/spinja

Automatically exported from code.google.com/p/spinja

Primary LanguageJava

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"
"http://www.w3.org/TR/html4/loose.dtd">

<html>
<head>
    <style type="text/css">
    body {font-family: "lucida grande", lucida, verdana, sans-serif;}
    body {margin: 20px 50px 0px 50px;}
    body {background: #f5f5f5;}
    body {line-height: 140%;}
    a, a:link, a:visited {text-decoration: none; color: #0082ae;}
    a:hover {border-color: #0082ae; border-width: 0 0 1px 0; border-style: solid;}
    h1, h2, h3 {color: #A00; padding-top:15px;}
    h1, h2 {border-bottom: 2px solid #a1a5a9; padding-bottom: 6px;}
    h3 {font-style:italic; font-size:100%;}
    code, pre {font-size: 110%;}
    table {border-collapse: collapse; }
    td {border:1px solid #a1a5a9; padding:10px; font-size: 90%; vertical-align: top;}
    span.lastchanged {color: #AAA; font-size: 20%;}
    </style>

    <meta http-equiv="content-type" content="text/html; charset=iso-8859-1">
    <meta name="author" content="Theo Ruys">
    <meta name="keywords" content="SpinJa, model checking, Promela, SPIN">
    <meta name="description" content="readme documentation for SpinJa">
    <title>SpinJa - README.html documentation</title>
</head>

<body>

<h1>SpinJa</h1>

<a href="">SpinJa</a> is a model checker for Promela, written 
in Java. Promela is the modelling language for the 
<a href="http://spinroot.com">SPIN</a> model checker.
SpinJa supports a large subset of the Promela language.

<p>
SpinJa can be used to check for the absence of deadlocks, assertions, 
liveness properties and LTL properties (via never claims).
SpinJa verification mode can exploit (nested) depth first search or
breadth first search. Bitstate hashing and hash compaction modes 
are also supported. Furthermore, SPIN's partial order reduction 
and statement merging are implemented. 
SpinJa can also be used for simulation: random, interactive or guided.

<p>
SpinJa is designed to behave similarly to SPIN, but to be more 
easily extendible and reusable. From the start we have committed 
ourselves to Java and a clean object-oriented approach. 
Despite the fact that SpinJa uses a layered object-oriented design 
and is written in Java, SpinJa is not slow: benchmark experiments have 
shown that on average it is about 3 times slower than the 
highly optimized SPIN which uses C as implementation language.

<p>
SpinJa has been developed within the <a href="http://fmt.cs.utwente.nl/">
Formal Methods and Tools</a> (FMT) group of the 
<a href="http://www.utwente.nl">University of Twente</a>.<br>
SpinJa is now an open-source project at 
<a href="http://code.google.com/p/spinja/">
http://code.google.com/p/spinja/</a>.<br>
SpinJa is released under the
<a href="http://www.apache.org/licenses/LICENSE-2.0.html">
Apache 2.0 license</a>. 

<h2>Files</h2>

The <em>binary</em> distribution of SpinJa (<code>spinja-0.9-bin.zip</code>)
contains the following files:

<ul>
    <li><code>README.html</code>: this file</li>
    <li><code>spinja.jar</code>: SpinJa binary and library</li>
    <li><code>spinja.sh</code>: SpinJa shell script</li>
    <li><code>doc/CHANGES.txt</code>: revision history for SpinJa</li>
    <li><code>doc/LICENSE.txt</code>: Apache 2.0 license</li>
    <li><code>tests/</code>: directory with Promela examples</li>
</ul>

The <em>source</em> distribution of SpinJa (<code>spinja-0.9-src.zip</code>)
contains two additional directories:
<ul>
    <li><code>lib/</code>: additional Java library: the 
                           <a href="https://javacc.dev.java.net/">JavaCC</a> 
                           compiler generator</li>
    <li><code>src/</code>: complete Java source of SpinJa</li>
</ul>

<p>
SpinJa has been developed under Windows XP. <br>
SpinJa 0.9 has been compiled under Mac OS X 10.5.8 using Java
version 1.6.0_17.

<p>
As SpinJa is an open-source project at Google code, you can also use SVN to 
retrieve the latest version from <a href="http://code.google.com/p/spinja/">
http://code.google.com/p/spinja/</a>.

<h2>Compilation</h2>

<em>
<small>
If you have downloaded the binary distribution of SpinJa, you can
skip this section: the binary distribution already contains a complete
<code>spinja.jar</code> file.
</small>
</em>

<p>
To automatically build the <code>spinja.jar</code>, you need <code>ant</code>, 
the Java build tool.
<pre>
    cd ./src/
    ant build
</pre>
This will build <code>spinja.jar</code> into the <code>../build/</code> 
directory.

<p>
To generate the API documentation of SpinJa do the following:
<pre>
    cd ./src/
    ant javadocs
</pre>
This will build the Javadoc documentation of the SpinJa library into the
<code>../doc/api/</code> directory.

<h2>Installation</h2>

SpinJa requires Java version 1.5 or later.<br>
The only file that is needed to run SpinJa is <code>spinja.jar</code>.
This Java jar file contains the SpinJa Promela compiler and all
utility code that is needed by the generated checker.

<p>
In the following we assume that <code>spinja.jar</code> resides in 
the current working directory. It is easier, however, to have 
<code>spinja.jar</code> available in Java's <code>CLASSPATH</code> 
environment variable.

<h2>Running SpinJa</h2>

<p>
Like SPIN, verification with SpinJa is a three-step process: generation,
compilation and execution:
<ol>
    <li><em>Generation.</em>
        Use <code>spinja.jar</code> to generate a Java verification 
        program for the Promela model <code>sample.prom</code>:
        <pre>    java -cp spinja.jar spinja.Compile <i>[options]</i> sample.prom</pre>
        This generates the Java program
        <code>./spinja/PanModel.java</code>.<br>
        The optional <code><i>options</i></code> are SpinJa's 
        <em>compiler</em> options (see below).
    </li><br>
    <li><em>Compilation.</em>
        Use the <code>javac</code> compiler to compile the generated Java
        program to Java bytecode:
        <pre>    javac -cp spinja.jar:. spinja/PanModel.java</pre>
        This will generate several Java class files in 
        <code>./spinja/</code>. 
    </li><br>
    <li><em>Execution.</em>
        Run the compiled Java program which checks the original Promela model:
        <pre>    java -cp spinja.jar:. spinja.PanModel <i>[options]</i></pre>
        The optional <code><i>options</i></code> are SpinJa's 
        <em>model checking</em> options (see below).
    </li>
</ol>

After the verification, one can safely remove the generated
<code>./spinja/</code> directory.

<h3>Shell script</h3>

To ease these tedious steps, a shell script is provided, which automates 
the whole process: <code>spinja.sh</code>. Verification of
<code>sample.prom</code> is now straightforward:
<pre>    ./spinja.sh <i>[options]</i> sample.prom</pre>
The optional <code><i>options</i></code> will be passed to the 
SpinJa model checker. 

<h3>SpinJa compiler options</h3>

For the generation of the Java program with <code>spinja.Compile</code>, 
SpinJa supports the following options to tune the generation process:

<ul>
    <li><code>-n<i>text</i></code>:
        sets the name of the model to <code><i>text</i></code>
        (default: <code>Pan</code>). </li>
    <li><code>-o3</code>:
        disables statement merging</li>
    <li><code>-s<i>text</i></code>:
        sets the output directory for the sourcecode to 
        <code><i>text</i></code>
        (default: <code>spinja</code>)</li>
    <li><code>-v</code>:
        shows diagnostic information on the compilation process.</li>
</ul>

<h3>SpinJa model checker options</h3>

The generated SpinJa model checker supports several run-time options
to guide the verification or simulation of the original Promela model.

<p>
<b>Verification.</b>
In <em>verification</em> mode, SpinJa understands the following options
to steer the verification process.
<em>
<small>
Note that most options coincide with SPIN's C compiler options or 
pan's run-time options.
</small>
</em>

<ul>
    <li><code>-a</code>: 
        checks for acceptance cycles</li>
    <li><code>-A</code>: 
        assert statements will be ignored</li>
    <li><code>-b</code>: 
        exceeding the depth limit is considered an error</li>
    <li><code>-c<i>n</i></code>:
        stops the verification after N errors (default: 1); 
        when set to 0, SpinJa will not stop for any error</li>
    <li><code>-DARRAY</code>:
        uses a hash table that uses array lists instead of probing</li>
    <li><code>-DBFS</code>:
        uses a Breadth First Search algorithm</li>
    <li><code>-DBITSTATE</code>:
        uses bitstate hashing for storing states (approximate search)</li>
    <li><code>-DHC</code>:
        uses hash compaction for storing states (approximate search)</li>
    <li><code>-DNOREDUCE</code>:
        disables the partial order reduction algoritm</li>
    <li><code>-E</code>:
        ignore invalid end states (deadlocks)</li>
    <li><code>-k<i>n</i></code>:
        sets N bits per state when using bitstate hashing (default: 3)</li>
    <li><code>-m<i>n</i></code>:
        sets the maximum search depth (default: 10000)</li>
    <li><code>-N</code>:
        ignores any never claim (if present)</li>
    <li><code>-v</code>:
        prints the version number and exits</li>
    <li><code>-w<i>n</i></code>:
        sets the number of entries in the hash table to 2^N (default: 21)</li>
</ul>

<p>
<b>Simulation.</b>
Instead of a verification run, one can also use SpinJa's generated 
application for simulation. Like SPIN, SpinJa supports three 
simulation modes.

<ul>
    <li><code>-DGRANDOM</code>:
        uses a randomizer to guide the search</li>
    <li><code>-DGTRAIL</code>:
        uses the generated trail-file to guide the search</li>
    <li><code>-DGUSER</code>:
        uses user input to guide the search</li>
</ul>

<h3>Java options</h3>

To configure the Java VM which runs the SpinJa model checker, one can 
supply additional options to the <code>java</code> program.

The <code>spinja.sh</code> script, for example, currently uses 
the following Java options:
<ul>
    <li><code>-Xss16m</code>: sets thread stack size to 16Mb,</li>
    <li><code>-Xms256m</code>: sets the initial heap size to 256Mb,</li>
    <li><code>-Xmx1024m</code>: sets the maximum heap size to 1024Mb.</li>
</ul>
Note that without such settings for the Java VM, Spinja 
would not be able to verify very large Promela models.

<h2>Known issues</h2>

<h3>Promela</h3>

SpinJa does not yet support the full Promela language. <br>
Currently, the following aspects of Promela are <em>not</em> yet implemented:
<ul>
    <li><code>unless</code> statement</li>
    <li>user defined types (i.e., <code>typedef</code> definitions)</li>
    <li>communication: 
        sorted send (<code>!!</code>),
        random receive (<code>??</code>), 
        non-removing receive (<code>&lt;</code>..<code>&gt;</code>)</li>
    <li>labels within <code>d_step</code> and communication 
        within <code>d_step</code></li>
    <li>remote references: <code>@</code><i>labelname</i> and
        <code>:</code><i>varname</i></li>
    <li>special variables: <code>np_</code>, <code>pc_value</code>,
        <code>enabled</code></li>
    <li>Promela's C code extensions (e.g., <code>c_code</code>, 
        <code>c_expr</code>)</li>
</ul>

<h3>Bugs</h3>

<ul>
    <li>When the hash table gets full (i.e., the parameter for 
        <code>-w</code> was too low), the automatic resizing of the 
        hash table is not always working properly. 
        <em>Work around:</em> rerun the model checker with a 
        higher value for <code>-w</code>. </li>
</ul>

<h2>License</h2>

SpinJa is released under the
<a href="http://www.apache.org/licenses/LICENSE-2.0.html">
Apache 2.0 license</a>.

<h2>People</h2>

<ul>
    <li>Marc de Jonge: principal designer of SpinJa 0.8.</li>
    <li>Theo Ruys: supervisor of SpinJa project.</li>
</ul>

Communication about SpinJa is controlled through SpinJa's project at Google 
code: <a href="http://code.google.com/p/spinja/">http://code.google.com/p/spinja/</a>.

<h2>Further Reading</h2>

For general documentation on Promela and the model checker SPIN,
please consult the <a href="spinroot.com">SPIN</a> website, which
hosts a wealth of information on the subjects.

More specific information on the design and implementation of SpinJa 
can be found in the following two MSc theses.

<ul>
    <li> Marc de Jonge.<br>
        <a href="http://fmt.cs.utwente.nl/msc-completed/jonge-msc-thesis.pdf">
        The SpinJ Model Checker - A fast, extensible, object-oriented 
        model checker</a>.<br>
        MSc Thesis, University of Twente, The Netherlands, September 2008.<br>
        <em>This MSc thesis describes the design and implementation
        of SpinJa 0.8 in detail.</em>
    </li><br>
    <li>Mark Kattenbelt.<br>
        <a href="http://fmt.cs.utwente.nl/msc-completed/kattenbelt-msc-thesis.pdf">
        Towards an Explicit-State Model Checking Framework</a>.<br>
        MSc Thesis, University of Twente, The Netherlands, September 2006.<br>
        <em>This MSc this presents a layered, object-oriented design for
        explicit-state model checkers. SpinJa's design is based on 
        this approach.</em>
    </li>
</ul>

<h2>Version history</h2>

<table>
<tr><td>0.9</td>   
    <td>2010.04.10</td>     
    <td>First public release of SpinJa (binary and <em>source</em>).</td>
</tr>
<tr><td>0.8</td>   
    <td>2008.09.10</td>     
    <td>Initial version of SpinJa after MSc Project of Marc de Jonge.</td>
</tr>
</table>

<br><p>
<span class="lastchanged">
This page has last been updated by Theo Ruys on Saturday, 10 April 2010.
</span>

</body>
</html>