README

AVRA v1.3.0 - Assember for the Atmel AVR microcontroller family

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. Please read below for for information.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program; see the file "COPYING". If not, visit http://www.gnu.org or write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. You can also contact the authors of AVRA to receive a copy of the COPYING file.

Atmel, AVR, AVR Studio, Intel, Windows are registered enterprises, brands and registered trademarks. The mentioned companies have no relation to AVRA and are therefore not responslible for any problems that occur when using AVRA. Many thanks for your products, support and efforts.

AVRA is an assembler for Atmel AVR microcontrollers, and it is almost compatible with Atmel’s own assembler AVRASM32. The programming principles and conceptions are based on the ANSI programming language "C".

The initial version of AVRA was written by John Anders Haugum. He released all versions until v0.7. All later versions were released by Tobias Weber.

There are some differences between the original Atmel assembler AVRASM32 and AVRA. Basically AVRA is designed to replace AVRASM32 without special changes in your current Atmel AVR Studio enviroment. Command line options have been adapted as far as it was possible until now. Jumping to fault containing line directly by double-clicking on the error message in the output window does work as with AVRASM32.

Since AVRA is written in ANSI C, it should be possible to compile it on most system platforms. If you have problems compiling AVRA, please leave a message on the sourceforge message board or send a mail to the authors of AVRA.

To install avra you should copy the avra-executable to an apropriate location. To compile you should rename the appropriate makefile, and perform a make (use smake for Amiga SAS/C, and nmake for Mickeysoft visual c++).

To compile avra you need gcc and the automake utilities. These will create a ./configure script that evaluates your system enviroment. To get the AVRA executable, you have to issue the following commands:

aclocal autoconf automake -a ./configure make && make install

avra can be copied any apropriate directory. If you are using the source distribution a make install will do the same.

If you received the Windows binary package, look into the \bin directory where you can find avra.exe. This should be copied to any apropriate location. You can also overwrite AVRASM32.EXE in your Atmel AVR Studio. If you want to compile it yourself you could download then OpenWatcom C/C++ Toolchain for windows and create a new project and add the C and H files to it and compile.

If you recieved the Apple OS X binary package, look into the bin directory this file is compiled universal and should run on intel 32 and 64 bit and powerpc. If you want to compile it yourself go to the src directory and invoke make -f makefiles/Makefiles.osx and then the executable should be created.

Since avra 1.1 there is a possibility to supress certain warnings. Currently only register reassignment warnings can be supressed.

Example: avra -W NoRegDef

AVRA offers a number of directives that are not part of Atmel’s assembler. These directives should help you creating versatile code that can be designed more modular.

To define a constant, use ".define". This does the same thing as ".equ", it is just a little more C style. Keep in mind that AVRA is not case sensitive. Do not mix ".def" and ".define", because ".def" is used to assign registers only. This is due to backward compatibility to Atmel’s AVRASM32. Here is an example on how .define can be used.

Now "network" is set to the value 1. You may want to assemble a specific part of your code depeding on a define or switch setting. You can test your defined word on existence (.ifdef and .ifndef) as well as on the value it represents. The following code shows a way to prevent error messages due to testing undefined constants. Conditional directives must always end with an .endif directive.

The three lines in the last example set the default value of "network". In the next example, you see how we can use default values. If a constant has not defined previously, it is set to zero. Now you can test wether e.g. network support is included into the assemby process.

In the second part of the above listing you see the use of .else, which defines the part of the condition that is being executed if the equation of the preceding .if statement is not equal. You can also use the else statement to test another equasion. For that purpose use .elif, which means "else if". Always close this conditional part with ".endif"

This directive can be used to throw errors if a part in the code has reached that should not be reached. The following example shows how we can stop the assembly process if a particular value has not been previously set.

The ouput to the list file can be paused by this two directives. After avra discovers a .nolist while assembling, it stops output to the list file. After a .list directive is detected, it continues the normal list file output.

By default, any file that is included from within the source file must either be a single filename or a complete absolute path. With the directive .includepath you can set an additional include path . Furthermore you can set as many include paths as you want. Be sure not no use same filename in separate includes, because then it is no longer clear which one avra should take.

To avoid multiple inclusions of include files, you may use some pre- processor directives. See example file stack.asm that is being included into the main programm file as well as in other include files.

If you like to implement compiler build time and date into your program, you can make use of some sepcial tags that avra supports.

This line will then assembled by avra into:

You may also create a self defined serial number with meta tags:

The %TAG% is translated before any other parsing happens. The real output can be found in the list file.

Sometimes you have to work with 16 bit or greater variables stored in 8 bit registers. The enhanced macro support allows you to write short and flexible macros that simplify access to big variables. The extended mode is active, as soon as you use parenthesis like this "[ ]" to wrap macro parameters.

Values representing more than 8 Bits are usualy kept in a set of byte wide registers. To simplify 16 Bit or greater operations, I added a new language definitions. Words can be written as r16:r17, whereas register r16 contains the higher part and register r17 the lower part of this 16 Bit value.

There are 3 data types that can be used. They will be added as character separated by one underline character.

16 Bit Source and Destionation registers dst and src

Within the parenthesis, the two words src and dst are interpreted as YH:YL and ZH:ZL. Normal code outside of the macro parameter parenthesis can still make use of these special key words "src" and "dst".

To simplify the parameters in the demonstration below, we need to redefine some registers.

If we substract 16 Bit values stored in a, higher byte and b, lower byte with that in c:d, we usually have to use the following command sequence:

Now we can do the following steps to simplify 16 or more Bit manipulations

Like in you are used to C functions, you can write macros for different parameter lists. If you would like to have a versatile macro, you can specify a unique macro for each parameter situation. See the next sample.

One problem you may have experienced, is that labels defined within macros are defined twice if you call the macro for example two times. Now you can use labels for macro loops. Loops within macros must end with _%. the "%" symbol is replaced by a running number.

Some errors and warnings may confuse you a little bit so we will try to clear some frequently asked questions about such cases.

This warning occurs if a value exceeds the byte or word value of a assignment. Read the comment posted by Jim Galbraith:

The expression (~0x80) is a Bitwise Not operation. This operator returns the input expression with all its bits inverted. If 0x80 represents -128, then 0x7f, or +127 should be ok. If this is considered as a 32-bit expression (AVRA internal representation), then it appears to be more like oxffffffff-0x80 or 0xffffffff^{0x80. The result would then be 0xffffff7f. The assembler would then have to be told or it would have to decide, based on context, how much significance to assign to the higher bits. I have also encountered such conditions with various assemblers, including AVRA. To make sure the assembler does what I really want, I use a construct like 0xff-0x80 or 0xff}0x80. This way the bit significance cannot extend beyond bit-7 and there cannot be any misunderstanding.

The reason why using it within data segment is forbidden is, that you cannot set ram content at assembly time. The values must be programmed into ROM area and at boot read from ROM into RAM. This is up to the user code. You can only allocate memory for your variables using labels and the .byte directive.

This directive cannot be used in code or eeprom region because this only allocates memory without assgning distinct values to it. Please use .db or .dw instead.

If you get an "Internal assembler error" please contact the project maintainer by sending him a code example and a description of your working enviroment.

This section provides thoughts of the avra internal design. I have to admit that the code of avra is anything else than clean and optimized. To increase the code readability I will try to give you some standards that should improve quality. The following standards are similar to what GNU proposes.

Tab space is always 2 spaces. The Tab character (ascii 9) is not used. if,while,for are always opened on the same line but closed on the next line. The closing bracket is in the same column as the first letter of the loop directive.

We would like to thank the following people for giving contributions, patches and bug reports, as well as suggestions and new ideas.

And all the anonymous people who submitted patches!

Thank you for your work and support.