kolibrios-fun/contrib/toolchain/avra/doc/README.html
Serhii Sakhno 7c539d8e19 upload avra
git-svn-id: svn://kolibrios.org@6400 a494cfbc-eb01-0410-851d-a64ba20cac60
2016-04-13 17:05:47 +00:00

760 lines
31 KiB
HTML
Raw Blame History

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN"
"http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
<meta name="generator" content="AsciiDoc 8.5.3" />
<link rel="stylesheet" href="./xhtml11.css" type="text/css" />
<link rel="stylesheet" href="./xhtml11-quirks.css" type="text/css" />
<link rel="stylesheet" href="./layout1.css" type="text/css" />
<script type="text/javascript">
/*<![CDATA[*/
window.onload = function(){asciidoc.footnotes();}
/*]]>*/
</script>
<script type="text/javascript" src="./asciidoc-xhtml11.js"></script>
<title>README</title>
</head>
<body>
<div id="layout-banner">
<div id="layout-title">AVRA</div>
<div id="layout-description">Assember for the Atmel AVR microcontroller family</div>
</div>
<table>
<tr valign="top">
<td id="layout-menu">
<div>&#187;<a href="index.html">Home</a></div>
<div>&#187;<a href="README.html">Readme</a></div>
<div>&#187;<a href="downloads.html">Downloads</a></div>
<div>&#187;<a href="ChangeLog.html">ChangeLog</a></div>
<div>&#187;<a href="http://sourceforge.net/projects/avra">Sourceforge</a></div>
<div id="page-source">&#187;<a href="README.txt">Page&nbsp;Source</a></div>
</td>
<td>
<div id="layout-content">
<div id="header">
<h1>README</h1>
</div>
<div id="content">
<h2 id="_general_and_licensing_information">General and licensing information</h2>
<div class="sectionbody">
<div class="paragraph"><p>AVRA v1.3.0 - Assember for the Atmel AVR microcontroller family</p></div>
<h3 id="_licensing_information">Licensing information</h3><div style="clear:left"></div>
<div class="paragraph"><p>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.</p></div>
<h3 id="_disclaimer">Disclaimer</h3><div style="clear:left"></div>
<div class="paragraph"><p>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.</p></div>
<h3 id="_gnu_general_public_license">GNU General Public License</h3><div style="clear:left"></div>
<div class="paragraph"><p>You should have received a copy of the GNU General Public License
along with this program; see the file "COPYING". If not, visit
<a href="http://www.gnu.org">http://www.gnu.org</a> 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.</p></div>
<h3 id="_trademarks_and_copyright">Trademarks and copyright</h3><div style="clear:left"></div>
<div class="paragraph"><p>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.</p></div>
</div>
<h2 id="_introducion">Introducion</h2>
<div class="sectionbody">
<div class="paragraph"><p>AVRA is an assembler for Atmel AVR microcontrollers, and it is almost
compatible with Atmel&#8217;s own assembler AVRASM32. The programming
principles and conceptions are based on the ANSI programming language "C".</p></div>
<div class="paragraph"><p>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.</p></div>
<h3 id="_differences_between_avra_and_avrasm32">Differences between AVRA and AVRASM32</h3><div style="clear:left"></div>
<div class="paragraph"><p>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.</p></div>
<h3 id="_the_differences_in_detail">The differences in detail</h3><div style="clear:left"></div>
<div class="ulist"><ul>
<li>
<p>
Support for some extra preprocessor directives.
</p>
<div class="literalblock">
<div class="content">
<pre><tt>.define, .undef, .ifdef, .ifndef, .if, .else, .endif, .elif, .elseif, .warning</tt></pre>
</div></div>
</li>
<li>
<p>
Not all command line options are supported.
Specifying an eeprom file (-e) is not supported. All eeprom data is
put out into a file called program.eep.hex and always Intel hex
format. Other hex file formats than Intel are currently not supported.
</p>
</li>
<li>
<p>
Forward references not supported for .ifdef and .ifndef directives.
This makes sure, that directives like .ifdef and .undef are working
properly. If you are familiar with the C programming language, you
should get easily into AVRA. See chapter "Programming techniques" for
more information about how to write proper code.
</p>
</li>
<li>
<p>
Enhanced macro support
AVRA has some new features for writing flexible macros. This should
increase the ability to reuse code e.g. build your own library.
</p>
</li>
<li>
<p>
Debugging support
AVRA creates a coff file everytime the assembly was sucessful. This
file allows AVR Studio or any coff compatible debugger to simulate
or emulate the program.
</p>
</li>
<li>
<p>
Meta tags for assembly time
This helps you tracking versions of your software and can also be
used to generate customer specific serial numbers.
</p>
</li>
</ul></div>
<h3 id="_compatibility">Compatibility</h3><div style="clear:left"></div>
<div class="paragraph"><p>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.</p></div>
</div>
<h2 id="_installation">Installation</h2>
<div class="sectionbody">
<div class="paragraph"><p>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++).</p></div>
<h3 id="_linux">Linux</h3><div style="clear:left"></div>
<div class="paragraph"><p>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:</p></div>
<div class="paragraph"><p>aclocal
autoconf
automake -a
./configure
make &amp;&amp; make install</p></div>
<h3 id="_amigaos">AmigaOS</h3><div style="clear:left"></div>
<div class="paragraph"><p>avra can be copied any apropriate directory. If you are using the source
distribution a <em>make install</em> will do the same.</p></div>
<h3 id="_microsoft_windows">Microsoft Windows</h3><div style="clear:left"></div>
<div class="paragraph"><p>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.</p></div>
<h3 id="_apple_os_x">Apple OS X</h3><div style="clear:left"></div>
<div class="paragraph"><p>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 <tt>make -f
makefiles/Makefiles.osx</tt> and then the executable should be created.</p></div>
</div>
<h2 id="_synopsis">Synopsis</h2>
<div class="sectionbody">
<h3 id="_command_line_usage">Command line usage</h3><div style="clear:left"></div>
<div class="literalblock">
<div class="content">
<pre><tt>usage: AVRA [-f][O|M|I|G] output file type
[-o &lt;filename&gt;] output file name
[-l &lt;filename&gt;] generate list file
[-m &lt;mapfile&gt;] generate map file
[--define &lt;symbol&gt;[=&lt;value&gt;]] [--includedir &lt;dir&gt;] [--listmac]
[--max_errors &lt;number&gt;] [--devices] [--version]
[-h] [--help] general help
[-W NoRegDef] supress register redefinition warnings
&lt;file to assemble&gt;</tt></pre>
</div></div>
<h3 id="_parameter_list">Parameter list</h3><div style="clear:left"></div>
<div class="literalblock">
<div class="content">
<pre><tt>--listfile -l : Create list file
--mapfile -m : Create map file
--define -D : Define symbol.
--includedir -I : Additional include dirs.
--listmac : List macro expansion in listfile.
--max_errors : Maximum number of errors before exit
(default: 10)
--devices : List out supported devices.
--version : Version information.
--help, -h : This help text.</tt></pre>
</div></div>
<h3 id="_warning_supression">Warning supression</h3><div style="clear:left"></div>
<div class="paragraph"><p>Since avra 1.1 there is a possibility to supress certain warnings.
Currently only register reassignment warnings can be supressed.</p></div>
<div class="paragraph"><p>Example: avra -W NoRegDef</p></div>
</div>
<h2 id="_programming_techniques">Programming techniques</h2>
<div class="sectionbody">
<h3 id="_using_directives">Using directives</h3><div style="clear:left"></div>
<div class="paragraph"><p>AVRA offers a number of directives that are not part of Atmel&#8217;s
assembler. These directives should help you creating versatile code that
can be designed more modular.</p></div>
<h4 id="_directive_define">Directive .define</h4>
<div class="paragraph"><p>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&#8217;s
AVRASM32. Here is an example on how .define can be used.</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>.define network 1</tt></pre>
</div></div>
<div class="paragraph"><p>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.</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>.ifndef network
.define network 0
.endif</tt></pre>
</div></div>
<h4 id="_directive_if_and_else">Directive .if and .else</h4>
<div class="paragraph"><p>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.</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>.if network = 1
.include "include\tcpip.asm"
.else
.include "include\dummynet.asm"
.endif</tt></pre>
</div></div>
<div class="paragraph"><p>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"</p></div>
<h4 id="_directive_error">Directive .error</h4>
<div class="paragraph"><p>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.</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>.ifndef network
.error "network is not configured!" ;the assembler stops here</tt></pre>
</div></div>
<h4 id="_directive_nolist_and_list">Directive .nolist and .list</h4>
<div class="paragraph"><p>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.</p></div>
<h4 id="_directive_includepath">Directive .includepath</h4>
<div class="paragraph"><p>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.</p></div>
</div>
<h2 id="_using_include_files">Using include files</h2>
<div class="sectionbody">
<div class="paragraph"><p>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.</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>.ifndef _STACK_ASM_
.define _STACK_ASM_</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.include "include/config.inc"</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>; *** stack macro ***</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.dseg
m_stack: .byte __stack_size__
.cseg</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.macro stack_setup
load [v:w,m_stack + __stack_size__]
outp [SPREG,v:w]
.endm</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.endif ; avoid multiple inclusion of stack.asm</tt></pre>
</div></div>
<h3 id="_using_build_date_meta_tags">Using build date meta tags</h3><div style="clear:left"></div>
<div class="paragraph"><p>If you like to implement compiler build time and date into your
program, you can make use of some sepcial tags that avra supports.</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>%MINUTE% is being replaced by the current minute (00-59)
%HOUR% is being replaced by the current hour (00-23)
%DAY% is being replaced by the current day of month (01-31)
%MONTH% is being replaced by the current month (01-12)
%YEAR% is being replaced by the current year (2004-9999)</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>buildtime: .db "Release date %DAY%.%MONTH%.%YEAR% %HOUR%:%MINUTE%"</tt></pre>
</div></div>
<div class="paragraph"><p>This line will then assembled by avra into:</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>buildtime: .db "Release date 10.05.2004 19:54"</tt></pre>
</div></div>
<div class="paragraph"><p>You may also create a self defined serial number with meta tags:</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>.define serialnumber %DAY% + %MONTH%*31 + (%YEAR% - 2000) *31*12</tt></pre>
</div></div>
<div class="paragraph"><p>The %TAG% is translated before any other parsing happens. The real
output can be found in the list file.</p></div>
</div>
<h2 id="_macro_features">Macro features</h2>
<div class="sectionbody">
<div class="paragraph"><p>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.</p></div>
<h3 id="_auto_type_conversion_for_macros">Auto type conversion for macros</h3><div style="clear:left"></div>
<div class="paragraph"><p>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.</p></div>
<h4 id="_macro_data_types">Macro data types</h4>
<div class="paragraph"><p>There are 3 data types that can be used. They will be added as character
separated by one underline character.</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>immediate values _i
registers _8,_16,_24,_32,_40,_48,_56,_64
void parameter _v</tt></pre>
</div></div>
<div class="paragraph"><p>16 Bit Source and Destionation registers <em>dst</em> and <em>src</em></p></div>
<div class="literalblock">
<div class="content">
<pre><tt>src = YH:YL
dst = ZH:ZL</tt></pre>
</div></div>
<div class="paragraph"><p>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".</p></div>
<h4 id="_examples_for_automatic_type_conversion">Examples for automatic type conversion</h4>
<div class="paragraph"><p>To simplify the parameters in the demonstration below, we need to
redefine some registers.</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>.def a = r16 ; general purpose registers
.def b = r17
.def c = r18
.def d = r19</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.def w = r20 ; working register
.def v = r21 ; working register</tt></pre>
</div></div>
<div class="paragraph"><p>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:</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>sub b,d
sbc a,c</tt></pre>
</div></div>
<div class="paragraph"><p>Now we can do the following steps to simplify 16 or more Bit manipulations</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>.macro subs
.message "no parameters specified"
.endm</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.macro subs_16_16
sub @1,@3
sbc @0,@2
.endm</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.macro subs_16_8
sub @1,@2
sbci @0,0
.endm</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>;now we can write a 16 Bit subraction as:</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>subs [a:b,c:d]</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>;or for calculating 16 minus 8 Bit</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>subs [a:b,c]</tt></pre>
</div></div>
<h3 id="_overloading_macros">Overloading macros</h3><div style="clear:left"></div>
<div class="paragraph"><p>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.</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>.macro load</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>; this message is shown if you use the macro within your code
; specifying no parameters. If your macro allows the case where
; no parameters are given, exchange .message with your code.</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.message "no parameters specified"
.endm</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>; Here we define the macro "load" for the case it is being used
; with two registers as first parameter and a immediate (constant)
; value as second parameter.</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.macro load_16_i
ldi @0,high(@2)
ldi @1,low(@2)
.endm</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>; the same case, but now with a 32 bit register value as first
; parameter</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.macro load_32_i
ldi @0,BYTE4(@4)
ldi @1,BYTE3(@4)
ldi @2,high(@4)
ldi @3,low(@4)
.endm</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>; Now let's see how these macros are being used in the code</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>load [a:b,15] ;uses macro load_16_i to load immediate</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>load [a:b:c:d,15] ;uses macro load_32_i to load immediate</tt></pre>
</div></div>
<h3 id="_more_examples">More examples</h3><div style="clear:left"></div>
<div class="literalblock">
<div class="content">
<pre><tt>.dseg
counter .byte 2
.cseg</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.macro poke
.message "no parameters"
.endm</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.macro poke_i_16_i
ldi @1,high(@3)
sts @0+0,@1
ldi @2,low(@3)
sts @0+1,@2
.endm</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.macro poke_i_i
ldi w,@1
sts @0+0,w
.endm</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.macro poke_i_v_i
ldi w,high(@3)
sts @0+0,w
ldi w,low(@3)
sts @0+1,w
.endm</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.macro poke_i_v_v_v_i
ldi w,high(@3)
sts @0+0,w
ldi w,low(@3)
sts @0+1,w
ldi w,BYTE3(@3)
sts @0+2,w
ldi w,BYTE4(@3)
sts @0+3,w
.endm</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>; this writes '9999' into the memory at 'counter'
; uses only the working register for transfering the values.</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>poke [counter,w:w,9999]</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>; works same as above, but the transferred value '9999' is also
; kept in the pair of register a:b</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>poke [counter,a:b,9999]</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>; in my design 'w' is always working reg. which implies that
; it cannot be used for normal variables. The following example
; uses poke_i_i because the parameter contains two immediate values.</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>poke [counter,9999] ;uses poke_i_i</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>; to be able to choose between a 8,16 or 32 Bit operation, you just
; add a void parameter.</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>poke [counter,,9999] ;uses poke_i_v_i</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>; and the same for 32 Bit pokes</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>poke [counter,,,,9999] ;uses poke_i_v_v_v_i</tt></pre>
</div></div>
<h3 id="_loops_within_macros">Loops within macros</h3><div style="clear:left"></div>
<div class="paragraph"><p>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 <em>_%</em>. the
"%" symbol is replaced by a running number.</p></div>
<h4 id="_loop_example">Loop example</h4>
<div class="literalblock">
<div class="content">
<pre><tt>; Definition of the macro</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>.macro write_8_8
write_%:
st Z+,@0
dec @1
brne write_%
.endm</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>; Use in user code</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>write [a,b]
write [c,d]</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>; After assembling this code, the result looks like this</tt></pre>
</div></div>
<div class="literalblock">
<div class="content">
<pre><tt>write_1:
st Z+,a
dec b
brne write_1
write_2:
st Z+,c
dec d
brne write_2</tt></pre>
</div></div>
</div>
<h2 id="_warnings_and_errors">Warnings and Errors</h2>
<div class="sectionbody">
<div class="paragraph"><p>Some errors and warnings may confuse you a little bit so we will try to
clear some frequently asked questions about such cases.</p></div>
<h3 id="_constant_out_of_range">Constant out of range</h3><div style="clear:left"></div>
<div class="paragraph"><p>This warning occurs if a value exceeds the byte or word value of a assignment.
Read the comment posted by Jim Galbraith:</p></div>
<div class="paragraph"><p>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<sup>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</sup>0x80.
This way the bit significance cannot extend beyond bit-7 and
there cannot be any misunderstanding.</p></div>
<h3 id="_can_8217_t_use_db_directive_in_data_segment">Can&#8217;t use .DB directive in data segment</h3><div style="clear:left"></div>
<div class="paragraph"><div class="title">DB and .DW is only used to assign constant data in eeprom or code space.</div><p>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.</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>.dseg
my_string: .byte 15</tt></pre>
</div></div>
<h3 id="_byte_directive">BYTE directive</h3><div style="clear:left"></div>
<div class="paragraph"><div class="title">BYTE directive can only be used in data segment (.DSEG)</div><p>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.</p></div>
<h3 id="_internal_assembler_error">Internal assembler error</h3><div style="clear:left"></div>
<div class="paragraph"><p>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.</p></div>
</div>
<h2 id="_avra_internals">AVRA internals</h2>
<div class="sectionbody">
<div class="paragraph"><p>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.</p></div>
<h3 id="_coding_standards">Coding standards</h3><div style="clear:left"></div>
<div class="paragraph"><p>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.</p></div>
<div class="literalblock">
<div class="content">
<pre><tt>Example:</tt></pre>
</div></div>
<div class="listingblock">
<div class="content">
<pre><tt> while(i &gt; 0) {
do_something();
}</tt></pre>
</div></div>
</div>
<h2 id="_credits">Credits</h2>
<div class="sectionbody">
<div class="paragraph"><p>We would like to thank the following people for giving contributions,
patches and bug reports, as well as suggestions and new ideas.</p></div>
<div class="listingblock">
<div class="content">
<pre><tt> Jon Anders Haugum (project founder)
Burkhard Arenfeld (release 1.2.0)
Tobias Weber (old maintainer)
Jerry Jacobs (release 1.3.0)
Bernt Hembre
Nils Str<74>m
Roberto Biancardi
Qwerty Jones
Ben Hitchcock (Maker of the mac port)
Daniel Drotos
Laurence Boyd II
Varuzhan Danielyan
Laurence Turner
Eugene R. O'Bryan
Dmitry Dicky
Bob Harris (Maker of coff support)
Tobias Weber (enhanced macro support)
Lesha Bogdanow
Jim Galbraith
Mark Brinicombe
Igor Nikolayenko
Peter Hettkamp
Herb Poppe
David Burke
Alexey Pavluchenko
Alan Probandt
Mariusz Matuszek
Arne Rossius
Marti Tichacek
Patrick Parity
Johannes Overmann
Roland Riegel
Peter Katzmann
Donald D. Davis</tt></pre>
</div></div>
<div class="paragraph"><p>And all the anonymous people who submitted patches!</p></div>
<div class="paragraph"><p>Thank you for your work and support.</p></div>
</div>
<h2 id="_references">References</h2>
<div class="sectionbody">
<div class="literalblock">
<div class="content">
<pre><tt>http://www.suprafluid.com/avra
http://www.avrfreaks.de
http://www.atmel.com</tt></pre>
</div></div>
</div>
</div>
<div id="footnotes"><hr /></div>
<div id="footer">
<div id="footer-text">
Version 1.3.0<br />
Last updated 2010-06-28 16:04:26 CEST
</div>
<div id="footer-badges">
<a href="http://validator.w3.org/check?uri=referer">
<img style="border:0;width:88px;height:31px"
src="http://www.w3.org/Icons/valid-xhtml11-blue"
alt="Valid XHTML 1.1" height="31" width="88" />
</a>
<a href="http://jigsaw.w3.org/css-validator/">
<img style="border:0;width:88px;height:31px"
src="http://jigsaw.w3.org/css-validator/images/vcss-blue"
alt="Valid CSS!" />
</a>
<a href="http://www.mozilla.org/products/firefox/">
<img style="border:none; width:110px; height:32px;"
src="http://www.spreadfirefox.com/community/images/affiliates/Buttons/110x32/safer.gif"
alt="Get Firefox!" />
</a>
</div>
</div>
</div>
</td>
</tr>
</table>
</body>
</html>