Skip to content

Latest commit

 

History

History
135 lines (108 loc) · 5.17 KB

Insight-Embedding.md

File metadata and controls

135 lines (108 loc) · 5.17 KB
layout toc_group link_title permalink
docs-experimental
insight
Embedding Insight into Applications
/tools/graalvm-insight/embedding/

Embedding Insight into Applications

Embedding Insight into Java

GraalVM languages (languages implemented with the Truffle framework, i.e., JavaScript, Python, Ruby, R) can be embedded into custom Java applications via Polyglot Context API. GraalVM Insight can also be controlled via the same API. For example:

final Engine engine = context.getEngine();
Instrument instrument = engine.getInstruments().get("insight");
Function<Source, AutoCloseable> access = instrument.lookup(Function.class);
AutoCloseable handle = access.apply(agentSrc);

Obtain Engine for Context and ask for the insight instrument. Then create Source with the GraalVM Insight script and apply it while obtaining its instrumentation handle. Use handle.close() to disable all the script's instrumentations when when no longer needed.

Ignoring Internal Scripts

Often one wants to treat certain code written in a dynamic language as a priviledged one. Imagine various bindings to OS concepts or other features of one's application. Such scripts are better to remain blackboxed and hidden from GraalVM Insight instrumentation capabilities.

To hide priviledged scripts from sight, mark them as internal. By default GraalVM Insight ignores and does not process internal scripts.

Extending Functionality of Insight Scripts

When embedding GraalVM Insight into a Java application, you can make additional objects available to the Insight scripts being evaluated. For example:

@TruffleInstrument.Registration(
    id = "meaningOfWorld", name = "Meaning Of World", version = "demo",
    services = { Insight.SymbolProvider.class }
)
public final class MeaningOfWorldInstrument extends TruffleInstrument {
    @Override
    protected void onCreate(Env env) {
        Map<String, Integer> symbols = Collections.singletonMap("meaning", 42);
        Insight.SymbolProvider provider = () -> symbols;
        env.registerService(provider);
    }
}

The previous Java code creates an instrument which registers a new symbol meaning to every Insight script evaluated. Each script can then reference it and use it, for example, to limit the number of method invocations:

insight.on('enter', (ctx, frames) => { if (--meaning <= 0) throw 'Stop!' }, { roots : true });

It is possible to expose simple values, as well as complex objects. See the javadoc for more detailed information. Note that instrumentation can alter many aspects of program execution and are not subject to any security sandbox.

Embedding Insight into Node.js

The Insight Manual shows many examples of using GraalVM Insight with node. However most of them rely on the command line option --insight and do not benefit from the dynamic nature of the tool. The next example shows how to create an admin server.

Save this code to adminserver.js:

function initialize(insight, require) {
    const http = require("http");
    const srv = http.createServer((req, res) => {
        let method = req.method;
        if (method === 'POST') {
            var data = '';
            req.on('data', (chunk) => {
                data += chunk.toString();
            });
            req.on('end', () => {
                const fn = new Function('insight', data);
                try {
                    fn(insight);
                    res.write('GraalVM Insight hook activated\n');
                } finally {
                    res.end();
                }
            });
        }
    });
    srv.listen(9999, () => console.log("Admin ready at 9999"));
}


let waitForRequire = function (event) {
  if (typeof process === 'object' && process.mainModule && process.mainModule.require) {
    insight.off('source', waitForRequire);
    initialize(insight, process.mainModule.require.bind(process.mainModule));
  }
};

insight.on('source', waitForRequire, { roots: true });

The program opens an HTTP server at port 9999 and listens for incoming scripts to be applied any time later. Invoke the application:

node --insight=adminserver.js yourapp.js
Admin ready at 9999

While it is running, connect to the admin port. Send in any GraalVM Insight script to it. For example, the following script is going to observe who calls process.exit:

curl --data \
'insight.on("enter", (ctx, frame) => { console.log(new Error("call to exit").stack); }, \
{ roots: true, rootNameFilter: "exit" });' \
-X POST http://localhost:9999/

When writing your own adminserver.js, pay attention to security. Only an authorized person should apply arbitrary hooks to your application. Do not open the admin server port to everybody.

What to Read Rext

To learn more about Insight and find some usecases, go to the Insight Manual. It starts with an obligatory HelloWorld example and then demonstrates more challenging tasks.