" "\n\t"
"inc __tmp_reg__" "\n\t"
"st %a1, __tmp_reg__" "\n\t"
"out __SREG__, %0" "\n\t"
: "=&r" (s)
: "e" (ptr)
);
}
Now every thing seems correct, but it isn’t really. The assembler code modifies the
variable, that ptr points to. The compiler will not recognize this and may keep its
value in any of the other registers. Not only does the compiler work with the wrong
value, but the assembler code does too. The C program may have modified the value
too, but the compiler didn’t update the memory location for optimization reasons. The
worst thing you can do in this case is:
{
uint8_t s;
asm volatile(
"in %0, __SREG__" "\n\t"
"cli" "\n\t"
"ld __tmp_reg__, %a1" "\n\t"
"inc __tmp_reg__" "\n\t"
"st %a1, __tmp_reg__" "\n\t"
"out __SREG__, %0" "\n\t"
: "=&r" (s)
: "e" (ptr)
: "memory"
);
}
The special clobber "memory" informs the compiler that the assembler code may mod-
ify any memory location. It forces the compiler to update all variables for which the
contents are currently held in a register before executing the assembler code. And of
course, everything has to be reloaded again after this code.
In most situations, a much better solution would be to declare the pointer destination
itself volatile:
This way, the compiler expects the value pointed to by ptr to be changed and will
load it whenever used and store it whenever modified.
Situations in which you need clobbers are very rare. In most cases there will be better
ways. Clobbered registers will force the compiler to store their values before and reload
them after your assembler code. Avoiding clobbers gives the compiler more freedom
while optimizing your code.
In order to reuse your assembler language parts, it is useful to define them as macros
and put them into include files. AVR Libc comes with a bunch of them, which could be
found in the directory avr/include. Using such include files may produce compiler
warnings, if they are used in modules, which are compiled in strict ANSI mode. To
avoid that, you can write __asm__ instead of asm and __volatile__ instead of
volatile. These are equivalent aliases.
Another problem with reused macros arises if you are using labels. In such
cases you may make use of the special pattern =, which is replaced by a unique
number on each asm statement. The following code had been taken from
avr/include/iomacros.h:
#define loop_until_bit_is_clear(port,bit) \
__asm__ __volatile__ ( \
"L_%=: " "sbic %0, %1" "\n\t" \
"rjmp L_%=" \
: /* no outputs */
: "I" (_SFR_IO_ADDR(port)),
"I" (bit)
)
When used for the first time, L_= may be translated to L_1404, the next usage might
create L_1405 or whatever. In any case, the labels became unique too.
Another option is to use Unix-assembler style numeric labels. They are explained in
How do I trace an assembler file in avr-gdb?. The above example would then look like:
#define loop_until_bit_is_clear(port,bit)
__asm__ __volatile__ (
"1: " "sbic %0, %1" "\n\t"
"rjmp 1b"
: /* no outputs */
: "I" (_SFR_IO_ADDR(port)),
"I" (bit)
)
Macro definitions will include the same assembler code whenever they are referenced.
This may not be acceptable for larger routines. In this case you may define a C stub
function, containing nothing other than your assembler code.
"L_dl1%=:" "\n\t"
"mov %A0, %A2" "\n\t"
"mov %B0, %B2" "\n"
"L_dl2%=:" "\n\t"
"sbiw %A0, 1" "\n\t"
"brne L_dl2%=" "\n\t"
"dec %1" "\n\t"
"brne L_dl1%=" "\n\t"
: "=&w" (cnt)
: "r" (ms), "r" (delay_count)
);
}
The purpose of this function is to delay the program execution by a specified number
of milliseconds using a counting loop. The global 16 bit variable delay_count must
contain the CPU clock frequency in Hertz divided by 4000 and must have been set
before calling this routine for the first time. As described in the clobber section, the
routine uses a local variable to hold a temporary value.
Another use for a local variable is a return value. The following function returns a 16
bit value read from two successive port addresses.
Note:
inw() is supplied by avr-libc.
By default AVR-GCC uses the same symbolic names of functions or variables in C and
assembler code. You can specify a different name for the assembler code by using a
special form of the asm statement:
This statement instructs the compiler to use the symbol name clock rather than value.
This makes sense only for external or static variables, because local variables do not
have symbolic names in the assembler code. However, local variables may be held in
registers.
void Count(void)
{
register unsigned char counter asm("r3");
The assembler instruction, "clr r3", will clear the variable counter. AVR-GCC will
not completely reserve the specified register. If the optimizer recognizes that the vari-
able will not be referenced any longer, the register may be re-used. But the compiler
is not able to check wether this register usage conflicts with any predefined register. If
you reserve too many registers in this way, the compiler may even run out of registers
during code generation.
In order to change the name of a function, you need a prototype declaration, because
the compiler will not accept the asm keyword in the function definition:
Calling the function Calc() will create assembler instructions to call the function
CALCULATE.
7.8 Links
For a more thorough discussion of inline assembly usage, see the gcc user
manual. The latest version of the gcc manual is always available here:
http://gcc.gnu.org/onlinedocs/
8.1 Introduction
So you keep reusing the same functions that you created over and over? Tired of cut and
paste going from one project to the next? Would you like to reduce your maintenance
overhead? Then you’re ready to create your own library! Code reuse is a very laudable
goal. With some upfront investment, you can save time and energy on future projects
by having ready-to-go libraries. This chapter describes some background information,
design considerations, and practical knowledge that you will need to create and use
your own libraries.
The compiler compiles a single high-level language file (C language, for example) into
a single object module file. The linker (ld) can only work with object modules to link
them together. Object modules are the smallest unit that the linker works with.
Typically, on the linker command line, you will specify a set of object modules (that
has been previously compiled) and then a list of libraries, including the Standard C
Library. The linker takes the set of object modules that you specify on the command
line and links them together. Afterwards there will probably be a set of "undefined
references". A reference is essentially a function call. An undefined reference is a
function call, with no defined function to match the call.
The linker will then go through the libraries, in order, to match the undefined references
with function definitions that are found in the libraries. If it finds the function that
matches the call, the linker will then link in the object module in which the function is
located. This part is important: the linker links in THE ENTIRE OBJECT MODULE in
which the function is located. Remember, the linker knows nothing about the functions
internal to an object module, other than symbol names (such as function names). The
smallest unit the linker works with is object modules.
When there are no more undefined references, the linker has linked everything and is
done and outputs the final application.
How the linker behaves is very important in designing a library. Ideally, you want to
design a library where only the functions that are called are the only functions to be
linked into the final application. This helps keep the code size to a minimum. In order
to do this, with the way the linker works, is to only write one function per code module.
This will compile to one function per object module. This is usually a very different
way of doing things than writing an application!
There are always exceptions to the rule. There are generally two cases where you
would want to have more than one function per object module.
The first is when you have very complementary functions that it doesn’t make much
sense to split them up. For example, malloc() and free(). If someone is going to use
malloc(), they will very likely be using free() (or at least should be using free()). In this
case, it makes more sense to aggregate those two functions in the same object module.
The second case is when you want to have an Interrupt Service Routine (ISR) in your
library that you want to link in. The problem in this case is that the linker looks for
unresolved references and tries to resolve them with code in libraries. A reference is
the same as a function call. But with ISRs, there is no function call to initiate the ISR.
The ISR is placed in the Interrupt Vector Table (IVT), hence no call, no reference,
and no linking in of the ISR. In order to do this, you have to trick the linker in a way.
Aggregate the ISR, with another function in the same object module, but have the other
function be something that is required for the user to call in order to use the ISR, like
perhaps an initialization function for the subsystem, or perhaps a function that enables
the ISR in the first place.
The librarian program is called ar (for "archiver") and is found in the GNU Binutils
project. This program will have been built for the AVR target and will therefore be
named avr-ar.
The job of the librarian program is simple: aggregate a list of object modules into a
single library (archive) and create an index for the linker to use. The name that you
create for the library filename must follow a specific pattern: libname.a. The name part
is the unique part of the filename that you create. It makes it easier if the name part
relates to what the library is about. This name part must be prefixed by "lib", and it
must have a file extension of .a, for "archive". The reason for the special form of the
filename is for how the library gets used by the toolchain, as we will see later on.
Note:
The filename is case-sensitive. Use a lowercase "lib" prefix, and a lowercase ".a"
as the file extension.
The r command switch tells the program to insert the object modules into the archive
with replacement. The c command line switch tells the program to create the archive.
And the s command line switch tells the program to write an object-file index into the
archive, or update an existing one. This last switch is very important as it helps the
linker to find what it needs to do its job.
Note:
The command line switches are case sensitive! There are uppercase switches that
have completely different actions.
MFile and the WinAVR distribution contain a Makefile Template that includes the
necessary command lines to build a library. You will have to manually modify the
template to switch it over to build a library instead of an application.
See the GNU Binutils manual for more information on the ar program.
To use a library, use the -l switch on your linker command line. The string immedi-
ately following the -l is the unique part of the library filename that the linker will link
in. For example, if you use:
-lm
libm.a
-lprintf_flt
libprintf_flt.a
This is why naming your library is so important when you create it!
The linker will search libraries in the order that they appear on the command line.
Whichever function is found first that matches the undefined reference, it will be linked
in.
There are also command line switches that tell GCC which directory to look in (-L)
for the libraries that are specified to be linke in with -l.
See the GNU Binutils manual for more information on the GNU linker (ld) program.
9 Benchmarks
The results below can only give a rough estimate of the resources necessary for using
certain library functions. There is a number of factors which can both increase or
reduce the effort required:
• Expenses for preparation of operands and their stack are not considered.
• In the table, the size includes all additional functions (for example, function to
multiply two integers) but they are only linked from the library.
• Different versions of the compiler can give a significant difference in code size
and execution time. For example, the dtostre() function, compiled with avr-gcc
3.4.6, requires 930 bytes. After transition to avr-gcc 4.2.3, the size become 1088
bytes.
The table contains the number of MCU clocks to calculate a function with a given
argument(s). The main reason of a big difference between Avr2 and Avr4 is a hardware
multiplication.
Function Avr2 Avr4
__addsf3 (1.234, 5.678) 113 108
__mulsf3 (1.234, 5.678) 375 138
__divsf3 (1.234, 5.678) 466 465
acos (0.54321) 4411 2455
asin (0.54321) 4517 2556
atan (0.54321) 4710 2271
atan2 (1.234, 5.678) 5270 2857
cbrt (1.2345) 2684 2555
ceil (1.2345) 177 177
cos (1.2345) 3387 1671
cosh (1.2345) 4922 2979
exp (1.2345) 4708 2765
fdim (5.678, 1.234) 111 111
floor (1.2345) 180 180
fmax (1.234, 5.678) 39 37
fmin (1.234, 5.678) 35 35
fmod (5.678, 1.234) 131 131
frexp (1.2345, 0) 42 41
hypot (1.234, 5.678) 1341 866
ldexp (1.2345, 6) 42 42
log (1.2345) 4142 2134
log10 (1.2345) 4498 2260
modf (1.2345, 0) 433 429
pow (1.234, 5.678) 9293 5047
round (1.2345) 150 150
sin (1.2345) 3353 1653
sinh (1.2345) 4946 3003
sqrt (1.2345) 494 492
tan (1.2345) 4381 2426
tanh (1.2345) 5126 3173
trunc (1.2345) 178 178
10.1 Introduction
C language was designed to be a portable language. There two main types of port-
ing activities: porting an application to a different platform (OS and/or processor),
10.2 Registers
IO header files contain identifiers for all the register names and bit names for a par-
ticular processor. IAR has individual header files for each processor and they must be
included when registers are being used in the code. For example:
#include <iom169.h>
Note:
IAR does not always use the same register names or bit names that are used in the
AVR datasheet.
AVR GCC also has individual IO header files for each processor. However, the ac-
tual processor type is specified as a command line flag to the compiler. (Using the
-mmcu=processor flag.) This is usually done in the Makefile. This allows you to
specify only a single header file for any processor type:
#include <avr/io.h>
Note:
The forward slash in the <avr/io.h> file name that is used to separate subdirecto-
ries can be used on Windows distributions of the toolchain and is the recommended
method of including this file.
The compiler knows the processor type and through the single header file above, it can
pull in and include the correct individual IO header file. This has the advantage that you
only have to specify one generic header file, and you can easily port your application
to another processor type without having to change every file to include the new IO
header file.
The AVR toolchain tries to adhere to the exact names of the registers and names of
the bits found in the AVR datasheet. There may be some descrepencies between the
register names found in the IAR IO header files and the AVR GCC IO header files.
As mentioned above, the C language Standard, strangely, does not specify a standard
way of declaring and defining an ISR. Hence, every compiler seems to have their own
special way of doing so.
IAR declares an ISR like so:
#pragma vector=TIMER0_OVF_vect
__interrupt void MotorPWMBottom()
{
// code
}
ISR(PCINT1_vect)
{
//code
}
AVR GCC uses the ISR macro to define an ISR. This macro requries the header file:
#include <avr/interrupt.h>
The names of the various interrupt vectors are found in the individual processor IO
header files that you must include with <avr/io.h>.
Note:
The names of the interrupt vectors in AVR GCC has been changed to match the
names of the vectors in IAR. This significantly helps in porting applications from
IAR to AVR GCC.
The C language was not designed for Harvard architecture processors with separate
memory spaces. This means that there are various non-standard ways to define a vari-
able whose data resides in the Program Memory (Flash).
IAR uses a non-standard keyword to declare a variable in Program Memory:
Note:
See the GCC User Manual for more information about Variable Attributes.
#include <avr/pgmspace.h>
.
.
.
int mydata[] PROGMEM = ....
Note:
The PROGMEM macro expands to the Variable Attribute of progmem. This
macro requires that you include <avr/pgmspace.h>. This is the canonical
method for defining a variable in Program Space.
This code snippet checks for the IAR compiler or for the GCC compiler and defines a
macro FLASH_DECLARE(x) that will declare a variable in Program Memory using
the appropriate method based on the compiler that is being used. Then you would used
it like so:
void main(void)
{
//...
}
Note:
See the GCC User Manual for more information on Function Attributes.
In AVR GCC, a prototype for main() is required so you can declare the function at-
tribute to specify that the main() function is of type "noreturn". Then, define main() as
normal. Note that the return type for main() is now void.
The IAR compiler allows a user to lock general registers from r15 and down by using
compiler options and this keyword syntax:
This line locks r14 for use only when explicitly referenced in your code thorugh the var
name "filteredTimeSinceCommutation". This means that the compiler cannot dispose
of it at its own will.
To do this in AVR GCC, do this:
Note:
Do not reserve r0 or r1 as these are used internally by the compiler for a temporary
register and for a zero value.
Locking registers is not recommended in AVR GCC as it removes this register
from the control of the compiler, which may make code generation worse. Use at
your own risk.
20. Why does the compiler compile an 8-bit operation that uses bitwise operators
into a 16-bit operation in assembly?
27. Why are interrupts re-enabled in the middle of writing the stack pointer?
31. I am using floating point math. Why is the compiled code so big? Why does my
code not work?
33. Why are some addresses of the EEPROM corrupted (usually address zero)?
35. On a device with more than 128 KiB of flash, how to make function pointers
work?
uint8_t flag;
...
ISR(SOME_vect) {
flag = 1;
}
...
while (flag == 0) {
...
}
the compiler will typically access flag only once, and optimize further accesses com-
pletely away, since its code path analysis shows that nothing inside the loop could
change the value of flag anyway. To tell the compiler that this variable could be
changed outside the scope of its code path analysis (e. g. from within an interrupt
routine), the variable needs to be declared like:
In order to access the mathematical functions that are declared in <math.h>, the
linker needs to be told to also link the mathematical library, libm.a.
Typically, system libraries like libm.a are given to the final C compiler command
line that performs the linking step by adding a flag -lm at the end. (That is, the initial
lib and the filename suffix from the library are written immediately after a -l flag. So
for a libfoo.a library, -lfoo needs to be provided.) This will make the linker
search the library in a path known to the system.
An alternative would be to specify the full path to the libm.a file at the same place
on the command line, i. e. after all the object files (∗.o). However, since this re-
quires knowledge of where the build system will exactly find those library files, this is
deprecated for system libraries.
Back to FAQ Index.
The method of early initialization (MCUCR, WDTCR or anything else) is different (and
more flexible) in the current version. Basically, write a small assembler file which
looks like this:
;; begin xram.S
#include <avr/io.h>
.section .init1,"ax",@progbits
;; end xram.S
Assemble it, link the resulting xram.o with other files in your program, and this piece
of code will be inserted in initialization code, which is run right after reset. See the
linker script for comments about the new .initN sections (which one to use, etc.).
The advantage of this method is that you can insert any initialization code you want
(just remember that this is very early startup -- no stack and no __zero_reg__ yet),
and no program memory space is wasted if this feature is not used.
There should be no need to modify linker scripts anymore, except for some very spe-
cial cases. It is best to leave __stack at its default value (end of internal SRAM
-- faster, and required on some devices like ATmega161 because of errata), and add
-Wl,-Tdata,0x801100 to start the data section above the stack.
For more information on using sections, see Memory Sections. There is also an ex-
ample for Using Sections in C Code. Note that in C code, any such function would
preferably be placed into section .init3 as the code in .init2 ensures the internal register
__zero_reg__ is already cleared.
Back to FAQ Index.
When performing low-level output work, which is a very central point in microcon-
troller programming, it is quite common that a particular bit needs to be set or cleared
in some IO register. While the device documentation provides mnemonic names for
the various bits in the IO registers, and the AVR device-specific IO definitions reflect
these names in definitions for numerical constants, a way is needed to convert a bit
number (usually within a byte register) into a byte value that can be assigned directly
to the register. However, sometimes the direct bit numbers are needed as well (e. g. in
an SBI() instruction), so the definitions cannot usefully be made as byte values in the
first place.
So in order to access a particular bit number as a byte value, use the _BV() macro.
Of course, the implementation of this macro is just the usual bit shift (which is done
by the compiler anyway, thus doesn’t impose any run-time penalty), so the following
applies:
However, using the macro often makes the program better readable.
"BV" stands for "bit value", in case someone might ask you. :-)
Example: clock timer 2 with full IO clock (CS2x = 0b001), toggle OC2 output on
compare match (COM2x = 0b01), and clear timer on compare match (CTC2 = 1). Make
OC2 (PD7) an output.
TCCR2 = _BV(COM20)|_BV(CTC2)|_BV(CS20);
DDRD = _BV(PD7);
Basically yes, C++ is supported (assuming your compiler has been configured and
compiled to support it, of course). Source files ending in .cc, .cpp or .C will automati-
cally cause the compiler frontend to invoke the C++ compiler. Alternatively, the C++
compiler could be explicitly called by the name avr-c++.
However, there’s currently no support for libstdc++, the standard support library
needed for a complete C++ implementation. This imposes a number of restrictions on
the C++ programs that can be compiled. Among them are:
• Obviously, none of the C++ related standard functions, classes, and template
classes are available.
• The operators new and delete are not implemented, attempting to use them
will cause the linker to complain about undefined external references. (This
could perhaps be fixed.)
• Some of the supplied include files are not C++ safe, i. e. they need to be wrapped
into
extern "C" { . . . }
• Exceptions are not supported. Since exceptions are enabled by default in the
C++ frontend, they explicitly need to be turned off using -fno-exceptions
in the compiler options. Failing this, the linker will complain about an undefined
external reference to __gxx_personality_sj0.
Note:
Recent versions of GCC are now smart enough to detect this situation, and revert
variables that are explicitly initialized to 0 to the .bss section. Still, other compilers
might not do that optimization, and as the C standard guarantees the initialization,
it is safe to rely on it.
Some of the timer-related 16-bit IO registers use a temporary register (called TEMP in
the Atmel datasheet) to guarantee an atomic access to the register despite the fact that
two separate 8-bit IO transfers are required to actually move the data. Typically, this
includes access to the current timer/counter value register (TCNTn), the input capture
register (ICRn), and write access to the output compare registers (OCRnM). Refer to
the actual datasheet for each device’s set of registers that involves the TEMP register.
When accessing one of the registers that use TEMP from the main application, and
possibly any other one from within an interrupt routine, care must be taken that no
access from within an interrupt context could clobber the TEMP register data of an
in-progress transaction that has just started elsewhere.
To protect interrupt routines against other interrupt routines, it’s usually best to use the
ISR() macro when declaring the interrupt function, and to ensure that interrupts are still
disabled when accessing those 16-bit timer registers.
Within the main program, access to those registers could be encapsulated in calls to the
cli() and sei() macros. If the status of the global interrupt flag before accessing one of
those registers is uncertain, something like the following example code can be used.
uint16_t
read_timer1(void)
{
uint8_t sreg;
uint16_t val;
sreg = SREG;
cli();
val = TCNT1;
SREG = sreg;
return val;
}
Which works. When you do the same thing but replace the address of the port by its
macro name, like this:
Note:
For C programs, rather use the standard C bit operators instead, so the above would
be expressed as PORTB |= (1 << 7). The optimizer will take care to trans-
form this into a single SBI instruction, assuming the operands allow for this.
When compiling a program with both optimization (-O) and debug information (-g)
which is fortunately possible in avr-gcc, the code watched in the debugger is opti-
mized code. While it is not guaranteed, very often this code runs with the exact same
optimizations as it would run without the -g switch.
This can have unwanted side effects. Since the compiler is free to reorder code ex-
ecution as long as the semantics do not change, code is often rearranged in order to
make it possible to use a single branch instruction for conditional operations. Branch
instructions can only cover a short range for the target PC (-63 through +64 words from
the current PC). If a branch instruction cannot be used directly, the compiler needs to
work around it by combining a skip instruction together with a relative jump (rjmp)
instruction, which will need one additional word of ROM.
Another side effect of optimization is that variable usage is restricted to the area of code
where it is actually used. So if a variable was placed in a register at the beginning of
some function, this same register can be re-used later on if the compiler notices that the
first variable is no longer used inside that function, even though the variable is still in
lexical scope. When trying to examine the variable in avr-gdb, the displayed result
will then look garbled.
So in order to avoid these side effects, optimization can be turned off while debugging.
However, some of these optimizations might also have the side effect of uncovering
bugs that would otherwise not be obvious, so it must be noted that turning off opti-
mization can easily change the bug pattern. In most cases, you are better off leaving
optimizations enabled while debugging.
Back to FAQ Index.
When using the -g compiler option, avr-gcc only generates line number and other
debug information for C (and C++) files that pass the compiler. Functions that don’t
have line number information will be completely skipped by a single step command
in gdb. This includes functions linked from a standard library, but by default also
functions defined in an assembler source file, since the -g compiler switch does not
apply to the assembler.
So in order to debug an assembler input file (possibly one that has to be passed through
the C preprocessor), it’s the assembler that needs to be told to include line-number
information into the output file. (Other debug information like data types and variable
allocation cannot be generated, since unlike a compiler, the assembler basically doesn’t
know about this.) This is done using the (GNU) assembler option --gstabs.
Example:
When the assembler is not called directly but through the C compiler frontend
(either implicitly by passing a source file ending in .S, or explicitly using -x
assembler-with-cpp), the compiler frontend needs to be told to pass the
--gstabs option down to the assembler. This is done using -Wa,--gstabs.
Please take care to only pass this option when compiling an assembler input file. Oth-
erwise, the assembler code that results from the C compilation stage will also get line
number information, which confuses the debugger.
Note:
You can also use -Wa,-gstabs since the compiler will add the extra ’-’ for
you.
Example:
Also note that the debugger might get confused when entering a piece of code that has
a non-local label before, since it then takes this label as the name of a new function that
appears to have been entered. Thus, the best practice to avoid this confusion is to only
use non-local labels when declaring a new function, and restrict anything else to local
labels. Local labels consist just of a number only. References to these labels consist
of the number, followed by the letter b for a backward reference, or f for a forward
reference. These local labels may be re-used within the source file, references will pick
the closest label with the same number and given direction.
Example:
1: pop YH
pop YL
pop r18
pop r17
pop r16
ret
#include <inttypes.h>
#include <avr/io.h>
void
set_bits_func_wrong (volatile uint8_t port, uint8_t mask)
{
port |= mask;
}
void
return (0);
}
The first function will generate object code which is not even close to what is intended.
The major problem arises when the function is called. When the compiler sees this call,
it will actually pass the value of the PORTB register (using an IN instruction), instead
of passing the address of PORTB (e.g. memory mapped io addr of 0x38, io port 0x18
for the mega128). This is seen clearly when looking at the disassembly of the call:
So, the function, once called, only sees the value of the port register and knows nothing
about which port it came from. At this point, whatever object code is generated for
the function by the compiler is irrelevant. The interested reader can examine the full
disassembly to see that the function’s body is completely fubar.
The second function shows how to pass (by reference) the memory mapped address of
the io port to the function so that you can read and write to it in the function. Here’s
the object code generated for the function call:
You can clearly see that 0x0038 is correctly passed for the address of the io port.
Looking at the disassembled object code for the body of the function, we can see that
the function is indeed performing the operation we intended:
void
set_bits_func_correct (volatile uint8_t *port, uint8_t mask)
{
f8: fc 01 movw r30, r24
*port |= mask;
fa: 80 81 ld r24, Z
fc: 86 2b or r24, r22
fe: 80 83 st Z, r24
}
100: 08 95 ret
Notice that we are accessing the io port via the LD and ST instructions.
The port parameter must be volatile to avoid a compiler warning.
Note:
Because of the nature of the IN and OUT assembly instructions, they can not be
used inside the function when passing the port in this way. Readers interested in
the details should consult the Instruction Set datasheet.
Finally we come to the macro version of the operation. In this contrived example, the
macro is the most efficient method with respect to both execution speed and code size:
Of course, in a real application, you might be doing a lot more in your function which
uses a passed by reference io port address and thus the use of a function over a macro
could save you some code space, but still at a cost of execution speed.
Care should be taken when such an indirect port access is going to one of the 16-bit
IO registers where the order of write access is critical (like some timer registers). All
versions of avr-gcc up to 3.3 will generate instructions that use the wrong access order
in this situation (since with normal memory operands where the order doesn’t matter,
this sometimes yields shorter code).
See http://mail.nongnu.org/archive/html/avr-libc-dev/2003-01/msg00044.html
for a possible workaround.
avr-gcc versions after 3.3 have been fixed in a way where this optimization will be
disabled if the respective pointer variable is declared to be volatile, so the correct
behaviour for 16-bit IO ports can be forced that way.
Back to FAQ Index.
• Data types:
char is 8 bits, int is 16 bits, long is 32 bits, long long is 64 bits, float and
double are 32 bits (this is the only supported floating point format), pointers
are 16 bits (function pointers are word addresses, to allow addressing up to 128K
program memory space). There is a -mint8 option (see Options for the C
compiler avr-gcc) to make int 8 bits, but that is not supported by avr-libc and
violates C standards (int must be at least 16 bits). It may be removed in a future
release.
If too many, those that don’t fit are passed on the stack.
Return values: 8-bit in r24 (not r25!), 16-bit in r25:r24, up to 32 bits in r22-r25, up to
64 bits in r18-r25. 8-bit return values are zero/sign-extended to 16 bits by the called
function (unsigned char is more efficient than signed char - just clr r25).
Arguments to functions with variable argument lists (printf etc.) are all passed on stack,
and char is extended to int.
Warning:
There was no such alignment before 2000-07-01, including the old patches for
gcc-2.95.2. Check your old assembler subroutines, and adjust them accordingly.
There are times when you may need an array of strings which will never be modified.
In this case, you don’t want to waste ram storing the constant strings. The most obvious
(and incorrect) thing to do is this:
#include <avr/pgmspace.h>
The result is not what you want though. What you end up with is the array stored in
ROM, while the individual strings end up in RAM (in the .data section).
To work around this, you need to do something like this:
#include <avr/pgmspace.h>
Looking at the disassembly of the resulting object file we see that array is in flash as
such:
00000026 <array>:
26: 2e 00 .word 0x002e ; ????
28: 2a 00 .word 0x002a ; ????
0000002a <bar>:
2a: 42 61 72 00 Bar.
0000002e <foo>:
2e: 46 6f 6f 00 Foo.
This code reads the pointer to the desired string from the ROM table array into a
register pair.
The value of i (in r22:r23) is doubled to accommodate for the word offset required to
access array[], then the address of array (0x26) is added, by subtracting the negated
address (0xffda). The address of variable p is computed by adding its offset within the
stack frame (33) to the Y pointer register, and memcpy_P is called.
strcpy_P(buf, p);
82: 69 a1 ldd r22, Y+33 ; 0x21
84: 7a a1 ldd r23, Y+34 ; 0x22
86: ce 01 movw r24, r28
88: 01 96 adiw r24, 0x01 ; 1
8a: 0c d0 rcall .+24 ; 0xa4
This will finally copy the ROM string into the local buffer buf.
Variable p (located at Y+33) is read, and passed together with the address of buf (Y+1)
to strcpy_P. This will copy the string from ROM to buf.
Note that when using a compile-time constant index, omitting the first step (reading
the pointer from ROM via memcpy_P) usually remains unnoticed, since the compiler
would then optimize the code for accessing array at compile-time.
Back to FAQ Index.
Well, there is no universal answer to this question; it depends on what the external
RAM is going to be used for.
Basically, the bit SRE (SRAM enable) in the MCUCR register needs to be set in order
to enable the external memory interface. Depending on the device to be used, and
the application details, further registers affecting the external memory operation like
XMCRA and XMCRB, and/or further bits in MCUCR might be configured. Refer to the
datasheet for details.
If the external RAM is going to be used to store the variables from the C program
(i. e., the .data and/or .bss segment) in that memory area, it is essential to set up the
external memory interface early during the device initialization so the initialization of
these variable will take place. Refer to How to modify MCUCR or WDTCR early? for
a description how to do this using few lines of assembler code, or to the chapter about
memory sections for an example written in C.
The explanation of malloc() contains a discussion about the use of internal RAM vs.
external RAM in particular with respect to the various possible locations of the heap
(area reserved for malloc()). It also explains the linker command-line options that are
required to move the memory regions away from their respective standard locations in
internal RAM.
Finally, if the application simply wants to use the additional RAM for private data
storage kept outside the domain of the C compiler (e. g. through a char ∗ variable
initialized directly to a particular address), it would be sufficient to defer the initial-
ization of the external RAM interface to the beginning of main(), so no tweaking of
the .init3 section is necessary. The same applies if only the heap is going to be located
there, since the application start-up code does not affect the heap.
It is not recommended to locate the stack in external RAM. In general, accessing exter-
nal RAM is slower than internal RAM, and errata of some AVR devices even prevent
this configuration from working properly at all.
Back to FAQ Index.
There’s a common misconception that larger numbers behind the -O option might auto-
matically cause "better" optimization. First, there’s no universal definition for "better",
with optimization often being a speed vs. code size trade off. See the detailed discus-
sion for which option affects which part of the code generation.
A test case was run on an ATmega128 to judge the effect of compiling the library itself
using different optimization levels. The following table lists the results. The test case
consisted of around 2 KB of strings to sort. Test #1 used qsort() using the standard
library strcmp(), test #2 used a function that sorted the strings by their size (thus had
two calls to strlen() per invocation).
When comparing the resulting code size, it should be noted that a floating point version
of fvprintf() was linked into the binary (in order to print out the time elapsed) which
is entirely not affected by the different optimization levels, and added about 2.5 KB to
the code.
Optimization Size of .text Time for test #1 Time for test #2
flags
-O3 6898 903 µs 19.7 ms
-O2 6666 972 µs 20.1 ms
-Os 6618 955 µs 20.1 ms
-Os 6474 972 µs 20.1 ms
-mcall-prologues
(The difference between 955 µs and 972 µs was just a single timer-tick, so take this
with a grain of salt.)
So generally, it seems -Os -mcall-prologues is the most universal "best" opti-
mization level. Only applications that need to get the last few percent of speed benefit
from using -O3.
Back to FAQ Index.
First, the code should be put into a new named section. This is done with a section
attribute:
In this example, .bootloader is the name of the new section. This attribute needs to be
placed after the prototype of any function to force the function into the new section.
To relocate the section to a fixed address the linker flag --section-start is used.
This option can be passed to the linker using the -Wl compiler option:
-Wl,--section-start=.bootloader=0x1E000
The name after section-start is the name of the section to be relocated. The number
after the section name is the beginning address of the named section.
Back to FAQ Index.
Well, certain odd problems arise out of the situation that the AVR devices as shipped
by Atmel often come with a default fuse bit configuration that doesn’t match the user’s
expectations. Here is a list of things to care for:
• All devices that have an internal RC oscillator ship with the fuse enabled that
causes the device to run off this oscillator, instead of an external crystal. This
often remains unnoticed until the first attempt is made to use something critical
in timing, like UART communication.
• The ATmega128 ships with the fuse enabled that turns this device into AT-
mega103 compatibility mode. This means that some ports are not fully usable,
and in particular that the internal SRAM is located at lower addresses. Since by
default, the stack is located at the top of internal SRAM, a program compiled for
an ATmega128 running on such a device will immediately crash upon the first
function call (or rather, upon the first function return).
• Devices with a JTAG interface have the JTAGEN fuse programmed by default.
This will make the respective port pins that are used for the JTAG interface un-
available for regular IO.
By default, all strings are handled as all other initialized variables: they occupy RAM
(even though the compiler might warn you when it detects write attempts to these RAM
locations), and occupy the same amount of flash ROM so they can be initialized to the
actual string by startup code. The compiler can optimize multiple identical strings into
a single one, but obviously only for one compilation unit (i. e., a single C source file).
That way, any string literal will be a valid argument to any C function that expects a
const char ∗ argument.
Of course, this is going to waste a lot of SRAM. In Program Space String Utilities, a
method is described how such constant data can be moved out to flash ROM. How-
ever, a constant string located in flash ROM is no longer a valid argument to pass to a
function that expects a const char ∗-type string, since the AVR processor needs
the special instruction LPM to access these strings. Thus, separate functions are needed
that take this into account. Many of the standard C library functions have equivalents
available where one of the string arguments can be located in flash ROM. Private func-
tions in the applications need to handle this, too. For example, the following can be
used to implement simple debugging messages that will be sent through a UART:
#include <inttypes.h>
#include <avr/io.h>
#include <avr/pgmspace.h>
int
uart_putchar(char c)
{
if (c == ’\n’)
uart_putchar(’\r’);
loop_until_bit_is_set(USR, UDRE);
UDR = c;
return 0; /* so it could be used for fdevopen(), too */
}
void
debug_P(const char *addr)
{
char c;
int
main(void)
{
ioinit(); /* initialize UART, ... */
debug_P(PSTR("foo was here\n"));
return 0;
}
Note:
By convention, the suffix _P to the function name is used as an indication that
this function is going to accept a "program-space string". Note also the use of the
PSTR() macro.
11.21 Why does the compiler compile an 8-bit operation that uses
bitwise operators into a 16-bit operation in assembly?
The bitwise "not" operator (∼) will also promote the value in mask to an int. To keep
it an 8-bit value, typecast before the "not" operator:
You can simply run avr-nm on your output (ELF) file. Run it with the -n option, and
it will sort the symbols numerically (by default, they are sorted alphabetically).
Look for the symbol _end, that’s the first address in RAM that is not allocated by
a variable. (avr-gcc internally adds 0x800000 to all data/bss variable addresses, so
please ignore this offset.) Then, the run-time initialization code initializes the stack
pointer (by default) to point to the last available address in (internal) SRAM. Thus, the
region between _end and the end of SRAM is what is available for stack. (If your
application uses malloc(), which e. g. also can happen inside printf(), the heap for
dynamic memory is also located there. See Memory Areas and Using malloc().)
The amount of stack required for your application cannot be determined that easily.
For example, if you recursively call a function and forget to break that recursion, the
amount of stack required is infinite. :-) You can look at the generated assembler code
(avr-gcc ... -S), there’s a comment in each generated assembler file that tells
you the frame size for each generated function. That’s the amount of stack required for
this function, you have to add up that for all functions where you know that the calls
could be nested.
Back to FAQ Index.
While some small AVRs are not directly supported by the C compiler since they do not
have a RAM-based stack (and some do not even have RAM at all), it is possible anyway
to use the general-purpose registers as a RAM replacement since they are mapped into
the data memory region.
Bruce D. Lightner wrote an excellent description of how to do this, and offers this
together with a toolkit on his web page:
http://lightner.net/avr/ATtinyAvrGcc.html
Back to FAQ Index.
It’s a known problem of the MS-DOS FAT file system. Since the FAT file system has
only a granularity of 2 seconds for maintaining a file’s timestamp, and it seems that
some MS-DOS derivative (Win9x) perhaps rounds up the current time to the next sec-
ond when calculating the timestamp of an updated file in case the current time cannot
be represented in FAT’s terms, this causes a situation where make sees a "file coming
from the future".
Since all make decisions are based on file timestamps, and their dependencies, make
warns about this situation.
Solution: don’t use inferior file systems / operating systems. Neither Unix file systems
nor HPFS (aka NTFS) do experience that problem.
Workaround: after saving the file, wait a second before starting make. Or simply
ignore the warning. If you are paranoid, execute a make clean all to make sure
everything gets rebuilt.
In networked environments where the files are accessed from a file server, this message
can also happen if the file server’s clock differs too much from the network client’s
clock. In this case, the solution is to use a proper time keeping protocol on both sys-
tems, like NTP. As a workaround, synchronize the client’s clock frequently with the
server’s clock.
Back to FAQ Index.
Usually, each interrupt has its own interrupt flag bit in some control register, indicating
the specified interrupt condition has been met by representing a logical 1 in the respec-
tive bit position. When working with interrupt handlers, this interrupt flag bit usually
gets cleared automatically in the course of processing the interrupt, sometimes by just
calling the handler at all, sometimes (e. g. for the U[S]ART) by reading a particular
hardware register that will normally happen anyway when processing the interrupt.
From the hardware’s point of view, an interrupt is asserted as long as the respective bit
is set, while global interrupts are enabled. Thus, it is essential to have the bit cleared
before interrupts get re-enabled again (which usually happens when returning from an
interrupt handler).
Only few subsystems require an explicit action to clear the interrupt request when using
interrupt handlers. (The notable exception is the TWI interface, where clearing the
interrupt indicates to proceed with the TWI bus hardware handshake, so it’s never done
automatically.)
However, if no normal interrupt handlers are to be used, or in order to make extra
sure any pending interrupt gets cleared before re-activating global interrupts (e. g.
simply use
TIFR = _BV(TOV0);
Basically, fuses are just a bit in a special EEPROM area. For technical reasons, erased
E[E]PROM cells have all bits set to the value 1, so unprogrammed fuses also have a
logical 1. Conversely, programmed fuse cells read out as bit value 0.
Back to FAQ Index.
When setting up space for local variables on the stack, the compiler generates code like
this:
in r28,__SP_L__
in r29,__SP_H__
sbiw r28,20
in __tmp_reg__,__SREG__
cli
out __SP_H__,r29
out __SREG__,__tmp_reg__
out __SP_L__,r28
/* prologue end (size=10) */
It reads the current stack pointer value, decrements it by the required amount of bytes,
then disables interrupts, writes back the high part of the stack pointer, writes back
the saved SREG (which will eventually re-enable interrupts if they have been enabled
before), and finally writes the low part of the stack pointer.
At the first glance, there’s a race between restoring SREG, and writing SPL. However,
after enabling interrupts (either explicitly by setting the I flag, or by restoring it as part
of the entire SREG), the AVR hardware executes (at least) the next instruction still with
interrupts disabled, so the write to SPL is guaranteed to be executed with interrupts
disabled still. Thus, the emitted sequence ensures interrupts will be disabled only for
the minimum time required to guarantee the integrity of this operation.
Back to FAQ Index.
The GNU linker avr-ld cannot handle binary data directly. However, there’s a com-
panion tool called avr-objcopy. This is already known from the output side: it’s
used to extract the contents of the linked ELF file into an Intel Hex load file.
avr-objcopy can create a relocatable object file from arbitrary binary input, like
This will create a file named foo.o, with the contents of foo.bin. The contents
will default to section .data, and two symbols will be created named _binary_-
foo_bin_start and _binary_foo_bin_end. These symbols can be referred
to inside a C source to access these data.
If the goal is to have those data go to flash ROM (similar to having used the PROGMEM
attribute in C source code), the sections have to be renamed while copying, and it’s also
useful to set the section flags:
Note that all this could be conveniently wired into a Makefile, so whenever foo.bin
changes, it will trigger the recreation of foo.o, and a subsequent relink of the final
ELF file.
Below are two Makefile fragments that provide rules to convert a .txt file to an object
file, and to convert a .bin file to an object file:
$(OBJDIR)/%.o : %.txt
@echo Converting $<
@cp $(<) $(*).tmp
@echo -n 0 | tr 0 ’\000’ >> $(*).tmp
@$(OBJCOPY) -I binary -O elf32-avr \
--rename-section .data=.progmem.data,contents,alloc,load,readonly,data \
--redefine-sym _binary_$*_tmp_start=$* \
--redefine-sym _binary_$*_tmp_end=$*_end \
--redefine-sym _binary_$*_tmp_size=$*_size_sym \
$(*).tmp $(@)
@echo "extern const char" $(*)"[] PROGMEM;" > $(*).h
@echo "extern const char" $(*)_end"[] PROGMEM;" >> $(*).h
@echo "extern const char" $(*)_size_sym"[];" >> $(*).h
@echo "#define $(*)_size ((int)$(*)_size_sym)" >> $(*).h
@rm $(*).tmp
$(OBJDIR)/%.o : %.bin
@echo Converting $<
@$(OBJCOPY) -I binary -O elf32-avr \
--rename-section .data=.progmem.data,contents,alloc,load,readonly,data \
--redefine-sym _binary_$*_bin_start=$* \
--redefine-sym _binary_$*_bin_end=$*_end \
--redefine-sym _binary_$*_bin_size=$*_size_sym \
$(<) $(@)
@echo "extern const char" $(*)"[] PROGMEM;" > $(*).h
@echo "extern const char" $(*)_end"[] PROGMEM;" >> $(*).h
@echo "extern const char" $(*)_size_sym"[];" >> $(*).h
@echo "#define $(*)_size ((int)$(*)_size_sym)" >> $(*).h
The canonical way to perform a software reset of non-XMega AVR’s is to use the
watchdog timer. Enable the watchdog timer to the shortest timeout setting, then go into
an infinite, do-nothing loop. The watchdog will then reset the processor.
XMega parts have a specific bit RST_SWRST_bm in the RST.CTRL register, that gen-
erates a hardware reset. RST_SWRST_bm is protected by the XMega Configuration
Change Protection system.
The reason why using the watchdog timer or RST_SWRST_bm is preferable over jump-
ing to the reset vector, is that when the watchdog or RST_SWRST_bm resets the AVR,
the registers will be reset to their known, default settings. Whereas jumping to the reset
vector will leave the registers in their previous state, which is generally not a good idea.
CAUTION! Older AVRs will have the watchdog timer disabled on a reset. For these
older AVRs, doing a soft reset by enabling the watchdog is easy, as the watchdog will
then be disabled after the reset. On newer AVRs, once the watchdog is enabled, then it
stays enabled, even after a reset! For these newer AVRs a function needs to be added
to the .init3 section (i.e. during the startup code, before main()) to disable the watchdog
early enough so it does not continually reset the AVR.
Here is some example code that creates a macro that can be called to perform a soft
reset:
#include <avr/wdt.h>
...
#define soft_reset() \
do \
{ \
wdt_enable(WDTO_15MS); \
for(;;) \
{ \
} \
} while(0)
For newer AVRs (such as the ATmega1281) also add this function to your code to then
disable the watchdog after a reset (e.g., after a soft reset):
#include <avr/wdt.h>
...
// Function Pototype
void wdt_init(void) __attribute__((naked)) __attribute__((section(".init3")));
...
// Function Implementation
void wdt_init(void)
{
MCUSR = 0;
wdt_disable();
return;
}
You are not linking in the math library from AVR-LibC. GCC has a library that is used
for floating point operations, but it is not optimized for the AVR, and so it generates big
code, or it could be incorrect. This can happen even when you are not using any floating
point math functions from the Standard C library, but you are just doing floating point
math operations.
When you link in the math library from AVR-LibC, those routines get replaced by
hand-optimized AVR assembly and it produces much smaller code.
See I get "undefined reference to..." for functions like "sin()" for more details on how
to link in the math library.
Back to FAQ Index.
Reentrant code means the ability for a piece of code to be called simultaneously from
two or more threads. Attention to re-enterability is needed when using a multi-tasking
operating system, or when using interrupts since an interrupt is really a temporary
thread.
The code generated natively by gcc is reentrant. But, only some of the libraries in
avr-libc are explicitly reentrant, and some are known not to be reentrant. In general,
any library call that reads and writes global variables (including I/O registers) is not
reentrant. This is because more than one thread could read or write the same storage at
the same time, unaware that other threads are doing the same, and create inconsistent
and/or erroneous results.
A library call that is known not to be reentrant will work if it is used only within one
thread and no other thread makes use of a library call that shares common storage with
it.
Below is a table of library calls with known issues.
Note:
It’s not clear one would ever want to do character input simultaneously from more
than one thread anyway, but these entries are included for completeness.
An effort will be made to keep this table up to date if any new issues are discovered or
introduced.
Back to FAQ Index.
The two most common reason for EEPROM corruption is either writing to the EEP-
ROM beyond the datasheet endurance specification, or resetting the AVR while an
EEPROM write is in progress.
EEPROM writes can take up to tens of milliseconds to complete. So that the CPU
is not tied up for that long of time, an internal state-machine handles EEPROM write
requests. The EEPROM state-machine expects to have all of the EEPROM registers
setup, then an EEPROM write request to start the process. Once the EEPROM state-
machine has started, changing EEPROM related registers during an EEPROM write
is guaranteed to corrupt the EEPROM write process. The datasheet always shows the
proper way to tell when a write is in progress, so that the registers are not changed by
the user’s program. The EEPROM state-machine will always complete the write in
progress unless power is removed from the device.
As with all EEPROM technology, if power fails during an EEPROM write the state of
the byte being written is undefined.
In older generation AVRs the EEPROM Address Register (EEAR) is initialized to zero
on reset, be it from Brown Out Detect, Watchdog or the Reset Pin. If an EEPROM
write has just started at the time of the reset, the write will be completed, but now
at address zero instead of the requested address. If the reset occurs later in the write
process both the requested address and address zero may be corrupted.
To distinguish which AVRs may exhibit the corrupt of address zero while a write is
in process during a reset, look at the "initial value" section for the EEPROM Address
Register. If EEAR shows the initial value as 0x00 or 0x0000, then address zero and
possibly the one being written will be corrupted. Newer parts show the initial value as
"undefined", these will not corrupt address zero during a reset (unless it was address
zero that was being written).
EEPROMs have limited write endurance. The datasheet specifies the number of EEP-
ROM writes that are guaranteed to function across the full temperature specification of
the AVR, for a given byte. A read should always be performed before a write, to see
if the value in the EEPROM actually needs to be written, so not to cause unnecessary
EEPROM wear.
AVRs use a paging mechanism for doing EEPROM writes. This is almost entirely
transparent to the user with one exception: When a byte is written to the EEPROM,
the entire EEPROM page is also transparently erased and (re)written, which will cause
wear to bytes that the programmer did not explicitly write. If it is desired to extend
EEPROM write lifetimes, in an attempt not to exceed the datasheet EEPROM write
endurance specification for a given byte, then writes must be in multiples of the EEP-
ROM page size, and not sequential bytes. The EEPROM write page size varies with
the device. The EEPROM page size is found in the datasheet section on Memory Pro-
gramming, generally before the Electrical Specifications near the end of the datasheet.
The failure mechanism for an overwritten byte/page is generally one of "stuck" bits,
i. e. a bit will stay at a one or zero state regardless of the byte written. Also a write
followed by a read may return the correct data, but the data will change with the passage
of time, due the EEPROM’s inability to hold a charge from the excessive write wear.
Back to FAQ Index.
Some AVR datasheets give the following formula for calculating baud rates:
(F_CPU/(UART_BAUD_RATE*16L)-1)
Unfortunately that formula does not work with all combinations of clock speeds and
baud rates due to integer truncation during the division operator.
When doing integer division it is usually better to round to the nearest integer, rather
than to the lowest. To do this add 0.5 (i. e. half the value of the denominator) to the
numerator before the division, resulting in the formula:
This is also the way it is implemented in <util/setbaud.h>: Helper macros for baud
rate calculations.
Back to FAQ Index.
11.36 On a device with more than 128 KiB of flash, how to make
function pointers work?
Function pointers beyond the "magical" 128 KiB barrier(s) on larger devices are sup-
posed to be resolved through so-called trampolines by the linker, so the actual pointers
used in the code can remain 16 bits wide.
In order for this to work, the option -mrelax must be given on the compiler
command-line that is used to link the final ELF file. (Older compilers did not im-
plement this option for the AVR, use -Wl,--relax instead.)
The default behaviour for most of these tools is to install every thing under the
/usr/local directory. In order to keep the AVR tools separate from the base
system, it is usually better to install everything into /usr/local/avr. If the
/usr/local/avr directory does not exist, you should create it before trying to
install anything. You will need root access to install there. If you don’t have root
access to the system, you can alternatively install in your home directory, for exam-
ple, in $HOME/local/avr. Where you install is a completely arbitrary decision, but
should be consistent for all the tools.
You specify the installation directory by using the --prefix=dir option with the
configure script. It is important to install all the AVR tools in the same directory
or some of the tools will not work correctly. To ensure consistency and simplify the
discussion, we will use $PREFIX to refer to whatever directory you wish to install in.
You can set this as an environment variable if you wish as such (using a Bourne-like
shell):
$ PREFIX=$HOME/local/avr
$ export PREFIX
Note:
Be sure that you have your PATH environment variable set to search the direc-
tory you install everything in before you start installing anything. For example, if
you use --prefix=$PREFIX, you must have $PREFIX/bin in your exported
PATH. As such:
$ PATH=$PATH:$PREFIX/bin
$ export PATH
Warning:
If you have CC set to anything other than avr-gcc in your environment, this will
cause the configure script to fail. It is best to not have CC set at all.
Note:
It is usually the best to use the latest released version of each of the tools.
• GNU Binutils
http://sources.redhat.com/binutils/
Installation
• GCC
http://gcc.gnu.org/
Installation
• AVR Libc
http://savannah.gnu.org/projects/avr-libc/
Installation
You can develop programs for AVR devices without the following tools. They may or
may not be of use for you.
• AVRDUDE
http://savannah.nongnu.org/projects/avrdude/
Installation
Usage Notes
• GDB
http://sources.redhat.com/gdb/
Installation
• SimulAVR
http://savannah.gnu.org/projects/simulavr/
Installation
• AVaRICE
http://avarice.sourceforge.net/
Installation
The binutils package provides all the low-level utilities needed in building and ma-
nipulating object files. Once installed, your environment will have an AVR assembler
(avr-as), linker (avr-ld), and librarian (avr-ar and avr-ranlib). In addi-
tion, you get tools which extract data from object files (avr-objcopy), dissassem-
ble object file information (avr-objdump), and strip information from object files
(avr-strip). Before we can build the C compiler, these tools need to be in place.
Download and unpack the source files:
Note:
Replace <version> with the version of the package you downloaded.
If you obtained a gzip compressed file (.gz), use gunzip instead of bunzip2.
$ mkdir obj-avr
$ cd obj-avr
The next step is to configure and build the tools. This is done by supplying arguments
to the configure script that enable the AVR-specific options.
If you don’t specify the --prefix option, the tools will get installed in the
/usr/local hierarchy (i.e. the binaries will get installed in /usr/local/bin,
the info pages get installed in /usr/local/info, etc.) Since these tools are chang-
ing frequently, It is preferrable to put them in a location that is easily removed.
When configure is run, it generates a lot of messages while it determines what
is available on your operating system. When it finishes, it will have created several
Makefiles that are custom tailored to your platform. At this point, you can build the
project.
$ make
Note:
BSD users should note that the project’s Makefile uses GNU make syntax.
This means FreeBSD users may need to build the tools by using gmake.
If the tools compiled cleanly, you’re ready to install them. If you specified a destination
that isn’t owned by your account, you’ll need root access to install them. To install:
$ make install
You should now have the programs from binutils installed into $PREFIX/bin. Don’t
forget to set your PATH environment variable before going to build avr-gcc.
Note:
The official version of binutils might lack support for recent AVR
devices. A patch that adds more AVR types can be found at
http://www.freebsd.org/cgi/cvsweb.cgi/ports/devel/avr-binutils/files/patch-new
You must install avr-binutils and make sure your path is set properly before in-
stalling avr-gcc.
To save your self some download time, you can alternatively download only the
gcc-core-<version>.tar.bz2 and gcc-c++-<version>.tar.bz2
parts of the gcc. Also, if you don’t need C++ support, you only need the core part
and should only enable the C language support.
Note:
Early versions of these tools did not support C++.
The stdc++ libs are not included with C++ for AVR due to the size limitations of
the devices.
The official version of GCC might lack support for recent AVR
devices. A patch that adds more AVR types can be found at
http://www.freebsd.org/cgi/cvsweb.cgi/ports/devel/avr-gcc/files/patch-newdevic
You must install avr-binutils, avr-gcc and make sure your path is set properly be-
fore installing avr-libc.
Note:
If you have obtained the latest avr-libc from cvs, you will have to run the
bootstrap script before using either of the build methods described below.
12.7 AVRDUDE
Note:
It has been ported to windows (via MinGW or cygwin), Linux and Solaris. Other
Unix systems should be trivial to port to.
avrdude is part of the FreeBSD ports system. To install it, simply do the following:
# cd /usr/ports/devel/avrdude
# make install
Note:
Installation into the default location usually requires root permissions. However,
running the program only requires access permissions to the appropriate ppi(4)
device.
Building and installing on other systems should use the configure system, as such:
Note:
If you are planning on using avr-gdb, you will probably want to install either
simulavr or avarice since avr-gdb needs one of these to run as a a remote target
backend.
12.9 SimulAVR
Note:
You might want to have already installed avr-binutils, avr-gcc and avr-libc if you
want to have the test programs built in the simulavr source.
12.10 AVaRICE
Note:
These install notes are not applicable to avarice-1.5 or older. You probably don’t
want to use anything that old anyways since there have been many improvements
and bug fixes since the 1.5 release.
$ cd obj-avr
$ ../configure --prefix=$PREFIX
$ make
$ make install
Note:
AVaRICE uses the BFD library for accessing various binary file formats. You
may need to tell the configure script where to find the lib and headers for the link
to work. This is usually done by invoking the configure script like this (Replace
<hdr_path> with the path to the bfd.h file on your system. Replace <lib_-
path> with the path to libbfd.a on your system.):
Building and installing the toolchain under Windows requires more effort because all
of the tools required for building, and the programs themselves, are mainly designed
for running under a POSIX environment such as Unix and Linux. Windows does not
natively provide such an environment.
There are two projects available that provide such an environment, Cygwin and MinG-
W/MSYS. There are advantages and disadvantages to both. Cygwin provides a very
complete POSIX environment that allows one to build many Linux based tools from
source with very little or no source modifications. However, POSIX functionality is
provided in the form of a DLL that is linked to the application. This DLL has to be
redistributed with your application and there are issues if the Cygwin DLL already ex-
ists on the installation system and different versions of the DLL. On the other hand,
MinGW/MSYS can compile code as native Win32 applications. However, this means
that programs designed for Unix and Linux (i.e. that use POSIX functionality) will not
compile as MinGW/MSYS does not provide that POSIX layer for you. Therefore most
programs that compile on both types of host systems, usually must provide some sort
of abstraction layer to allow an application to be built cross-platform.
MinGW/MSYS does provide somewhat of a POSIX environment that allows you to
build Unix and Linux applications as they woud normally do, with a configure
step and a make step. Cygwin also provides such an environment. This means that
building the AVR toolchain is very similar to how it is built in Linux, described above.
The main differences are in what the PATH environment variable gets set to, pathname
differences, and the tools that are required to build the projects under Windows. We’ll
take a look at the tools next.
These are the tools that are currently used to build WinAVR 20070525 (or later). This
list may change, either the version of the tools, or the tools themselves, as improve-
• MinGW/MSYS
<http://downloads.sourceforge.net/mingw/MinGW-5.1.4.exe?use_-
mirror=superb-east>
– Put MinGW-5.1.4.exe in it’s own directory (for example:
C:\MinGWSetup)
– Run MinGW-5.1.4.exe
– Select "Download and install"
– Select "Current" package.
– Select type of install: Full.
• Edit c:\msys\1.0\msys.bat
Change line (should be line 41):
if EXIST rxvt.exe goto startrxvt
to:
rem if EXIST rxvt.exe goto startrxvt
to remark out this line. Doing this will cause MSYS to always use the bash shell
and not the rxvt shell.
Note:
The order of the next three is important. Install MSYS Developer toolkit before
the autotools.
• automake 1.8.2
• Install Cygwin
– Install everything, all users, UNIX line endings. This will take a ∗long∗
time. A fat internet pipe is highly recommended. It is also recommended
that you download all to a directory first, and then install from that directory
to your machine.
Note:
GMP is a prequisite for building MPFR. Build GMP first.
– Version 4.2.3
– <http://gmplib.org/>
– Build script:
./configure 2>&1 | tee gmp-configure.log
make 2>&1 | tee gmp-make.log
make check 2>&1 | tee gmp-make-check.log
make install 2>&1 | tee gmp-make-install.log
• Install Doxygen
– Version 1.5.6
– <http://www.stack.nl/∼dimitri/doxygen/>
– Download and install.
• Install NetPBM
– Version 10.27.0
– From the GNUWin32 project: <http://gnuwin32.sourceforge.net/packages.html>
– Download and install.
• Install fig2dev
– Version 3.2 Patchlevel 5
– From WinFig 2.2: <http://www.schmidt-web-berlin.de/winfig/>
– Unzip the download file and install fig2dev.exe in a location of your choice.
• Install MiKTeX
– Version 2.7
– <http://miktex.org/>
– Download and install.
• Install Ghostscript
– Version 8.63
– <http://www.cs.wisc.edu/∼ghost/>
– Download and install.
– In the \bin subdirectory of the installaion, copy gswin32c.exe to gs.exe.
• Set the TEMP and TMP environment variables to c:\temp or to the short file-
name version. This helps to avoid NTVDM errors during building.
All directories in the PATH enviornment variable should be specified using their short
filename (8.3) version. This will also help to avoid NTVDM errors during building.
These short filenames can be specific to each machine.
Build the tools below in MSYS.
• Binutils
– Make
make all html install install-html 2>&1 | tee binutils-make.log
• GCC
* /usr/bin
* /bin
* /mingw/bin
* c:/cygwin/bin
* <install directory>/bin
– Configure
CFLAGS=-D__USE_MINGW_ACCESS \
../gcc-$version/configure \
--prefix=$installdir \
--target=$target \
--enable-languages=c,c++ \
--with-dwarf2 \
--enable-win32-registry=WinAVR-$release \
--disable-nls \
--with-gmp=/usr/local \
--with-mpfr=/usr/local \
--enable-doc \
--disable-libssp \
2>&1 | tee $package-configure.log
– Make
make all html install 2>&1 | tee $package-make.log
– Manually copy the HTML documentation from the source code tree to the
installation tree.
• avr-libc
./configure \
--host=avr \
--prefix=$installdir \
--enable-doc \
--disable-versioned-doc \
--enable-html-doc \
--enable-pdf-doc \
--enable-man-doc \
--mandir=$installdir/man \
--datadir=$installdir \
2>&1 | tee $package-configure.log
– Make
make all install 2>&1 | tee $package-make.log
• AVRDUDE
– Configure
./configure \
--prefix=$installdir \
--datadir=$installdir \
--sysconfdir=$installdir/bin \
--enable-doc \
--disable-versioned-doc \
2>&1 | tee $package-configure.log
– Make
make -k all install 2>&1 | tee $package-make.log
• Insight/GDB
– Make
make all install 2>&1 | tee $package-make.log
• SRecord
* /mingw/bin
* c:/cygwin/bin
* <install directory>/bin
– Configure
./configure \
--prefix=$installdir \
--infodir=$installdir/info \
--mandir=$installdir/man \
2>&1 | tee $package-configure.log
– Make
make all install 2>&1 | tee $package-make.log
• AVaRICE
– Configure
../$archivedir/configure \
--prefix=$installdir \
--datadir=$installdir/doc \
--mandir=$installdir/man \
--infodir=$installdir/info \
2>&1 | tee avarice-configure.log
– Make
make all install 2>&1 | tee avarice-make.log
• SimulAVR
– Make
make -k all install 2>&1 | tee simulavr-make.log
make pdf install-pdf 2>&1 | tee simulavr-pdf-make.log
• -mmcu=architecture
• -mmcu=MCU type
The following MCU types are currently understood by avr-gcc. The table matches
them against the corresponding avr-gcc architecture name, and shows the preprocessor
symbol declared by the -mmcu option.
• -morder1
• -morder2
• -mint8
Assume int to be an 8-bit integer. Note that this is not really supported by
avr-libc, so it should normally not be used. The default is to use 16-bit integers.
• -mno-interrupts
Generates code that changes the stack pointer without disabling interrupts. Normally,
the state of the status register SREG is saved in a temporary register, interrupts are
disabled while changing the stack pointer, and SREG is restored.
Specifying this option will define the preprocessor macro __NO_INTERRUPTS__ to
the value 1.
• -mcall-prologues
Use subroutines for function prologue/epilogue. For complex functions that use many
registers (that needs to be saved/restored on function entry/exit), this saves some space
at the cost of a slightly increased execution time.
• -mtiny-stack
• -mno-tablejump
• -mshort-calls
Use rjmp/rcall (limited range) on >8K devices. On avr2 and avr4 architec-
tures (less than 8 KB or flash memory), this is always the case. On avr3 and avr5
architectures, calls and jumps to targets outside the current function will by default use
jmp/call instructions that can cover the entire address range, but that require more
flash ROM and execution time.
• -mrtl
Dump the internal compilation result called "RTL" into comments in the generated
assembler code. Used for debugging avr-gcc.
• -msize
Dump the address, size, and relative cost of each statement into comments in the gen-
erated assembler code. Used for debugging avr-gcc.
• -mdeb
The following general gcc options might be of some interest to AVR users.
• -On
• -Wa,assembler-options
• -Wl,linker-options
• -g
• -ffreestanding
Assume a "freestanding" environment as per the C standard. This turns off automatic
builtin functions (though they can still be reached by prepending __builtin_ to
the actual function name). It also makes the compiler not complain when main()
is declared with a void return type which makes some sense in a microcontroller
environment where the application cannot meaningfully provide a return value to its
environment (in most cases, main() won’t even return anyway). However, this also
turns off all optimizations normally done by the compiler which assume that functions
known by a certain name behave as described by the standard. E. g., applying the
function strlen() to a literal string will normally cause the compiler to immediately
replace that call by the actual length of the string, while with -ffreestanding, it
will always call strlen() at run-time.
• -funsigned-char
Make any unqualfied char type an unsigned char. Without this option, they default to
a signed char.
• -funsigned-bitfields
Make any unqualified bitfield type unsigned. By default, they are signed.
• -fshort-enums
Allocate to an enum type only as many bytes as it needs for the declared range of
possible values. Specifically, the enum type will be equivalent to the smallest integer
type which has enough room.
• -fpack-struct
• -fno-jump-tables
Do not generate tablejump instructions. By default, jump tables can be used to op-
timize switch statements. When turned off, sequences of compare statements are
used instead. Jump tables are usually faster to execute on average, but in particular for
switch statements, where most of the jumps would go to the default label, they might
waste a bit of flash memory.
NOTE: The tablejump instructions use the LPM assembler instruction for access to
jump tables. Always use -fno-jump-tables switch, if compiling a bootloader for
devices with more than 64 KB of code memory.
• -mmcu=architecture
• -mmcu=MCU name
avr-as understands the same -mmcu= options as avr-gcc. By default, avr2 is assumed,
but this can be altered by using the appropriate .arch pseudo-instruction inside the
assembler source file.
• -mall-opcodes
Turns off opcode checking for the actual MCU type, and allows any possible AVR
opcode to be assembled.
• -mno-skip-bug
• -mno-wrap
For RJMP/RCALL instructions, don’t allow the target address to wrap around for de-
vices that have more than 8 KB of memory.
• --gstabs
Generate .stabs debugging symbols for assembler source lines. This enables avr-gdb
to trace through assembler source files. This option must not be used when assembling
sources that have been generated by the C compiler; these files already contain the
appropriate line number information from the C source files.
• -a[cdhlmns=file]
The various sub-options can be combined into a single -a option list; =file must be the
last one in that case.
Remember that assembler options can be passed from the C compiler frontend using
-Wa (see above), so in order to include the C source code into the assembler listing in
file foo.lst, when compiling foo.c, the following compiler command-line can be
used:
In order to pass an assembler file through the C preprocessor first, and have the assem-
bler generate line number debugging information for it, the following command can be
used:
Note that on Unix systems that have case-distinguishing file systems, specifying a file
name with the suffix .S (upper-case letter S) will make the compiler automatically
assume -x assembler-with-cpp, while using .s would pass the file directly to
the assembler (no preprocessing done).
While there are no machine-specific options for avr-ld, a number of the standard op-
tions might be of interest to AVR users.
• -lname
Locate the archive library named libname.a, and use it to resolve currently
unresolved symbols from it. The library is searched along a path that con-
sists of builtin pathname entries that have been specified at compile time (e. g.
/usr/local/avr/lib on Unix systems), possibly extended by pathname entries
as specified by -L options (that must precede the -l options on the command-line).
• -Lpath
• --defsym symbol=expr
• -M
• -Map mapfile
• --cref
Output a cross reference table to the map file (in case -Map is also present), or to
stdout.
• --section-start sectionname=org
• -Tbss org
• -Tdata org
• -Ttext org
• -T scriptfile
Use scriptfile as the linker script, replacing the default linker script. De-
fault linker scripts are stored in a system-specific location (e. g. under
/usr/local/avr/lib/ldscripts on Unix systems), and consist of the AVR
architecture name (avr2 through avr5) with the suffix .x appended. They describe how
the various memory sections will be linked together.
By default, all unknown non-option arguments on the avr-gcc command-line (i. e.,
all filename arguments that don’t have a suffix that is handled by avr-gcc) are passed
straight to the linker. Thus, all files ending in .o (object files) and .a (object libraries)
are provided to the linker.
System libraries are usually not passed by their explicit filename but rather using the
-l option which uses an abbreviated form of the archive filename (see above). avr-
libc ships two system libraries, libc.a, and libm.a. While the standard library
libc.a will always be searched for unresolved references when the linker is started
using the C compiler frontend (i. e., there’s always at least one implied -lc option),
the mathematics library libm.a needs to be explicitly requested using -lm. See also
the entry in the FAQ explaining this.
Conventionally, Makefiles use the make macro LDLIBS to keep track of -l (and
possibly -L) options that should only be appended to the C compiler command-line
when linking the final binary. In contrast, the macro LDFLAGS is used to store other
command-line options to the C compiler that should be passed as options during the
linking stage. The difference is that options are placed early on the command-line,
while libraries are put at the end since they are to be used to resolve global symbols
that are still unresolved at this point.
Specific linker flags can be passed from the C compiler command-line using the -Wl
compiler option, see above. This option requires that there be no spaces in the appended
linker option, while some of the linker options above (like -Map or --defsym) would
require a space. In these situations, the space can be replaced by an equal sign as
well. For example, the following command-line can be used to compile foo.c into an
executable, and also produce a link map that contains a cross-reference list in the file
foo.map:
See the explanation of the data section for why 0x800000 needs to be added to the
actual value. Note that the stack will still remain in internal RAM, through the symbol
__stack that is provided by the run-time startup code. This is probably a good idea
anyway (since internal RAM access is faster), and even required for some early devices
that had hardware bugs preventing them from using a stack in external RAM. Note
also that the heap for malloc() will still be placed after all the variables in the data
section, so in this situation, no stack/heap collision can occur.
In order to relocate the stack from its default location at the top of interns RAM, the
value of the symbol __stack can be changed on the linker command-line. As the
linker is typically called from the compiler frontend, this can be achieved using a com-
piler option like
-Wl,--defsym=__stack=0x8003ff
The above will make the code use stack space from RAM address 0x3ff downwards.
The amount of stack space available then depends on the bottom address of internal
RAM for a particular device. It is the responsibility of the application to ensure the
stack does not grow out of bounds, as well as to arrange for the stack to not collide
with variable allocations made by the compiler (sections .data and .bss).
14 Compiler optimization
Programs contain sequences of statements, and a naive compiler would execute them
exactly in the order as they are written. But an optimizing compiler is free to re-
order the statements - or even parts of them - if the resulting "net effect" is the same.
The "measure" of the "net effect" is what the standard calls "side effects", and is ac-
complished exclusively through accesses (reads and writes) to variables qualified as
volatile. So, as long as all volatile reads and writes are to the same addresses and
in the same order (and writes write the same values), the program is correct, regard-
less of other operations in it. (One important point to note here is, that time duration
between consecutive volatile accesses is not considered at all.)
Unfortunately, there are also operations which are not covered by volatile accesses.
An example of this in avr-gcc/avr-libc are the cli() and sei() macros defined in
<avr/interrupt.h>, which convert directly to the respective assembler mnemonics
through the __asm__() statement. These don’t constitute a variable access at all,
not even volatile, so the compiler is free to move them around. Although there is
a "volatile" qualifier which can be attached to the __asm__() statement, its effect on
(re)ordering is not clear from the documentation (and is more likely only to prevent
complete removal by the optimiser), as it (among other) states:
Note that even a volatile asm instruction can be moved relative to other code, including
across jump instructions. [...] Similarly, you can’t expect a sequence of volatile asm
instructions to remain perfectly consecutive.
See also:
http://gcc.gnu.org/onlinedocs/gcc-4.3.4/gcc/Extended-Asm.html
There is another mechanism which can be used to achieve something similar: memory
barriers. This is accomplished through adding a special "memory" clobber to the inline
asm statement, and ensures that all variables are flushed from registers to memory
before the statement, and then re-read after the statement. The purpose of memory
barriers is slightly different than to enforce code ordering: it is supposed to ensure that
there are no variables "cached" in registers, so that it is safe to change the content of
registers e.g. when switching context in a multitasking OS (on "big" processors with
out-of-order execution they also imply usage of special instructions which force the
processor into "in-order" state (this is not the case of AVRs)).
However, memory barrier works well in ensuring that all volatile accesses before and
after the barrier occur in the given order with respect to the barrier. However, it does
not ensure the compiler moving non-volatile-related statements across the barrier. Peter
Dannegger provided a nice example of this effect:
cli();
ivar = val;
sei();
}
00000112 <test2>:
112: bc 01 movw r22, r24
114: f8 94 cli
116: 8f ef ldi r24, 0xFF ; 255
118: 9f ef ldi r25, 0xFF ; 255
11a: 0e 94 96 00 call 0x12c ; 0x12c <__udivmodhi4>
11e: 70 93 01 02 sts 0x0201, r23
122: 60 93 00 02 sts 0x0200, r22
126: 78 94 sei
128: 08 95 ret
where the potentially slow division is moved across cli(), resulting in interrupts to be
disabled longer than intended. Note, that the volatile access occurs in order with respect
to cli() or sei(); so the "net effect" required by the standard is achieved as intended, it
is "only" the timing which is off. However, for most of embedded applications, timing
is an important, sometimes critical factor.
See also:
https://www.mikrocontroller.net/topic/65923
avrdude is a program that is used to update or read the flash and EEPROM memories
of Atmel AVR microcontrollers on FreeBSD Unix. It supports the Atmel serial pro-
gramming protocol using the PC’s parallel port and can upload either a raw binary file
or an Intel Hex format file. It can also be used in an interactive mode to individually
update EEPROM cells, fuse bits, and/or lock bits (if their access is supported by the
Atmel serial programming protocol.) The main flash instruction memory of the AVR
can also be programmed in interactive mode, however this is not very useful because
one can only turn bits off. The only way to turn flash bits on is to erase the entire
memory (using avrdude’s -e option).
avrdude is part of the FreeBSD ports system. To install it, simply do the following:
# cd /usr/ports/devel/avrdude
# make install
Once installed, avrdude can program processors using the contents of the .hex file
specified on the command line. In this example, the file main.hex is burned into the
flash memory:
The -p 2313 option lets avrdude know that we are operating on an AT90S2313
chip. This option specifies the device id and is matched up with the device of the same
id in avrdude’s configuration file ( /usr/local/etc/avrdude.conf ). To list
valid parts, specify the -v option. The -e option instructs avrdude to perform a
chip-erase before programming; this is almost always necessary before programming
the flash. The -m flash option indicates that we want to upload data into the flash
memory, while -i main.hex specifies the name of the input file.
The EEPROM is uploaded in the same way, the only difference is that you would use
-m eeprom instead of -m flash.
To use interactive mode, use the -t option:
# avrdude -p 2313 -t
avrdude: AVR device initialized and ready to accept instructions
avrdude: Device signature = 0x1e9101
avrdude>
avrdude> ?
>>> ?
Valid commands:
Use the ’part’ command to display valid memory types for use with the
’dump’ and ’write’ commands.
avrdude>
Release numbers consist of three parts, a major number, a minor number, and a revision
number, each separated by a dot.
The major number is currently 1 (and has always been). It will only be bumped in case
a new version offers a major change in the API that is not backwards compatible.
In the past (up to 1.6.x), even minor numbers have been used to indicate "stable" re-
leases, and odd minor numbers have been reserved for development branches/versions.
As the latter has never really been used, and maintaining a stable branch that eventually
became effectively the same as the development version has proven to be just a cum-
bersome and tedious job, this scheme has given up in early 2010, so starting with 1.7.0,
every minor number will be used. Minor numbers will be bumped upon judgement of
the development team, whenever it seems appropriate, but at least in cases where some
API was changed.
Starting with version 1.4.0, a file <avr/version.h> indicates the library version of an
installed library tree.
The information in this section is only relevant to AVR Libc developers and can be
ignored by end users.
Note:
In what follows, I assume you know how to use SVN and how to checkout multiple
source trees in a single directory without having them clobber each other. If you
don’t know how to do this, you probably shouldn’t be making releases or cutting
branches.
The following steps should be taken to cut a branch in SVN (assuming $username is
set to your savannah username):
2. Update the NEWS file with pending release number and commit to SVN trunk:
Change Changes since avr-libc-<last_release>: to Changes in avr-libc-<this_-
relelase>.
10. Perform a ’make distcheck’ and make sure it succeeds. This will create the
snapshot source tarball. This should be considered the first release candidate.
11. Upload the snapshot tarball to savannah.
12. Update the bug tracker interface on Savannah: Bugs —> Edit field values —>
Release / Fixed Release
13. Announce the branch and the branch tag to the avr-libc-dev list so other devel-
opers can checkout the branch.
A stable release will only be done on a branch, not from the SVN trunk.
The following steps should be taken when making a release:
1. Make sure the source tree you are working from is on the correct branch:
svn switch svn+ssh://$username.savannah.nongnu.org/avr-libc/branches/avr-lib
<minor>-branch
3. Update the gnu tool chain version requirements in the README and commit to
SVN.
4. Update the ChangeLog file to note the release and commit to SVN on the branch:
Add Released avr-libc-<this_release>.
5. Update the NEWS file with pending release number and commit to SVN:
Change Changes since avr-libc-<last_release>: to Changes in avr-libc-<this_-
relelase>:.
7. Perform a ’make distcheck’ and make sure it succeeds. This will create the
source tarball.
11. Update the bug tracker interface on Savannah: Bugs —> Edit field values —>
Release / Fixed Release
12. Generate the latest documentation and upload to savannah.
The following hypothetical diagram should help clarify version and branch relation-
ships.
17 Acknowledgments
This document tries to tie together the labors of a large group of people. Without
these individuals’ efforts, we wouldn’t have a terrific, free set of tools to develop AVR
projects. We all owe thanks to:
• The GCC Team, which produced a very capable set of development tools for an
amazing number of platforms and processors.
• Rich Neswold for writing the original avr-tools document (which he graciously
allowed to be merged into this document) and his improvements to the demo
project.
• Theodore A. Roth for having been a long-time maintainer of many of the tools
(AVR-Libc, the AVR port of GDB, AVaRICE, uisp, avrdude).
• All the people who currently maintain the tools, and/or have submitted sugges-
tions, patches and bug reports. (See the AUTHORS files of the various tools.)
• And lastly, all the users who use the software. If nobody used the software, we
would probably not be very motivated to continue to develop it. Keep those bug
reports coming. ;-)
18 Todo List
Group avr_boot From email with Marek: On smaller devices (all except AT-
mega64/128), __SPM_REG is in the I/O space, accessible with the shorter "in"
and "out" instructions - since the boot loader has a limited size, this could be an
important optimization.
19 Deprecated List
Global <globalScope>::cbi(port, bit)
Global <globalScope>::enable_external_int(mask)
Global <globalScope>::inb(port)
Global <globalScope>::inp(port)
Global <globalScope>::INTERRUPT(signame)
20 Module Index
20.1 Modules
div_t 368
ldiv_t 368
22 File Index
alloca.h ??
assert.h 369
atoi.S 369
atol.S 369
atomic.h 369
boot.h 370
cpufunc.h 376
crc16.h 376
ctype.h 377
defines.h ??
delay.h 377
delay_basic.h 378
deprecated.h ??
dtoa_conv.h ??
eedef.h ??
eeprom.h ??
errno.h 378
fdevopen.c 378
ffs.S 379
ffsl.S 379
ffsll.S 379
fuse.h 379
hd44780.h ??
ina90.h ??
interrupt.h 379
inttypes.h 380
io.h 382
iocompat.h ??
lcd.h ??
lock.h 382
math.h 383
memccpy.S 387
memchr.S 387
memchr_P.S 387
memcmp.S 387
memcmp_P.S 387
memcmp_PF.S 387
memcpy.S 387
memcpy_P.S 387
memmem.S 387
memmove.S 387
memrchr.S 387
memrchr_P.S 387
memset.S 387
parity.h 387
pgmspace.h 388
portpins.h ??
power.h 399
project.h ??
setbaud.h 399
setjmp.h 400
sfr_defs.h ??
signal.h ??
signature.h 400
sleep.h 400
stdint.h 401
stdio.h 404
stdio_private.h ??
stdlib.h 405
stdlib_private.h ??
strcasecmp.S 409
strcasecmp_P.S 409
strcasestr.S 409
strcat.S 409
strcat_P.S 409
strchr.S 409
strchr_P.S 409
strchrnul.S 409
strchrnul_P.S 409
strcmp.S 409
strcmp_P.S 409
strcpy.S 409
strcpy_P.S 409
strcspn.S 409
strcspn_P.S 409
strdup.c 409
string.h 410
strlcat.S 413
strlcat_P.S 413
strlcpy.S 413
strlcpy_P.S 413
strlen.S 413
strlen_P.S 413
strlwr.S 413
strncasecmp.S 413
strncasecmp_P.S 413
strncat.S 413
strncat_P.S 413
strncmp.S 413
strncmp_P.S 413
strncpy.S 413
strncpy_P.S 413
strnlen.S 413
strnlen_P.S 413
strpbrk.S 413
strpbrk_P.S 413
strrchr.S 413
strrchr_P.S 413
strrev.S 413
strsep.S 413
strsep_P.S 413
strspn.S 413
strspn_P.S 413
strstr.S 413
strstr_P.S 413
strtok.c 413
strtok_P.c 414
strtok_r.S 414
strtok_rP.S 414
strupr.S 414
util/twi.h 414
compat/twi.h ??
uart.h ??
version.h ??
wdt.h 415
xtoa_fast.h ??
23 Module Documentation
Functions
Allocate __size bytes of space in the stack frame of the caller. This temporary space is
automatically freed when the function that called alloca() returns to its caller. Avr-libc
defines the alloca() as a macro, which is translated into the inlined __builtin_-
alloca() function. The fact that the code is inlined, means that it is impossible to
take the address of this function, or to change its behaviour by linking with a different
library.
Returns:
alloca() returns a pointer to the beginning of the allocated space. If the allocation
causes stack overflow, program behaviour is undefined.
Warning:
Defines
• #define assert(expression)
#include <assert.h>
__ASSERT_USE_STDERR
before including the <assert.h> header file. By default, only abort() will be called
to halt the application.
Parameters:
expression Expression to test for.
The assert() macro tests the given expression and if it is false, the calling process is
terminated. A diagnostic message is written to stderr and the function abort() is called,
effectively terminating the program.
If expression is true, the assert() macro does nothing.
The assert() macro may be removed at compile time by defining NDEBUG as a macro
(e.g., by using the compiler option -DNDEBUG).
These functions perform character classification. They return true or false status de-
pending whether the character passed to the function falls into the function’s classifi-
cation (i.e. isdigit() returns true if its argument is any value ’0’ though ’9’, inclusive).
If the input is not an unsigned char value, all of this function return false.
This realization permits all possible values of integer argument. The toascii() function
clears all highest bits. The tolower() and toupper() functions return an input argument
as is, if it is not an unsigned char value.
#include <ctype.h>
Checks whether c is a 7-bit unsigned char value that fits into the ASCII character set.
Checks for any printable character which is not a space or an alphanumeric character.
Checks for white-space characters. For the avr-libc library, these are:
space, form-feed (’\f’), newline (’\n’), carriage return (’\r’), horizontal tab (’\t’), and
vertical tab (’\v’).
Converts c to a 7-bit unsigned char value that fits into the ASCII character set, by
clearing the high-order bits.
Warning:
Many people will be unhappy if you use this function. This function will convert
accented letters into random characters.
Defines
• #define EDOM 33
• #define ERANGE 34
#include <errno.h>
Some functions in the library set the global variable errno when an error occurs. The
file, <errno.h>, provides symbolic names for various error codes.
Warning:
The errno global variable is not safe to use in a threaded or multi-task system. A
race condition can occur if a task is interrupted between the call which sets error
and when the task examines errno. If another task changes errno during this
time, the result will be incorrect for the interrupted task.
Domain error.
Range error.
For C++, these are only included if __STDC_LIMIT_MACROS is defined before in-
cluding <inttypes.h>.
#include <inttypes.h>
This header file includes the exact-width integer definitions from <stdint.h>, and
extends them with additional facilities provided by the implementation.
Currently, the extensions include two additional integer types that could hold a "far"
pointer (i.e. a code pointer that can address more than 64 KB), as well as standard
names for all printf and scanf formatting options that are supported by the <stdio.h>:
Standard IO facilities. As the library does not support the full range of conversion
specifiers from ISO 9899:1999, only those conversions that are actually implemented
will be listed here.
The idea behind these conversion macros is that, for each of the types defined by
<stdint.h>, a macro will be supplied that portably allows formatting an object of that
type in printf() or scanf() operations. Example:
#include <inttypes.h>
uint8_t smallval;
int32_t longval;
...
printf("The hexadecimal value of smallval is %" PRIx8
", the decimal value of longval is %" PRId32 ".\n",
smallval, longval);
Defines
Functions
#include <math.h>
Notes:
• In order to access the functions delcared herein, it is usually also required to
additionally link against the library libm.a. See also the related FAQ entry.
• Math functions do not raise exceptions and do not change the errno vari-
able. Therefore the majority of them are declared with const attribute, for
better optimization by GCC.
INFINITY constant.
The constant e.
NAN constant.
The acos() function computes the principal value of the arc cosine of __x. The
returned value is in the range [0, pi] radians. A domain error occurs for arguments not
in the range [-1, +1].
The asin() function computes the principal value of the arc sine of
__x. The returned value is in the range [-pi/2, pi/2] radians. A domain error occurs for
arguments not in the range [-1, +1].
The atan() function computes the principal value of the arc tangent of __x. The
returned value is in the range [-pi/2, pi/2] radians.
The atan2() function computes the principal value of the arc tangent of __y
/ __x, using the signs of both arguments to determine the quadrant of the return value.
The returned value is in the range [-pi, +pi] radians.
The ceil() function returns the smallest integral value greater than or equal to __x,
expressed as a floating-point number.
The copysign() function returns __x but with the sign of __y. They work even if __x
or __y are NaN or zero.
The fabs() function computes the absolute value of a floating-point number __x.
The fdim() function returns max(__x - __y, 0). If __x or __y or both are NaN, NaN is
returned.
The floor() function returns the largest integral value less than or equal to __x,
expressed as a floating-point number.
The fma() function performs floating-point multiply-add. This is the operation (__x ∗
__y) + __z, but the intermediate result is not rounded to the destination type. This can
sometimes improve the precision of a calculation.
The fmax() function returns the greater of the two values __x and __y. If an argument
is NaN, the other argument is returned. If both arguments are NaN, NaN is returned.
The fmin() function returns the lesser of the two values __x and __y. If an argument
is NaN, the other argument is returned. If both arguments are NaN, NaN is returned.
The frexp() function breaks a floating-point number into a normalized fraction and an
integral power of 2. It stores the integer in the int object pointed to by __pexp.
If __x is a normal float point number, the frexp() function returns the value v, such that
v has a magnitude in the interval [1/2, 1) or zero, and __x equals v times 2 raised to
the power __pexp. If __x is zero, both parts of the result are zero. If __x is not a finite
number, the frexp() returns __x as is and stores 0 by __pexp.
Note:
This implementation permits a zero pointer as a directive to skip a storing the
exponent.
The isfinite() function returns a nonzero value if __x is finite: not plus or minus
infinity, and not NaN.
The function isinf() returns 1 if the argument __x is positive infinity, -1 if __x is
negative infinity, and 0 otherwise.
Note:
The GCC 4.3 can replace this function with inline code that returns the 1 value for
both infinities (gcc bug #35509).
The function isnan() returns 1 if the argument __x represents a "not-a-number" (NaN)
object, otherwise 0.
The log10() function returns the logarithm of argument __x to base 10.
Returns:
The rounded long integer value. If __x is not a finite number or an overflow was,
this realization returns the LONG_MIN value (0x80000000).
The lround() function rounds __x to the nearest integer, but rounds
halfway cases away from zero (instead of to the nearest even integer). This function is
similar to round() function, but it differs in type of return value and in that an overflow
is possible.
Returns:
The rounded long integer value. If __x is not a finite number or an overflow was,
this realization returns the LONG_MIN value (0x80000000).
The modf() function breaks the argument __x into integral and fractional parts,
each of which has the same sign as the argument. It stores the integral part as a double
in the object pointed to by __iptr.
The modf() function returns the signed fractional part of __x.
Note:
This implementation skips writing by zero pointer. However, the GCC 4.3 can
replace this function with inline code that does not permit to use NULL address
for the avoiding of storing.
The function pow() returns the value of __x to the exponent __y.
The round() function rounds __x to the nearest integer, but rounds halfway cases
away from zero (instead of to the nearest even integer). Overflow is impossible.
Returns:
The rounded value. If __x is an integral or infinite, __x itself is returned. If __x is
NaN, then NaN is returned.
The signbit() function returns a nonzero value if the value of __x has its sign bit set.
This is not the same as ‘__x < 0.0’, because IEEE 754 floating point allows zero to be
signed. The comparison ‘-0.0 < 0.0’ is false, but ‘signbit (-0.0)’ will return a nonzero
value.
Note:
This function does not belong to the C standard definition.
The trunc() function rounds __x to the nearest integer not larger in absolute value.
Functions
While the C language has the dreaded goto statement, it can only be used to jump to
a label in the same (local) function. In order to jump directly to another (non-local)
function, the C library provides the setjmp() and longjmp() functions. setjmp() and
longjmp() are useful for dealing with errors and interrupts encountered in a low-level
subroutine of a program.
Note:
setjmp() and longjmp() make programs hard to understand and maintain. If possi-
ble, an alternative should be used.
longjmp() can destroy changes made to global register variables (see How to per-
manently bind a variable to a register?).
#include <setjmp.h>
jmp_buf env;
while (1)
{
... main processing loop which calls foo() some where ...
}
}
...
if (err)
{
longjmp (env, 1);
}
}
#include <setjmp.h>
longjmp() restores the environment saved by the last call of setjmp() with the corre-
sponding __jmpb argument. After longjmp() is completed, program execution contin-
ues as if the corresponding call of setjmp() had just returned the value __ret.
Note:
longjmp() cannot cause 0 to be returned. If longjmp() is invoked with a second
argument of 0, 1 will be returned instead.
Parameters:
__jmpb Information saved by a previous call to setjmp().
__ret Value to return to the caller of setjmp().
Returns:
This function never returns.
#include <setjmp.h>
setjmp() saves the stack context/environment in __jmpb for later use by longjmp(). The
stack context will be invalidated if the function which called setjmp() returns.
Parameters:
__jmpb Variable of type jmp_buf which holds the stack information such that
the environment can be restored.
Returns:
setjmp() returns 0 if returning directly, and non-zero when returning from
longjmp() using the saved context.
Integer types being usually fastest having at least the specified width
Types designating integer data capable of representing any value of any integer type in
the corresponding signed or unsigned category
#include <stdint.h>
Note:
This type is not available when the compiler option -mint8 is in effect.
Note:
This type is not available when the compiler option -mint8 is in effect.
Note:
This type is not available when the compiler option -mint8 is in effect.
Note:
This type is not available when the compiler option -mint8 is in effect.
Note:
This type is not available when the compiler option -mint8 is in effect.
Note:
This type is not available when the compiler option -mint8 is in effect.
Defines
Functions
#include <stdio.h>
Introduction to the Standard IO facilities This file declares the standard IO facili-
ties that are implemented in avr-libc. Due to the nature of the underlying hardware,
only a limited subset of standard IO is implemented. There is no actual file implementa-
tion available, so only device IO can be performed. Since there’s no operating system,
the application needs to provide enough details about their devices in order to make
them usable by the standard IO facilities.
Due to space constraints, some functionality has not been implemented at all (like some
of the printf conversions that have been left out). Nevertheless, potential users of
this implementation should be warned: the printf and scanf families of functions,
although usually associated with presumably simple things like the famous "Hello,
world!" program, are actually fairly complex which causes their inclusion to eat up
a fair amount of code space. Also, they are not fast due to the nature of interpreting
the format string at run-time. Whenever possible, resorting to the (sometimes non-
standard) predetermined conversion facilities that are offered by avr-libc will usually
cost much less in terms of speed and code size.
Tunable options for code size vs. feature set In order to allow programmers a code
size vs. functionality tradeoff, the function vfprintf() which is the heart of the printf
family can be selected in different flavours using linker options. See the documentation
of vfprintf() for a detailed description. The same applies to vfscanf() and the scanf
family of functions.
Outline of the chosen API The standard streams stdin, stdout, and stderr are
provided, but contrary to the C standard, since avr-libc has no knowledge about appli-
cable devices, these streams are not already pre-initialized at application startup. Also,
since there is no notion of "file" whatsoever to avr-libc, there is no function fopen()
that could be used to associate a stream to some device. (See note 1.) Instead, the
function fdevopen() is provided to associate a stream to a device, where the device
needs to provide a function to send a character, to receive a character, or both. There
is no differentiation between "text" and "binary" streams inside avr-libc. Character \n
is sent literally down to the device’s put() function. If the device requires a carriage
return (\r) character to be sent before the linefeed, its put() routine must implement
this (see note 2).
As an alternative method to fdevopen(), the macro fdev_setup_stream() might be used
to setup a user-supplied FILE structure.
It should be noted that the automatic conversion of a newline character into a carriage
return - newline sequence breaks binary transfers. If binary transfers are desired, no
automatic conversion should be performed, but instead any string that aims to issue a
Format strings in flash ROM All the printf and scanf family functions come
in two flavours: the standard name, where the format string is expected to be in SRAM,
as well as a version with the suffix "_P" where the format string is expected to reside
in the flash ROM. The macro PSTR (explained in <avr/pgmspace.h>: Program Space
Utilities) becomes very handy for declaring these format strings.
Example
#include <stdio.h>
static int
uart_putchar(char c, FILE *stream)
{
if (c == ’\n’)
uart_putchar(’\r’, stream);
loop_until_bit_is_set(UCSRA, UDRE);
UDR = c;
return 0;
}
int
main(void)
{
init_uart();
stdout = &mystdout;
printf("Hello, world!\n");
return 0;
}
This example uses the initializer form FDEV_SETUP_STREAM() rather than the
function-like fdev_setup_stream(), so all data initialization happens during C start-up.
If streams initialized that way are no longer needed, they can be destroyed by first
calling the macro fdev_close(), and then destroying the object itself. No call to fclose()
should be issued for these streams. While calling fclose() itself is harmless, it will cause
an undefined reference to free() and thus cause the linker to link the malloc module into
the application.
Notes
Note 1:
It might have been possible to implement a device abstraction that is compatible
with fopen() but since this would have required to parse a string, and to take all
the information needed either out of this string, or out of an additional table that
would need to be provided by the application, this approach was not taken.
Note 2:
This basically follows the Unix approach: if a device such as a terminal needs
special handling, it is in the domain of the terminal device driver to provide this
functionality. Thus, a simple function suitable as put() for fdevopen() that
talks to a UART interface might look like this:
int
uart_putchar(char c, FILE *stream)
{
if (c == ’\n’)
uart_putchar(’\r’);
loop_until_bit_is_set(UCSRA, UDRE);
UDR = c;
return 0;
}
Note 3:
This implementation has been chosen because the cost of maintaining an alias
is considerably smaller than the cost of maintaining full copies of each stream.
This macro retrieves a pointer to user defined data from a FILE stream object.
This macro inserts a pointer to user defined data into a FILE stream object.
The user data can be useful for tracking state in the put and get functions supplied to
the fdevopen() function.
Initializer for a user-supplied stdio stream. This macro acts similar to fdev_setup_-
stream(), but it is to be used as the initializer of a variable of type FILE.
The remaining arguments are to be used as explained in fdev_setup_stream().
Setup a user-supplied buffer as an stdio stream. This macro takes a user-supplied buffer
stream, and sets it up as a stream that is valid for stdio operations, similar to one that
has been obtained dynamically from fdevopen(). The buffer to setup must be of type
FILE.
The arguments put and get are identical to those that need to be passed to fdevopen().
The rwflag argument can take one of the values _FDEV_SETUP_READ, _FDEV_-
SETUP_WRITE, or _FDEV_SETUP_RW, for read, write, or read/write intent, respec-
tively.
Note:
No assignments to the standard streams will be performed by fdev_setup_stream().
If standard streams are to be used, these need to be assigned by the user. See also
under Running stdio without malloc().
FILE is the opaque structure that is passed around between the various standard IO
functions.
The macro getchar reads a character from stdin. Return values and error
handling is identical to fgetc().
Stream destined for error output. Unless specifically assigned, identical to stdout.
If stderr should point to another stream, the result of another fdevopen() must
be explicitly assigned to it without closing the previous stderr (since this would also
close stdout).
Stream that will be used as an input stream by the simplified functions that don’t take
a stream argument.
The first stream opened with read intent using fdevopen() will be assigned to
stdin.
Stream that will be used as an output stream by the simplified functions that don’t
take a stream argument.
The first stream opened with write intent using fdevopen() will be assigned to both,
stdin, and stderr.
This function closes stream, and disallows and further IO to and from it.
When using fdevopen() to setup the stream, a call to fclose() is needed in order to free
the internal resources allocated.
If the stream has been set up using fdev_setup_stream() or FDEV_SETUP_-
STREAM(), use fdev_close() instead.
It currently always returns 0 (for success).
one character from the device, passed as an int type. If an error occurs when trying
to read from the device, it shall return _FDEV_ERR. If an end-of-file condition was
reached while reading from the device, _FDEV_EOF shall be returned.
If both functions are provided, the stream is opened with read and write intent.
The first stream opened with read intent is assigned to stdin, and the first one opened
with write intent is assigned to both, stdout and stderr.
fdevopen() uses calloc() (und thus malloc()) in order to allocate the storage for the new
stream.
Note:
If the macro __STDIO_FDEVOPEN_COMPAT_12 is declared before including
<stdio.h>, a function prototype for fdevopen() will be chosen that is backwards
compatible with avr-libc version 1.2 and before. This is solely intented for pro-
viding a simple migration path without the need to immediately change all source
code. Do not use for new code.
Test the end-of-file flag of stream. This flag can only be cleared by a call to
clearerr().
Test the error flag of stream. This flag can only be cleared by a call to clearerr().
Flush stream.
This is a null operation provided for source-code compatibility only, as the standard IO
implementation currently does not perform any buffering.
Read at most
size - 1 bytes from stream, until a newline character was encountered, and store
the characters in the buffer pointed to by str. Unless an error was encountered while
reading, the string will then be terminated with a NUL character.
If an error was encountered, the function returns NULL and sets the error flag of
stream, which can be tested using ferror(). Otherwise, a pointer to the string will
be returned.
Variant of fprintf() that uses a fmt string that resides in program memory.
The function fputc sends the character c (though given as type int) to stream. It
returns the character, or EOF in case an error occurred.
23.9.3.14 size_t fread (void ∗ __ptr, size_t __size, size_t __nmemb, FILE ∗
__stream)
Read nmemb objects, size bytes each, from stream, to the buffer pointed to by
ptr.
Returns the number of objects successfully read, i. e. nmemb unless an input error
occured or end-of-file was encountered. feof() and ferror() must be used to distinguish
between these two conditions.
The function fscanf performs formatted input, reading the input data from
stream.
See vfscanf() for details.
23.9.3.17 size_t fwrite (const void ∗ __ptr, size_t __size, size_t __nmemb, FILE
∗ __stream)
Write nmemb objects, size bytes each, to stream. The first byte of the first object
is referenced by ptr.
Returns the number of objects successfully written, i. e. nmemb unless an output error
occured.
Similar to fgets() except that it will operate on stream stdin, and the
trailing newline (if any) will not be stored in the string. It is the caller’s responsibility
to provide enough storage to hold the characters read.
Variant of printf() that uses a fmt string that resides in program memory.
Write the string pointed to by str, and a trailing newline character, to stdout.
23.9.3.25 int snprintf (char ∗ __s, size_t __n, const char ∗ __fmt, ...)
23.9.3.26 int snprintf_P (char ∗ __s, size_t __n, const char ∗ __fmt, ...)
Variant of snprintf() that uses a fmt string that resides in program memory.
Variant of sprintf() that uses a fmt string that resides in program memory.
23.9.3.29 int sscanf (const char ∗ __buf, const char ∗ __fmt, ...)
The function sscanf performs formatted input, reading the input data from the
buffer pointed to by buf.
See vfscanf() for details.
23.9.3.30 int sscanf_P (const char ∗ __buf, const char ∗ __fmt, ...)
23.9.3.32 int vfprintf (FILE ∗ __stream, const char ∗ __fmt, va_list __ap)
rather than on the left with blanks or zeros. A - overrides a 0 if both are
given.
– ’ ’ (space) A blank should be left before a positive number produced by a
signed conversion (d, or i).
– + A sign must always be placed before a number produced by a signed
conversion. A + overrides a space if both are used.
• An optional decimal digit string specifying a minimum field width. If the con-
verted value has fewer characters than the field width, it will be padded with
spaces on the left (or right, if the left-adjustment flag has been given) to fill out
the field width.
• An optional precision, in the form of a period . followed by an optional digit
string. If the digit string is omitted, the precision is taken as zero. This gives the
minimum number of digits to appear for d, i, o, u, x, and X conversions, or the
maximum number of characters to be printed from a string for s conversions.
• An optional l or h length modifier, that specifies that the argument for the d, i,
o, u, x, or X conversion is a "long int" rather than int. The h is ignored,
as "short int" is equivalent to int.
• A character that specifies the type of conversion to be applied.
• diouxX The int (or appropriate variant) argument is converted to signed decimal
(d and i), unsigned octal (o), unsigned decimal (u), or unsigned hexadecimal
(x and X) notation. The letters "abcdef" are used for x conversions; the letters
"ABCDEF" are used for X conversions. The precision, if any, gives the minimum
number of digits that must appear; if the converted value requires fewer digits, it
is padded on the left with zeros.
• p The void ∗ argument is taken as an unsigned integer, and converted similarly
as a %#x command would do.
• c The int argument is converted to an "unsigned char", and the resulting
character is written.
• s The "char ∗" argument is expected to be a pointer to an array of character
type (pointer to a string). Characters from the array are written up to (but not
including) a terminating NUL character; if a precision is specified, no more than
the number specified are written. If a precision is given, no null character need
be present; if the precision is not specified, or is greater than the size of the array,
the array must contain a terminating NUL character.
• % A % is written. No argument is converted. The complete conversion specifica-
tion is "%%".
In no case does a non-existent or small field width cause truncation of a numeric field;
if the result of a conversion is wider than the field width, the field is expanded to contain
the conversion result.
Since the full implementation of all the mentioned features becomes fairly large, three
different flavours of vfprintf() can be selected using linker options. The default vf-
printf() implements all the mentioned functionality except floating point conversions.
A minimized version of vfprintf() is available that only implements the very basic in-
teger and string conversion facilities, but only the # additional option can be specified
using conversion flags (these flags are parsed correctly from the format specification,
but then simply ignored). This version can be requested using the following compiler
options:
-Wl,-u,vfprintf -lprintf_min
If the full functionality including the floating point conversions is required, the follow-
ing options should be used:
Limitations:
• The specified width and precision can be at most 255.
Notes:
• For floating-point conversions, if you link default or minimized version of
vfprintf(), the symbol ? will be output and double argument will be skiped.
So you output below will not be crashed. For default version the width field
and the "pad to left" ( symbol minus ) option will work in this case.
• The hh length modifier is ignored (char argument is promouted to int).
More exactly, this realization does not check the number of h symbols.
• But the ll length modifier will to abort the output, as this realization does
not operate long long arguments.
• The variable width or precision field (an asterisk ∗ symbol) is not realized
and will to abort the output.
23.9.3.33 int vfprintf_P (FILE ∗ __stream, const char ∗ __fmt, va_list __ap)
Variant of vfprintf() that uses a fmt string that resides in program memory.
23.9.3.34 int vfscanf (FILE ∗ stream, const char ∗ fmt, va_list ap)
Formatted input. This function is the heart of the scanf family of functions.
Characters are read from stream and processed in a way described by fmt. Conversion
results will be assigned to the parameters passed via ap.
The format string fmt is scanned for conversion specifications. Anything that doesn’t
comprise a conversion specification is taken as text that is matched literally against
the input. White space in the format string will match any white space in the data
(including none), all other characters match only itself. Processing is aborted as soon as
the data and format string no longer match, or there is an error or end-of-file condition
on stream.
Most conversions skip leading white space before starting the actual conversion.
Conversions are introduced with the character %. Possible options can follow the %:
• a ∗ indicating that the conversion should be performed but the conversion result
is to be discarded; no parameters will be processed from ap,
• the character h indicating that the argument is a pointer to short int (rather
than int),
• the 2 characters hh indicating that the argument is a pointer to char (rather than
int).
• the character l indicating that the argument is a pointer to long int (rather
than int, for integer type conversions), or a pointer to double (for floating
point conversions),
• c Matches a sequence of width count characters (default 1); the next pointer must
be a pointer to char, and there must be enough room for all the characters (no
terminating NUL is added). The usual skip of leading white space is suppressed.
To skip white space first, use an explicit space in the format.
open bracket or the circumflex; any other position will end the set. The hyphen
character - is also special; when placed between two other characters, it adds all
intervening characters to the set. To include a hyphen, make it the last character
before the final close bracket. For instance, [∧ ]0-9-] means the set of every-
thing except close bracket, zero through nine, and hyphen. The string ends with
the appearance of a character not in the (or, with a circumflex, in) set or when
the field width runs out. Note that usage of this conversion enlarges the stack
expense.
• p Matches a pointer value (as printed by p in printf()); the next pointer must be
a pointer to void.
• n Nothing is expected; instead, the number of characters consumed thus far from
the input is stored through the next pointer, which must be a pointer to int. This
is not a conversion, although it can be suppressed with the ∗ flag.
These functions return the number of input items assigned, which can be fewer than
provided for, or even zero, in the event of a matching failure. Zero indicates that, while
there was input available, no conversions were assigned; typically this is due to an
invalid input character, such as an alphabetic character for a d conversion. The value
EOF is returned if an input failure occurs before any conversion such as an end-of-file
occurs. If an error or end-of-file occurs after conversion has begun, the number of
conversions which were successfully completed is returned.
By default, all the conversions described above are available except the floating-point
conversions and the width is limited to 255 characters. The float-point conversion will
be available in the extended version provided by the library libscanf_flt.a. Also
in this case the width is not limited (exactly, it is limited to 65535 characters). To link
a program against the extended version, use the following compiler flags in the link
stage:
A third version is available for environments that are tight on space. In addition to
the restrictions of the standard one, this version implements no %[ specification. This
version is provided in the library libscanf_min.a, and can be requested using the
following options in the link stage:
23.9.3.35 int vfscanf_P (FILE ∗ __stream, const char ∗ __fmt, va_list __ap)
The function vscanf performs formatted input from stream stdin, taking a
variable argument list as in vfscanf().
See vfscanf() for details.
23.9.3.38 int vsnprintf (char ∗ __s, size_t __n, const char ∗ __fmt, va_list ap)
23.9.3.39 int vsnprintf_P (char ∗ __s, size_t __n, const char ∗ __fmt, va_list ap)
Variant of vsnprintf() that uses a fmt string that resides in program memory.
23.9.3.40 int vsprintf (char ∗ __s, const char ∗ __fmt, va_list ap)
Like sprintf() but takes a variable argument list for the arguments.
23.9.3.41 int vsprintf_P (char ∗ __s, const char ∗ __fmt, va_list ap)
Variant of vsprintf() that uses a fmt string that resides in program memory.
Data Structures
• struct div_t
• struct ldiv_t
Defines
Typedefs
Functions
Variables
• size_t __malloc_margin
• char ∗ __malloc_heap_start
• char ∗ __malloc_heap_end
Note that these functions are not located in the default library, libc.a, but in the
mathematical library, libm.a. So when linking the application, the -lm option needs
to be specified.
• char ∗ dtostre (double __val, char ∗__s, unsigned char __prec, unsigned char
__flags)
• char ∗ dtostrf (double __val, signed char __width, unsigned char __prec, char
∗__s)
• #define DTOSTR_ALWAYS_SIGN 0x01
• #define DTOSTR_PLUS_SIGN 0x02
• #define DTOSTR_UPPERCASE 0x04
#include <stdlib.h>
This file declares some basic C macros and functions as defined by the ISO standard,
plus some AVR-specific extensions.
Note:
The abs() and labs() functions are builtins of gcc.
The atof() function converts the initial portion of the string pointed to by nptr to
double representation.
It is equivalent to calling
strtod(nptr, (char **)0);
Convert a string to an integer. The atoi() function converts the initial portion of the
string pointed to by s to integer representation. In contrast to
this function does not detect overflow (errno is not changed and the result value is
not predictable), uses smaller memory (flash and stack) and works more quickly.
Convert a string to a long integer. The atol() function converts the initial portion of the
string pointed to by s to long integer representation. In contrast to
this function does not detect overflow (errno is not changed and the result value is
not predictable), uses smaller memory (flash and stack) and works more quickly.
23.10.4.6 void∗ bsearch (const void ∗ __key, const void ∗ __base, size_t
__nmemb, size_t __size, int(∗)(const void ∗, const void ∗) __compar)
The bsearch() function searches an array of nmemb objects, the initial member of
which is pointed to by base, for a member that matches the object pointed to by key.
The size of each member of the array is specified by size.
The contents of the array should be in ascending sorted order according to the compar-
ison function referenced by compar. The compar routine is expected to have two
arguments which point to the key object and to an array member, in that order, and
should return an integer less than, equal to, or greater than zero if the key object is
found, respectively, to be less than, to match, or be greater than the array member.
The bsearch() function returns a pointer to a matching member of the array, or a null
pointer if no match is found. If two members compare as equal, which member is
matched is unspecified.
Allocate nele elements of size each. Identical to calling malloc() using nele
∗ size as argument, except the allocated memory will be cleared to zero.
23.10.4.9 char∗ dtostre (double __val, char ∗ __s, unsigned char __prec,
unsigned char __flags)
23.10.4.10 char∗ dtostrf (double __val, signed char __width, unsigned char
__prec, char ∗ __s)
reach an infinite loop, thereby effectively halting all code processing. Before entering
the infinite loop, interrupts are globally disabled.
In a C++ context, global destructors will be called before halting execution.
The free() function causes the allocated memory referenced by ptr to be made
available for future allocations. If ptr is NULL, no action occurs.
Convert an integer to a string. The function itoa() converts the integer value from val
into an ASCII representation that will be stored under s. The caller is responsible for
providing sufficient storage in s.
Note:
The minimal size of the buffer s depends on the choice of radix. For example, if
the radix is 2 (binary), you need to supply a buffer with a minimal length of 8 ∗
sizeof (int) + 1 characters, i.e. one character for each bit plus one for the string
terminator. Using a larger radix will require a smaller minimal buffer size.
Warning:
Conversion is done using the radix as base, which may be a number between 2
(binary conversion) and up to 36. If radix is greater than 10, the next digit after
’9’ will be the letter ’a’.
If radix is 10 and val is negative, a minus sign will be prepended.
The itoa() function returns the pointer passed as s.
The labs() function computes the absolute value of the long integer i.
Note:
The abs() and labs() functions are builtins of gcc.
23.10.4.16 char∗ ltoa (long int __val, char ∗ __s, int __radix)
Convert a long integer to a string. The function ltoa() converts the long integer value
from val into an ASCII representation that will be stored under s. The caller is
responsible for providing sufficient storage in s.
Note:
The minimal size of the buffer s depends on the choice of radix. For example,
if the radix is 2 (binary), you need to supply a buffer with a minimal length of 8
∗ sizeof (long int) + 1 characters, i.e. one character for each bit plus one for the
string terminator. Using a larger radix will require a smaller minimal buffer size.
Warning:
Conversion is done using the radix as base, which may be a number between 2
(binary conversion) and up to 36. If radix is greater than 10, the next digit after
’9’ will be the letter ’a’.
If radix is 10 and val is negative, a minus sign will be prepended.
The ltoa() function returns the pointer passed as s.
The malloc() function allocates size bytes of memory. If malloc() fails, a NULL
pointer is returned.
Note that malloc() does not initialize the returned memory to zero bytes.
See the chapter about malloc() usage for implementation details.
The qsort() function sorts an array of nmemb objects, the initial member of which is
pointed to by base. The size of each object is specified by size. The contents of the
array base are sorted in ascending order according to a comparison function pointed to
by compar, which requires two arguments pointing to the objects being compared.
The comparison function must return an integer less than, equal to, or greater than zero
if the first argument is considered to be respectively less than, equal to, or greater than
the second.
Variant of rand() that stores the context in the user-supplied variable located at ctx
instead of a static library variable so the function becomes re-entrant.
Variant of random() that stores the context in the user-supplied variable located at
ctx instead of a static library variable so the function becomes re-entrant.
The realloc() function tries to change the size of the region allocated at ptr to the
new size value. It returns a pointer to the new region. The returned pointer might be
the same as the old pointer, or a pointer to a completely different region.
The contents of the returned region up to either the old or the new size value (whatever
is less) will be identical to the contents of the old region, even in case a new region had
to be allocated.
It is acceptable to pass ptr as NULL, in which case realloc() will behave identical to
malloc().
If the new memory cannot be allocated, realloc() returns NULL, and the region at ptr
will not be changed.
The strtod() function converts the initial portion of the string pointed to by nptr to
double representation.
The expected form of the string is an optional plus ( ’+’ ) or minus sign ( ’-’ )
followed by a sequence of digits optionally containing a decimal-point character, op-
tionally followed by an exponent. An exponent consists of an ’E’ or ’e’, followed
by an optional plus or minus sign, followed by a sequence of digits.
Leading white-space characters in the string are skipped.
The strtod() function returns the converted value, if any.
If endptr is not NULL, a pointer to the character after the last character used in the
conversion is stored in the location referenced by endptr.
If no conversion is performed, zero is returned and the value of nptr is stored in the
location referenced by endptr.
If the correct value would cause overflow, plus or minus INFINITY is returned (ac-
cording to the sign of the value), and ERANGE is stored in errno. If the correct value
would cause underflow, zero is returned and ERANGE is stored in errno.
23.10.4.27 long strtol (const char ∗ __nptr, char ∗∗ __endptr, int __base)
The strtol() function converts the string in nptr to a long value. The conversion is
done according to the given base, which must be between 2 and 36 inclusive, or be the
special value 0.
The string may begin with an arbitrary amount of white space (as determined by iss-
pace()) followed by a single optional ’+’ or ’-’ sign. If base is zero or 16, the string
may then include a "0x" prefix, and the number will be read in base 16; otherwise, a
zero base is taken as 10 (decimal) unless the next character is ’0’, in which case it is
taken as 8 (octal).
The remainder of the string is converted to a long value in the obvious manner, stopping
at the first character which is not a valid digit in the given base. (In bases above 10, the
letter ’A’ in either upper or lower case represents 10, ’B’ represents 11, and so forth,
with ’Z’ representing 35.)
If endptr is not NULL, strtol() stores the address of the first invalid character in
∗endptr. If there were no digits at all, however, strtol() stores the original value of
nptr in endptr. (Thus, if ∗nptr is not ’\0’ but ∗∗endptr is ’\0’ on return, the
entire string was valid.)
The strtol() function returns the result of the conversion, unless the value would under-
flow or overflow. If no conversion could be performed, 0 is returned. If an overflow or
underflow occurs, errno is set to ERANGE and the function return value is clamped
to LONG_MIN or LONG_MAX, respectively.
23.10.4.28 unsigned long strtoul (const char ∗ __nptr, char ∗∗ __endptr, int
__base)
23.10.4.29 char∗ ultoa (unsigned long int __val, char ∗ __s, int __radix)
Convert an unsigned long integer to a string. The function ultoa() converts the unsigned
long integer value from val into an ASCII representation that will be stored under s.
The caller is responsible for providing sufficient storage in s.
Note:
The minimal size of the buffer s depends on the choice of radix. For example, if
the radix is 2 (binary), you need to supply a buffer with a minimal length of 8 ∗
sizeof (unsigned long int) + 1 characters, i.e. one character for each bit plus one
for the string terminator. Using a larger radix will require a smaller minimal buffer
size.
Warning:
Conversion is done using the radix as base, which may be a number between 2
(binary conversion) and up to 36. If radix is greater than 10, the next digit after
’9’ will be the letter ’a’.
The ultoa() function returns the pointer passed as s.
23.10.4.30 char∗ utoa (unsigned int __val, char ∗ __s, int __radix)
Convert an unsigned integer to a string. The function utoa() converts the unsigned
integer value from val into an ASCII representation that will be stored under s. The
caller is responsible for providing sufficient storage in s.
Note:
The minimal size of the buffer s depends on the choice of radix. For example, if
the radix is 2 (binary), you need to supply a buffer with a minimal length of 8 ∗
sizeof (unsigned int) + 1 characters, i.e. one character for each bit plus one for the
string terminator. Using a larger radix will require a smaller minimal buffer size.
Warning:
Conversion is done using the radix as base, which may be a number between 2
(binary conversion) and up to 36. If radix is greater than 10, the next digit after
’9’ will be the letter ’a’.
The utoa() function returns the pointer passed as s.
malloc() tunable.
malloc() tunable.
malloc() tunable.
Defines
• #define _FFS(x)
Functions
#include <string.h>
Note:
If the strings you are working on resident in program space (flash), you will need to
use the string functions described in <avr/pgmspace.h>: Program Space Utilities.
This macro finds the first (least significant) bit set in the input value.
This macro is very similar to the function ffs() except that it evaluates its argument at
compile-time, so it should only be applied to compile-time constant expressions where
it will reduce to a constant itself. Application of this macro to expressions that are not
constant at compile-time is not recommended, and might result in a huge amount of
code generated.
Returns:
The _FFS() macro returns the position of the first (least significant) bit set in the
word val, or 0 if no bits are set. The least significant bit is position 1. Only 16 bits
of argument are evaluted.
This function finds the first (least significant) bit set in the input value.
Returns:
The ffs() function returns the position of the first (least significant) bit set in the
word val, or 0 if no bits are set. The least significant bit is position 1.
Note:
For expressions that are constant at compile time, consider using the _FFS macro
instead.
23.11.3.4 void ∗ memccpy (void ∗ dest, const void ∗ src, int val, size_t len)
Copy memory area. The memccpy() function copies no more than len bytes from
memory area src to memory area dest, stopping when the character val is found.
Returns:
The memccpy() function returns a pointer to the next character in dest after val,
or NULL if val was not found in the first len characters of src.
23.11.3.5 void ∗ memchr (const void ∗ src, int val, size_t len)
Scan memory for a character. The memchr() function scans the first len bytes of the
memory area pointed to by src for the character val. The first byte to match val (inter-
preted as an unsigned character) stops the operation.
Returns:
The memchr() function returns a pointer to the matching byte or NULL if the
character does not occur in the given memory area.
23.11.3.6 int memcmp (const void ∗ s1, const void ∗ s2, size_t len)
Compare memory areas. The memcmp() function compares the first len bytes of the
memory areas s1 and s2. The comparision is performed using unsigned char operations.
Returns:
The memcmp() function returns an integer less than, equal to, or greater than zero
if the first len bytes of s1 is found, respectively, to be less than, to match, or be
greater than the first len bytes of s2.
Note:
Be sure to store the result in a 16 bit variable since you may get incorrect results if
you use an unsigned char or char due to truncation.
Warning:
This function is not -mint8 compatible, although if you only care about testing for
equality, this function should be safe to use.
23.11.3.7 void ∗ memcpy (void ∗ dest, const void ∗ src, size_t len)
Copy a memory area. The memcpy() function copies len bytes from memory area src
to memory area dest. The memory areas may not overlap. Use memmove() if the
memory areas do overlap.
Returns:
The memcpy() function returns a pointer to dest.
23.11.3.8 void ∗ memmem (const void ∗ s1, size_t len1, const void ∗ s2, size_t
len2)
The memmem() function finds the start of the first occurrence of the substring s2 of
length len2 in the memory area s1 of length len1.
Returns:
The memmem() function returns a pointer to the beginning of the substring, or
NULL if the substring is not found. If len2 is zero, the function returns s1.
23.11.3.9 void ∗ memmove (void ∗ dest, const void ∗ src, size_t len)
Copy memory area. The memmove() function copies len bytes from memory area src
to memory area dest. The memory areas may overlap.
Returns:
The memmove() function returns a pointer to dest.
23.11.3.10 void ∗ memrchr (const void ∗ src, int val, size_t len)
The memrchr() function is like the memchr() function, except that it searches
backwards from the end of the len bytes pointed to by src instead of forwards from
the front. (Glibc, GNU extension.)
Returns:
The memrchr() function returns a pointer to the matching byte or NULL if the
character does not occur in the given memory area.
Fill memory with a constant byte. The memset() function fills the first len bytes of the
memory area pointed to by dest with the constant byte val.
Returns:
The memset() function returns a pointer to the memory area dest.
Compare two strings ignoring case. The strcasecmp() function compares the two
strings s1 and s2, ignoring the case of the characters.
Returns:
The strcasecmp() function returns an integer less than, equal to, or greater than
zero if s1 is found, respectively, to be less than, to match, or be greater than
s2. A consequence of the ordering used by strcasecmp() is that if s1 is an initial
substring of s2, then s1 is considered to be "less than" s2.
The strcasestr() function finds the first occurrence of the substring s2 in the string
s1. This is like strstr(), except that it ignores case of alphabetic symbols in searching
for the substring. (Glibc, GNU extension.)
Returns:
The strcasestr() function returns a pointer to the beginning of the substring, or
NULL if the substring is not found. If s2 points to a string of zero length, the
function returns s1.
Concatenate two strings. The strcat() function appends the src string to the dest string
overwriting the ’\0’ character at the end of dest, and then adds a terminating ’\0’
character. The strings may not overlap, and the dest string must have enough space for
the result.
Returns:
The strcat() function returns a pointer to the resulting string dest.
Locate character in string. The strchr() function returns a pointer to the first occurrence
of the character val in the string src.
Here "character" means "byte" - these functions do not work with wide or multi-byte
characters.
Returns:
The strchr() function returns a pointer to the matched character or NULL if the
character is not found.
Returns:
The strchrnul() function returns a pointer to the matched character, or a pointer to
the null byte at the end of s (i.e., s+strlen(s)) if the character is not found.
Compare two strings. The strcmp() function compares the two strings s1 and s2.
Returns:
The strcmp() function returns an integer less than, equal to, or greater than zero
if s1 is found, respectively, to be less than, to match, or be greater than s2. A
consequence of the ordering used by strcmp() is that if s1 is an initial substring of
s2, then s1 is considered to be "less than" s2.
Copy a string. The strcpy() function copies the string pointed to by src (including the
terminating ’\0’ character) to the array pointed to by dest. The strings may not overlap,
and the destination string dest must be large enough to receive the copy.
Returns:
The strcpy() function returns a pointer to the destination string dest.
Note:
If the destination string of a strcpy() is not large enough (that is, if the programmer
was stupid/lazy, and failed to check the size before copying) then anything might
happen. Overflowing fixed length strings is a favourite cracker technique.
The strcspn() function calculates the length of the initial segment of s which consists
entirely of characters not in reject.
Returns:
The strcspn() function returns the number of characters in the initial segment of s
which are not in the string reject. The terminating zero is not considered as a
part of string.
Duplicate a string. The strdup() function allocates memory and copies into it the string
addressed by s1, including the terminating null character.
Warning:
The strdup() function calls malloc() to allocate the memory for the duplicated
string! The user is responsible for freeing the memory by calling free().
Returns:
The strdup() function returns a pointer to the resulting string dest. If malloc()
cannot allocate enough storage for the string, strdup() will return NULL.
Warning:
Be sure to check the return value of the strdup() function to make sure that the
function has succeeded in allocating the memory!
23.11.3.21 size_t strlcat (char ∗ dst, const char ∗ src, size_t siz)
Concatenate two strings. Appends src to string dst of size siz (unlike strncat(), siz is
the full size of dst, not space left). At most siz-1 characters will be copied. Always
NULL terminates (unless siz <= strlen(dst)).
Returns:
The strlcat() function returns strlen(src) + MIN(siz, strlen(initial dst)). If retval >=
siz, truncation occurred.
Appends src to string dst of size siz (unlike strncat(), siz is the full size of dst,
not space left). At most siz-1 characters will be copied. Always NULL terminates
(unless siz <= strlen(dst)).
Returns:
The strlcat() function returns strlen(src) + MIN(siz, strlen(initial dst)). If retval >=
siz, truncation occurred.
23.11.3.22 size_t strlcpy (char ∗ dst, const char ∗ src, size_t siz)
Copy a string. Copy src to string dst of size siz. At most siz-1 characters will be copied.
Always NULL terminates (unless siz == 0).
Returns:
The strlcpy() function returns strlen(src). If retval >= siz, truncation occurred.
Copy src to string dst of size siz. At most siz-1 characters will be copied.
Always NULL terminates (unless siz == 0).
Returns:
The strlcpy() function returns strlen(src). If retval >= siz, truncation occurred.
Calculate the length of a string. The strlen() function calculates the length of the string
src, not including the terminating ’\0’ character.
Returns:
The strlen() function returns the number of characters in src.
Convert a string to lower case. The strlwr() function will convert a string to lower
case. Only the upper case alphabetic characters [A .. Z] are converted. Non-alphabetic
characters will not be changed.
Returns:
The strlwr() function returns a pointer to the converted string.
23.11.3.25 int strncasecmp (const char ∗ s1, const char ∗ s2, size_t len)
Compare two strings ignoring case. The strncasecmp() function is similar to str-
casecmp(), except it only compares the first len characters of s1.
Returns:
The strncasecmp() function returns an integer less than, equal to, or greater than
zero if s1 (or the first len bytes thereof) is found, respectively, to be less than, to
match, or be greater than s2. A consequence of the ordering used by strncasecmp()
is that if s1 is an initial substring of s2, then s1 is considered to be "less than"
s2.
23.11.3.26 char ∗ strncat (char ∗ dest, const char ∗ src, size_t len)
Concatenate two strings. The strncat() function is similar to strcat(), except that only
the first n characters of src are appended to dest.
Returns:
The strncat() function returns a pointer to the resulting string dest.
23.11.3.27 int strncmp (const char ∗ s1, const char ∗ s2, size_t len)
Compare two strings. The strncmp() function is similar to strcmp(), except it only
compares the first (at most) n characters of s1 and s2.
Returns:
The strncmp() function returns an integer less than, equal to, or greater than zero
if s1 (or the first n bytes thereof) is found, respectively, to be less than, to match,
or be greater than s2.
23.11.3.28 char ∗ strncpy (char ∗ dest, const char ∗ src, size_t len)
Copy a string. The strncpy() function is similar to strcpy(), except that not more than n
bytes of src are copied. Thus, if there is no null byte among the first n bytes of src, the
result will not be null-terminated.
In the case where the length of src is less than that of n, the remainder of dest will be
padded with nulls.
Returns:
The strncpy() function returns a pointer to the destination string dest.
Determine the length of a fixed-size string. The strnlen function returns the number of
characters in the string pointed to by src, not including the terminating ’\0’ character,
but at most len. In doing this, strnlen looks only at the first len characters at src and
never beyond src+len.
Returns:
The strnlen function returns strlen(src), if that is less than len, or len if there is no
’\0’ character among the first len characters pointed to by src.
The strpbrk() function locates the first occurrence in the string s of any of the
characters in the string accept.
Returns:
The strpbrk() function returns a pointer to the character in s that matches one of
the characters in accept, or NULL if no such character is found. The terminating
zero is not considered as a part of string: if one or both args are empty, the result
will NULL.
Locate character in string. The strrchr() function returns a pointer to the last occurrence
of the character val in the string src.
Here "character" means "byte" - these functions do not work with wide or multi-byte
characters.
Returns:
The strrchr() function returns a pointer to the matched character or NULL if the
character is not found.
Reverse a string. The strrev() function reverses the order of the string.
Returns:
The strrev() function returns a pointer to the beginning of the reversed string.
Parse a string into tokens. The strsep() function locates, in the string referenced by
∗sp, the first occurrence of any character in the string delim (or the terminating ’\0’
character) and replaces it with a ’\0’. The location of the next character after the
delimiter character (or NULL, if the end of the string was reached) is stored in ∗sp. An
“empty” field, i.e. one caused by two adjacent delimiter characters, can be detected by
comparing the location referenced by the pointer returned in ∗sp to ’\0’.
Returns:
The strsep() function returns a pointer to the original value of ∗sp. If ∗sp is
initially NULL, strsep() returns NULL.
The strspn() function calculates the length of the initial segment of s which consists
entirely of characters in accept.
Returns:
The strspn() function returns the number of characters in the initial segment of
s which consist only of characters from accept. The terminating zero is not
considered as a part of string.
Locate a substring. The strstr() function finds the first occurrence of the substring s2
in the string s1. The terminating ’\0’ characters are not compared.
Returns:
The strstr() function returns a pointer to the beginning of the substring, or NULL
if the substring is not found. If s2 points to a string of zero length, the function
returns s1.
Parses the string s into tokens. strtok parses the string s into tokens. The first call
to strtok should have s as its first argument. Subsequent calls should have the first
argument set to NULL. If a token ends with a delimiter, this delimiting character is
overwritten with a ’\0’ and a pointer to the next character is saved for the next call to
strtok. The delimiter string delim may be different for each call.
Returns:
The strtok() function returns a pointer to the next token or NULL when no more
tokens are found.
Note:
strtok() is NOT reentrant. For a reentrant version of this function see strtok_-
r().
23.11.3.37 char ∗ strtok_r (char ∗ string, const char ∗ delim, char ∗∗ last)
Parses string into tokens. strtok_r parses string into tokens. The first call to strtok_r
should have string as its first argument. Subsequent calls should have the first argument
set to NULL. If a token ends with a delimiter, this delimiting character is overwritten
with a ’\0’ and a pointer to the next character is saved for the next call to strtok_r.
The delimiter string delim may be different for each call. last is a user allocated
char∗ pointer. It must be the same while parsing the same string. strtok_r is a reentrant
version of strtok().
Returns:
The strtok_r() function returns a pointer to the next token or NULL when no more
tokens are found.
Convert a string to upper case. The strupr() function will convert a string to upper
case. Only the lower case alphabetic characters [a .. z] are converted. Non-alphabetic
characters will not be changed.
Returns:
The strupr() function returns a pointer to the converted string. The pointer is the
same as that passed in since the operation is perform in place.
Defines
#include <avr/io.h>
#include <avr/boot.h>
The macros in this module provide a C language interface to the bootloader support
functionality of certain AVR processors. These macros are designed to work with all
sizes of flash memory.
Global interrupts are not automatically disabled for these macros. It is left up to the
programmer to do this. See the code example below. Also see the processor datasheet
for caveats on having global interrupts enabled during writing of the Flash.
Note:
Not all AVR processors provide bootloader support. See your processor datasheet
to see if it provides bootloader support.
Todo
From email with Marek: On smaller devices (all except ATmega64/128), __SPM_-
REG is in the I/O space, accessible with the shorter "in" and "out" instructions -
since the boot loader has a limited size, this could be an important optimization.
#include <inttypes.h>
#include <avr/interrupt.h>
#include <avr/pgmspace.h>
// Disable interrupts.
sreg = SREG;
cli();
eeprom_busy_wait ();
boot_page_erase (page);
boot_spm_busy_wait (); // Wait until the memory is erased.
uint16_t w = *buf++;
w += (*buf++) << 8;
boot_rww_enable ();
SREG = sreg;
}
Parameters:
lock_bits A mask of which Boot Loader Lock Bits to set.
Note:
In this context, a ’set bit’ will be written to a zero value. Note also that only BLBxx
bits can be programmed by this command.
For example, to disallow the SPM instruction from writing to the Boot Loader memory
section of flash, you would use this macro as such:
Note:
Like any lock bits, the Boot Loader Lock Bits, once set, cannot be cleared again
except by a chip erase which will in turn also erase the boot loader itself.
Value:
do { \
boot_spm_busy_wait(); \
eeprom_busy_wait(); \
boot_lock_bits_set (lock_bits); \
} while (0)
Same as boot_lock_bits_set() except waits for eeprom and spm operations to complete
before setting the lock bits.
Value:
(__extension__({ \
uint8_t __result; \
__asm__ __volatile__ \
( \
"sts %1, %2\n\t" \
"lpm %0, Z\n\t" \
: "=r" (__result) \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"r" ((uint8_t)(__BOOT_LOCK_BITS_SET)), \
"z" ((uint16_t)(address)) \
); \
__result; \
}))
Note:
The lock and fuse bits returned are the physical values, i.e. a bit returned as 0
means the corresponding fuse or lock bit is programmed.
Note:
address is a byte address in flash, not a word address.
Value:
do { \
boot_spm_busy_wait(); \
eeprom_busy_wait(); \
boot_page_erase (address); \
} while (0)
Same as boot_page_erase() except it waits for eeprom and spm operations to complete
before erasing the page.
Fill the bootloader temporary page buffer for flash address with data word.
Note:
The address is a byte address. The data is a word. The AVR writes data to the
buffer a word at a time, but addresses the buffer per byte! So, increment your
address by 2 between calls, and send 2 data bytes in a word format! The LSB of
the data is written to the lower address; the MSB of the data is written to the higher
address.
Value:
do { \
boot_spm_busy_wait(); \
eeprom_busy_wait(); \
boot_page_fill(address, data); \
} while (0)
Same as boot_page_fill() except it waits for eeprom and spm operations to complete
before filling the page.
Write the bootloader temporary page buffer to flash page that contains address.
Note:
address is a byte address in flash, not a word address.
Value:
do { \
boot_spm_busy_wait(); \
eeprom_busy_wait(); \
boot_page_write (address); \
} while (0)
Same as boot_page_write() except it waits for eeprom and spm operations to complete
before writing the page.
Value:
do { \
boot_spm_busy_wait(); \
eeprom_busy_wait(); \
boot_rww_enable(); \
} while (0)
Same as boot_rww_enable() except waits for eeprom and spm operations to complete
before enabling the RWW mameory.
Value:
(__extension__({ \
uint8_t __result; \
__asm__ __volatile__ \
( \
"sts %1, %2\n\t" \
"lpm %0, Z" "\n\t" \
: "=r" (__result) \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"r" ((uint8_t)(__BOOT_SIGROW_READ)), \
"z" ((uint16_t)(addr)) \
); \
__result; \
}))
Read the Signature Row byte at address. For some MCU types, this function can
also retrieve the factory-stored oscillator calibration bytes.
Parameter address can be 0-0x1f as documented by the datasheet.
Note:
The values are MCU type dependent.
Defines
• #define _NOP()
• #define _MemoryBarrier()
#include <avr/cpufunc.h>
This header file contains macros that access special functions of the AVR CPU which
do not fit into any of the other header files.
Defines
Functions
#include <avr/eeprom.h>
This header file declares the interface to some simple library routines suitable for han-
dling the data EEPROM contained in the AVR microcontrollers. The implementation
uses a simple polled mode interface. Applications that require interrupt-controlled
EEPROM access to ensure that no time will be wasted in spinloops will have to deploy
their own implementation.
Notes:
• In addition to the write functions there is a set of update ones. This functions read
each byte first and skip the burning if the old value is the same with new. The
scaning direction is from high address to low, to obtain quick return in common
cases.
• All of the read/write functions first make sure the EEPROM is ready to be ac-
cessed. Since this may cause long delays if a write operation is still pending,
time-critical applications should first poll the EEPROM e. g. using eeprom_-
is_ready() before attempting any actual I/O. But this functions are not wait until
SELFPRGEN in SPMCSR becomes zero. Do this manually, if your softwate
contains the Flash burning.
• For Xmega the EEPROM start address is 0, like other architectures. The reading
functions add the 0x2000 value to use EEPROM mapping into data space.
Returns:
Nothing.
Returns:
1 if EEPROM is ready for a new read/write operation, 0 if not.
23.14.3.1 void eeprom_read_block (void ∗ __dst, const void ∗ __src, size_t __n)
Read a block of __n bytes from EEPROM address __src to SRAM __dst.
Read one 32-bit double word (little endian) from EEPROM address __p.
Read one float value (little endian) from EEPROM address __p.
Read one 16-bit word (little endian) from EEPROM address __p.
Note:
The argument order is mismatch with common functions like strcpy().
Note:
The argument order is mismatch with common functions like strcpy().
The Fuse API allows a user to specify the fuse settings for the specific AVR device they
are compiling for. These fuse settings will be placed in a special section in the ELF
output file, after linking.
Programming tools can take advantage of the fuse information embedded in the ELF
file, by extracting this information and determining if the fuses need to be programmed
before programming the Flash and EEPROM memories. This also allows a single ELF
file to contain all the information needed to program an AVR.
To use the Fuse API, include the <avr/io.h> header file, which in turn automatically
includes the individual I/O header file and the <avr/fuse.h> file. These other two files
provides everything necessary to set the AVR fuses.
Fuse API
Each I/O header file must define the FUSE_MEMORY_SIZE macro which is defined
to the number of fuse bytes that exist in the AVR device.
A new type, __fuse_t, is defined as a structure. The number of fields in this structure
are determined by the number of fuse bytes in the FUSE_MEMORY_SIZE macro.
If FUSE_MEMORY_SIZE == 1, there is only a single field: byte, of type unsigned
char.
If FUSE_MEMORY_SIZE == 2, there are two fields: low, and high, of type unsigned
char.
If FUSE_MEMORY_SIZE == 3, there are three fields: low, high, and extended, of
type unsigned char.
If FUSE_MEMORY_SIZE > 3, there is a single field: byte, which is an array of
unsigned char with the size of the array being FUSE_MEMORY_SIZE.
A convenience macro, FUSEMEM, is defined as a GCC attribute for a custom-named
section of ".fuse".
A convenience macro, FUSES, is defined that declares a variable, __fuse, of type __-
fuse_t with the attribute defined by FUSEMEM. This variable allows the end user to
easily set the fuse data.
Note:
If a device-specific I/O header file has previously defined FUSEMEM, then FUSE-
MEM is not redefined. If a device-specific I/O header file has previously defined
FUSES, then FUSES is not redefined.
Each AVR device I/O header file has a set of defined macros which specify the actual
fuse bits available on that device. The AVR fuses have inverted values, logical 1 for
an unprogrammed (disabled) bit and logical 0 for a programmed (enabled) bit. The
defined macros for each individual fuse bit represent this in their definition by a bit-
wise inversion of a mask. For example, the FUSE_EESAVE fuse in the ATmega128 is
defined as:
Note:
The _BV macro creates a bit mask from a bit number. It is then inverted to repre-
sent logical values for a fuse memory byte.
To combine the fuse bits macros together to represent a whole fuse byte, use the bitwise
AND operator, like so:
Each device I/O header file also defines macros that provide default values for each fuse
byte that is available. LFUSE_DEFAULT is defined for a Low Fuse byte. HFUSE_-
DEFAULT is defined for a High Fuse byte. EFUSE_DEFAULT is defined for an Ex-
tended Fuse byte.
If FUSE_MEMORY_SIZE > 3, then the I/O header file defines macros that pro-
vide default values for each fuse byte like so: FUSE0_DEFAULT FUSE1_DEFAULT
FUSE2_DEFAULT FUSE3_DEFAULT FUSE4_DEFAULT ....
#include <avr/io.h>
FUSES =
{
.low = LFUSE_DEFAULT,
.high = (FUSE_BOOTSZ0 & FUSE_BOOTSZ1 & FUSE_EESAVE & FUSE_SPIEN & FUSE_JT
AGEN),
.extended = EFUSE_DEFAULT,
};
int main(void)
{
return 0;
}
#include <avr/io.h>
int main(void)
{
return 0;
}
If you are compiling in C++, you cannot use the designated intializers so you must do:
#include <avr/io.h>
FUSES =
{
LFUSE_DEFAULT, // .low
(FUSE_BOOTSZ0 & FUSE_BOOTSZ1 & FUSE_EESAVE & FUSE_SPIEN & FUSE_JTAGEN), /
/ .high
EFUSE_DEFAULT, // .extended
};
int main(void)
{
return 0;
}
However there are a number of caveats that you need to be aware of to use this API
properly.
Be sure to include <avr/io.h> to get all of the definitions for the API. The FUSES
macro defines a global variable to store the fuse data. This variable is assigned to its
own linker section. Assign the desired fuse values immediately in the variable initial-
ization.
The .fuse section in the ELF file will get its values from the initial variable assignment
ONLY. This means that you can NOT assign values to this variable in functions and the
new values will not be put into the ELF .fuse section.
The global variable is declared in the FUSES macro has two leading underscores,
which means that it is reserved for the "implementation", meaning the library, so it
will not conflict with a user-named variable.
You must initialize ALL fields in the __fuse_t structure. This is because the fuse bits
in all bytes default to a logical 1, meaning unprogrammed. Normal uninitialized data
defaults to all locgial zeros. So it is vital that all fuse bytes are initialized, even with
default data. If they are not, then the fuse bits may not programmed to the desired
settings.
Be sure to have the -mmcu=device flag in your compile command line and your linker
command line to have the correct device selected and to have the correct I/O header
file included when you include <avr/io.h>.
You can print out the contents of the .fuse section in the ELF file by using this command
line:
The section contents shows the address on the left, then the data going from lower
address to a higher address, left to right.
The global interrupt flag is maintained in the I bit of the status register (SREG).
Handling interrupts frequently requires attention regarding atomic access to objects
that could be altered by code running within an interrupt context, see <util/atomic.h>.
Frequently, interrupts are being disabled for periods of time in order to perform certain
operations without being disturbed; see Problems with reordering code for things to be
taken into account with respect to compiler optimizations.
• #define sei()
• #define cli()
ISR attributes
• #define ISR_BLOCK
• #define ISR_NOBLOCK
• #define ISR_NAKED
• #define ISR_ALIASOF(target_vector)
#include <avr/interrupt.h>
ISR(ADC_vect)
{
// user code here
}
Refer to the chapter explaining assembler programming for an explanation about inter-
rupt routines written solely in assembler language.
#include <avr/interrupt.h>
ISR(BADISR_vect)
{
// user code here
}
Nested interrupts The AVR hardware clears the global interrupt flag in SREG be-
fore entering an interrupt vector. Thus, normally interrupts will remain disabled inside
the handler until the handler exits, where the RETI instruction (that is emitted by the
compiler as part of the normal function epilogue for an interrupt handler) will even-
tually re-enable further interrupts. For that reason, interrupt handlers normally do not
nest. For most interrupt handlers, this is the desired behaviour, for some it is even
required in order to prevent infinitely recursive interrupts (like UART interrupts, or
level-triggered external interrupts). In rare circumstances though it might be desired to
re-enable the global interrupt flag as early as possible in the interrupt handler, in order
to not defer any other interrupt more than absolutely needed. This could be done using
an sei() instruction right at the beginning of the interrupt handler, but this still leaves
few instructions inside the compiler-generated function prologue to run with global in-
terrupts disabled. The compiler can be instructed to insert an SEI instruction right at
the beginning of an interrupt handler by declaring the handler the following way:
ISR(XXX_vect, ISR_NOBLOCK)
{
...
}
where XXX_vect is the name of a valid interrupt vector for the MCU type in question,
as explained below.
Two vectors sharing the same code In some circumstances, the actions to be taken
upon two different interrupts might be completely identical so a single implementa-
tion for the ISR would suffice. For example, pin-change interrupts arriving from two
different ports could logically signal an event that is independent from the actual port
(and thus interrupt vector) where it happened. Sharing interrupt vector code can be
accomplished using the ISR_ALIASOF() attribute to the ISR macro:
ISR(PCINT0_vect)
{
...
// Code to handle the event.
}
ISR(PCINT1_vect, ISR_ALIASOF(PCINT0_vect));
Note:
There is no body to the aliased ISR.
Note that the ISR_ALIASOF() feature requires GCC 4.2 or above (or a patched version
of GCC 4.1.x). See the documentation of the ISR_ALIAS() macro for an implementa-
tion which is less elegant but could be applied to all compiler versions.
Empty interrupt service routines In rare circumstances, in interrupt vector does not
need any code to be implemented at all. The vector must be declared anyway, so when
the interrupt triggers it won’t execute the BADISR_vect code (which by default restarts
the application).
This could for example be the case for interrupts that are solely enabled for the purpose
of getting the controller out of sleep_mode().
A handler for such an interrupt vector can be declared using the EMPTY_-
INTERRUPT() macro:
EMPTY_INTERRUPT(ADC_vect);
Note:
There is no body to this macro.
ISR(TIMER1_OVF_vect, ISR_NAKED)
{
PORTB |= _BV(0); // results in SBI which does not affect SREG
reti();
}
Choosing the vector: Interrupt vector names The interrupt is chosen by supplying
one of the symbols in following table.
There are currently two different styles present for naming the vectors. One form uses
names starting with SIG_, followed by a relatively verbose but arbitrarily chosen name
describing the interrupt vector. This has been the only available style in avr-libc up to
version 1.2.x.
Starting with avr-libc version 1.4.0, a second style of interrupt vector names has been
added, where a short phrase for the vector description is followed by _vect. The
short phrase matches the vector name as described in the datasheet of the respective
device (and in Atmel’s XML files), with spaces replaced by an underscore and other
non-alphanumeric characters dropped. Using the suffix _vect is intented to improve
portability to other C compilers available for the AVR that use a similar naming con-
vention.
The historical naming style might become deprecated in a future release, so it is not
recommended for new projects.
Note:
The ISR() macro cannot really spell-check the argument passed to them. Thus, by
misspelling one of the names below in a call to ISR(), a function will be created
that, while possibly being usable as an interrupt function, is not actually wired into
the interrupt vector table. The compiler will generate a warning if it detects a sus-
piciously looking name of a ISR() function (i.e. one that after macro replacement
does not start with "__vector_").
#include <avr/interrupt.h>
This is a vector which is aliased to __vector_default, the vector executed when an ISR
fires with no accompanying ISR handler. This may be used along with the ISR() macro
to create a catch-all for undefined but used ISRs for debugging purposes.
#include <avr/interrupt.h>
Disables all interrupts by clearing the global interrupt mask. This function actually
compiles into a single line of assembly, so there is no function call overhead. However,
the macro also implies a memory barrier which can cause additional loss of optimiza-
tion.
In order to implement atomic access to multi-byte objects, consider using the macros
from <util/atomic.h>, rather than implementing them manually with cli() and sei().
#include <avr/interrupt.h>
Defines an empty interrupt handler function. This will not generate any prolog or
epilog code and will only return from the ISR. Do not define a function body as this
will define it for you. Example:
EMPTY_INTERRUPT(ADC_vect);
#include <avr/interrupt.h>
Introduces an interrupt handler function (interrupt service routine) that runs with global
interrupts initially disabled by default with no attributes specified.
The attributes are optional and alter the behaviour and resultant generated code of the
interrupt routine. Multiple attributes may be used for a single function, with a space
seperating each attribute.
#include <avr/interrupt.h>
Aliases a given vector to another one in the same manner as the ISR_ALIASOF at-
tribute for the ISR() macro. Unlike the ISR_ALIASOF attribute macro however, this is
compatible for all versions of GCC rather than just GCC version 4.2 onwards.
Note:
This macro creates a trampoline function for the aliased macro. This will result in
a two cycle penalty for the aliased vector compared to the ISR the vector is aliased
to, due to the JMP/RJMP opcode used.
Deprecated
Example:
ISR(INT0_vect)
{
PORTB = 42;
}
ISR_ALIAS(INT1_vect, INT0_vect);
#include <avr/interrupt.h>
The ISR is linked to another ISR, specified by the vect parameter. This is compatible
with GCC 4.2 and greater only.
Use this attribute in the attributes parameter of the ISR macro.
# include <avr/interrupt.h>
Identical to an ISR with no attributes specified. Global interrupts are initially disabled
by the AVR hardware when entering the ISR, without the compiler modifying this state.
Use this attribute in the attributes parameter of the ISR macro.
# include <avr/interrupt.h>
ISR is created with no prologue or epilogue code. The user code is responsible for
preservation of the machine state including the SREG register, as well as placing a
reti() at the end of the interrupt routine.
Use this attribute in the attributes parameter of the ISR macro.
# include <avr/interrupt.h>
ISR runs with global interrupts initially enabled. The interrupt enable flag is activated
by the compiler as early as possible within the ISR to ensure minimal processing delay
for nested interrupts.
This may be used to create nested ISRs, however care should be taken to avoid stack
overflows, or to avoid infinitely entering the ISR for those cases where the AVR hard-
ware does not clear the respective interrupt flag before entering the ISR.
Use this attribute in the attributes parameter of the ISR macro.
#include <avr/interrupt.h>
Returns from an interrupt routine, enabling global interrupts. This should be the last
command executed before leaving an ISR defined with the ISR_NAKED attribute.
This macro actually compiles into a single line of assembly, so there is no function call
overhead.
#include <avr/interrupt.h>
Enables interrupts by setting the global interrupt mask. This function actually compiles
into a single line of assembly, so there is no function call overhead. However, the macro
also implies a memory barrier which can cause additional loss of optimization.
In order to implement atomic access to multi-byte objects, consider using the macros
from <util/atomic.h>, rather than implementing them manually with cli() and sei().
#include <avr/interrupt.h>
Introduces an interrupt handler function that runs with global interrupts initially dis-
abled.
This is the same as the ISR macro without optional attributes.
Deprecated
This header file includes the apropriate IO definitions for the device that has been
specified by the -mmcu= compiler command-line switch. This is done by divert-
ing to the appropriate file <avr/ioXXXX.h> which should never be included di-
rectly. Some register names common to all AVR devices are defined directly within
<avr/common.h>, which is included in <avr/io.h>, but most of the details
come from the respective include file.
Note that this file always includes the following files:
#include <avr/sfr_defs.h>
#include <avr/portpins.h>
#include <avr/common.h>
#include <avr/version.h>
See <avr/sfr_defs.h>: Special function registers for more details about that header file.
Included are definitions of the IO register set and their respective bit values as specified
in the Atmel documentation. Note that inconsistencies in naming conventions, so even
identical functions sometimes get different names on different devices.
Also included are the specific names useable for interrupt function definitions as docu-
mented here.
Finally, the following macros are defined:
• RAMEND
The last on-chip RAM address.
• XRAMEND
The last possible RAM location that is addressable. This is equal to RAMEND
for devices that do not allow for external RAM. For devices that allow external
RAM, this will be larger than RAMEND.
• E2END
The last EEPROM address.
• FLASHEND
The last byte address in the Flash program space.
• SPM_PAGESIZE
For devices with bootloader support, the flash pagesize (in bytes) to be used for
the SPM instruction.
• E2PAGESIZE
The size of the EEPROM page.
The Lockbit API allows a user to specify the lockbit settings for the specific AVR
device they are compiling for. These lockbit settings will be placed in a special section
in the ELF output file, after linking.
Programming tools can take advantage of the lockbit information embedded in the
ELF file, by extracting this information and determining if the lockbits need to be
programmed after programming the Flash and EEPROM memories. This also allows a
single ELF file to contain all the information needed to program an AVR.
To use the Lockbit API, include the <avr/io.h> header file, which in turn automatically
includes the individual I/O header file and the <avr/lock.h> file. These other two files
provides everything necessary to set the AVR lockbits.
Lockbit API
Each I/O header file may define up to 3 macros that controls what kinds of lockbits are
available to the user.
If __LOCK_BITS_EXIST is defined, then two lock bits are available to the user and 3
mode settings are defined for these two bits.
If __BOOT_LOCK_BITS_0_EXIST is defined, then the two BLB0 lock bits are avail-
able to the user and 4 mode settings are defined for these two bits.
If __BOOT_LOCK_BITS_1_EXIST is defined, then the two BLB1 lock bits are avail-
able to the user and 4 mode settings are defined for these two bits.
If __BOOT_LOCK_APPLICATION_TABLE_BITS_EXIST is defined then two lock
bits are available to set the locking mode for the Application Table Section (which is
used in the XMEGA family).
If __BOOT_LOCK_APPLICATION_BITS_EXIST is defined then two lock bits are
available to set the locking mode for the Application Section (which is used in the
XMEGA family).
If __BOOT_LOCK_BOOT_BITS_EXIST is defined then two lock bits are available
to set the locking mode for the Boot Loader Section (which is used in the XMEGA
family).
The AVR lockbit modes have inverted values, logical 1 for an unprogrammed (dis-
abled) bit and logical 0 for a programmed (enabled) bit. The defined macros for each
individual lock bit represent this in their definition by a bit-wise inversion of a mask.
For example, the LB_MODE_3 macro is defined as:
To combine the lockbit mode macros together to represent a whole byte, use the bitwise
AND operator, like so:
<avr/lock.h> also defines a macro that provides a default lockbit value: LOCKBITS_-
DEFAULT which is defined to be 0xFF.
See the AVR device specific datasheet for more details about these lock bits and the
available mode settings.
A convenience macro, LOCKMEM, is defined as a GCC attribute for a custom-named
section of ".lock".
A convenience macro, LOCKBITS, is defined that declares a variable, __lock, of type
unsigned char with the attribute defined by LOCKMEM. This variable allows the end
user to easily set the lockbit data.
Note:
If a device-specific I/O header file has previously defined LOCKMEM, then
LOCKMEM is not redefined. If a device-specific I/O header file has previously
defined LOCKBITS, then LOCKBITS is not redefined. LOCKBITS is currently
known to be defined in the I/O header files for the XMEGA devices.
#include <avr/io.h>
int main(void)
{
return 0;
}
Or:
#include <avr/io.h>
int main(void)
{
return 0;
}
However there are a number of caveats that you need to be aware of to use this API
properly.
Be sure to include <avr/io.h> to get all of the definitions for the API. The LOCKBITS
macro defines a global variable to store the lockbit data. This variable is assigned to
its own linker section. Assign the desired lockbit values immediately in the variable
initialization.
The .lock section in the ELF file will get its values from the initial variable assignment
ONLY. This means that you can NOT assign values to this variable in functions and the
new values will not be put into the ELF .lock section.
The global variable is declared in the LOCKBITS macro has two leading underscores,
which means that it is reserved for the "implementation", meaning the library, so it will
not conflict with a user-named variable.
You must initialize the lockbit variable to some meaningful value, even if it is the de-
fault value. This is because the lockbits default to a logical 1, meaning unprogrammed.
Normal uninitialized data defaults to all locgial zeros. So it is vital that all lockbits
are initialized, even with default data. If they are not, then the lockbits may not pro-
grammed to the desired settings and can possibly put your device into an unrecoverable
state.
Be sure to have the -mmcu=device flag in your compile command line and your linker
command line to have the correct device selected and to have the correct I/O header
file included when you include <avr/io.h>.
You can print out the contents of the .lock section in the ELF file by using this command
line:
Defines
Typedefs
Functions
#include <avr/io.h>
#include <avr/pgmspace.h>
The functions in this module provide interfaces for a program to access data stored in
program space (flash memory) of the device. In order to use these functions, the target
device must support either the LPM or ELPM instructions.
Note:
These functions are an attempt to provide some compatibility with header files
that come with IAR C, to make porting applications between different compilers
easier. This is not 100% compatibility though (GCC does not have full support for
multiple address spaces yet).
If you are working with strings which are completely based in ram, use the stan-
dard string functions described in <string.h>: Strings.
If possible, put your constant tables in the lower 64 KB and use pgm_read_byte_-
near() or pgm_read_word_near() instead of pgm_read_byte_far() or pgm_read_-
word_far() since it is more efficient that way, and you can still use the upper 64K
for executable code. All functions that are suffixed with a _P require their ar-
guments to be in the lower 64 KB of the flash ROM, as they do not use ELPM
instructions. This is normally not a big concern as the linker setup arranges any
program space constants declared using the macros from this header file so they
are placed right after the interrupt vectors, and in front of any executable code.
However, it can become a problem if there are too many of these constants, or for
bootloaders on devices with more than 64 KB of ROM. All these functions will not
work in that situation.
For Xmega devices, make sure the NVM controller command register (NVM.CMD
or NVM_CMD) is set to 0x00 (NOP) before using any of these functions.
Read a byte from the program space with a 16-bit (near) address.
Note:
The address is a byte address. The address is in the program space.
Read a byte from the program space with a 32-bit (far) address.
Note:
The address is a byte address. The address is in the program space.
Read a byte from the program space with a 16-bit (near) address.
Note:
The address is a byte address. The address is in the program space.
Read a double word from the program space with a 16-bit (near) address.
Note:
The address is a byte address. The address is in the program space.
Read a double word from the program space with a 32-bit (far) address.
Note:
The address is a byte address. The address is in the program space.
Read a double word from the program space with a 16-bit (near) address.
Note:
The address is a byte address. The address is in the program space.
Read a float from the program space with a 16-bit (near) address.
Note:
The address is a byte address. The address is in the program space.
Read a float from the program space with a 32-bit (far) address.
Note:
The address is a byte address. The address is in the program space.
Read a float from the program space with a 16-bit (near) address.
Note:
The address is a byte address. The address is in the program space.
Read a word from the program space with a 16-bit (near) address.
Note:
The address is a byte address. The address is in the program space.
Read a word from the program space with a 32-bit (far) address.
Note:
The address is a byte address. The address is in the program space.
Read a word from the program space with a 16-bit (near) address.
Note:
The address is a byte address. The address is in the program space.
23.19.3.1 prog_char
23.19.3.2 prog_int16_t
23.19.3.3 prog_int32_t
23.19.3.4 prog_int64_t
Note:
This type is not available when the compiler option -mint8 is in effect.
23.19.3.5 prog_int8_t
23.19.3.6 prog_uchar
23.19.3.7 prog_uint16_t
23.19.3.8 prog_uint32_t
23.19.3.9 prog_uint64_t
Note:
This type is not available when the compiler option -mint8 is in effect.
23.19.3.10 prog_uint8_t
23.19.3.11 prog_void
Type of a "void" object located in flash ROM. Does not make much sense by itself,
but can be used to declare a "void ∗" object in flash ROM.
23.19.4.1 void ∗ memccpy_P (void ∗ dest, PGM_VOID_P src, int val, size_t
len)
Scan flash memory for a character. The memchr_P() function scans the first len bytes
of the flash memory area pointed to by s for the character val. The first byte to match
val (interpreted as an unsigned character) stops the operation.
Returns:
The memchr_P() function returns a pointer to the matching byte or NULL if the
character does not occur in the given memory area.
23.19.4.3 int memcmp_P (const void ∗ s1, PGM_VOID_P s2, size_t len)
Compare memory areas. The memcmp_P() function compares the first len bytes of
the memory areas s1 and flash s2. The comparision is performed using unsigned char
operations.
Returns:
The memcmp_P() function returns an integer less than, equal to, or greater than
zero if the first len bytes of s1 is found, respectively, to be less than, to match, or
be greater than the first len bytes of s2.
23.19.4.4 int memcmp_PF (const void ∗ s1, uint_farptr_t s2, size_t len)
Compare memory areas. The memcmp_PF() function compares the first len bytes
of the memory areas s1 and flash s2. The comparision is performed using unsigned
char operations. It is an equivalent of memcmp_P() function, except that it is capable
working on all FLASH including the exteded area above 64kB.
Returns:
The memcmp_PF() function returns an integer less than, equal to, or greater than
zero if the first len bytes of s1 is found, respectively, to be less than, to match, or
be greater than the first len bytes of s2.
The memcpy_P() function is similar to memcpy(), except the src string resides in
program space.
Returns:
The memcpy_P() function returns a pointer to dest.
Copy a memory block from flash to SRAM. The memcpy_PF() function is similar to
memcpy(), except the data is copied from the program space and is addressed using a
far pointer
Parameters:
dst A pointer to the destination buffer
src A far pointer to the origin of data in flash memory
n The number of bytes to be copied
Returns:
The memcpy_PF() function returns a pointer to dst. The contents of RAMPZ SFR
are undefined when the function returns
23.19.4.7 void ∗ memmem_P (const void ∗ s1, size_t len1, PGM_VOID_P s2,
size_t len2)
The memrchr_P() function is like the memchr_P() function, except that it searches
backwards from the end of the len bytes pointed to by src instead of forwards from
the front. (Glibc, GNU extension.)
Returns:
The memrchr_P() function returns a pointer to the matching byte or NULL if the
character does not occur in the given memory area.
Compare two strings ignoring case. The strcasecmp_P() function compares the two
strings s1 and s2, ignoring the case of the characters.
Parameters:
s1 A pointer to a string in the devices SRAM.
s2 A pointer to a string in the devices Flash.
Returns:
The strcasecmp_P() function returns an integer less than, equal to, or greater than
zero if s1 is found, respectively, to be less than, to match, or be greater than s2.
A consequence of the ordering used by strcasecmp_P() is that if s1 is an initial
substring of s2, then s1 is considered to be "less than" s2.
Compare two strings ignoring case. The strcasecmp_PF() function compares the two
strings s1 and s2, ignoring the case of the characters
Parameters:
s1 A pointer to the first string in SRAM
s2 A far pointer to the second string in Flash
Returns:
The strcasecmp_PF() function returns an integer less than, equal to, or greater than
zero if s1 is found, respectively, to be less than, to match, or be greater than s2.
The contents of RAMPZ SFR are undefined when the function returns
The strcat_P() function is similar to strcat() except that the src string must be located
in program space (flash).
Returns:
The strcat() function returns a pointer to the resulting string dest.
Concatenates two strings. The strcat_PF() function is similar to strcat() except that the
src string must be located in program space (flash) and is addressed using a far pointer
Parameters:
dst A pointer to the destination string in SRAM
src A far pointer to the string to be appended in Flash
Returns:
The strcat_PF() function returns a pointer to the resulting string dst. The contents
of RAMPZ SFR are undefined when the function returns
Locate character in program space string. The strchr_P() function locates the first
occurrence of val (converted to a char) in the string pointed to by s in program space.
The terminating null character is considered to be part of the string.
The strchr_P() function is similar to strchr() except that s is pointer to a string in
program space.
Returns:
The strchr_P() function returns a pointer to the matched character or NULL if the
character is not found.
Returns:
The strchrnul_P() function returns a pointer to the matched character, or a pointer
to the null byte at the end of s (i.e., s+strlen(s)) if the character is not found.
Returns:
The strcmp_P() function returns an integer less than, equal to, or greater than zero
if s1 is found, respectively, to be less than, to match, or be greater than s2. A
consequence of the ordering used by strcmp_P() is that if s1 is an initial substring
of s2, then s1 is considered to be "less than" s2.
Compares two strings. The strcmp_PF() function is similar to strcmp() except that s2
is a far pointer to a string in program space
Parameters:
s1 A pointer to the first string in SRAM
s2 A far pointer to the second string in Flash
Returns:
The strcmp_PF() function returns an integer less than, equal to, or greater than
zero if s1 is found, respectively, to be less than, to match, or be greater than s2.
The contents of RAMPZ SFR are undefined when the function returns
The strcpy_P() function is similar to strcpy() except that src is a pointer to a string in
program space.
Returns:
The strcpy_P() function returns a pointer to the destination string dest.
Duplicate a string. The strcpy_PF() function is similar to strcpy() except that src is a
far pointer to a string in program space
Parameters:
dst A pointer to the destination string in SRAM
src A far pointer to the source string in Flash
Returns:
The strcpy_PF() function returns a pointer to the destination string dst. The con-
tents of RAMPZ SFR are undefined when the funcion returns
Returns:
The strcspn_P() function returns the number of characters in the initial segment of
s which are not in the string reject. The terminating zero is not considered as a
part of string.
Concatenate two strings. The strlcat_P() function is similar to strlcat(), except that the
src string must be located in program space (flash).
Appends src to string dst of size siz (unlike strncat(), siz is the full size of dst,
not space left). At most siz-1 characters will be copied. Always NULL terminates
(unless siz <= strlen(dst)).
Returns:
The strlcat_P() function returns strlen(src) + MIN(siz, strlen(initial dst)). If retval
>= siz, truncation occurred.
Concatenate two strings. The strlcat_PF() function is similar to strlcat(), except that the
src string must be located in program space (flash) and is addressed using a far pointer
Appends src to string dst of size n (unlike strncat(), n is the full size of dst, not space
left). At most n-1 characters will be copied. Always NULL terminates (unless n <=
strlen(dst))
Parameters:
dst A pointer to the destination string in SRAM
src A far pointer to the source string in Flash
n The total number of bytes allocated to the destination string
Returns:
The strlcat_PF() function returns strlen(src) + MIN(n, strlen(initial dst)). If retval
>= n, truncation occurred. The contents of RAMPZ SFR are undefined when the
funcion returns
Copy a string from progmem to RAM. Copy src to string dst of size siz. At most
siz-1 characters will be copied. Always NULL terminates (unless siz == 0). The
strlcpy_P() function is similar to strlcpy() except that the src is pointer to a string in
memory space.
Returns:
The strlcpy_P() function returns strlen(src). If retval >= siz, truncation occurred.
Copy a string from progmem to RAM. Copy src to string dst of size siz. At most siz-1
characters will be copied. Always NULL terminates (unless siz == 0).
Returns:
The strlcpy_PF() function returns strlen(src). If retval >= siz, truncation occurred.
The contents of RAMPZ SFR are undefined when the function returns
The strlen_P() function is similar to strlen(), except that src is a pointer to a string in
program space.
Returns:
The strlen() function returns the number of characters in src.
Obtain the length of a string. The strlen_PF() function is similar to strlen(), except that
s is a far pointer to a string in program space
Parameters:
s A far pointer to the string in flash
Returns:
The strlen_PF() function returns the number of characters in s. The contents of
RAMPZ SFR are undefined when the function returns
Parameters:
s1 A pointer to a string in the devices SRAM.
s2 A pointer to a string in the devices Flash.
n The maximum number of bytes to compare.
Returns:
The strncasecmp_P() function returns an integer less than, equal to, or greater
than zero if s1 (or the first n bytes thereof) is found, respectively, to be less
than, to match, or be greater than s2. A consequence of the ordering used by
strncasecmp_P() is that if s1 is an initial substring of s2, then s1 is considered to
be "less than" s2.
Parameters:
s1 A pointer to a string in SRAM
s2 A far pointer to a string in Flash
n The maximum number of bytes to compare
Returns:
The strncasecmp_PF() function returns an integer less than, equal to, or greater
than zero if s1 (or the first n bytes thereof) is found, respectively, to be less than,
to match, or be greater than s2. The contents of RAMPZ SFR are undefined when
the function returns
Concatenate two strings. The strncat_P() function is similar to strncat(), except that the
src string must be located in program space (flash).
Returns:
The strncat_P() function returns a pointer to the resulting string dest.
Concatenate two strings. The strncat_PF() function is similar to strncat(), except that
the src string must be located in program space (flash) and is addressed using a far
pointer
Parameters:
dst A pointer to the destination string in SRAM
src A far pointer to the source string in Flash
n The maximum number of bytes to append
Returns:
The strncat_PF() function returns a pointer to the resulting string dst. The contents
of RAMPZ SFR are undefined when the function returns
The strncmp_P() function is similar to strcmp_P() except it only compares the first (at
most) n characters of s1 and s2.
Returns:
The strncmp_P() function returns an integer less than, equal to, or greater than zero
if s1 (or the first n bytes thereof) is found, respectively, to be less than, to match,
or be greater than s2.
Compare two strings with limited length. The strncmp_PF() function is similar to
strcmp_PF() except it only compares the first (at most) n characters of s1 and s2
Parameters:
s1 A pointer to the first string in SRAM
s2 A far pointer to the second string in Flash
n The maximum number of bytes to compare
Returns:
The strncmp_PF() function returns an integer less than, equal to, or greater than
zero if s1 (or the first n bytes thereof) is found, respectively, to be less than, to
match, or be greater than s2. The contents of RAMPZ SFR are undefined when
the function returns
The strncpy_P() function is similar to strcpy_P() except that not more than n bytes
of src are copied. Thus, if there is no null byte among the first n bytes of src, the result
will not be null-terminated.
In the case where the length of src is less than that of n, the remainder of dest will be
padded with nulls.
Returns:
The strncpy_P() function returns a pointer to the destination string dest.
Parameters:
dst A pointer to the destination string in SRAM
src A far pointer to the source string in Flash
n The maximum number of bytes to copy
Returns:
The strncpy_PF() function returns a pointer to the destination string dst. The con-
tents of RAMPZ SFR are undefined when the function returns
Returns:
The strnlen_P function returns strlen_P(src), if that is less than len, or len if
there is no ’\0’ character among the first len characters pointed to by src.
Parameters:
s A far pointer to the string in Flash
len The maximum number of length to return
Returns:
The strnlen_PF function returns strlen_P(s), if that is less than len, or len if there
is no ’\0’ character among the first len characters pointed to by s. The contents of
RAMPZ SFR are undefined when the function returns
The strpbrk_P() function locates the first occurrence in the string s of any of the
characters in the flash string accept. This function is similar to strpbrk() except that
accept is a pointer to a string in program space.
Returns:
The strpbrk_P() function returns a pointer to the character in s that matches one of
the characters in accept, or NULL if no such character is found. The terminating
zero is not considered as a part of string: if one or both args are empty, the result
will NULL.
Locate character in string. The strrchr_P() function returns a pointer to the last occur-
rence of the character val in the flash string s.
Returns:
The strrchr_P() function returns a pointer to the matched character or NULL if the
character is not found.
Parse a string into tokens. The strsep_P() function locates, in the string referenced
by ∗sp, the first occurrence of any character in the string delim (or the terminating
’\0’ character) and replaces it with a ’\0’. The location of the next character after the
delimiter character (or NULL, if the end of the string was reached) is stored in ∗sp. An
“empty” field, i.e. one caused by two adjacent delimiter characters, can be detected by
comparing the location referenced by the pointer returned in ∗sp to ’\0’. This function
is similar to strsep() except that delim is a pointer to a string in program space.
Returns:
The strsep_P() function returns a pointer to the original value of ∗sp. If ∗sp is
initially NULL, strsep_P() returns NULL.
Returns:
The strspn_P() function returns the number of characters in the initial segment of
s which consist only of characters from accept. The terminating zero is not
considered as a part of string.
Locate a substring. The strstr_P() function finds the first occurrence of the substring
s2 in the string s1. The terminating ’\0’ characters are not compared. The strstr_P()
function is similar to strstr() except that s2 is pointer to a string in program space.
Returns:
The strstr_P() function returns a pointer to the beginning of the substring, or NULL
if the substring is not found. If s2 points to a string of zero length, the function
returns s1.
Locate a substring. The strstr_PF() function finds the first occurrence of the substring
s2 in the string s1. The terminating ’\0’ characters are not compared. The strstr_PF()
function is similar to strstr() except that s2 is a far pointer to a string in program space.
Returns:
The strstr_PF() function returns a pointer to the beginning of the substring, or
NULL if the substring is not found. If s2 points to a string of zero length, the
function returns s1. The contents of RAMPZ SFR are undefined when the func-
tion returns
Parses the string into tokens. strtok_P() parses the string s into tokens. The first call
to strtok_P() should have s as its first argument. Subsequent calls should have the first
argument set to NULL. If a token ends with a delimiter, this delimiting character is
overwritten with a ’\0’ and a pointer to the next character is saved for the next call to
strtok_P(). The delimiter string delim may be different for each call.
The strtok_P() function is similar to strtok() except that delim is pointer to a string in
program space.
Returns:
The strtok_P() function returns a pointer to the next token or NULL when no more
tokens are found.
Note:
strtok_P() is NOT reentrant. For a reentrant version of this function see strtok_rP().
Parses string into tokens. The strtok_rP() function parses string into tokens. The
first call to strtok_rP() should have string as its first argument. Subsequent calls should
have the first argument set to NULL. If a token ends with a delimiter, this delimiting
character is overwritten with a ’\0’ and a pointer to the next character is saved for the
next call to strtok_rP(). The delimiter string delim may be different for each call.
last is a user allocated char∗ pointer. It must be the same while parsing the same
string. strtok_rP() is a reentrant version of strtok_P().
The strtok_rP() function is similar to strtok_r() except that delim is pointer to a string
in program space.
Returns:
The strtok_rP() function returns a pointer to the next token or NULL when no more
tokens are found.
Many AVRs contain a Power Reduction Register (PRR) or Registers (PRRx) that allow
you to reduce power consumption by disabling or enabling various on-board peripher-
als as needed.
There are many macros in this header file that provide an easy interface to enable or
disable on-board peripherals to reduce power. See the table below.
Note:
Not all AVR devices have a Power Reduction Register (for example the AT-
mega128). On those devices without a Power Reduction Register, these macros
are not available.
Not all AVR devices contain the same peripherals (for example, the LCD inter-
face), or they will be named differently (for example, USART and USART0).
Please consult your device’s datasheet, or the header file, to find out which macros
are applicable to your device.
Some of the newer AVRs contain a System Clock Prescale Register (CLKPR) that
allows you to decrease the system clock frequency and the power consumption when
the need for processing power is low. Below are two macros and an enumerated type
that can be used to interface to the Clock Prescale Register.
Note:
Not all AVR devices have a Clock Prescale Register. On those devices without a
Clock Prescale Register, these macros are not available.
typedef enum
{
clock_div_1 = 0,
clock_div_2 = 1,
clock_div_4 = 2,
clock_div_8 = 3,
clock_div_16 = 4,
clock_div_32 = 5,
clock_div_64 = 6,
clock_div_128 = 7,
clock_div_256 = 8,
clock_div_1_rc = 15, // ATmega128RFA1 only
} clock_div_t;
clock_prescale_set(x)
Set the clock prescaler register select bits, selecting a system clock division setting.
This function is inlined, even if compiler optimizations are disabled.
The type of x is clock_div_t.
clock_prescale_get()
Gets and returns the clock prescaler register setting. The return type is clock_div_t.
If _SFR_ASM_COMPAT is not defined, C programs can use names like PORTA directly
in C expressions (also on the left side of assignment operators) and GCC will do the
right thing (use short I/O instructions if possible). The __SFR_OFFSET definition is
not used in any way in this case.
Define _SFR_ASM_COMPAT as 1 to make these names work as simple constants (ad-
dresses of the I/O registers). This is necessary when included in preprocessed assem-
bler (∗.S) source files, so it is done automatically if __ASSEMBLER__ is defined. By
default, all addresses are defined as if they were memory addresses (used in lds/sts
instructions). To use these addresses in in/out instructions, you must subtract 0x20
from them.
For more backwards compatibility, insert the following at the start of your old assem-
bler source file:
#define __SFR_OFFSET 0
This automatically subtracts 0x20 from I/O space addresses, but it’s a hack, so it is
recommended to change your source: wrap such addresses in macros defined here, as
shown below. After this is done, the __SFR_OFFSET definition is no longer necessary
and can be removed.
Real example - this code could be used in a boot loader that is portable between devices
with SPMCR at different addresses.
#if _SFR_IO_REG_P(SPMCR)
out _SFR_IO_ADDR(SPMCR), r24
#else
sts _SFR_MEM_ADDR(SPMCR), r24
#endif
Modules
Bit manipulation
When working with microcontrollers, many tasks usually consist of controlling internal
peripherals, or external peripherals that are connected to the device. The entire IO
address space is made available as memory-mapped IO, i.e. it can be accessed using
all the MCU instructions that are applicable to normal data memory. For most AVR
devices, the IO register space is mapped into the data memory address space with an
offset of 0x20 since the bottom of this space is reserved for direct access to the MCU
registers. (Actual SRAM is available only behind the IO register area, starting at some
specific address depending on the device.)
For example the user can access memory-mapped IO registers as if they were globally
defined variables like this:
PORTA = 0x33;
unsigned char foo = PINA;
The compiler will choose the correct instruction sequence to generate based on the
address of the register being accessed.
The advantage of using the memory-mapped registers in C programs is that it makes
the programs more portable to other C compilers for the AVR platform.
Note that special care must be taken when accessing some of the 16-bit timer IO reg-
isters where access from both the main program and within an interrupt context can
happen. See Why do some 16-bit timer registers sometimes get trashed?.
Access to the AVR single bit set and clear instructions are provided via the standard C
bit manipulation commands. The sbi and cbi macros are no longer directly supported.
sbi (sfr,bit) can be replaced by sfr |= _BV(bit) .
i.e.: sbi(PORTB, PB1); is now PORTB |= _BV(PB1);
This actually is more flexible than having sbi directly, as the optimizer will use a hard-
ware sbi if appropriate, or a read/or/write operation if not appropriate. You do not need
to keep track of which registers sbi/cbi will operate on.
Likewise, cbi (sfr,bit) is now sfr &= ∼(_BV(bit));
#include <avr/io.h>
Note:
The bit shift is performed by the compiler which then inserts the result into the
code. Thus, there is no run-time overhead when using _BV().
#include <avr/io.h>
Test whether bit bit in IO register sfr is clear. This will return non-zero if the bit is
clear, and a 0 if the bit is set.
#include <avr/io.h>
Test whether bit bit in IO register sfr is set. This will return a 0 if the bit is clear,
and non-zero if the bit is set.
#include <avr/io.h>
#include <avr/io.h>
The <avr/signature.h> header file allows the user to automatically and easily include
the device’s signature data in a special section of the final linked ELF file.
This value can then be used by programming software to compare the on-device signa-
ture with the signature recorded in the ELF file to look for a match before programming
the device.
#include <avr/signature.h>
This will declare a constant unsigned char array and it is initialized with the three
signature bytes, MSB first, that are defined in the device I/O header file. This array is
then placed in the .signature section in the resulting linked ELF file.
The three signature bytes that are used to initialize the array are these defined macros
in the device I/O header file, from MSB to LSB: SIGNATURE_2, SIGNATURE_1,
SIGNATURE_0.
This header file should only be included once in an application.
Functions
#include <avr/sleep.h>
Use of the SLEEP instruction can allow an application to reduce its power comsump-
tion considerably. AVR devices can be put into different sleep modes. Refer to the
datasheet for the details relating to the device you are using.
There are several macros provided in this header file to actually put the device into
sleep mode. The simplest way is to optionally set the desired sleep mode using set_-
sleep_mode() (it usually defaults to idle mode where the CPU is put on sleep but
all peripheral clocks are still running), and then call sleep_mode(). This macro
automatically sets the sleep enable bit, goes to sleep, and clears the sleep enable bit.
Example:
#include <avr/sleep.h>
...
set_sleep_mode(<mode>);
sleep_mode();
Note that unless your purpose is to completely lock the CPU (until a hardware reset),
interrupts need to be enabled before going to sleep.
As the sleep_mode() macro might cause race conditions in some situations,
the individual steps of manipulating the sleep enable (SE) bit, and actually issuing
the SLEEP instruction, are provided in the macros sleep_enable(), sleep_-
disable(), and sleep_cpu(). This also allows for test-and-sleep scenarios that
take care of not missing the interrupt that will awake the device from sleep.
Example:
#include <avr/interrupt.h>
#include <avr/sleep.h>
...
set_sleep_mode(<mode>);
cli();
if (some_condition)
{
sleep_enable();
sei();
sleep_cpu();
sleep_disable();
}
sei();
This sequence ensures an atomic test of some_condition with interrupts being dis-
abled. If the condition is met, sleep mode will be prepared, and the SLEEP instruction
will be scheduled immediately after an SEI instruction. As the intruction right after
the SEI is guaranteed to be executed before an interrupt could trigger, it is sure the
device will really be put to sleep.
Some devices have the ability to disable the Brown Out Detector (BOD) before going
to sleep. This will also reduce power while sleeping. If the specific AVR device has this
ability then an additional macro is defined: sleep_bod_disable(). This macro
generates inlined assembly code that will correctly implement the timed sequence for
disabling the BOD before sleeping. However, there is a limited number of cycles af-
ter the BOD has been disabled that the device can be put into sleep mode, otherwise
the BOD will not truly be disabled. Recommended practice is to disable the BOD
(sleep_bod_disable()), set the interrupts (sei()), and then put the device to
sleep (sleep_cpu()), like so:
#include <avr/interrupt.h>
#include <avr/sleep.h>
...
set_sleep_mode(<mode>);
cli();
if (some_condition)
{
sleep_enable();
sleep_bod_disable();
sei();
sleep_cpu();
sleep_disable();
}
sei();
Put the device into sleep mode. The SE bit must be set beforehand, and it is
recommended to clear it afterwards.
Put the device in sleep mode. How the device is brought out of sleep
mode depends on the specific mode selected with the set_sleep_mode() function. See
the data sheet for your device for more details.
Set the SE (sleep enable) bit.
Defines
#include <avr/version.h>
This header file defines macros that contain version numbers and strings describing the
current version of avr-libc.
The version number itself basically consists of three pieces that are separated by a
dot: the major number, the minor number, and the revision number. For development
versions (which use an odd minor number), the string representation additionally gets
the date code (YYYYMMDD) appended.
This file will also be included by <avr/io.h>. That way, portable tests can be
implemented using <avr/io.h> that can be used in code that wants to remain
backwards-compatible to library versions prior to the date when the library version
API had been added, as referenced but undefined C preprocessor macros automatically
evaluate to 0.
Defines
#include <avr/wdt.h>
This header file declares the interface to some inline macros handling the watchdog
timer present in many AVR devices. In order to prevent the watchdog timer configura-
tion from being accidentally altered by a crashing application, a special timed sequence
is required in order to change it. The macros within this header file handle the required
sequence automatically before changing any value. Interrupts will be disabled during
the manipulation.
Note:
Depending on the fuse configuration of the particular device, further restrictions
might apply, in particular it might be disallowed to turn off the watchdog timer.
Note that for newer devices (ATmega88 and newer, effectively any AVR that has the op-
tion to also generate interrupts), the watchdog timer remains active even after a system
reset (except a power-on condition), using the fastest prescaler value (approximately
15 ms). It is therefore required to turn off the watchdog early during program startup,
the datasheet recommends a sequence like the following:
#include <stdint.h>
#include <avr/wdt.h>
void get_mcusr(void) \
__attribute__((naked)) \
__attribute__((section(".init3")));
void get_mcusr(void)
{
mcusr_mirror = MCUSR;
MCUSR = 0;
wdt_disable();
}
Value:
__asm__ __volatile__ ( \
"in __tmp_reg__, __SREG__" "\n\t" \
"cli" "\n\t" \
"out %0, %1" "\n\t" \
"out %0, __zero_reg__" "\n\t" \
"out __SREG__,__tmp_reg__" "\n\t" \
: /* no outputs */ \
: "I" (_SFR_IO_ADDR(_WD_CONTROL_REG)), \
"r" ((uint8_t)(_BV(_WD_CHANGE_BIT) | _BV(WDE))) \
: "r0" \
)
Disable the watchdog timer, if possible. This attempts to turn off the Enable bit in the
watchdog control register. See the datasheet for details.
Value:
__asm__ __volatile__ ( \
"in __tmp_reg__,__SREG__" "\n\t" \
"cli" "\n\t" \
"wdr" "\n\t" \
"out %0,%1" "\n\t" \
"out __SREG__,__tmp_reg__" "\n\t" \
"out %0,%2" \
: /* no outputs */ \
: "I" (_SFR_IO_ADDR(_WD_CONTROL_REG)), \
"r" (_BV(_WD_CHANGE_BIT) | _BV(WDE)), \
"r" ((uint8_t) ((value & 0x08 ? _WD_PS3_MASK : 0x00) | \
_BV(WDE) | (value & 0x07)) ) \
: "r0" \
)
Enable the watchdog timer, configuring it for expiry after timeout (which is a com-
bination of the WDP0 through WDP2 bits to write into the WDTCR register; For those
devices that have a WDTCSR register, it uses the combination of the WDP0 through
WDP3 bits).
See also the symbolic constants WDTO_15MS et al.
See WDT0_15MS
Symbolic constants for the watchdog timeout. Since the watchdog timer
is based on a free-running RC oscillator, the times are approximate only and apply to
a supply voltage of 5 V. At lower supply voltages, the times will increase. For older
devices, the times will be as large as three times when operating at Vcc = 3 V, while
the newer devices (e. g. ATmega128, ATmega8) only experience a negligible change.
Possible timeout values are: 15 ms, 30 ms, 60 ms, 120 ms, 250 ms, 500 ms, 1 s, 2 s.
(Some devices also allow for 4 s and 8 s.) Symbolic constants are formed by the prefix
WDTO_, followed by the time.
Example that would select a watchdog timer expiry of approximately 500 ms:
wdt_enable(WDTO_500MS);
See WDT0_15MS
See WDT0_15MS
See WDT0_15MS
See WDT0_15MS
See WDT0_15MS
WDT0_15MS
Defines
• #define ATOMIC_BLOCK(type)
• #define NONATOMIC_BLOCK(type)
• #define ATOMIC_RESTORESTATE
• #define ATOMIC_FORCEON
• #define NONATOMIC_RESTORESTATE
• #define NONATOMIC_FORCEOFF
#include <util/atomic.h>
Note:
The macros in this header file require the ISO/IEC 9899:1999 ("ISO C99") feature
of for loop variables that are declared inside the for loop itself. For that reason, this
header file can only be used if the standard level of the compiler (option --std=) is
set to either c99 or gnu99.
The macros in this header file deal with code blocks that are guaranteed to be excuted
Atomically or Non-Atmomically. The term "Atomic" in this context refers to the un-
ability of the respective code to be interrupted.
These macros operate via automatic manipulation of the Global Interrupt Status (I) bit
of the SREG register. Exit paths from both block types are all managed automatically
without the need for special considerations, i. e. the interrupt status will be restored to
the same value it has been when entering the respective block.
A typical example that requires atomic access is a 16 (or more) bit variable that is
shared between the main execution path and an ISR. While declaring such a variable
as volatile ensures that the compiler will not optimize accesses to it away, it does not
guarantee atomic access to it. Assuming the following example:
#include <inttypes.h>
#include <avr/interrupt.h>
#include <avr/io.h>
ISR(TIMER1_OVF_vect)
{
ctr--;
}
...
int
main(void)
{
...
ctr = 0x200;
start_timer();
while (ctr != 0)
// wait
;
...
}
There is a chance where the main context will exit its wait loop when the variable ctr
just reached the value 0xFF. This happens because the compiler cannot natively access
a 16-bit variable atomically in an 8-bit CPU. So the variable is for example at 0x100,
the compiler then tests the low byte for 0, which succeeds. It then proceeds to test the
high byte, but that moment the ISR triggers, and the main context is interrupted. The
ISR will decrement the variable from 0x100 to 0xFF, and the main context proceeds.
It now tests the high byte of the variable which is (now) also 0, so it concludes the
variable has reached 0, and terminates the loop.
Using the macros from this header file, the above code can be rewritten like:
#include <inttypes.h>
#include <avr/interrupt.h>
#include <avr/io.h>
#include <util/atomic.h>
ISR(TIMER1_OVF_vect)
{
ctr--;
}
...
int
main(void)
{
...
ctr = 0x200;
start_timer();
sei();
uint16_t ctr_copy;
do
{
ATOMIC_BLOCK(ATOMIC_FORCEON)
{
ctr_copy = ctr;
}
}
while (ctr_copy != 0);
...
}
This will install the appropriate interrupt protection before accessing variable ctr,
so it is guaranteed to be consistently tested. If the global interrupt state were uncer-
tain before entering the ATOMIC_BLOCK, it should be executed with the parameter
ATOMIC_RESTORESTATE rather than ATOMIC_FORCEON.
See Problems with reordering code for things to be taken into account with respect to
compiler optimizations.
blocks of code while maintaining the atomic access of the other sections of the parent
ATOMIC_BLOCK.
Two possible macro parameters are permitted, NONATOMIC_RESTORESTATE and
NONATOMIC_FORCEOFF.
Functions
#include <util/crc16.h>
This header file provides a optimized inline functions for calculating cyclic redundancy
checks (CRC) using common polynomials.
References:
See the Dallas Semiconductor app note 27 for 8051 assembler example and general
CRC optimization suggestions. The table on the last page of the app note is the key to
understanding these implementations.
Jack Crenshaw’s "Implementing CRCs" article in the January 1992 isue of Embedded
Systems Programming. This may be difficult to find, but it explains CRC’s in very clear
and concise terms. Well worth the effort to obtain a copy.
A typical application would look like:
int
checkcrc(void)
{
uint8_t crc = 0, i;
uint16_t
crc16_update(uint16_t crc, uint8_t a)
{
int i;
crc ^= a;
return crc;
}
Note:
Although the CCITT polynomial is the same as that used by the Xmodem protocol,
they are quite different. The difference is in how the bits are shifted through the
alorgithm. Xmodem shifts the MSB of the CRC and the input first, while CCITT
shifts the LSB of the CRC and the input first.
uint16_t
crc_ccitt_update (uint16_t crc, uint8_t data)
{
data ^= lo8 (crc);
data ^= data << 4;
uint8_t
_crc_ibutton_update(uint8_t crc, uint8_t data)
{
uint8_t i;
return crc;
}
uint16_t
crc_xmodem_update (uint16_t crc, uint8_t data)
{
int i;
return crc;
}
Functions
Note:
As an alternative method, it is possible to pass the F_CPU macro down to the com-
piler from the Makefile. Obviously, in that case, no #define statement should
be used.
The functions in this header file are wrappers around the basic busy-wait functions from
<util/delay_basic.h>. They are meant as convenience functions where actual time
values can be specified rather than a number of cycles to wait for. The idea behind is
that compile-time constant expressions will be eliminated by compiler optimization so
floating-point expressions can be used to calculate the number of delay cycles needed
based on the CPU frequency passed by the macro F_CPU.
Note:
In order for these functions to work as intended, compiler optimizations must be
enabled, and the delay time must be an expression that is a known constant at
compile-time. If these requirements are not met, the resulting delay will be much
longer (and basically unpredictable), and applications that otherwise do not use
floating-point calculations will experience severe code bloat by the floating-point
library routines linked into the application.
The functions available allow the specification of microsecond, and millisecond delays
directly, using the application-supplied macro F_CPU as the CPU clock frequency (in
Hertz).
The macro F_CPU is supposed to be defined to a constant defining the CPU clock
frequency (in Hertz).
The maximal possible delay is 262.14 ms / F_CPU in MHz.
When the user request delay which exceed the maximum possible one, _delay_ms()
provides a decreased resolution functionality. In this mode _delay_ms() will work with
a resolution of 1/10 ms, providing delays up to 6.5535 seconds (independent from CPU
frequency). The user will not be informed about decreased resolution.
Functions
#include <util/delay_basic.h>
The functions in this header file implement simple delay loops that perform a busy-
waiting. They are typically used to facilitate short delays in the program execution.
They are implemented as count-down loops with a well-known CPU cycle count per
loop iteration. As such, no other processing can occur simultaneously. It should be
kept in mind that the functions described here do not disable interrupts.
In general, for long delays, the use of hardware timers is much preferrable, as they
free the CPU, and allow for concurrent processing of other events while the timer is
running. However, in particular for very short delays, the overhead of setting up a
hardware timer is too much compared to the overall delay time.
Two inline functions are provided for the actual delay algorithms.
Delay loop using an 8-bit counter __count, so up to 256 iterations are possible.
(The value 256 would have to be passed as 0.) The loop executes three CPU cycles per
iteration, not including the overhead the compiler needs to setup the counter register.
Thus, at a CPU speed of 1 MHz, delays of up to 768 microseconds can be achieved.
Defines
• #define parity_even_bit(val)
#include <util/parity.h>
This header file contains optimized assembler code to calculate the parity bit for a byte.
Value:
(__extension__({ \
unsigned char __t; \
__asm__ ( \
"mov __tmp_reg__,%0" "\n\t" \
"swap %0" "\n\t" \
"eor %0,__tmp_reg__" "\n\t" \
"mov __tmp_reg__,%0" "\n\t" \
Returns:
1 if val has an odd number of bits set.
Defines
• #define BAUD_TOL 2
• #define UBRR_VALUE
• #define UBRRL_VALUE
• #define UBRRH_VALUE
• #define USE_2X 0
This header file requires that on entry values are already defined for F_CPU and BAUD.
In addition, the macro BAUD_TOL will define the baud rate tolerance (in percent) that
is acceptable during the calculations. The value of BAUD_TOL will default to 2 %.
This header file defines macros suitable to setup the UART baud rate prescaler registers
of an AVR. All calculations are done using the C preprocessor. Including this header
file causes no other side effects so it is possible to include this file more than once
(supposedly, with different values for the BAUD parameter), possibly even within the
same function.
Assuming that the requested BAUD is valid for the given F_CPU then the macro
UBRR_VALUE is set to the required prescaler value. Two additional macros are pro-
vided for the low and high bytes of the prescaler, respectively: UBRRL_VALUE is set
to the lower byte of the UBRR_VALUE and UBRRH_VALUE is set to the upper byte.
An additional macro USE_2X will be defined. Its value is set to 1 if the desired BAUD
rate within the given tolerance could only be achieved by setting the U2X bit in the
UART configuration. It will be defined to 0 if U2X is not needed.
Example usage:
#include <avr/io.h>
static void
uart_9600(void)
{
#define BAUD 9600
#include <util/setbaud.h>
UBRRH = UBRRH_VALUE;
UBRRL = UBRRL_VALUE;
#if USE_2X
UCSRA |= (1 << U2X);
#else
UCSRA &= ~(1 << U2X);
#endif
}
static void
uart_38400(void)
{
#undef BAUD // avoid compiler warning
#define BAUD 38400
#include <util/setbaud.h>
UBRRH = UBRRH_VALUE;
UBRRL = UBRRL_VALUE;
#if USE_2X
UCSRA |= (1 << U2X);
#else
UCSRA &= ~(1 << U2X);
#endif
}
In this example, two functions are defined to setup the UART to run at 9600 Bd, and
38400 Bd, respectively. Using a CPU clock of 4 MHz, 9600 Bd can be achieved with
an acceptable tolerance without setting U2X (prescaler 25), while 38400 Bd require
U2X to be set (prescaler 12).
TWSR values
Mnemonics:
TW_MT_xxx - master transmitter
TW_MR_xxx - master receiver
TW_ST_xxx - slave transmitter
TW_SR_xxx - slave receiver
• #define TW_READ 1
• #define TW_WRITE 0
#include <util/twi.h>
This header file contains bit mask definitions for use with the AVR TWI interface.
SLA+R address
Value:
(_BV(TWS7)|_BV(TWS6)|_BV(TWS5)|_BV(TWS4)|\
_BV(TWS3))
The lower 3 bits of TWSR are reserved on the ATmega163. The 2 LSB carry the
prescaler bits on the newer ATmegas.
SLA+W address
// Do some work...
Note:
Be careful when you use these functions. If you already have a different interrupt
enabled, you could inadvertantly disable it by enabling another intterupt.
Obsolete IO macros
Back in a time when AVR-GCC and avr-libc could not handle IO port access in the di-
rect assignment form as they are handled now, all IO port access had to be done through
specific macros that eventually resulted in inline assembly instructions performing the
desired action.
These macros became obsolete, as reading and writing IO ports can be done by simply
using the IO port name in an expression, and all bit manipulation (including those on
IO ports) can be done using generic C bit manipulation operators.
The macros in this group simulate the historical behaviour. While they are supposed to
be applied to IO ports, the emulation actually uses standard C methods, so they could
be applied to arbitrary memory locations as well.
This header file contains several items that used to be available in previous versions of
this library, but have eventually been deprecated over time.
#include <compat/deprecated.h>
These items are supplied within that header file for backward compatibility reasons
only, so old source code that has been written for previous library versions could easily
be maintained until its end-of-life. Use of any of these items in new code is strongly
discouraged.
Deprecated
Deprecated
This macro gives access to the GIMSK register (or EIMSK register if using an AVR
Mega device or GICR register for others). Although this macro is essentially the same
as assigning to the register, it does adapt slightly to the type of device being used. This
macro is unavailable if none of the registers listed above are defined.
Deprecated
Deprecated
Value:
Deprecated
Introduces an interrupt handler function that runs with global interrupts initially en-
abled. This allows interrupt handlers to be interrupted.
As this macro has been used by too many unsuspecting people in the past, it has been
deprecated, and will be removed in a future version of the library. Users who want to
legitimately re-enable interrupts in their interrupt handlers as quickly as possible are
encouraged to explicitly declare their handlers as described above.
Deprecated
Deprecated
Deprecated
Deprecated
This function modifies the timsk register. The value you pass via ints is device
specific.
This is an attempt to provide some compatibility with header files that come with IAR
C, to make porting applications between different compilers easier. No 100% compat-
ibility though.
Note:
For actual documentation, please see the IAR manual.
Modules
Various small demo projects are provided to illustrate several aspects of using the open-
source utilities for the AVR controller series. It should be kept in mind that these de-
mos serve mainly educational purposes, and are normally not directly suitable for use
in any production environment. Usually, they have been kept as simple as sufficient to
demonstrate one particular feature.
The simple project is somewhat like the "Hello world!" application for a microcon-
troller, about the most simple project that can be done. It is explained in good detail,
to allow the reader to understand the basic concepts behind using the tools on an AVR
microcontroller.
The more sophisticated demo project builds on top of that simple project, and adds
some controls to it. It touches a number of avr-libc’s basic concepts on its way.
A comprehensive example on using the standard IO facilities intends to explain that
complex topic, using a practical microcontroller peripheral setup with one RS-232 con-
nection, and an HD44780-compatible industry-standard LCD display.
The Example using the two-wire interface (TWI) project explains the use of the two-
wire hardware interface (also known as "I2C") that is present on many AVR controllers.
Finally, the Combining C and assembly source files demo shows how C and assem-
bly language source files can collaborate within one project. While the overall project
is managed by a C program part for easy maintenance, time-critical parts are written
directly in manually optimized assembly language for shortest execution times possi-
ble. Naturally, this kind of project is very closely tied to the hardware design, thus it is
custom-tailored to a particular controller type and peripheral setup. As an alternative to
the assembly-language solution, this project also offers a C-only implementation (de-
ploying the exact same peripheral setup) based on a more sophisticated (and thus more
expensive) but pin-compatible controller.
While the simple demo is meant to run on about any AVR setup possible where a
LED could be connected to the OCR1[A] output, the large and stdio demos are mainly
targeted to the Atmel STK500 starter kit, and the TWI example requires a controller
where some 24Cxx two-wire EEPPROM can be connected to. For the STK500 demos,
the default CPU (either an AT90S8515 or an ATmega8515) should be removed from
its socket, and the ATmega16 that ships with the kit should be inserted into socket
SCKT3100A3. The ATmega16 offers an on-board ADC that is used in the large demo,
and all AVRs with an ADC feature a different pinout than the industry-standard com-
patible devices.
In order to fully utilize the large demo, a female 10-pin header with cable, connecting
to a 10 kOhm potentiometer will be useful.
For the stdio demo, an industry-standard HD44780-compatible LCD display of at least
16x1 characters will be needed. Among other things, the LCD4Linux project page
describes many things around these displays, including common pinouts.
The incoming PWM pulse train is fed into PB4. It will generate a pin change interrupt
there on eache edge of the incoming signal.
The outgoing PWM is generated through OC0B of timer channel 0 (PB1). For demon-
stration purposes, a LED should be connected to that pin (like, one of the LEDs of an
STK500).
The controllers run on their internal calibrated RC oscillators, 1.2 MHz on the AT-
tiny13, and 1.0 MHz on the ATtiny45.
23.37.2.1 asmdemo.c
After the usual include files, two variables are defined. The first one, pwm_incoming
is used to communicate the most recent pulse width detected by the incoming PWM
decoder up to the main loop.
The second variable actually only constitutes of a single bit, intbits.pwm_-
received. This bit will be set whenever the incoming PWM decoder has updated
pwm_incoming.
Both variables are marked volatile to ensure their readers will always pick up an up-
dated value, as both variables will be set by interrupt service routines.
The function ioinit() initializes the microcontroller peripheral devices. In partic-
ular, it starts timer 0 to generate the outgoing PWM signal on OC0B. Setting OCR0A
to 255 (which is the TOP value of timer 0) is used to generate a timer 0 overflow A
interrupt on the ATtiny13. This interrupt is used to inform the incoming PWM decoder
that the counting direction of channel 0 is just changing from up to down. Likewise, an
overflow interrupt will be generated whenever the countdown reached BOTTOM (value
0), where the counter will again alter its counting direction to upwards. This informa-
tion is needed in order to know whether the current counter value of TCNT0 is to be
evaluated from bottom or top.
Further, ioinit() activates the pin-change interrupt PCINT0 on any edge of PB4.
Finally, PB1 (OC0B) will be activated as an output pin, and global interrupts are being
enabled.
In the ATtiny45 setup, the C code contains an ISR for PCINT0. At each pin-change
interrupt, it will first be analyzed whether the interrupt was caused by a rising or a
falling edge. In case of the rising edge, timer 1 will be started with a prescaler of 16
after clearing the current timer value. Then, at the falling edge, the current timer value
will be recorded (and timer 1 stopped), the pin-change interrupt will be suspended, and
the upper layer will be notified that the incoming PWM measurement data is available.
Function main() first initializes the hardware by calling ioinit(), and then waits
until some incoming PWM value is available. If it is, the output PWM will be adjusted
by computing the relative value of the incoming PWM. Finally, the pin-change interrupt
is re-enabled, and the CPU is put to sleep.
23.37.2.2 project.h
In order for the interrupt service routines to be as fast as possible, some of the CPU
registers are set aside completely for use by these routines, so the compiler would not
use them for C code. This is arranged for in project.h.
The file is divided into one section that will be used by the assembly source code, and
another one to be used by C code. The assembly part is distinguished by the prepro-
cessing macro __ASSEMBLER__ (which will be automatically set by the compiler
front-end when preprocessing an assembly-language file), and it contains just macros
that give symbolic names to a number of CPU registers. The preprocessor will then
replace the symbolic names by their right-hand side definitions before calling the as-
sembler.
In C code, the compiler needs to see variable declarations for these objects. This is
done by using declarations that bind a variable permanently to a CPU register (see
How to permanently bind a variable to a register?). Even in case the C code never
has a need to access these variables, declaring the register binding that way causes the
compiler to not use these registers in C code at all.
The flags variable needs to be in the range of r16 through r31 as it is the target of a
load immediate (or SER) instruction that is not applicable to the entire register file.
23.37.2.3 isrs.S
This file is a preprocessed assembly source file. The C preprocessor will be run by
the compiler front-end first, resolving all #include, #define etc. directives. The
resulting program text will then be passed on to the assembler.
As the C preprocessor strips all C-style comments, preprocessed assembly source files
can have both, C-style (/∗ ... ∗/, // ...) as well as assembly-style (; ...)
comments.
At the top, the IO register definition file avr/io.h and the project declaration file
project.h are included. The remainder of the file is conditionally assembled only if
the target MCU type is an ATtiny13, so it will be completely ignored for the ATtiny45
option.
Next are the two interrupt service routines for timer 0 compare A match (timer 0 hits
TOP, as OCR0A is set to 255) and timer 0 overflow (timer 0 hits BOTTOM). As dis-
cussed above, these are kept as short as possible. They only save SREG (as the flags
will be modified by the INC instruction), increment the counter_hi variable which
forms the high part of the current time counter (the low part is formed by querying
TCNT0 directly), and clear or set the variable flags, respectively, in order to note
the current counting direction. The RETI instruction terminates these interrupt service
routines. Total cycle count is 8 CPU cycles, so together with the 4 CPU cycles needed
for interrupt setup, and the 2 cycles for the RJMP from the interrupt vector to the han-
dler, these routines will require 14 out of each 256 CPU cycles, or about 5 % of the
overall CPU time.
The pin-change interrupt PCINT0 will be handled in the final part of this file. The
basic algorithm is to quickly evaluate the current system time by fetching the current
timer value of TCNT0, and combining it with the overflow part in counter_hi. If
the counter is currently counting down rather than up, the value fetched from TCNT0
must be negated. Finally, if this pin-change interrupt was triggered by a rising edge,
the time computed will be recorded as the start time only. Then, at the falling edge,
this start time will be subracted from the current time to compute the actual pulse width
seen (left in pwm_incoming), and the upper layers are informed of the new value by
setting bit 0 in the intbits flags. At the same time, this pin-change interrupt will be
disabled so no new measurement can be performed until the upper layer had a chance
to process the current value.
At this point, you should have the GNU tools configured, built, and installed on your
system. In this chapter, we present a simple example of using the GNU tools in an AVR
project. After reading this chapter, you should have a better feel as to how the tools are
used and how a Makefile can be configured.
This project will use the pulse-width modulator (PWM) to ramp an LED on and off every
two seconds. An AT90S2313 processor will be used as the controller. The circuit for
this demonstration is shown in the schematic diagram. If you have a development kit,
you should be able to use it, rather than build the circuit, for this project.
Note:
Meanwhile, the AT90S2313 became obsolete. Either use its successor, the (pin-
compatible) ATtiny2313 for the project, or perhaps the ATmega8 or one of its
successors (ATmega48/88/168) which have become quite popular since the origi-
nal demo project had been established. For all these more modern devices, it is no
longer necessary to use an external crystal for clocking as they ship with the inter-
nal 1 MHz oscillator enabled, so C1, C2, and Q1 can be omitted. Normally, for
this experiment, the external circuitry on /RESET (R1, C3) can be omitted as well,
leaving only the AVR, the LED, the bypass capacitor C4, and perhaps R2. For the
ATmega8/48/88/168, use PB1 (pin 15 at the DIP-28 package) to connect the LED
to. Additionally, this demo has been ported to many different other AVRs. The lo-
cation of the respective OC pin varies between different AVRs, and it is mandated
by the AVR hardware.
VCC
IC1
R1 (SCK)PB7 19
1 RESET 18
.01uf
(MISO)PB6
4mhz
20K C2 17
C3
(MOSI)PB5
Q1
4 16 LED5MM
C1 XTAL2 PB4 R2* D1
18pf (OCI)PB3 15
5 XTAL1 14
PB2 See note [8]
18pf 20 VCC (AIN1)PB1 13
12 GND
10 GND (AIN0)PB0
.1uf
11
C4
(ICP)PD6
(T1)PD5 9
GND 8
(T0)PD4
(INT1)PD3 7
GND (INT0)PD2 6
(TXD)PD1 3
(RXD)PD0 2
AT90S2313P
The source code is given in demo.c. For the sake of this example, create a file called
demo.c containing this source code. Some of the more important parts of the code
are:
Note [1]:
As the AVR microcontroller series has been developed during the past years, new
features have been added over time. Even though the basic concepts of the timer/-
counter1 are still the same as they used to be back in early 2001 when this sim-
ple demo was written initially, the names of registers and bits have been changed
slightly to reflect the new features. Also, the port and pin mapping of the output
compare match 1A (or 1 for older devices) pin which is used to control the LED
varies between different AVRs. The file iocompat.h tries to abstract between
all this differences using some preprocessor #ifdef statements, so the actual pro-
gram itself can operate on a common set of symbolic names. The macros defined
by that file are:
• OCR the name of the OCR register used to control the PWM (usually either
OCR1 or OCR1A)
• DDROC the name of the DDR (data direction register) for the OC output
• OC1 the pin number of the OC1[A] output within its port
• TIMER1_TOP the TOP value of the timer used for the PWM (1023 for 10-bit
PWMs, 255 for devices that can only handle an 8-bit PWM)
Note [2]:
ISR() is a macro that marks the function as an interrupt routine. In this case, the
function will get called when timer 1 overflows. Setting up interrupts is explained
in greater detail in <avr/interrupt.h>: Interrupts.
Note [3]:
The PWM is being used in 10-bit mode, so we need a 16-bit variable to remember
the current value.
Note [4]:
This section determines the new value of the PWM.
Note [5]:
Here’s where the newly computed value is loaded into the PWM register. Since
we are in an interrupt routine, it is safe to use a 16-bit assignment to the register.
Outside of an interrupt, the assignment should only be performed with interrupts
disabled if there’s a chance that an interrupt routine could also access this register
(or another register that uses TEMP), see the appropriate FAQ entry.
Note [6]:
This routine gets called after a reset. It initializes the PWM and enables interrupts.
Note [7]:
The main loop of the program does nothing -- all the work is done by the interrupt
routine! The sleep_mode() puts the processor on sleep until the next interrupt,
to conserve power. Of course, that probably won’t be noticable as we are still
driving a LED, it is merely mentioned here to demonstrate the basic principle.
Note [8]:
Early AVR devices saturate their outputs at rather low currents when sourcing cur-
rent, so the LED can be connected directly, the resulting current through the LED
will be about 15 mA. For modern parts (at least for the ATmega 128), however
Atmel has drastically increased the IO source capability, so when operating at 5
V Vcc, R2 is needed. Its value should be about 150 Ohms. When operating the
circuit at 3 V, it can still be omitted though.
/*
* ----------------------------------------------------------------------------
* "THE BEER-WARE LICENSE" (Revision 42):
* <[email protected]> wrote this file. As long as you retain this notice you
* can do whatever you want with this stuff. If we meet some day, and you think
* this stuff is worth it, you can buy me a beer in return. Joerg Wunsch
* ----------------------------------------------------------------------------
*
* Simple AVR demonstration. Controls a LED that can be directly
* connected from OC1/OC1A to GND. The brightness of the LED is
* controlled with the PWM. After each period of the PWM, the PWM
* value is either incremented or decremented, that’s all.
*
* $Id: demo.c 1637 2008-03-17 21:49:41Z joerg_wunsch $
*/
#include <inttypes.h>
#include <avr/io.h>
#include <avr/interrupt.h>
#include <avr/sleep.h>
break;
case DOWN:
if (--pwm == 0)
direction = UP;
break;
}
void
ioinit (void) /* Note [6] */
{
/* Timer 1 is 10-bit PWM (8-bit PWM on some ATtinys). */
TCCR1A = TIMER1_PWM_INIT;
/*
* Start timer 1.
*
* NB: TCCR1A and TCCR1B could actually be the same register, so
* take care to not clobber it.
*/
TCCR1B |= TIMER1_CLOCKSOURCE;
/*
* Run any device-dependent timer 1 setup hook if present.
*/
#if defined(TIMER1_SETUP_HOOK)
TIMER1_SETUP_HOOK();
#endif
int
main (void)
{
ioinit ();
return (0);
}
This first thing that needs to be done is compile the source. When compiling, the
compiler needs to know the processor type so the -mmcu option is specified. The
-Os option will tell the compiler to optimize the code for efficient space usage (at the
possible expense of code execution speed). The -g is used to embed debug info. The
debug info is useful for disassemblies and doesn’t end up in the .hex files, so I usually
specify it. Finally, the -c tells the compiler to compile and stop -- don’t link. This
demo is small enough that we could compile and link in one step. However, real-world
projects will have several modules and will typically need to break up the building of
the project into several compiles and one link.
The compilation will create a demo.o file. Next we link it into a binary called
demo.elf.
It is important to specify the MCU type when linking. The compiler uses the -mmcu
option to choose start-up files and run-time libraries that get linked together. If this
option isn’t specified, the compiler defaults to the 8515 processor environment, which
is most certainly what you didn’t want.
Now we have a binary file. Can we do anything useful with it (besides put it into the
processor?) The GNU Binutils suite is made up of many useful tools for manipulating
object files that get generated. One tool is avr-objdump, which takes information
from the object file and displays it in many useful ways. Typing the command by itself
will cause it to list out its options.
For instance, to get a feel of the application’s size, the -h option can be used. The
output of this option shows how much space is used in each of the sections (the .stab
and .stabstr sections hold the debugging information and won’t make it into the ROM
file).
An even more useful option is -S. This option disassembles the binary file and inter-
sperses the source code in the output! This method is much better, in my opinion, than
using the -S with the compiler because this listing includes routines from the libraries
and the vector table contents. Also, all the "fix-ups" have been satisfied. In other words,
the listing generated by this option reflects the actual code that the processor will run.
Sections:
Idx Name Size VMA LMA File off Algn
0 .text 000000fa 00000000 00000000 00000074 2**1
CONTENTS, ALLOC, LOAD, READONLY, CODE
1 .bss 00000003 00800060 00800060 0000016e 2**0
ALLOC
2 .stab 00000b88 00000000 00000000 00000170 2**2
CONTENTS, READONLY, DEBUGGING
3 .stabstr 0000078d 00000000 00000000 00000cf8 2**0
CONTENTS, READONLY, DEBUGGING
00000000 <__vectors>:
0: 12 c0 rjmp .+36 ; 0x26 <__ctors_end>
2: 76 c0 rjmp .+236 ; 0xf0 <__bad_interrupt>
4: 75 c0 rjmp .+234 ; 0xf0 <__bad_interrupt>
6: 74 c0 rjmp .+232 ; 0xf0 <__bad_interrupt>
8: 73 c0 rjmp .+230 ; 0xf0 <__bad_interrupt>
a: 72 c0 rjmp .+228 ; 0xf0 <__bad_interrupt>
c: 71 c0 rjmp .+226 ; 0xf0 <__bad_interrupt>
e: 70 c0 rjmp .+224 ; 0xf0 <__bad_interrupt>
10: 1a c0 rjmp .+52 ; 0x46 <__vector_8>
12: 6e c0 rjmp .+220 ; 0xf0 <__bad_interrupt>
14: 6d c0 rjmp .+218 ; 0xf0 <__bad_interrupt>
16: 6c c0 rjmp .+216 ; 0xf0 <__bad_interrupt>
18: 6b c0 rjmp .+214 ; 0xf0 <__bad_interrupt>
1a: 6a c0 rjmp .+212 ; 0xf0 <__bad_interrupt>
1c: 69 c0 rjmp .+210 ; 0xf0 <__bad_interrupt>
1e: 68 c0 rjmp .+208 ; 0xf0 <__bad_interrupt>
20: 67 c0 rjmp .+206 ; 0xf0 <__bad_interrupt>
22: 66 c0 rjmp .+204 ; 0xf0 <__bad_interrupt>
24: 65 c0 rjmp .+202 ; 0xf0 <__bad_interrupt>
00000026 <__ctors_end>:
26: 11 24 eor r1, r1
28: 1f be out 0x3f, r1 ; 63
2a: cf e5 ldi r28, 0x5F ; 95
2c: d4 e0 ldi r29, 0x04 ; 4
2e: de bf out 0x3e, r29 ; 62
30: cd bf out 0x3d, r28 ; 61
00000032 <__do_clear_bss>:
32: 10 e0 ldi r17, 0x00 ; 0
34: a0 e6 ldi r26, 0x60 ; 96
36: b0 e0 ldi r27, 0x00 ; 0
38: 01 c0 rjmp .+2 ; 0x3c <.do_clear_bss_start>
0000003a <.do_clear_bss_loop>:
3a: 1d 92 st X+, r1
0000003c <.do_clear_bss_start>:
3c: a3 36 cpi r26, 0x63 ; 99
3e: b1 07 cpc r27, r17
40: e1 f7 brne .-8 ; 0x3a <.do_clear_bss_loop>
00000046 <__vector_8>:
#include "iocompat.h" /* Note [1] */
case DOWN:
if (--pwm == 0)
9c: 20 91 61 00 lds r18, 0x0061
a0: 30 91 62 00 lds r19, 0x0062
a4: 21 50 subi r18, 0x01 ; 1
a6: 30 40 sbci r19, 0x00 ; 0
a8: 30 93 62 00 sts 0x0062, r19
ac: 20 93 61 00 sts 0x0061, r18
b0: 21 15 cp r18, r1
b2: 31 05 cpc r19, r1
b4: 11 f7 brne .-60 ; 0x7a <__vector_8+0x34>
direction = UP;
b6: 10 92 60 00 sts 0x0060, r1
ba: df cf rjmp .-66 ; 0x7a <__vector_8+0x34>
000000c4 <ioinit>:
void
ioinit (void) /* Note [6] */
{
/* Timer 1 is 10-bit PWM (8-bit PWM on some ATtinys). */
TCCR1A = TIMER1_PWM_INIT;
c4: 83 e8 ldi r24, 0x83 ; 131
c6: 8f bd out 0x2f, r24 ; 47
* Start timer 1.
*
* NB: TCCR1A and TCCR1B could actually be the same register, so
* take care to not clobber it.
*/
TCCR1B |= TIMER1_CLOCKSOURCE;
c8: 8e b5 in r24, 0x2e ; 46
ca: 81 60 ori r24, 0x01 ; 1
cc: 8e bd out 0x2e, r24 ; 46
#if defined(TIMER1_SETUP_HOOK)
TIMER1_SETUP_HOOK();
#endif
000000de <main>:
int
main (void)
{
ioinit ();
de: f2 df rcall .-28 ; 0xc4 <ioinit>
000000f0 <__bad_interrupt>:
f0: 87 cf rjmp .-242 ; 0x0 <__vectors>
000000f2 <exit>:
ASSEMBLY_CLIB_SECTION
.global _U(exit)
.type _U(exit), "function"
_U(exit):
cli
f2: f8 94 cli
XJMP _U(_exit)
f4: 00 c0 rjmp .+0 ; 0xf6 <_exit>
000000f6 <_exit>:
f6: f8 94 cli
000000f8 <__stop_program>:
f8: ff cf rjmp .-2 ; 0xf8 <__stop_program>
avr-objdump is very useful, but sometimes it’s necessary to see information about
the link that can only be generated by the linker. A map file contains this information.
A map file is useful for monitoring the sizes of your code and data. It also shows where
modules are loaded and which modules were loaded from libraries. It is yet another
view of your application. To get a map file, I usually add -Wl,-Map,demo.map to
my link command. Relink the application using the following command to generate
demo.map (a portion of which is shown below).
.rela.plt
*(.rela.plt)
The .text segment (where program instructions are stored) starts at location 0x0.
*(.fini2)
*(.fini2)
*(.fini1)
*(.fini1)
*(.fini0)
.fini0 0x000000f6 0x4 /usr/local/lib/gcc/avr/4.3.4/avr4/libgcc.a(
_exit.o)
*(.fini0)
0x000000fa _etext = .
The last address in the .text segment is location 0x114 ( denoted by _etext ), so the
instructions use up 276 bytes of FLASH.
The .data segment (where initialized static variables are stored) starts at location 0x60,
which is the first address after the register bank on an ATmega8 processor.
The next available address in the .data segment is also location 0x60, so the application
has no initialized data.
The .bss segment (where uninitialized data is stored) starts at location 0x60.
The next available address in the .bss segment is location 0x63, so the application uses
3 bytes of uninitialized data.
The .eeprom segment (where EEPROM variables are stored) starts at location 0x0.
The next available address in the .eeprom segment is also location 0x0, so there aren’t
any EEPROM variables.
We have a binary of the application, but how do we get it into the processor? Most (if
not all) programmers will not accept a GNU executable as an input file, so we need to
do a little more processing. The next step is to extract portions of the binary and save
the information into .hex files. The GNU utility that does this is called avr-objcopy.
The ROM contents can be pulled from our project’s binary and put into the file
demo.hex using the following command:
:1000000012C076C075C074C073C072C071C070C0B9
:100010001AC06EC06DC06CC06BC06AC069C068C0D9
:1000200067C066C065C011241FBECFE5D4E0DEBF47
:10003000CDBF10E0A0E6B0E001C01D92A336B1072D
:10004000E1F74DD056C01F920F920FB60F921124B8
:100050002F933F938F93809160008823C1F4209168
:100060006100309162002F5F3F4F30936200209318
:10007000610083E02F3F380709F13BBD2ABD8F9116
:100080003F912F910F900FBE0F901F9018958130C8
:1000900029F02091610030916200EFCF2091610042
:1000A0003091620021503040309362002093610013
:1000B0002115310511F710926000DFCF81E08093A8
:1000C0006000DBCF83E88FBD8EB581608EBD1BBC29
:1000D0001ABC82E087BB84E089BF78940895F2DF80
:1000E00085B7806885BF889585B78F7785BFF8CF3E
:0A00F00087CFF89400C0F894FFCF0A
:00000001FF
The -j option indicates that we want the information from the .text and .data segment
extracted. If we specify the EEPROM segment, we can generate a .hex file that can be
used to program the EEPROM:
Rather than type these commands over and over, they can all be placed in a make file.
To build the demo project using make, save the following in a file called Makefile.
Note:
This Makefile can only be used as input for the GNU version of make.
PRG = demo
OBJ = demo.o
#MCU_TARGET = at90s2313
#MCU_TARGET = at90s2333
#MCU_TARGET = at90s4414
#MCU_TARGET = at90s4433
#MCU_TARGET = at90s4434
#MCU_TARGET = at90s8515
#MCU_TARGET = at90s8535
#MCU_TARGET = atmega128
#MCU_TARGET = atmega1280
#MCU_TARGET = atmega1281
#MCU_TARGET = atmega1284p
#MCU_TARGET = atmega16
#MCU_TARGET = atmega163
#MCU_TARGET = atmega164p
#MCU_TARGET = atmega165
#MCU_TARGET = atmega165p
#MCU_TARGET = atmega168
#MCU_TARGET = atmega169
#MCU_TARGET = atmega169p
#MCU_TARGET = atmega2560
#MCU_TARGET = atmega2561
#MCU_TARGET = atmega32
#MCU_TARGET = atmega324p
#MCU_TARGET = atmega325
#MCU_TARGET = atmega3250
#MCU_TARGET = atmega329
#MCU_TARGET = atmega3290
#MCU_TARGET = atmega48
#MCU_TARGET =atmega64
#MCU_TARGET =atmega640
#MCU_TARGET =atmega644
#MCU_TARGET =atmega644p
#MCU_TARGET =atmega645
#MCU_TARGET =atmega6450
#MCU_TARGET =atmega649
#MCU_TARGET =atmega6490
MCU_TARGET = atmega8
#MCU_TARGET = atmega8515
#MCU_TARGET = atmega8535
#MCU_TARGET = atmega88
#MCU_TARGET = attiny2313
#MCU_TARGET = attiny24
#MCU_TARGET = attiny25
#MCU_TARGET = attiny26
#MCU_TARGET = attiny261
#MCU_TARGET = attiny44
#MCU_TARGET = attiny45
#MCU_TARGET = attiny461
#MCU_TARGET = attiny84
#MCU_TARGET = attiny85
#MCU_TARGET = attiny861
OPTIMIZE = -O2
DEFS =
LIBS =
CC = avr-gcc
OBJCOPY = avr-objcopy
OBJDUMP = avr-objdump
$(PRG).elf: $(OBJ)
$(CC) $(CFLAGS) $(LDFLAGS) -o $@ $^ $(LIBS)
# dependency:
demo.o: demo.c iocompat.h
clean:
rm -rf *.o $(PRG).elf *.eps *.png *.pdf *.bak
rm -rf *.lst *.map $(EXTRA_CLEAN_FILES)
lst: $(PRG).lst
%.lst: %.elf
$(OBJDUMP) -h -S $< > $@
hex: $(PRG).hex
bin: $(PRG).bin
srec: $(PRG).srec
%.hex: %.elf
$(OBJCOPY) -j .text -j .data -O ihex $< $@
%.srec: %.elf
$(OBJCOPY) -j .text -j .data -O srec $< $@
%.bin: %.elf
$(OBJCOPY) -j .text -j .data -O binary $< $@
ehex: $(PRG)_eeprom.hex
ebin: $(PRG)_eeprom.bin
esrec: $(PRG)_eeprom.srec
%_eeprom.hex: %.elf
$(OBJCOPY) -j .eeprom --change-section-lma .eeprom=0 -O ihex $< $@ \
|| { echo empty $@ not generated; exit 0; }
%_eeprom.srec: %.elf
$(OBJCOPY) -j .eeprom --change-section-lma .eeprom=0 -O srec $< $@ \
|| { echo empty $@ not generated; exit 0; }
%_eeprom.bin: %.elf
$(OBJCOPY) -j .eeprom --change-section-lma .eeprom=0 -O binary $< $@ \
|| { echo empty $@ not generated; exit 0; }
# Every thing below here is used by avr-libc’s build system and can be ignored
# by the casual user.
FIG2DEV = fig2dev
EXTRA_CLEAN_FILES = *.hex *.bin *.srec
eps: $(PRG).eps
png: $(PRG).png
pdf: $(PRG).pdf
%.eps: %.fig
$(FIG2DEV) -L eps $< $@
%.pdf: %.fig
$(FIG2DEV) -L pdf $< $@
%.png: %.fig
$(FIG2DEV) -L png $< $@
This project extends the basic idea of the simple project to control a LED with a PWM
output, but adds methods to adjust the LED brightness. It employs a lot of the basic
concepts of avr-libc to achieve that goal.
Understanding this project assumes the simple project has been understood in full, as
well as being acquainted with the basic hardware concepts of an AVR microcontroller.
The demo is set up in a way so it can be run on the ATmega16 that ships with the
STK500 development kit. The only external part needed is a potentiometer attached to
the ADC. It is connected to a 10-pin ribbon cable for port A, both ends of the poten-
tiometer to pins 9 (GND) and 10 (VCC), and the wiper to pin 1 (port A0). A bypass
capacitor from pin 1 to pin 9 (like 47 nF) is recommendable.
The coloured patch cables are used to provide various interconnections. As there are
only four of them in the STK500, there are two options to connect them for this demo.
The second option for the yellow-green cable is shown in parenthesis in the table.
Alternatively, the "squid" cable from the JTAG ICE kit can be used if available.
The following picture shows the alternate wiring where LED1 is connected but SW2 is
not:
As an alternative, this demo can also be run on the popular ATmega8 controller, or its
successor ATmega88 as well as the ATmega48 and ATmega168 variants of the latter.
These controllers do not have a port named "A", so their ADC inputs are located on
port C instead, thus the potentiometer needs to be attached to port C. Likewise, the
OC1A output is not on port D pin 5 but on port B pin 1 (PB1). Thus, the above
cabling scheme needs to be changed so that PB1 connects to the LED0 pin. (PD6
remains unconnected.) When using the STK500, use one of the jumper cables for this
connection. All other port D pins should be connected the same way as described for
the ATmega16 above.
When not using an STK500 starter kit, attach the LEDs through some resistor to Vcc
(low-active LEDs), and attach pushbuttons from the respective input pins to GND. The
internal pull-up resistors are enabled for the pushbutton pins, so no external resistors
are needed.
Finally, the demo has been ported to the ATtiny2313 as well. As this AVR does not
offer an ADC, everything related to handling the ADC is disabled in the code for that
MCU type. Also, port D of this controller type only features 6 pins, so the 1-second
flash LED had to be moved from PD6 to PD4. (PD4 is used as the ADC control button
on the other MCU types, but that is not needed here.) OC1A is located at PB3 on this
device.
The MCU_TARGET macro in the Makefile needs to be adjusted appropriately for the
alternative controller types.
The flash ROM and RAM consumption of this demo are way below the resources
of even an ATmega48, and still well within the capabilities of an ATtiny2313. The
major advantage of experimenting with the ATmega16 (in addition that it ships together
with an STK500 anyway) is that it can be debugged online via JTAG. Likewise, the
ATmega48/88/168 and ATtiny2313 devices can be debugged through debugWire, using
the Atmel JTAG ICE mkII or the low-cost AVR Dragon.
Note that in the explanation below, all port/pin names are applicable to the ATmega16
setup.
PD6 will be toggled with each internal clock tick (approx. 10 ms). PD7 will flash once
per second.
PD0 and PD1 are configured as UART IO, and can be used to connect the demo kit to
a PC (9600 Bd, 8N1 frame format). The demo application talks to the serial port, and
it can be controlled from the serial port.
PD2 through PD4 are configured as inputs, and control the application unless control
has been taken over by the serial port. Shorting PD2 to GND will decrease the current
PWM value, shorting PD3 to GND will increase it.
While PD4 is shorted to GND, one ADC conversion for channel 0 (ADC input is on
PA0) will be triggered each internal clock tick, and the resulting value will be used as
the PWM value. So the brightness of the LED follows the analog input value on PC0.
VAREF on the STK500 should be set to the same value as VCC.
When running in serial control mode, the function of the watchdog timer can be demon-
strated by typing an ‘r’. This will make the demo application run in a tight loop without
retriggering the watchdog so after some seconds, the watchdog will reset the MCU.
This situation can be figured out on startup by reading the MCUCSR register.
The current value of the PWM is backed up in an EEPROM cell after about 3 seconds
of idle time after the last change. If that EEPROM cell contains a reasonable (i. e.
non-erased) value at startup, it is taken as the initial value for the PWM. This virtually
preserves the last value across power cycles. By not updating the EEPROM immme-
diately but only after a timeout, EEPROM wear is reduced considerably compared to
immediately writing the value at each change.
This section explains the ideas behind individual parts of the code. The source code
has been divided into numbered parts, and the following subsections explain each of
these parts.
The intflags structure demonstrates a way to allocate bit variables in memory. Each
of the interrupt service routines just sets one bit within that structure, and the applica-
tion’s main loop then monitors the bits in order to act appropriately.
Like all variables that are used to communicate values between an interrupt service
routine and the main application, it is declared volatile.
The variable ee_pwm is not a variable in the classical C sense that could be used as an
lvalue or within an expression to obtain its value. Instead, the
__attribute__((section(".eeprom")))
marks it as belonging to the EEPROM section. This section is merely used as a place-
holder so the compiler can arrange for each individual variable’s location in EEPROM.
The compiler will also keep track of initial values assigned, and usually the Makefile
is arranged to extract these initial values into a separate load file (largedemo_-
eeprom.∗ in this case) that can be used to initialize the EEPROM.
The actual EEPROM IO must be performed manually.
Similarly, the variable mcucsr is kept in the .noinit section in order to prevent it from
being cleared upon application startup.
The ISR to handle timer 1’s overflow interrupt arranges for the software clock. While
timer 1 runs the PWM, it calls its overflow handler rather frequently, so the TMR1_-
SCALE value is used as a postscaler to reduce the internal software clock frequency
further. If the software clock triggers, it sets the tmr_int bitfield, and defers all
further tasks to the main loop.
The ADC ISR just fetches the value from the ADC conversion, disables the ADC
interrupt again, and announces the presence of the new value in the adc_int bitfield.
The interrupt is kept disabled while not needed, because the ADC will also be triggered
by executing the SLEEP instruction in idle mode (which is the default sleep mode).
Another option would be to turn off the ADC completely here, but that increases the
ADC’s startup time (not that it would matter much for this application).
The function set_pwm() propagates the new PWM value to the PWM, performing
range checking. When the value has been changed, the new percentage will be an-
nounced on the serial link. The current value is mirrored in the variable pwm so others
can use it in calculations. In order to allow for a simple calculation of a percentage
value without requiring floating-point mathematics, the maximal value of the PWM is
restricted to 1000 rather than 1023, so a simple division by 10 can be used. Due to the
nature of the human eye, the difference in LED brightness between 1000 and 1023 is
not noticable anyway.
At the start of main(), a variable mode is declared to keep the current mode of
operation. An enumeration is used to improve the readability. By default, the compiler
would allocate a variable of type int for an enumeration. The packed attribute declarator
instructs the compiler to use the smallest possible integer type (which would be an 8-bit
type here).
After some initialization actions, the application’s main loop follows. In an embedded
application, this is normally an infinite loop as there is nothing an application could
"exit" into anyway.
At the beginning of the loop, the watchdog timer will be retriggered. If that timer is
not triggered for about 2 seconds, it will issue a hardware reset. Care needs to be taken
that no code path blocks longer than this, or it needs to frequently perform watchdog
resets of its own. An example of such a code path would be the string IO functions: for
an overly large string to print (about 2000 characters at 9600 Bd), they might block for
too long.
The loop itself then acts on the interrupt indication bitfields as appropriate, and will
eventually put the CPU on sleep at its end to conserve power.
The first interrupt bit that is handled is the (software) timer, at a frequency of approx-
imately 100 Hz. The CLOCKOUT pin will be toggled here, so e. g. an oscilloscope
can be used on that pin to measure the accuracy of our software clock. Then, the LED
flasher for LED2 ("We are alive"-LED) is built. It will flash that LED for about 50
ms, and pause it for another 950 ms. Various actions depending on the operation mode
follow. Finally, the 3-second backup timer is implemented that will write the PWM
value back to EEPROM once it is not changing anymore.
The ADC interrupt will just adjust the PWM value only.
Finally, the UART Rx interrupt will dispatch on the last character received from the
UART.
All the string literals that are used as informational messages within main() are
placed in program memory so no SRAM needs to be allocated for them. This is done
by using the PSTR macro, and passing the string to printstr_p().
This project illustrates how to use the standard IO facilities (stdio) provided by this
library. It assumes a basic knowledge of how the stdio subsystem is used in standard C
applications, and concentrates on the differences in this library’s implementation that
mainly result from the differences of the microcontroller environment, compared to a
hosted environment of a standard computer.
This demo is meant to supplement the documentation, not to replace it.
The demo is set up in a way so it can be run on the ATmega16 that ships with the
STK500 development kit. The UART port needs to be connected to the RS-232 "spare"
port by a jumper cable that connects PD0 to RxD and PD1 to TxD. The RS-232 channel
is set up as standard input (stdin) and standard output (stdout), respectively.
In order to have a different device available for a standard error channel (stderr), an
industry-standard LCD display with an HD44780-compatible LCD controller has been
chosen. This display needs to be connected to port A of the STK500 in the following
way:
Port Header Function
A0 1 LCD D4
A1 2 LCD D5
A2 3 LCD D6
A3 4 LCD D7
A4 5 LCD R/∼W
A5 6 LCD E
A6 7 LCD RS
A7 8 unused
GND 9 GND
VCC 10 Vcc
The LCD controller is used in 4-bit mode, including polling the "busy" flag so the
R/∼W line from the LCD controller needs to be connected. Note that the LCD con-
troller has yet another supply pin that is used to adjust the LCD’s contrast (V5). Typ-
ically, that pin connects to a potentiometer between Vcc and GND. Often, it might
work to just connect that pin to GND, while leaving it unconnected usually yields an
unreadable display.
Port A has been chosen as 7 pins are needed to connect the LCD, yet all other ports are
already partially in use: port B has the pins for in-system programming (ISP), port C
has the ports for JTAG (can be used for debugging), and port D is used for the UART
connection.
23.40.3.1 stdiodemo.c
As usual, include files go first. While conventionally, system header files (those in
angular brackets < ... >) go before application-specific header files (in double quotes),
defines.h comes as the first header file here. The main reason is that this file defines
the value of F_CPU which needs to be known before including <utils/delay.h>.
The function ioinit() summarizes all hardware initialization tasks. As this function
is declared to be module-internal only (static), the compiler will notice its simplic-
ity, and with a reasonable optimization level in effect, it will inline that function. That
needs to be kept in mind when debugging, because the inlining might cause the debug-
ger to "jump around wildly" at a first glance when single-stepping.
The definitions of uart_str and lcd_str set up two stdio streams. The initial-
ization is done using the FDEV_SETUP_STREAM() initializer template macro, so a
static object can be constructed that can be used for IO purposes. This initializer macro
takes three arguments, two function macros to connect the corresponding output and
input functions, respectively, the third one describes the intent of the stream (read,
write, or both). Those functions that are not required by the specified intent (like the
input function for lcd_str which is specified to only perform output operations) can
be given as NULL.
The stream uart_str corresponds to input and output operations performed over the
RS-232 connection to a terminal (e.g. from/to a PC running a terminal program), while
the lcd_str stream provides a method to display character data on the LCD text
display.
The function delay_1s() suspends program execution for approximately one sec-
ond. This is done using the _delay_ms() function from <util/delay.h>
which in turn needs the F_CPU macro in order to adjust the cycle counts. As the
_delay_ms() function has a limited range of allowable argument values (depending
on F_CPU), a value of 10 ms has been chosen as the base delay which would be safe
for CPU frequencies of up to about 26 MHz. This function is then called 100 times to
accomodate for the actual one-second delay.
In a practical application, long delays like this one were better be handled by a hardware
timer, so the main CPU would be free for other tasks while waiting, or could be put on
sleep.
At the beginning of main(), after initializing the peripheral devices, the default stdio
streams stdin, stdout, and stderr are set up by using the existing static FILE
stream objects. While this is not mandatory, the availability of stdin and stdout
allows to use the shorthand functions (e.g. printf() instead of fprintf()), and
stderr can mnemonically be referred to when sending out diagnostic messages.
Just for demonstration purposes, stdin and stdout are connected to a stream that
will perform UART IO, while stderr is arranged to output its data to the LCD text
display.
Finally, a main loop follows that accepts simple "commands" entered via the RS-232
connection, and performs a few simple actions based on the commands.
First, a prompt is sent out using printf_P() (which takes a program space string).
The string is read into an internal buffer as one line of input, using fgets(). While it
would be also possible to use gets() (which implicitly reads from stdin), gets()
has no control that the user’s input does not overflow the input buffer provided so it
should never be used at all.
If fgets() fails to read anything, the main loop is left. Of course, normally the main
loop of a microcontroller application is supposed to never finish, but again, for demon-
strational purposes, this explains the error handling of stdio. fgets() will return
NULL in case of an input error or end-of-file condition on input. Both these condi-
tions are in the domain of the function that is used to establish the stream, uart_-
putchar() in this case. In short, this function returns EOF in case of a serial line
"break" condition (extended start condition) has been recognized on the serial line.
Common PC terminal programs allow to assert this condition as some kind of out-of-
band signalling on an RS-232 connection.
When leaving the main loop, a goodbye message is sent to standard error output (i.e. to
the LCD), followed by three dots in one-second spacing, followed by a sequence that
will clear the LCD. Finally, main() will be terminated, and the library will add an
infinite loop, so only a CPU reset will be able to restart the application.
There are three "commands" recognized, each determined by the first letter of the line
entered (converted to lower case):
• The ’q’ (quit) command has the same effect of leaving the main loop.
• The ’l’ (LCD) command takes its second argument, and sends it to the LCD.
• The ’u’ (UART) command takes its second argument, and sends it back to the
UART connection.
Command recognition is done using sscanf() where the first format in the format
string just skips over the command itself (as the assignment suppression modifier ∗ is
given).
23.40.3.2 defines.h
The F_CPU macro defines the CPU clock frequency, to be used in delay loops, as well
as in the UART baud rate calculation.
The macro UART_BAUD defines the RS-232 baud rate. Depending on the actual CPU
frequency, only a limited range of baud rates can be supported.
The remaining macros customize the IO port and pins used for the HD44780 LCD
driver. Each definition consists of a letter naming the port this pin is attached to, and a
respective bit number. For accessing the data lines, only the first data line gets its own
macro (line D4 on the HD44780, lines D0 through D3 are not used in 4-bit mode), all
other data lines are expected to be in ascending order next to D4.
23.40.3.3 hd44780.h
This file describes the public interface of the low-level LCD driver that interfaces to
the HD44780 LCD controller. Public functions are available to initialize the controller
into 4-bit mode, to wait for the controller’s busy bit to be clear, and to read or write one
byte from or to the controller.
As there are two different forms of controller IO, one to send a command or receive
the controller status (RS signal clear), and one to send or receive data to/from the
controller’s SRAM (RS asserted), macros are provided that build on the mentioned
function primitives.
Finally, macros are provided for all the controller commands to allow them to be used
symbolically. The HD44780 datasheet explains these basic functions of the controller
in more detail.
23.40.3.4 hd44780.c
23.40.3.5 lcd.h
This function declares the public interface of the higher-level (character IO) LCD
driver.
23.40.3.6 lcd.c
The implementation of the higher-level LCD driver. This driver builds on top of the
HD44780 low-level LCD controller driver, and offers a character IO interface suitable
for direct use by the standard IO facilities. Where the low-level HD44780 driver deals
with setting up controller SRAM addresses, writing data to the controller’s SRAM,
and controlling display functions like clearing the display, or moving the cursor, this
high-level driver allows to just write a character to the LCD, in the assumption this will
somehow show up on the display.
Control characters can be handled at this level, and used to perform specific actions
on the LCD. Currently, there is only one control character that is being dealt with: a
newline character (\n) is taken as an indication to clear the display and set the cursor
into its initial position upon reception of the next character, so a "new line" of text
can be displayed. Therefore, a received newline character is remembered until more
characters have been sent by the application, and will only then cause the display to be
cleared before continuing. This provides a convenient abstraction where full lines of
text can be sent to the driver, and will remain visible at the LCD until the next line is
to be displayed.
Further control characters could be implemented, e. g. using a set of escape sequences.
That way, it would be possible to implement self-scrolling display lines etc.
The public function lcd_init() first calls the initialization entry point of the lower-
level HD44780 driver, and then sets up the LCD in a way we’d like to (display cleared,
23.40.3.7 uart.h
Public interface definition for the RS-232 UART driver, much like in lcd.h except there
is now also a character input function available.
As the RS-232 input is line-buffered in this example, the macro RX_BUFSIZE deter-
mines the size of that buffer.
23.40.3.8 uart.c
∧
• u (control-U, ASCII NAK) deletes the entire input buffer
∧
• w (control-W, ASCII ETB) deletes the previous input word, delimited by white
space
∧
• r (control-R, ASCII DC2) sends a \r, then reprints the buffer (refresh)
The function uart_init() takes care of all hardware initialization that is required to
put the UART into a mode with 8 data bits, no parity, one stop bit (commonly referred
to as 8N1) at the baud rate configured in defines.h. At low CPU clock frequencies, the
U2X bit in the UART is set, reducing the oversampling from 16x to 8x, which allows
for a 9600 Bd rate to be achieved with tolerable error using the default 1 MHz RC
oscillator.
The public function uart_putchar() again has suitable arguments for direct use
by the stdio stream interface. It performs the \n into \r\n translation by recursively
calling itself when it sees a \n character. Just for demonstration purposes, the \a
(audible bell, ASCII BEL) character is implemented by sending a string to stderr,
so it will be displayed on the LCD.
The public function uart_getchar() implements the line editor. If there are char-
acters available in the line buffer (variable rxp is not NULL), the next character will
be returned from the buffer without any UART interaction.
If there are no characters inside the line buffer, the input loop will be entered. Charac-
ters will be read from the UART, and processed accordingly. If the UART signalled a
framing error (FE bit set), typically caused by the terminal sending a line break con-
dition (start condition held much longer than one character period), the function will
return an end-of-file condition using _FDEV_EOF. If there was a data overrun condi-
tion on input (DOR bit set), an error condition will be returned as _FDEV_ERR.
Line editing characters are handled inside the loop, potentially modifying the buffer
status. If characters are attempted to be entered beyond the size of the line buffer, their
reception is refused, and a \a character is sent to the terminal. If a \r or \n character is
seen, the variable rxp (receive pointer) is set to the beginning of the buffer, the loop is
left, and the first character of the buffer will be returned to the application. (If no other
characters have been entered, this will just be the newline character, and the buffer is
marked as being exhausted immediately again.)
Some newer devices of the ATmega series contain builtin support for interfacing the
microcontroller to a two-wire bus, called TWI. This is essentially the same called I2C
by Philips, but that term is avoided in Atmel’s documentation due to patenting issues.
For the original Philips documentation, see
http://www.semiconductors.philips.com/buses/i2c/index.html
The two-wire interface consists of two signal lines named SDA (serial data) and SCL
(serial clock) (plus a ground line, of course). All devices participating in the bus are
connected together, using open-drain driver circuitry, so the wires must be terminated
using appropriate pullup resistors. The pullups must be small enough to recharge
the line capacity in short enough time compared to the desired maximal clock fre-
quency, yet large enough so all drivers will not be overloaded. There are formulas in
the datasheet that help selecting the pullups.
Devices can either act as a master to the bus (i. e., they initiate a transfer), or as a
slave (they only act when being called by a master). The bus is multi-master capable,
and a particular device implementation can act as either master or slave at different
times. Devices are addressed using a 7-bit address (coordinated by Philips) transfered
as the first byte after the so-called start condition. The LSB of that byte is R/∼W, i. e.
it determines whether the request to the slave is to read or write data during the next
cycles. (There is also an option to have devices using 10-bit addresses but that is not
covered by this example.)
The ATmega TWI hardware supports both, master and slave operation. This example
will only demonstrate how to use an AVR microcontroller as TWI master. The imple-
mentation is kept simple in order to concentrate on the steps that are required to talk to
a TWI slave, so all processing is done in polled-mode, waiting for the TWI interface to
indicate that the next processing step is due (by setting the TWINT interrupt bit). If it
is desired to have the entire TWI communication happen in "background", all this can
be implemented in an interrupt-controlled way, where only the start condition needs to
be triggered from outside the interrupt routine.
There is a variety of slave devices available that can be connected to a TWI bus. For the
purpose of this example, an EEPROM device out of the industry-standard 24Cxx series
has been chosen (where xx can be one of 01, 02, 04, 08, or 16) which are available from
various vendors. The choice was almost arbitrary, mainly triggered by the fact that an
EEPROM device is being talked to in both directions, reading and writing the slave
device, so the example will demonstrate the details of both.
Usually, there is probably not much need to add more EEPROM to an ATmega system
that way: the smallest possible AVR device that offers hardware TWI support is the
ATmega8 which comes with 512 bytes of EEPROM, which is equivalent to an 24C04
device. The ATmega128 already comes with twice as much EEPROM as the 24C16
would offer. One exception might be to use an externally connected EEPROM device
that is removable; e. g. SDRAM PC memory comes with an integrated TWI EEPROM
that carries the RAM configuration information.
Note [1]
The header file <util/twi.h> contains some macro definitions for symbolic con-
stants used in the TWI status register. These definitions match the names used in the
Atmel datasheet except that all names have been prefixed with TW_.
Note [2]
The clock is used in timer calculations done by the compiler, for the UART baud rate
and the TWI clock rate.
Note [3]
The address assigned for the 24Cxx EEPROM consists of 1010 in the upper four bits.
The following three bits are normally available as slave sub-addresses, allowing to
operate more than one device of the same type on a single bus, where the actual sub-
address used for each device is configured by hardware strapping. However, since the
next data packet following the device selection only allows for 8 bits that are used as
an EEPROM address, devices that require more than 8 address bits (24C04 and above)
"steal" subaddress bits and use them for the EEPROM cell address bits 9 to 11 as re-
quired. This example simply assumes all subaddress bits are 0 for the smaller devices,
so the E0, E1, and E2 inputs of the 24Cxx must be grounded.
Note [3a]
EEPROMs of type 24C32 and above cannot be addressed anymore even with the sub-
address bit trick. Thus, they require the upper address bits being sent separately on the
bus. When activating the WORD_ADDRESS_16BIT define, the algorithm implements
that auxiliary address byte transmission.
Note [4]
For slow clocks, enable the 2 x U[S]ART clock multiplier, to improve the baud rate
error. This will allow a 9600 Bd communication using the standard 1 MHz calibrated
RC oscillator. See also the Baud rate tables in the datasheets.
Note [5]
The datasheet explains why a minimum TWBR value of 10 should be maintained when
running in master mode. Thus, for system clocks below 3.6 MHz, we cannot run the
bus at the intented clock rate of 100 kHz but have to slow down accordingly.
Note [6]
This function is used by the standard output facilities that are utilized in this example
for debugging and demonstration purposes.
Note [7]
In order to shorten the data to be sent over the TWI bus, the 24Cxx EEPROMs support
multiple data bytes transfered within a single request, maintaining an internal address
counter that is updated after each data byte transfered successfully. When reading
data, one request can read the entire device memory if desired (the counter would wrap
around and start back from 0 when reaching the end of the device).
Note [8]
When reading the EEPROM, a first device selection must be made with write intent
(R/∼W bit set to 0 indicating a write operation) in order to transfer the EEPROM ad-
dress to start reading from. This is called master transmitter mode. Each completion
of a particular step in TWI communication is indicated by an asserted TWINT bit in
TWCR. (An interrupt would be generated if allowed.) After performing any actions
that are needed for the next communication step, the interrupt condition must be man-
ually cleared by setting the TWINT bit. Unlike with many other interrupt sources, this
would even be required when using a true interrupt routine, since as soon as TWINT is
re-asserted, the next bus transaction will start.
Note [9]
Since the TWI bus is multi-master capable, there is potential for a bus contention when
one master starts to access the bus. Normally, the TWI bus interface unit will detect this
situation, and will not initiate a start condition while the bus is busy. However, in case
two masters were starting at exactly the same time, the way bus arbitration works, there
is always a chance that one master could lose arbitration of the bus during any transmit
operation. A master that has lost arbitration is required by the protocol to immediately
cease talking on the bus; in particular it must not initiate a stop condition in order to not
corrupt the ongoing transfer from the active master. In this example, upon detecting a
lost arbitration condition, the entire transfer is going to be restarted. This will cause a
new start condition to be initiated, which will normally be delayed until the currently
active master has released the bus.
Note [10]
Next, the device slave is going to be reselected (using a so-called repeated start con-
dition which is meant to guarantee that the bus arbitration will remain at the current
master) using the same slave address (SLA), but this time with read intent (R/∼W bit
set to 1) in order to request the device slave to start transfering data from the slave to
the master in the next packet.
Note [11]
If the EEPROM device is still busy writing one or more cells after a previous write
request, it will simply leave its bus interface drivers at high impedance, and does not
respond to a selection in any way at all. The master selecting the device will see the
high level at SDA after transfering the SLA+R/W packet as a NACK to its selection
request. Thus, the select process is simply started over (effectively causing a repeated
start condition), until the device will eventually respond. This polling procedure is
recommended in the 24Cxx datasheet in order to minimize the busy wait time when
writing. Note that in case a device is broken and never responds to a selection (e. g.
since it is no longer present at all), this will cause an infinite loop. Thus the maximal
number of iterations made until the device is declared to be not responding at all, and
an error is returned, will be limited to MAX_ITER.
Note [12]
This is called master receiver mode: the bus master still supplies the SCL clock, but the
device slave drives the SDA line with the appropriate data. After 8 data bits, the master
responds with an ACK bit (SDA driven low) in order to request another data transfer
from the slave, or it can leave the SDA line high (NACK), indicating to the slave that
it is going to stop the transfer now. Assertion of ACK is handled by setting the TWEA
bit in TWCR when starting the current transfer.
Note [13]
The control word sent out in order to initiate the transfer of the next data packet is
initially set up to assert the TWEA bit. During the last loop iteration, TWEA is de-
asserted so the client will get informed that no further transfer is desired.
Note [14]
Except in the case of lost arbitration, all bus transactions must properly be terminated
by the master initiating a stop condition.
Note [15]
Writing to the EEPROM device is simpler than reading, since only a master transmitter
mode transfer is needed. Note that the first packet after the SLA+W selection is always
considered to be the EEPROM address for the next operation. (This packet is exactly
the same as the one above sent before starting to read the device.) In case a master
transmitter mode transfer is going to send more than one data packet, all following
packets will be considered data bytes to write at the indicated address. The internal
address pointer will be incremented after each write operation.
Note [16]
24Cxx devices can become write-protected by strapping their ∼WC pin to logic high.
(Leaving it unconnected is explicitly allowed, and constitutes logic low level, i. e. no
write protection.) In case of a write protected device, all data transfer attempts will be
NACKed by the device. Note that some devices might not implement this.
Data Fields
• int quot
• int rem
The Quotient.
The Remainder.
The documentation for this struct was generated from the following file:
• stdlib.h
Data Fields
• long quot
• long rem
The Quotient.
The Remainder.
The documentation for this struct was generated from the following file:
• stdlib.h
25 File Documentation
Defines
• #define assert(expression)
Defines
• #define _UTIL_ATOMIC_H_ 1
• #define ATOMIC_BLOCK(type)
• #define NONATOMIC_BLOCK(type)
• #define ATOMIC_RESTORESTATE
• #define ATOMIC_FORCEON
• #define NONATOMIC_RESTORESTATE
• #define NONATOMIC_FORCEOFF
Defines
• #define _AVR_BOOT_H_ 1
• #define BOOTLOADER_SECTION __attribute__ ((section (".bootloader")))
• #define __COMMON_ASB RWWSB
• #define __COMMON_ASRE RWWSRE
• #define BLB12 5
• #define BLB11 4
• #define BLB02 3
• #define BLB01 2
• #define boot_spm_interrupt_enable() (__SPM_REG |= (uint8_t)_BV(SPMIE))
• #define boot_spm_interrupt_disable() (__SPM_REG &= (uint8_t)∼_-
BV(SPMIE))
• #define boot_is_spm_interrupt() (__SPM_REG & (uint8_t)_BV(SPMIE))
• #define boot_rww_busy() (__SPM_REG & (uint8_t)_BV(__COMMON_ASB))
• #define boot_spm_busy() (__SPM_REG & (uint8_t)_BV(__SPM_ENABLE))
• #define boot_spm_busy_wait() do{}while(boot_spm_busy())
• #define __BOOT_PAGE_ERASE (_BV(__SPM_ENABLE) | _BV(PGERS))
• #define __BOOT_PAGE_WRITE (_BV(__SPM_ENABLE) | _BV(PGWRT))
• #define __BOOT_PAGE_FILL _BV(__SPM_ENABLE)
• #define __BOOT_RWW_ENABLE (_BV(__SPM_ENABLE) | _BV(__-
COMMON_ASRE))
• #define __boot_page_fill_normal(address, data)
• #define __boot_page_fill_alternate(address, data)
• #define __boot_page_fill_extended(address, data)
• #define __boot_page_erase_normal(address)
• #define __boot_page_erase_alternate(address)
• #define __boot_page_erase_extended(address)
• #define __boot_page_write_normal(address)
• #define __boot_page_write_alternate(address)
• #define __boot_page_write_extended(address)
• #define __boot_rww_enable()
• #define __boot_rww_enable_alternate()
• #define __boot_lock_bits_set(lock_bits)
• #define __boot_lock_bits_set_alternate(lock_bits)
• #define GET_LOW_FUSE_BITS (0x0000)
• #define GET_LOCK_BITS (0x0001)
• #define GET_EXTENDED_FUSE_BITS (0x0002)
• #define GET_HIGH_FUSE_BITS (0x0003)
• #define boot_lock_fuse_bits_get(address)
• #define __BOOT_SIGROW_READ (_BV(__SPM_ENABLE) | _BV(SIGRD))
• #define boot_signature_byte_get(addr)
• #define boot_page_fill(address, data) __boot_page_fill_normal(address, data)
• #define boot_page_erase(address) __boot_page_erase_normal(address)
• #define boot_page_write(address) __boot_page_write_normal(address)
• #define boot_rww_enable() __boot_rww_enable()
• #define boot_lock_bits_set(lock_bits) __boot_lock_bits_set(lock_bits)
• #define boot_page_fill_safe(address, data)
• #define boot_page_erase_safe(address)
• #define boot_page_write_safe(address)
• #define boot_rww_enable_safe()
• #define boot_lock_bits_set_safe(lock_bits)
Value:
(__extension__({ \
uint8_t value = (uint8_t)(~(lock_bits)); \
__asm__ __volatile__ \
( \
"ldi r30, 1\n\t" \
"ldi r31, 0\n\t" \
"mov r0, %2\n\t" \
"sts %0, %1\n\t" \
"spm\n\t" \
: \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"r" ((uint8_t)(__BOOT_LOCK_BITS_SET)), \
"r" (value) \
: "r0", "r30", "r31" \
); \
}))
Value:
(__extension__({ \
uint8_t value = (uint8_t)(~(lock_bits)); \
__asm__ __volatile__ \
( \
"ldi r30, 1\n\t" \
"ldi r31, 0\n\t" \
"mov r0, %2\n\t" \
"sts %0, %1\n\t" \
"spm\n\t" \
".word 0xffff\n\t" \
"nop\n\t" \
: \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"r" ((uint8_t)(__BOOT_LOCK_BITS_SET)), \
"r" (value) \
: "r0", "r30", "r31" \
); \
}))
Value:
(__extension__({ \
__asm__ __volatile__ \
( \
"sts %0, %1\n\t" \
"spm\n\t" \
".word 0xffff\n\t" \
"nop\n\t" \
: \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"r" ((uint8_t)(__BOOT_PAGE_ERASE)), \
"z" ((uint16_t)(address)) \
); \
}))
Value:
(__extension__({ \
__asm__ __volatile__ \
( \
"movw r30, %A3\n\t" \
"sts %1, %C3\n\t" \
"sts %0, %2\n\t" \
"spm\n\t" \
: \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"i" (_SFR_MEM_ADDR(RAMPZ)), \
"r" ((uint8_t)(__BOOT_PAGE_ERASE)), \
"r" ((uint32_t)(address)) \
: "r30", "r31" \
); \
}))
Value:
(__extension__({ \
__asm__ __volatile__ \
( \
"sts %0, %1\n\t" \
"spm\n\t" \
: \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"r" ((uint8_t)(__BOOT_PAGE_ERASE)), \
"z" ((uint16_t)(address)) \
); \
}))
Value:
(__extension__({ \
__asm__ __volatile__ \
( \
"movw r0, %3\n\t" \
"sts %0, %1\n\t" \
"spm\n\t" \
".word 0xffff\n\t" \
"nop\n\t" \
"clr r1\n\t" \
: \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"r" ((uint8_t)(__BOOT_PAGE_FILL)), \
"z" ((uint16_t)(address)), \
"r" ((uint16_t)(data)) \
: "r0" \
); \
}))
Value:
(__extension__({ \
__asm__ __volatile__ \
( \
"movw r0, %4\n\t" \
"movw r30, %A3\n\t" \
"sts %1, %C3\n\t" \
"sts %0, %2\n\t" \
"spm\n\t" \
"clr r1\n\t" \
: \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"i" (_SFR_MEM_ADDR(RAMPZ)), \
"r" ((uint8_t)(__BOOT_PAGE_FILL)), \
"r" ((uint32_t)(address)), \
"r" ((uint16_t)(data)) \
: "r0", "r30", "r31" \
); \
}))
Value:
(__extension__({ \
__asm__ __volatile__ \
( \
"movw r0, %3\n\t" \
"sts %0, %1\n\t" \
"spm\n\t" \
"clr r1\n\t" \
: \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"r" ((uint8_t)(__BOOT_PAGE_FILL)), \
"z" ((uint16_t)(address)), \
"r" ((uint16_t)(data)) \
: "r0" \
); \
}))
Value:
(__extension__({ \
__asm__ __volatile__ \
( \
"sts %0, %1\n\t" \
"spm\n\t" \
".word 0xffff\n\t" \
"nop\n\t" \
: \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"r" ((uint8_t)(__BOOT_PAGE_WRITE)), \
"z" ((uint16_t)(address)) \
); \
}))
Value:
(__extension__({ \
__asm__ __volatile__ \
( \
"movw r30, %A3\n\t" \
"sts %1, %C3\n\t" \
"sts %0, %2\n\t" \
"spm\n\t" \
: \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"i" (_SFR_MEM_ADDR(RAMPZ)), \
"r" ((uint8_t)(__BOOT_PAGE_WRITE)), \
"r" ((uint32_t)(address)) \
: "r30", "r31" \
); \
}))
Value:
(__extension__({ \
__asm__ __volatile__ \
( \
"sts %0, %1\n\t" \
"spm\n\t" \
: \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"r" ((uint8_t)(__BOOT_PAGE_WRITE)), \
"z" ((uint16_t)(address)) \
); \
}))
Value:
(__extension__({ \
__asm__ __volatile__ \
( \
"sts %0, %1\n\t" \
"spm\n\t" \
: \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"r" ((uint8_t)(__BOOT_RWW_ENABLE)) \
); \
}))
Value:
(__extension__({ \
__asm__ __volatile__ \
( \
"sts %0, %1\n\t" \
"spm\n\t" \
".word 0xffff\n\t" \
"nop\n\t" \
: \
: "i" (_SFR_MEM_ADDR(__SPM_REG)), \
"r" ((uint8_t)(__BOOT_RWW_ENABLE)) \
); \
}))
Defines
• #define _AVR_CPUFUNC_H_ 1
• #define _NOP()
• #define _MemoryBarrier()
Functions
Defines
• #define __CTYPE_H_ 1
Functions
Defines
• #define _UTIL_DELAY_H_ 1
• #define __HAS_DELAY_CYCLES 1
• #define F_CPU 1000000UL
Functions
Defines
• #define _UTIL_DELAY_BASIC_H_ 1
Functions
Defines
• #define __ERRNO_H_ 1
• #define EDOM 33
• #define ERANGE 34
Variables
• int errno
Functions
Defines
• #define _AVR_FUSE_H_ 1
• #define FUSEMEM __attribute__((section (".fuse")))
• #define FUSES __fuse_t __fuse FUSEMEM
Defines
• #define sei()
• #define cli()
• #define SIGNAL(vector)
• #define EMPTY_INTERRUPT(vector)
• #define ISR_ALIAS(vector, target_vector)
• #define reti()
• #define BADISR_vect
ISR attributes
• #define ISR_BLOCK
• #define ISR_NOBLOCK
• #define ISR_NAKED
• #define ISR_ALIASOF(target_vector)
@{
Defines
Typedefs
Defines
• #define _AVR_LOCK_H_ 1
• #define LOCKMEM __attribute__((section (".lock")))
• #define LOCKBITS unsigned char __lock LOCKMEM
• #define LOCKBITS_DEFAULT (0xFF)
Defines
Functions
Defines
• #define __PGMSPACE_H_ 1
• #define __need_size_t
• #define __ATTR_PROGMEM__ __attribute__((__progmem__))
• #define PROGMEM __ATTR_PROGMEM__
• #define PSTR(s) ((const PROGMEM char ∗)(s))
• #define __LPM_classic__(addr)
• #define __LPM_enhanced__(addr)
• #define __LPM_word_classic__(addr)
• #define __LPM_word_enhanced__(addr)
• #define __LPM_dword_classic__(addr)
• #define __LPM_dword_enhanced__(addr)
• #define __LPM_float_classic__(addr)
• #define __LPM_float_enhanced__(addr)
• #define __LPM(addr) __LPM_classic__(addr)
• #define __LPM_word(addr) __LPM_word_classic__(addr)
• #define __LPM_dword(addr) __LPM_dword_classic__(addr)
• #define __LPM_float(addr) __LPM_float_classic__(addr)
• #define pgm_read_byte_near(address_short) __LPM((uint16_t)(address_short))
• #define pgm_read_word_near(address_short) __LPM_word((uint16_-
t)(address_short))
• #define pgm_read_dword_near(address_short) __LPM_dword((uint16_-
t)(address_short))
• #define pgm_read_float_near(address_short) __LPM_float((uint16_t)(address_-
short))
• #define __ELPM_classic__(addr)
• #define __ELPM_enhanced__(addr)
• #define __ELPM_xmega__(addr)
• #define __ELPM_word_classic__(addr)
• #define __ELPM_word_enhanced__(addr)
• #define __ELPM_word_xmega__(addr)
• #define __ELPM_dword_classic__(addr)
• #define __ELPM_dword_enhanced__(addr)
• #define __ELPM_dword_xmega__(addr)
• #define __ELPM_float_classic__(addr)
• #define __ELPM_float_enhanced__(addr)
• #define __ELPM_float_xmega__(addr)
• #define __ELPM(addr) __ELPM_classic__(addr)
Typedefs
Functions
Value:
(__extension__({ \
uint32_t __addr32 = (uint32_t)(addr); \
uint8_t __result; \
__asm__ \
( \
"out %2, %C1" "\n\t" \
"mov r31, %B1" "\n\t" \
"mov r30, %A1" "\n\t" \
"elpm" "\n\t" \
"mov %0, r0" "\n\t" \
: "=r" (__result) \
: "r" (__addr32), \
"I" (_SFR_IO_ADDR(RAMPZ)) \
: "r0", "r30", "r31" \
); \
__result; \
}))
Value:
(__extension__({ \
uint32_t __addr32 = (uint32_t)(addr); \
uint32_t __result; \
__asm__ \
( \
"out %2, %C1" "\n\t" \
"movw r30, %1" "\n\t" \
"elpm %A0, Z+" "\n\t" \
"elpm %B0, Z+" "\n\t" \
"elpm %C0, Z+" "\n\t" \
"elpm %D0, Z" "\n\t" \
: "=r" (__result) \
: "r" (__addr32), \
"I" (_SFR_IO_ADDR(RAMPZ)) \
: "r30", "r31" \
); \
__result; \
}))
Value:
(__extension__({ \
uint32_t __addr32 = (uint32_t)(addr); \
uint32_t __result; \
__asm__ \
( \
"in __tmp_reg__, %2" "\n\t" \
"out %2, %C1" "\n\t" \
"movw r30, %1" "\n\t" \
"elpm %A0, Z+" "\n\t" \
"elpm %B0, Z+" "\n\t" \
"elpm %C0, Z+" "\n\t" \
"elpm %D0, Z" "\n\t" \
"out %2, __tmp_reg__" \
: "=r" (__result) \
: "r" (__addr32), \
"I" (_SFR_IO_ADDR(RAMPZ)) \
: "r30", "r31" \
); \
__result; \
}))
Value:
(__extension__({ \
uint32_t __addr32 = (uint32_t)(addr); \
uint8_t __result; \
__asm__ \
( \
"out %2, %C1" "\n\t" \
"movw r30, %1" "\n\t" \
"elpm %0, Z+" "\n\t" \
: "=r" (__result) \
: "r" (__addr32), \
"I" (_SFR_IO_ADDR(RAMPZ)) \
: "r30", "r31" \
); \
__result; \
}))
Value:
(__extension__({ \
uint32_t __addr32 = (uint32_t)(addr); \
float __result; \
__asm__ \
( \
"out %2, %C1" "\n\t" \
"movw r30, %1" "\n\t" \
Value:
(__extension__({ \
uint32_t __addr32 = (uint32_t)(addr); \
float __result; \
__asm__ \
( \
"in __tmp_reg__, %2" "\n\t" \
"out %2, %C1" "\n\t" \
"movw r30, %1" "\n\t" \
"elpm %A0, Z+" "\n\t" \
"elpm %B0, Z+" "\n\t" \
"elpm %C0, Z+" "\n\t" \
"elpm %D0, Z" "\n\t" \
"out %2, __tmp_reg__" \
: "=r" (__result) \
: "r" (__addr32), \
"I" (_SFR_IO_ADDR(RAMPZ)) \
: "r30", "r31" \
); \
__result; \
}))
Value:
(__extension__({ \
uint32_t __addr32 = (uint32_t)(addr); \
uint16_t __result; \
__asm__ \
( \
"out %2, %C1" "\n\t" \
"mov r31, %B1" "\n\t" \
"mov r30, %A1" "\n\t" \
"elpm" "\n\t" \
"mov %A0, r0" "\n\t" \
"in r0, %2" "\n\t" \
Value:
(__extension__({ \
uint32_t __addr32 = (uint32_t)(addr); \
uint16_t __result; \
__asm__ \
( \
"out %2, %C1" "\n\t" \
"movw r30, %1" "\n\t" \
"elpm %A0, Z+" "\n\t" \
"elpm %B0, Z" "\n\t" \
: "=r" (__result) \
: "r" (__addr32), \
"I" (_SFR_IO_ADDR(RAMPZ)) \
: "r30", "r31" \
); \
__result; \
}))
Value:
(__extension__({ \
uint32_t __addr32 = (uint32_t)(addr); \
uint16_t __result; \
__asm__ \
( \
"in __tmp_reg__, %2" "\n\t" \
"out %2, %C1" "\n\t" \
"movw r30, %1" "\n\t" \
"elpm %A0, Z+" "\n\t" \
"elpm %B0, Z" "\n\t" \
"out %2, __tmp_reg__" \
: "=r" (__result) \
: "r" (__addr32), \
"I" (_SFR_IO_ADDR(RAMPZ)) \
: "r30", "r31" \
); \
__result; \
}))
Value:
(__extension__({ \
uint32_t __addr32 = (uint32_t)(addr); \
uint8_t __result; \
__asm__ \
( \
"in __tmp_reg__, %2" "\n\t" \
"out %2, %C1" "\n\t" \
"movw r30, %1" "\n\t" \
"elpm %0, Z+" "\n\t" \
"out %2, __tmp_reg__" \
: "=r" (__result) \
: "r" (__addr32), \
"I" (_SFR_IO_ADDR(RAMPZ)) \
: "r30", "r31" \
); \
__result; \
}))
Value:
(__extension__({ \
uint16_t __addr16 = (uint16_t)(addr); \
uint8_t __result; \
__asm__ \
( \
"lpm" "\n\t" \
"mov %0, r0" "\n\t" \
: "=r" (__result) \
: "z" (__addr16) \
: "r0" \
); \
__result; \
}))
Value:
(__extension__({ \
uint16_t __addr16 = (uint16_t)(addr); \
uint32_t __result; \
__asm__ \
( \
"lpm" "\n\t" \
"mov %A0, r0" "\n\t" \
"adiw r30, 1" "\n\t" \
"lpm" "\n\t" \
"mov %B0, r0" "\n\t" \
"adiw r30, 1" "\n\t" \
"lpm" "\n\t" \
"mov %C0, r0" "\n\t" \
"adiw r30, 1" "\n\t" \
"lpm" "\n\t" \
"mov %D0, r0" "\n\t" \
: "=r" (__result), "=z" (__addr16) \
: "1" (__addr16) \
: "r0" \
); \
__result; \
}))
Value:
(__extension__({ \
uint16_t __addr16 = (uint16_t)(addr); \
uint32_t __result; \
__asm__ \
( \
"lpm %A0, Z+" "\n\t" \
"lpm %B0, Z+" "\n\t" \
"lpm %C0, Z+" "\n\t" \
"lpm %D0, Z" "\n\t" \
: "=r" (__result), "=z" (__addr16) \
: "1" (__addr16) \
); \
__result; \
}))
Value:
(__extension__({ \
uint16_t __addr16 = (uint16_t)(addr); \
uint8_t __result; \
__asm__ \
( \
"lpm %0, Z" "\n\t" \
: "=r" (__result) \
: "z" (__addr16) \
); \
__result; \
}))
Value:
(__extension__({ \
uint16_t __addr16 = (uint16_t)(addr); \
float __result; \
__asm__ \
( \
"lpm" "\n\t" \
"mov %A0, r0" "\n\t" \
"adiw r30, 1" "\n\t" \
"lpm" "\n\t" \
"mov %B0, r0" "\n\t" \
"adiw r30, 1" "\n\t" \
"lpm" "\n\t" \
"mov %C0, r0" "\n\t" \
"adiw r30, 1" "\n\t" \
"lpm" "\n\t" \
"mov %D0, r0" "\n\t" \
: "=r" (__result), "=z" (__addr16) \
: "1" (__addr16) \
: "r0" \
); \
__result; \
}))
Value:
(__extension__({ \
uint16_t __addr16 = (uint16_t)(addr); \
float __result; \
__asm__ \
( \
"lpm %A0, Z+" "\n\t" \
"lpm %B0, Z+" "\n\t" \
"lpm %C0, Z+" "\n\t" \
"lpm %D0, Z" "\n\t" \
: "=r" (__result), "=z" (__addr16) \
: "1" (__addr16) \
); \
__result; \
}))
Value:
(__extension__({ \
uint16_t __addr16 = (uint16_t)(addr); \
uint16_t __result; \
__asm__ \
( \
"lpm" "\n\t" \
"mov %A0, r0" "\n\t" \
"adiw r30, 1" "\n\t" \
"lpm" "\n\t" \
"mov %B0, r0" "\n\t" \
: "=r" (__result), "=z" (__addr16) \
: "1" (__addr16) \
: "r0" \
); \
__result; \
}))
Value:
(__extension__({ \
uint16_t __addr16 = (uint16_t)(addr); \
uint16_t __result; \
__asm__ \
( \
"lpm %A0, Z+" "\n\t" \
"lpm %B0, Z" "\n\t" \
: "=r" (__result), "=z" (__addr16) \
: "1" (__addr16) \
); \
__result; \
}))
Value:
({ \
uint_farptr_t tmp; \
\
__asm__ __volatile__( \
\
"ldi %A0, lo8(%1)" "\n\t" \
"ldi %B0, hi8(%1)" "\n\t" \
"ldi %C0, hh8(%1)" "\n\t" \
"clr %D0" "\n\t" \
: \
"=d" (tmp) \
: \
"p" (&(var)) \
); \
tmp; \
})
Defines
• #define _AVR_POWER_H_ 1
• #define clock_prescale_get() (clock_div_t)(CLKPR & (uint8_-
t)((1<<CLKPS0)|(1<<CLKPS1)|(1<<CLKPS2)|(1<<CLKPS3)))
Enumerations
• enum clock_div_t {
clock_div_1 = 0, clock_div_2 = 1, clock_div_4 = 2, clock_div_8 = 3,
clock_div_16 = 4, clock_div_32 = 5, clock_div_64 = 6, clock_div_128 = 7,
clock_div_256 = 8 }
Functions
Defines
• #define BAUD_TOL 2
• #define UBRR_VALUE
• #define UBRRL_VALUE
• #define UBRRH_VALUE
• #define USE_2X 0
Defines
• #define __SETJMP_H_ 1
• #define __ATTR_NORETURN__ __attribute__((__noreturn__))
Functions
Defines
• #define _AVR_SIGNATURE_H_ 1
Defines
• #define _AVR_SLEEP_H_ 1
• #define _SLEEP_CONTROL_REG MCUCR
• #define _SLEEP_ENABLE_MASK _BV(SE)
Functions
Defines
• #define __USING_MINT8 0
• #define __CONCATenate(left, right) left ## right
• #define __CONCAT(left, right) __CONCATenate(left, right)
Typedefs
Defines
• #define _STDIO_H_ 1
• #define __need_NULL
• #define __need_size_t
• #define FILE struct __file
• #define stdin (__iob[0])
• #define stdout (__iob[1])
• #define stderr (__iob[2])
• #define EOF (-1)
• #define fdev_set_udata(stream, u) do { (stream)->udata = u; } while(0)
• #define fdev_get_udata(stream) ((stream)->udata)
• #define fdev_setup_stream(stream, put, get, rwflag)
• #define _FDEV_SETUP_READ __SRD
• #define _FDEV_SETUP_WRITE __SWR
• #define _FDEV_SETUP_RW (__SRD|__SWR)
• #define _FDEV_ERR (-1)
• #define _FDEV_EOF (-2)
• #define FDEV_SETUP_STREAM(put, get, rwflag)
• #define fdev_close()
• #define putc(__c, __stream) fputc(__c, __stream)
• #define putchar(__c) fputc(__c, stdout)
• #define getc(__stream) fgetc(__stream)
• #define getchar() fgetc(stdin)
• #define SEEK_SET 0
• #define SEEK_CUR 1
• #define SEEK_END 2
Functions
Data Structures
• struct div_t
• struct ldiv_t
Defines
• #define _STDLIB_H_ 1
• #define __need_NULL
• #define __need_size_t
• #define __need_wchar_t
• #define __ptr_t void ∗
• #define RAND_MAX 0x7FFF
Typedefs
Functions
Variables
• size_t __malloc_margin
• char ∗ __malloc_heap_start
• char ∗ __malloc_heap_end
Note that these functions are not located in the default library, libc.a, but in the
mathematical library, libm.a. So when linking the application, the -lm option needs
to be specified.
Defines
• #define _STRING_H_ 1
• #define __need_NULL
• #define __need_size_t
• #define __ATTR_PURE__ __attribute__((__pure__))
• #define _FFS(x)
Functions
Variables
• static char ∗ p
Functions
Defines
• #define _UTIL_TWI_H_ 1
TWSR values
Mnemonics:
TW_MT_xxx - master transmitter
TW_MR_xxx - master receiver
TW_ST_xxx - slave transmitter
TW_SR_xxx - slave receiver
• #define TW_READ 1
• #define TW_WRITE 0
Defines
• #define WDTO_60MS 2
• #define WDTO_120MS 3
• #define WDTO_250MS 4
• #define WDTO_500MS 5
• #define WDTO_1S 6
• #define WDTO_2S 7
• #define WDTO_4S 8
• #define WDTO_8S 9
__AVR_LIBC_MAJOR__ __LPM_word_classic__
avr_version, 301 pgmspace.h, 397
__AVR_LIBC_MINOR__ __LPM_word_enhanced__
avr_version, 301 pgmspace.h, 398
__AVR_LIBC_REVISION__ __boot_lock_bits_set
avr_version, 301 boot.h, 371
__AVR_LIBC_VERSION_STRING__ __boot_lock_bits_set_alternate
avr_version, 301 boot.h, 371
__AVR_LIBC_VERSION__ __boot_page_erase_alternate
avr_version, 301 boot.h, 372
__EEGET __boot_page_erase_extended
avr_eeprom, 236 boot.h, 372
__EEPUT __boot_page_erase_normal
avr_eeprom, 236 boot.h, 373
__ELPM_classic__ __boot_page_fill_alternate
pgmspace.h, 391 boot.h, 373
__ELPM_dword_enhanced__ __boot_page_fill_extended
pgmspace.h, 391 boot.h, 373
__ELPM_dword_xmega__ __boot_page_fill_normal
pgmspace.h, 391 boot.h, 374
__ELPM_enhanced__ __boot_page_write_alternate
pgmspace.h, 392 boot.h, 374
__ELPM_float_enhanced__ __boot_page_write_extended
pgmspace.h, 392 boot.h, 375
__ELPM_float_xmega__ __boot_page_write_normal
pgmspace.h, 393 boot.h, 375
__ELPM_word_classic__ __boot_rww_enable
pgmspace.h, 393 boot.h, 375
__ELPM_word_enhanced__ __boot_rww_enable_alternate
pgmspace.h, 394 boot.h, 376
__ELPM_word_xmega__ __compar_fn_t
pgmspace.h, 394 avr_stdlib, 204
__ELPM_xmega__ __malloc_heap_end
pgmspace.h, 395 avr_stdlib, 213
__LPM_classic__ __malloc_heap_start
pgmspace.h, 395 avr_stdlib, 213
__LPM_dword_classic__ __malloc_margin
pgmspace.h, 395 avr_stdlib, 213
__LPM_dword_enhanced__ _crc16_update
pgmspace.h, 396 util_crc, 310
__LPM_enhanced__ _crc_ccitt_update
pgmspace.h, 396 util_crc, 311
__LPM_float_classic__ _crc_ibutton_update
pgmspace.h, 397 util_crc, 311
__LPM_float_enhanced__ _crc_xmodem_update
pgmspace.h, 397 util_crc, 312
eeprom_write_word ferror
avr_eeprom, 238 avr_stdio, 191
EMPTY_INTERRUPT fflush
avr_interrupts, 262 avr_stdio, 191
enable_external_int ffs
deprecated_items, 324 avr_string, 215
EOF ffs.S, 379
avr_stdio, 187 ffsl
ERANGE avr_string, 215
avr_errno, 140 ffsl.S, 379
errno.h, 378 ffsll
Example using the two-wire interface avr_string, 215
(TWI), 363 ffsll.S, 379
exit fgetc
avr_stdlib, 206 avr_stdio, 191
exp fgets
avr_math, 162 avr_stdio, 191
expf FILE
avr_math, 157 avr_stdio, 189
floor
fabs avr_math, 163
avr_math, 162 floorf
fabsf avr_math, 157
avr_math, 157 fma
FAQ, 60 avr_math, 163
fclose fmaf
avr_stdio, 190 avr_math, 157
fdev_close fmax
avr_stdio, 187 avr_math, 163
fdev_get_udata fmaxf
avr_stdio, 188 avr_math, 157
fdev_set_udata fmin
avr_stdio, 188 avr_math, 163
FDEV_SETUP_STREAM fminf
avr_stdio, 188 avr_math, 157
fdev_setup_stream fmod
avr_stdio, 188 avr_math, 163
fdevopen fmodf
avr_stdio, 190 avr_math, 157
fdevopen.c, 378 fprintf
fdim avr_stdio, 192
avr_math, 162 fprintf_P
fdimf avr_stdio, 192
avr_math, 157 fputc
feof avr_stdio, 192
avr_stdio, 191 fputs
PRIu16 PRIxLEAST16
avr_inttypes, 146 avr_inttypes, 149
PRIu32 PRIXLEAST32
avr_inttypes, 146 avr_inttypes, 149
PRIu8 PRIxLEAST32
avr_inttypes, 146 avr_inttypes, 149
PRIuFAST16 PRIXLEAST8
avr_inttypes, 147 avr_inttypes, 149
PRIuFAST32 PRIxLEAST8
avr_inttypes, 147 avr_inttypes, 149
PRIuFAST8 PRIXPTR
avr_inttypes, 147 avr_inttypes, 149
PRIuLEAST16 PRIxPTR
avr_inttypes, 147 avr_inttypes, 149
PRIuLEAST32 prog_char
avr_inttypes, 147 avr_pgmspace, 275
PRIuLEAST8 prog_int16_t
avr_inttypes, 147 avr_pgmspace, 275
PRIuPTR prog_int32_t
avr_inttypes, 147 avr_pgmspace, 275
PRIX16 prog_int64_t
avr_inttypes, 147 avr_pgmspace, 275
PRIx16 prog_int8_t
avr_inttypes, 147 avr_pgmspace, 275
PRIX32 prog_uchar
avr_inttypes, 147 avr_pgmspace, 275
PRIx32 prog_uint16_t
avr_inttypes, 148 avr_pgmspace, 275
PRIX8 prog_uint32_t
avr_inttypes, 148 avr_pgmspace, 275
PRIx8 prog_uint64_t
avr_inttypes, 148 avr_pgmspace, 276
PRIXFAST16 prog_uint8_t
avr_inttypes, 148 avr_pgmspace, 276
PRIxFAST16 prog_void
avr_inttypes, 148 avr_pgmspace, 276
PRIXFAST32 PROGMEM
avr_inttypes, 148 avr_pgmspace, 274
PRIxFAST32 PSTR
avr_inttypes, 148 avr_pgmspace, 274
PRIXFAST8 PTRDIFF_MAX
avr_inttypes, 148 avr_stdint, 176
PRIxFAST8 PTRDIFF_MIN
avr_inttypes, 148 avr_stdint, 176
PRIXLEAST16 putc
avr_inttypes, 148 avr_stdio, 189
strcasecmp_P strcspn
avr_pgmspace, 278 avr_string, 220
strcasecmp_P.S, 409 strcspn.S, 409
strcasecmp_PF strcspn_P
avr_pgmspace, 278 avr_pgmspace, 281
strcasestr strcspn_P.S, 409
avr_string, 218 strdup
strcasestr.S, 409 avr_string, 220
strcasestr_P strdup.c, 409
avr_pgmspace, 279 string.h, 410
strcat strlcat
avr_string, 218 avr_string, 221
strcat.S, 409 strlcat.S, 413
strcat_P strlcat_P
avr_pgmspace, 279 avr_pgmspace, 282
strcat_P.S, 409 strlcat_P.S, 413
strcat_PF strlcat_PF
avr_pgmspace, 279 avr_pgmspace, 282
strchr strlcpy
avr_string, 219 avr_string, 221
strchr.S, 409 strlcpy.S, 413
strchr_P strlcpy_P
avr_pgmspace, 280 avr_pgmspace, 283
strchr_P.S, 409 strlcpy_P.S, 413
strchrnul strlcpy_PF
avr_string, 219 avr_pgmspace, 283
strchrnul.S, 409 strlen
strchrnul_P avr_string, 221
avr_pgmspace, 280 strlen.S, 413
strchrnul_P.S, 409 strlen_P
strcmp avr_pgmspace, 283
avr_string, 219 strlen_P.S, 413
strcmp.S, 409 strlen_PF
strcmp_P avr_pgmspace, 283
avr_pgmspace, 280 strlwr
strcmp_P.S, 409 avr_string, 222
strcmp_PF strlwr.S, 413
avr_pgmspace, 280 strncasecmp
strcpy avr_string, 222
avr_string, 219 strncasecmp.S, 413
strcpy.S, 409 strncasecmp_P
strcpy_P avr_pgmspace, 284
avr_pgmspace, 281 strncasecmp_P.S, 413
strcpy_P.S, 409 strncasecmp_PF
strcpy_PF avr_pgmspace, 284
avr_pgmspace, 281 strncat
avr_watchdog, 305
WDTO_60MS
avr_watchdog, 305
WDTO_8S
avr_watchdog, 305