Instrumentation
大约 16 分钟魔法
Instrumentation
Instrumentation [ˌɪnstrəmenˈteɪʃn] ，位于java.lang.instrument包，提供了一种在程序运行时修改字节码的机制，通过该API，我们可以在类被加载之前或之后进行修改、增强或替换，这种机制被称为“字节码增强技术”，进而实现AOP编程、方法耗时统计、性能分析、异常分析等诸多应用。
1. 基本原理
Java Instrumentation 基于 Java Agent 技术实现，通过 Java Agent，我们可以在 Java 虚拟机启动时，向 JVM 中注入一个代理程序，该代理程序可以在应用程序运行时获取和修改 Java 类的信息。
Java Agent 的核心是 premain 方法，这个方法会在 JVM 启动时执行，我们可以在这个方法中获取 Instrumentation 对象，并注册 ClassFileTransformer。ClassFileTransformer 会在类被加载时被调用，可以在这里对类字节码进行修改。
2. 如何使用
- 编写一个实现
premain(String agentArgs, Instrumentation inst)的agent类
import java.lang.instrument.ClassFileTransformer;
import java.lang.instrument.Instrumentation;
import java.security.ProtectionDomain;
public class MyAgent {
public static void premain(String agentArgs, Instrumentation inst) {
inst.addTransformer(new MyClassFileTransformer());
}
static class MyClassFileTransformer implements ClassFileTransformer {
// 如果要进行字节码操作则需要引入依赖如bytebuddy,javassist,asm等
public byte[] transform(ClassLoader loader, String className, Class<?> classBeingRedefined,ProtectionDomain protectionDomain, byte[] classfileBuffer) {
// TODO: 修改字节码
return classfileBuffer;
}
}
}
- 在pom.xml中增加插件，并指定Premain-Class为MyAgent
<build>
<finalName>agent-hello</finalName>
<plugins>
<plugin>
<groupId>org.apache.maven.plugins</groupId>
<artifactId>maven-jar-plugin</artifactId>
<version>3.2.2</version>
<configuration>
<archive>
<manifestEntries>
<Premain-Class>com.example.MyAgent</Premain-Class>
</manifestEntries>
</archive>
</configuration>
</plugin>
<plugin>
<groupId>org.apache.maven.plugins</groupId>
<artifactId>maven-compiler-plugin</artifactId>
<configuration>
<source>8</source>
<target>8</target>
</configuration>
</plugin>
</plugins>
</build>
- 打成jar包
mvn clean package ，在target下得到myagent.jar
如果解压myagent.jar应该会在META-INF下发现MANIFEST.MF文件，内容如下（重点是Premain-Class）
Manifest-Version: 1.0
Premain-Class: com.example.MethodCounterAgent
Build-Jdk-Spec: 1.8
Created-By: Maven JAR Plugin 3.2.2
- 在启动目标应用程序是添加vm参数
java -javaagent:/path/to/myagent.jar -jar YourApp.jar
3. 应用场景
- 性能监控与优化：通过修改类字节码，可以在运行时对方法进行计时、统计方法调用次数、消除不必要的方法调用等，以优化程序的性能。
- 安全监控：通过修改类字节码，可以在运行时对代码进行加固、保护敏感信息等，以增强程序的安全性。
- 调试与排查问题：通过修改类字节码，可以在运行时添加调试信息、进行程序分析等，以便排查问题和诊断程序。
- 动态生成类和对象：通过 Instrumentation，我们可以在程序运行时动态地生成新的类和对象，从而实现更灵活的程序设计和实现。
4. 具体应用
在 Java 中，有很多类库使用了 agent 技术来实现自己的功能。其中比较知名的包括：
- SkyWalking：一个开源的应用程序性能监测工具，使用 agent 技术来收集和分析应用程序性能数据。
- Arthas：一个开源的 Java 诊断工具，也是基于 agent 技术实现的。
- JRebel：一个热部署工具，同样使用了 agent 技术来实现修改应用程序的类加载机制。
除此之外，还有很多类库使用了 agent 技术来实现自己的功能，例如日志工具、安全工具、代码覆盖率工具等等
例如在SkyWalking的 skywalking-agent.jar中MANIFEST.MF
Manifest-Version: 1.0
...
Premain-Class: org.apache.skywalking.apm.agent.SkyWalkingAgent
...
在Arthas的arthas-agent.jar中MANIFEST.MF(这里有Agent-Class)
Manifest-Version: 1.0
...
Premain-Class: com.taobao.arthas.agent334.AgentBootstrap
...
Agent-Class: com.taobao.arthas.agent334.AgentBootstrap
...
premain 和 agentmain
premain 方法是 Java Agent 中最常用的方法，它会在 Java 应用程序启动时被调用。在 premain 方法中，我们可以获取 Instrumentation 对象，并注册 ClassFileTransformer 对象，用于对类字节码进行转换
import java.lang.instrument.Instrumentation;
public class MyAgent {
public static void premain(String agentArgs, Instrumentation inst) {
// 在这里注册 ClassFileTransformer 对象，用于对类字节码进行转换
}
}
premain 方法的实现必须满足以下要求：
- 必须是 public static void 类型的方法。
- 方法名必须为 premain。
- 方法参数中的 Instrumentation 对象用于在程序运行时获取和修改类信息
agentmain 方法也是一种 Java Agent 注册方法，它可以在程序运行时动态注册 Java Agent，从而实现动态修改类字节码
import java.lang.instrument.Instrumentation;
public class MyAgent {
public static void agentmain(String agentArgs, Instrumentation inst) {
// 在这里注册 ClassFileTransformer 对象，用于对类字节码进行转换
}
}
必须满足以下要求：
- 必须是 public static void 类型的方法。
- 方法名必须为 agentmain。
- 方法参数中的 Instrumentation 对象用于在程序运行时获取和修改类信息。
- 方法必须在运行时使用 VirtualMachine.attach 方法调用
agentmain 方法的调用需要通过 VirtualMachine.attach 方法实现。VirtualMachine.attach 方法是在 jdk.tools.jar 包中定义的，因此需要在启动时指定 jdk.tools.jar 包。
premain 方法是在程序启动时被调用的，而 agentmain 方法是在程序运行时动态注册 Java Agent 的方法
附: Instrumentation源码(JDK1.8)
Instrumentation.java
package java.lang.instrument;
import java.io.File;
import java.io.IOException;
import java.util.jar.JarFile;
/*
* Copyright 2003 Wily Technology, Inc.
*/
/**
* This class provides services needed to instrument Java
* programming language code.
* Instrumentation is the addition of byte-codes to methods for the
* purpose of gathering data to be utilized by tools.
* Since the changes are purely additive, these tools do not modify
* application state or behavior.
* Examples of such benign tools include monitoring agents, profilers,
* coverage analyzers, and event loggers.
*
* <P>
* There are two ways to obtain an instance of the
* <code>Instrumentation</code> interface:
*
* <ol>
* <li><p> When a JVM is launched in a way that indicates an agent
* class. In that case an <code>Instrumentation</code> instance
* is passed to the <code>premain</code> method of the agent class.
* </p></li>
* <li><p> When a JVM provides a mechanism to start agents sometime
* after the JVM is launched. In that case an <code>Instrumentation</code>
* instance is passed to the <code>agentmain</code> method of the
* agent code. </p> </li>
* </ol>
* <p>
* These mechanisms are described in the
* {@linkplain java.lang.instrument package specification}.
* <p>
* Once an agent acquires an <code>Instrumentation</code> instance,
* the agent may call methods on the instance at any time.
*
* @since 1.5
*/
public interface Instrumentation {
/**
* Registers the supplied transformer. All future class definitions
* will be seen by the transformer, except definitions of classes upon which any
* registered transformer is dependent.
* The transformer is called when classes are loaded, when they are
* {@linkplain #redefineClasses redefined}. and if <code>canRetransform</code> is true,
* when they are {@linkplain #retransformClasses retransformed}.
* See {@link java.lang.instrument.ClassFileTransformer#transform
* ClassFileTransformer.transform} for the order
* of transform calls.
* If a transformer throws
* an exception during execution, the JVM will still call the other registered
* transformers in order. The same transformer may be added more than once,
* but it is strongly discouraged -- avoid this by creating a new instance of
* transformer class.
* <P>
* This method is intended for use in instrumentation, as described in the
* {@linkplain Instrumentation class specification}.
*
* @param transformer the transformer to register
* @param canRetransform can this transformer's transformations be retransformed
* @throws java.lang.NullPointerException if passed a <code>null</code> transformer
* @throws java.lang.UnsupportedOperationException if <code>canRetransform</code>
* is true and the current configuration of the JVM does not allow
* retransformation ({@link #isRetransformClassesSupported} is false)
* @since 1.6
*/
void
addTransformer(ClassFileTransformer transformer, boolean canRetransform);
/**
* Registers the supplied transformer.
* <P>
* Same as <code>addTransformer(transformer, false)</code>.
*
* @param transformer the transformer to register
* @throws java.lang.NullPointerException if passed a <code>null</code> transformer
* @see #addTransformer(ClassFileTransformer,boolean)
*/
void
addTransformer(ClassFileTransformer transformer);
/**
* Unregisters the supplied transformer. Future class definitions will
* not be shown to the transformer. Removes the most-recently-added matching
* instance of the transformer. Due to the multi-threaded nature of
* class loading, it is possible for a transformer to receive calls
* after it has been removed. Transformers should be written defensively
* to expect this situation.
*
* @param transformer the transformer to unregister
* @return true if the transformer was found and removed, false if the
* transformer was not found
* @throws java.lang.NullPointerException if passed a <code>null</code> transformer
*/
boolean
removeTransformer(ClassFileTransformer transformer);
/**
* Returns whether or not the current JVM configuration supports retransformation
* of classes.
* The ability to retransform an already loaded class is an optional capability
* of a JVM.
* Retransformation will only be supported if the
* <code>Can-Retransform-Classes</code> manifest attribute is set to
* <code>true</code> in the agent JAR file (as described in the
* {@linkplain java.lang.instrument package specification}) and the JVM supports
* this capability.
* During a single instantiation of a single JVM, multiple calls to this
* method will always return the same answer.
* @return true if the current JVM configuration supports retransformation of
* classes, false if not.
* @see #retransformClasses
* @since 1.6
*/
boolean
isRetransformClassesSupported();
/**
* Retransform the supplied set of classes.
*
* <P>
* This function facilitates the instrumentation
* of already loaded classes.
* When classes are initially loaded or when they are
* {@linkplain #redefineClasses redefined},
* the initial class file bytes can be transformed with the
* {@link java.lang.instrument.ClassFileTransformer ClassFileTransformer}.
* This function reruns the transformation process
* (whether or not a transformation has previously occurred).
* This retransformation follows these steps:
* <ul>
* <li>starting from the initial class file bytes
* </li>
* <li>for each transformer that was added with <code>canRetransform</code>
* false, the bytes returned by
* {@link java.lang.instrument.ClassFileTransformer#transform transform}
* during the last class load or redefine are
* reused as the output of the transformation; note that this is
* equivalent to reapplying the previous transformation, unaltered;
* except that
* {@link java.lang.instrument.ClassFileTransformer#transform transform}
* is not called
* </li>
* <li>for each transformer that was added with <code>canRetransform</code>
* true, the
* {@link java.lang.instrument.ClassFileTransformer#transform transform}
* method is called in these transformers
* </li>
* <li>the transformed class file bytes are installed as the new
* definition of the class
* </li>
* </ul>
* <P>
*
* The order of transformation is described in the
* {@link java.lang.instrument.ClassFileTransformer#transform transform} method.
* This same order is used in the automatic reapplication of retransformation
* incapable transforms.
* <P>
*
* The initial class file bytes represent the bytes passed to
* {@link java.lang.ClassLoader#defineClass ClassLoader.defineClass} or
* {@link #redefineClasses redefineClasses}
* (before any transformations
* were applied), however they might not exactly match them.
* The constant pool might not have the same layout or contents.
* The constant pool may have more or fewer entries.
* Constant pool entries may be in a different order; however,
* constant pool indices in the bytecodes of methods will correspond.
* Some attributes may not be present.
* Where order is not meaningful, for example the order of methods,
* order might not be preserved.
*
* <P>
* This method operates on
* a set in order to allow interdependent changes to more than one class at the same time
* (a retransformation of class A can require a retransformation of class B).
*
* <P>
* If a retransformed method has active stack frames, those active frames continue to
* run the bytecodes of the original method.
* The retransformed method will be used on new invokes.
*
* <P>
* This method does not cause any initialization except that which would occur
* under the customary JVM semantics. In other words, redefining a class
* does not cause its initializers to be run. The values of static variables
* will remain as they were prior to the call.
*
* <P>
* Instances of the retransformed class are not affected.
*
* <P>
* The retransformation may change method bodies, the constant pool and attributes.
* The retransformation must not add, remove or rename fields or methods, change the
* signatures of methods, or change inheritance. These restrictions maybe be
* lifted in future versions. The class file bytes are not checked, verified and installed
* until after the transformations have been applied, if the resultant bytes are in
* error this method will throw an exception.
*
* <P>
* If this method throws an exception, no classes have been retransformed.
* <P>
* This method is intended for use in instrumentation, as described in the
* {@linkplain Instrumentation class specification}.
*
* @param classes array of classes to retransform;
* a zero-length array is allowed, in this case, this method does nothing
* @throws java.lang.instrument.UnmodifiableClassException if a specified class cannot be modified
* ({@link #isModifiableClass} would return <code>false</code>)
* @throws java.lang.UnsupportedOperationException if the current configuration of the JVM does not allow
* retransformation ({@link #isRetransformClassesSupported} is false) or the retransformation attempted
* to make unsupported changes
* @throws java.lang.ClassFormatError if the data did not contain a valid class
* @throws java.lang.NoClassDefFoundError if the name in the class file is not equal to the name of the class
* @throws java.lang.UnsupportedClassVersionError if the class file version numbers are not supported
* @throws java.lang.ClassCircularityError if the new classes contain a circularity
* @throws java.lang.LinkageError if a linkage error occurs
* @throws java.lang.NullPointerException if the supplied classes array or any of its components
* is <code>null</code>.
*
* @see #isRetransformClassesSupported
* @see #addTransformer
* @see java.lang.instrument.ClassFileTransformer
* @since 1.6
*/
void
retransformClasses(Class<?>... classes) throws UnmodifiableClassException;
/**
* Returns whether or not the current JVM configuration supports redefinition
* of classes.
* The ability to redefine an already loaded class is an optional capability
* of a JVM.
* Redefinition will only be supported if the
* <code>Can-Redefine-Classes</code> manifest attribute is set to
* <code>true</code> in the agent JAR file (as described in the
* {@linkplain java.lang.instrument package specification}) and the JVM supports
* this capability.
* During a single instantiation of a single JVM, multiple calls to this
* method will always return the same answer.
* @return true if the current JVM configuration supports redefinition of classes,
* false if not.
* @see #redefineClasses
*/
boolean
isRedefineClassesSupported();
/**
* Redefine the supplied set of classes using the supplied class files.
*
* <P>
* This method is used to replace the definition of a class without reference
* to the existing class file bytes, as one might do when recompiling from source
* for fix-and-continue debugging.
* Where the existing class file bytes are to be transformed (for
* example in bytecode instrumentation)
* {@link #retransformClasses retransformClasses}
* should be used.
*
* <P>
* This method operates on
* a set in order to allow interdependent changes to more than one class at the same time
* (a redefinition of class A can require a redefinition of class B).
*
* <P>
* If a redefined method has active stack frames, those active frames continue to
* run the bytecodes of the original method.
* The redefined method will be used on new invokes.
*
* <P>
* This method does not cause any initialization except that which would occur
* under the customary JVM semantics. In other words, redefining a class
* does not cause its initializers to be run. The values of static variables
* will remain as they were prior to the call.
*
* <P>
* Instances of the redefined class are not affected.
*
* <P>
* The redefinition may change method bodies, the constant pool and attributes.
* The redefinition must not add, remove or rename fields or methods, change the
* signatures of methods, or change inheritance. These restrictions maybe be
* lifted in future versions. The class file bytes are not checked, verified and installed
* until after the transformations have been applied, if the resultant bytes are in
* error this method will throw an exception.
*
* <P>
* If this method throws an exception, no classes have been redefined.
* <P>
* This method is intended for use in instrumentation, as described in the
* {@linkplain Instrumentation class specification}.
*
* @param definitions array of classes to redefine with corresponding definitions;
* a zero-length array is allowed, in this case, this method does nothing
* @throws java.lang.instrument.UnmodifiableClassException if a specified class cannot be modified
* ({@link #isModifiableClass} would return <code>false</code>)
* @throws java.lang.UnsupportedOperationException if the current configuration of the JVM does not allow
* redefinition ({@link #isRedefineClassesSupported} is false) or the redefinition attempted
* to make unsupported changes
* @throws java.lang.ClassFormatError if the data did not contain a valid class
* @throws java.lang.NoClassDefFoundError if the name in the class file is not equal to the name of the class
* @throws java.lang.UnsupportedClassVersionError if the class file version numbers are not supported
* @throws java.lang.ClassCircularityError if the new classes contain a circularity
* @throws java.lang.LinkageError if a linkage error occurs
* @throws java.lang.NullPointerException if the supplied definitions array or any of its components
* is <code>null</code>
* @throws java.lang.ClassNotFoundException Can never be thrown (present for compatibility reasons only)
*
* @see #isRedefineClassesSupported
* @see #addTransformer
* @see java.lang.instrument.ClassFileTransformer
*/
void
redefineClasses(ClassDefinition... definitions)
throws ClassNotFoundException, UnmodifiableClassException;
/**
* Determines whether a class is modifiable by
* {@linkplain #retransformClasses retransformation}
* or {@linkplain #redefineClasses redefinition}.
* If a class is modifiable then this method returns <code>true</code>.
* If a class is not modifiable then this method returns <code>false</code>.
* <P>
* For a class to be retransformed, {@link #isRetransformClassesSupported} must also be true.
* But the value of <code>isRetransformClassesSupported()</code> does not influence the value
* returned by this function.
* For a class to be redefined, {@link #isRedefineClassesSupported} must also be true.
* But the value of <code>isRedefineClassesSupported()</code> does not influence the value
* returned by this function.
* <P>
* Primitive classes (for example, <code>java.lang.Integer.TYPE</code>)
* and array classes are never modifiable.
*
* @param theClass the class to check for being modifiable
* @return whether or not the argument class is modifiable
* @throws java.lang.NullPointerException if the specified class is <code>null</code>.
*
* @see #retransformClasses
* @see #isRetransformClassesSupported
* @see #redefineClasses
* @see #isRedefineClassesSupported
* @since 1.6
*/
boolean
isModifiableClass(Class<?> theClass);
/**
* Returns an array of all classes currently loaded by the JVM.
*
* @return an array containing all the classes loaded by the JVM, zero-length if there are none
*/
@SuppressWarnings("rawtypes")
Class[]
getAllLoadedClasses();
/**
* Returns an array of all classes for which <code>loader</code> is an initiating loader.
* If the supplied loader is <code>null</code>, classes initiated by the bootstrap class
* loader are returned.
*
* @param loader the loader whose initiated class list will be returned
* @return an array containing all the classes for which loader is an initiating loader,
* zero-length if there are none
*/
@SuppressWarnings("rawtypes")
Class[]
getInitiatedClasses(ClassLoader loader);
/**
* Returns an implementation-specific approximation of the amount of storage consumed by
* the specified object. The result may include some or all of the object's overhead,
* and thus is useful for comparison within an implementation but not between implementations.
*
* The estimate may change during a single invocation of the JVM.
*
* @param objectToSize the object to size
* @return an implementation-specific approximation of the amount of storage consumed by the specified object
* @throws java.lang.NullPointerException if the supplied Object is <code>null</code>.
*/
long
getObjectSize(Object objectToSize);
/**
* Specifies a JAR file with instrumentation classes to be defined by the
* bootstrap class loader.
*
* <p> When the virtual machine's built-in class loader, known as the "bootstrap
* class loader", unsuccessfully searches for a class, the entries in the {@link
* java.util.jar.JarFile JAR file} will be searched as well.
*
* <p> This method may be used multiple times to add multiple JAR files to be
* searched in the order that this method was invoked.
*
* <p> The agent should take care to ensure that the JAR does not contain any
* classes or resources other than those to be defined by the bootstrap
* class loader for the purpose of instrumentation.
* Failure to observe this warning could result in unexpected
* behavior that is difficult to diagnose. For example, suppose there is a
* loader L, and L's parent for delegation is the bootstrap class loader.
* Furthermore, a method in class C, a class defined by L, makes reference to
* a non-public accessor class C$1. If the JAR file contains a class C$1 then
* the delegation to the bootstrap class loader will cause C$1 to be defined
* by the bootstrap class loader. In this example an <code>IllegalAccessError</code>
* will be thrown that may cause the application to fail. One approach to
* avoiding these types of issues, is to use a unique package name for the
* instrumentation classes.
*
* <p>
* <cite>The Java™ Virtual Machine Specification</cite>
* specifies that a subsequent attempt to resolve a symbolic
* reference that the Java virtual machine has previously unsuccessfully attempted
* to resolve always fails with the same error that was thrown as a result of the
* initial resolution attempt. Consequently, if the JAR file contains an entry
* that corresponds to a class for which the Java virtual machine has
* unsuccessfully attempted to resolve a reference, then subsequent attempts to
* resolve that reference will fail with the same error as the initial attempt.
*
* @param jarfile
* The JAR file to be searched when the bootstrap class loader
* unsuccessfully searches for a class.
*
* @throws NullPointerException
* If <code>jarfile</code> is <code>null</code>.
*
* @see #appendToSystemClassLoaderSearch
* @see java.lang.ClassLoader
* @see java.util.jar.JarFile
*
* @since 1.6
*/
void
appendToBootstrapClassLoaderSearch(JarFile jarfile);
/**
* Specifies a JAR file with instrumentation classes to be defined by the
* system class loader.
*
* When the system class loader for delegation (see
* {@link java.lang.ClassLoader#getSystemClassLoader getSystemClassLoader()})
* unsuccessfully searches for a class, the entries in the {@link
* java.util.jar.JarFile JarFile} will be searched as well.
*
* <p> This method may be used multiple times to add multiple JAR files to be
* searched in the order that this method was invoked.
*
* <p> The agent should take care to ensure that the JAR does not contain any
* classes or resources other than those to be defined by the system class
* loader for the purpose of instrumentation.
* Failure to observe this warning could result in unexpected
* behavior that is difficult to diagnose (see
* {@link #appendToBootstrapClassLoaderSearch
* appendToBootstrapClassLoaderSearch}).
*
* <p> The system class loader supports adding a JAR file to be searched if
* it implements a method named <code>appendToClassPathForInstrumentation</code>
* which takes a single parameter of type <code>java.lang.String</code>. The
* method is not required to have <code>public</code> access. The name of
* the JAR file is obtained by invoking the {@link java.util.zip.ZipFile#getName
* getName()} method on the <code>jarfile</code> and this is provided as the
* parameter to the <code>appendToClassPathForInstrumentation</code> method.
*
* <p>
* <cite>The Java™ Virtual Machine Specification</cite>
* specifies that a subsequent attempt to resolve a symbolic
* reference that the Java virtual machine has previously unsuccessfully attempted
* to resolve always fails with the same error that was thrown as a result of the
* initial resolution attempt. Consequently, if the JAR file contains an entry
* that corresponds to a class for which the Java virtual machine has
* unsuccessfully attempted to resolve a reference, then subsequent attempts to
* resolve that reference will fail with the same error as the initial attempt.
*
* <p> This method does not change the value of <code>java.class.path</code>
* {@link java.lang.System#getProperties system property}.
*
* @param jarfile
* The JAR file to be searched when the system class loader
* unsuccessfully searches for a class.
*
* @throws UnsupportedOperationException
* If the system class loader does not support appending a
* a JAR file to be searched.
*
* @throws NullPointerException
* If <code>jarfile</code> is <code>null</code>.
*
* @see #appendToBootstrapClassLoaderSearch
* @see java.lang.ClassLoader#getSystemClassLoader
* @see java.util.jar.JarFile
* @since 1.6
*/
void
appendToSystemClassLoaderSearch(JarFile jarfile);
/**
* Returns whether the current JVM configuration supports
* {@linkplain #setNativeMethodPrefix(ClassFileTransformer,String)
* setting a native method prefix}.
* The ability to set a native method prefix is an optional
* capability of a JVM.
* Setting a native method prefix will only be supported if the
* <code>Can-Set-Native-Method-Prefix</code> manifest attribute is set to
* <code>true</code> in the agent JAR file (as described in the
* {@linkplain java.lang.instrument package specification}) and the JVM supports
* this capability.
* During a single instantiation of a single JVM, multiple
* calls to this method will always return the same answer.
* @return true if the current JVM configuration supports
* setting a native method prefix, false if not.
* @see #setNativeMethodPrefix
* @since 1.6
*/
boolean
isNativeMethodPrefixSupported();
/**
* This method modifies the failure handling of
* native method resolution by allowing retry
* with a prefix applied to the name.
* When used with the
* {@link java.lang.instrument.ClassFileTransformer ClassFileTransformer},
* it enables native methods to be
* instrumented.
* <p>
* Since native methods cannot be directly instrumented
* (they have no bytecodes), they must be wrapped with
* a non-native method which can be instrumented.
* For example, if we had:
* <pre>
* native boolean foo(int x);</pre>
* <p>
* We could transform the class file (with the
* ClassFileTransformer during the initial definition
* of the class) so that this becomes:
* <pre>
* boolean foo(int x) {
* <i>... record entry to foo ...</i>
* return wrapped_foo(x);
* }
*
* native boolean wrapped_foo(int x);</pre>
* <p>
* Where <code>foo</code> becomes a wrapper for the actual native
* method with the appended prefix "wrapped_". Note that
* "wrapped_" would be a poor choice of prefix since it
* might conceivably form the name of an existing method
* thus something like "$$$MyAgentWrapped$$$_" would be
* better but would make these examples less readable.
* <p>
* The wrapper will allow data to be collected on the native
* method call, but now the problem becomes linking up the
* wrapped method with the native implementation.
* That is, the method <code>wrapped_foo</code> needs to be
* resolved to the native implementation of <code>foo</code>,
* which might be:
* <pre>
* Java_somePackage_someClass_foo(JNIEnv* env, jint x)</pre>
* <p>
* This function allows the prefix to be specified and the
* proper resolution to occur.
* Specifically, when the standard resolution fails, the
* resolution is retried taking the prefix into consideration.
* There are two ways that resolution occurs, explicit
* resolution with the JNI function <code>RegisterNatives</code>
* and the normal automatic resolution. For
* <code>RegisterNatives</code>, the JVM will attempt this
* association:
* <pre>{@code
* method(foo) -> nativeImplementation(foo)
* }</pre>
* <p>
* When this fails, the resolution will be retried with
* the specified prefix prepended to the method name,
* yielding the correct resolution:
* <pre>{@code
* method(wrapped_foo) -> nativeImplementation(foo)
* }</pre>
* <p>
* For automatic resolution, the JVM will attempt:
* <pre>{@code
* method(wrapped_foo) -> nativeImplementation(wrapped_foo)
* }</pre>
* <p>
* When this fails, the resolution will be retried with
* the specified prefix deleted from the implementation name,
* yielding the correct resolution:
* <pre>{@code
* method(wrapped_foo) -> nativeImplementation(foo)
* }</pre>
* <p>
* Note that since the prefix is only used when standard
* resolution fails, native methods can be wrapped selectively.
* <p>
* Since each <code>ClassFileTransformer</code>
* can do its own transformation of the bytecodes, more
* than one layer of wrappers may be applied. Thus each
* transformer needs its own prefix. Since transformations
* are applied in order, the prefixes, if applied, will
* be applied in the same order
* (see {@link #addTransformer(ClassFileTransformer,boolean) addTransformer}).
* Thus if three transformers applied
* wrappers, <code>foo</code> might become
* <code>$trans3_$trans2_$trans1_foo</code>. But if, say,
* the second transformer did not apply a wrapper to
* <code>foo</code> it would be just
* <code>$trans3_$trans1_foo</code>. To be able to
* efficiently determine the sequence of prefixes,
* an intermediate prefix is only applied if its non-native
* wrapper exists. Thus, in the last example, even though
* <code>$trans1_foo</code> is not a native method, the
* <code>$trans1_</code> prefix is applied since
* <code>$trans1_foo</code> exists.
*
* @param transformer
* The ClassFileTransformer which wraps using this prefix.
* @param prefix
* The prefix to apply to wrapped native methods when
* retrying a failed native method resolution. If prefix
* is either <code>null</code> or the empty string, then
* failed native method resolutions are not retried for
* this transformer.
* @throws java.lang.NullPointerException if passed a <code>null</code> transformer.
* @throws java.lang.UnsupportedOperationException if the current configuration of
* the JVM does not allow setting a native method prefix
* ({@link #isNativeMethodPrefixSupported} is false).
* @throws java.lang.IllegalArgumentException if the transformer is not registered
* (see {@link #addTransformer(ClassFileTransformer,boolean) addTransformer}).
*
* @since 1.6
*/
void
setNativeMethodPrefix(ClassFileTransformer transformer, String prefix);
}