版权声明：非商业目的可自由转载，转载请标明出处 https://blog.csdn.net/u010013573/article/details/85731161
Channel的Unsafe对Buffer的使用
- 在Channel的接口体系设计中，Channel只是与ChannelHandler，ChannelHandlerContext，ChannelPipeline完成对数据的处理，而不直接参与和底层socket进行数据交互，是通过一个内部Unsafe的接口设计来完成这项工作。
- 如对于数据读取，在Channel的实现中拿到的是Unsafe已经读取回来的数据，Unsafe将socket的字节数据读取后转为了对象Object msg的数据，或者将ServerSocket接收到的客户端socket连接，将socket对象交给Channel。具体为交给Channel所绑定的ChannelPipeline，从而开始Channel体系结构的数据流处理。
- 故Buffer包中相关缓存实现类主要是在Channel的Unsafe接口的实现类中引用，读取底层Java NIO的socket的数据，如SocketChannel的bytes：
如下为AbstractNioByteChannel的NioByteUnsafe的read方法：
AbstractNioByteChannel是NioSocketChannel的基类，主要是读取客户端发送过来的字节数据
@Override
public final void read() {
final ChannelConfig config = config();
if (shouldBreakReadReady(config)) {
clearReadPending();
return;
}
final ChannelPipeline pipeline = pipeline();
// 获取allocator，allocator是创建ByteBuf对象
final ByteBufAllocator allocator = config.getAllocator();
final RecvByteBufAllocator.Handle allocHandle = recvBufAllocHandle();
allocHandle.reset(config);
ByteBuf byteBuf = null;
boolean close = false;
try {
do {
// 从allocator获取一个ByteBuf对象实例
byteBuf = allocHandle.allocate(allocator);
// doReadBytes从底层socket读取数据到byteBuf
allocHandle.lastBytesRead(doReadBytes(byteBuf));
if (allocHandle.lastBytesRead() <= 0) {
// nothing was read. release the buffer.
byteBuf.release();
byteBuf = null;
close = allocHandle.lastBytesRead() < 0;
if (close) {
// There is nothing left to read as we received an EOF.
readPending = false;
}
break;
}
allocHandle.incMessagesRead(1);
readPending = false;
// 将byteBuf传给pipeline，从而开始数据处理
pipeline.fireChannelRead(byteBuf);
byteBuf = null;
} while (allocHandle.continueReading());
allocHandle.readComplete();
pipeline.fireChannelReadComplete();
if (close) {
closeOnRead(pipeline);
}
} catch (Throwable t) {
handleReadException(pipeline, byteBuf, t, close, allocHandle);
} finally {
// Check if there is a readPending which was not processed yet.
// This could be for two reasons:
// * The user called Channel.read() or ChannelHandlerContext.read() in channelRead(...) method
// * The user called Channel.read() or ChannelHandlerContext.read() in channelReadComplete(...) method
//
// See https://github.com/netty/netty/issues/2254
if (!readPending && !config.isAutoRead()) {
removeReadOp();
}
}
}
- 从以上代码可以看出是ByteBuf的allocator是从ChannelConfig中获取：
final ByteBufAllocator allocator = config.getAllocator();
Channel使用堆内存还是直接内存
-
allocator的实际类型代表了Channel具体是使用堆内存还是直接内存。这个可以通过用户代码通过childOption设置或者使用默认的：
- 用户代码传入：主要是通过childOption传入，则ServerBootstrap在接收到客户端连接并创建SocketChannel时，会根据初始化ServerBootstrap时的childOption设置，对该SocketChannel的config进行配置。
// 处理客户端请求的配置 serverBootstrap.childOption(ChannelOption.TCP_NODELAY, true); serverBootstrap.childOption(ChannelOption.SO_KEEPALIVE, true); // 客户端SocketChannel所使用的allocator的实现类 serverBootstrap.childOption(ChannelOption.ALLOCATOR, UnpooledByteBufAllocator.DEFAULT); serverBootstrap.childHandler(new NettyServerInitializer(webSocketService));-
默认实现，一般不需要通过用户代码传入，使用默认的即可，默认实现为：如果当前的运行平台是android，则使用unpooled，即不使用池化机制，因为android移动端，内存资源有限，不适合做缓存；其他平台则默认使用池化机制，提高性能，以空间换时间。其次是使用堆内存还是直接内存的问题，主要是根据：
（1）当前运行平台是否支持使用Java的unsafe来进行本地方法调用；
（2）程序的系统参数是否设置了io.netty.noPreferDirect=true。如果当前平台支持unsafe且io.netty.noPreferDirect=false或者没有设置，默认为false，则使用直接内存；否则使用堆内存。
这两个参数的默认值：
1. 是否运行通过底层api直接访问直接内存，默认：允许 -Dio.netty.noPreferDirect 2. 是否允许使用sun.misc.Unsafe，默认：允许；注意：使用sun的私有类库存在平台移植问题，另外sun.misc.Unsafe类是不安全的，如果操作失败，不是抛出异常，而是虚拟机core dump，不建议使用Unsafe -Dio.netty.noUnsafe源码分析如下：
1. public class DefaultChannelConfig implements ChannelConfig { ... private volatile ByteBufAllocator allocator = ByteBufAllocator.DEFAULT; ... } 2. public interface ByteBufAllocator { ByteBufAllocator DEFAULT = ByteBufUtil.DEFAULT_ALLOCATOR; } 3. DEFAULT_ALLOCATOR public final class ByteBufUtil { ... static final ByteBufAllocator DEFAULT_ALLOCATOR; static { // android则是unpooled，其他为pooled String allocType = SystemPropertyUtil.get( "io.netty.allocator.type", PlatformDependent.isAndroid() ? "unpooled" : "pooled"); allocType = allocType.toLowerCase(Locale.US).trim(); ByteBufAllocator alloc; if ("unpooled".equals(allocType)) { alloc = UnpooledByteBufAllocator.DEFAULT; logger.debug("-Dio.netty.allocator.type: {}", allocType); } else if ("pooled".equals(allocType)) { alloc = PooledByteBufAllocator.DEFAULT; logger.debug("-Dio.netty.allocator.type: {}", allocType); } else { alloc = PooledByteBufAllocator.DEFAULT; logger.debug("-Dio.netty.allocator.type: pooled (unknown: {})", allocType); } DEFAULT_ALLOCATOR = alloc; ... } ... } 4. UnpooledByteBufAllocator.DEFAULT：非池化机制默认alloctor /** * Default instance which uses leak-detection for direct buffers. */ public static final UnpooledByteBufAllocator DEFAULT = new UnpooledByteBufAllocator(PlatformDependent.directBufferPreferred()); 5. PooledByteBufAllocator.DEFAULT：池化机制默认allocator public static final PooledByteBufAllocator DEFAULT = new PooledByteBufAllocator(PlatformDependent.directBufferPreferred()); 6. 在4，5中，都调用了PlatformDependent.directBufferPreferred()，如果返回true，则使用直接内存，否则使用堆内存。PlatformDependent.directBufferPreferred()的底层实现如下： private static final Throwable UNSAFE_UNAVAILABILITY_CAUSE = unsafeUnavailabilityCause0(); private static final boolean DIRECT_BUFFER_PREFERRED = UNSAFE_UNAVAILABILITY_CAUSE == null && !SystemPropertyUtil.getBoolean("io.netty.noPreferDirect", false); unsafeUnavailabilityCause0的实现：判断当前平台是否支持使用Java的unsafe private static Throwable unsafeUnavailabilityCause0() { if (isAndroid()) { logger.debug("sun.misc.Unsafe: unavailable (Android)"); return new UnsupportedOperationException("sun.misc.Unsafe: unavailable (Android)"); } Throwable cause = PlatformDependent0.getUnsafeUnavailabilityCause(); if (cause != null) { return cause; } try { boolean hasUnsafe = PlatformDependent0.hasUnsafe(); logger.debug("sun.misc.Unsafe: {}", hasUnsafe ? "available" : "unavailable"); return hasUnsafe ? null : PlatformDependent0.getUnsafeUnavailabilityCause(); } catch (Throwable t) { logger.trace("Could not determine if Unsafe is available", t); // Probably failed to initialize PlatformDependent0. return new UnsupportedOperationException("Could not determine if Unsafe is available", t); } }