Commit Graph

264 Commits

Author SHA1 Message Date
Trustin Lee
10f7a31908 Fixed SpdySessionHandlerTest / Fixed NPE in EmbeddedChannel
- Some tests like SpdySessionHandlerTest accesses outbound buffer
  even before the outbound buffer is initialized by
  AbstractEmbeddedChannel's subclasses, leading to NPE at <init>.
  To fix this problem, subclasses now pass the outbound buffer as
  a constructor parameter to AbstractEmbeddedChannel.
2012-06-07 21:33:31 +09:00
Trustin Lee
994038975a Port HttpContentEncoder/Decoder to use EmbeddedStreamChannel / Cleanup
- Removed unused constructor parameter in AbstractChannel
- Re-enabled GZIP encoding in HTTP snoop example
2012-06-07 21:06:56 +09:00
Trustin Lee
3442ff90e8 Ported all tests in netty-codec to the new embedder 2012-06-07 20:17:26 +09:00
Trustin Lee
7bc10f2eba Replace codec embedder with EmbeddedChannel which can test any handlers
- Added EventExecutor.inEventLoop(Thread) and replaced executor identity
  comparison in DefaultChannelPipeline with it - more elegant IMO
- Removed the test classes that needs rewrite or is of no use
2012-06-07 19:39:37 +09:00
Trustin Lee
8701e24b9a Add back Channel(Inbound|Outbound)(Message|Stream)HandlerAdapter
- they are useful when creating an anonymous class
- Also added back CombinedChannelHandler with extra constraints
2012-06-07 17:49:45 +09:00
Trustin Lee
aab71ccd8a Remove Channel(Inbound|Outbound)HandlerAdapter which does nothing
- Thanks to the recent refactoring, Channel(Inbound|Outbound)Handler-
  Adapter ended up having empty body.  No need to keep.
2012-06-07 17:25:15 +09:00
Trustin Lee
09570db091 Fixed a bug where ChannelInitializer can trigger the same event twice 2012-06-07 17:06:13 +09:00
Trustin Lee
6b819fb993 Fix a NPE in DefaultChannelPipeline.write(...)
- Also re-enabled logging handler in the echo example
2012-06-07 17:00:21 +09:00
Trustin Lee
ea0c9cfe79 Post-overhaul fixes / Split LoggingHandler into three
- LoggingHandler now only logs state and operations
- StreamLoggingHandler and MessageLoggingHandler log the buffer content
- Added ChannelOperationHandlerAdapter
  - Used by WriteTimeoutHandler
2012-06-07 16:56:21 +09:00
Trustin Lee
5e93d206ff Overhaul - Split ChannelHandler & Merge ChannelHandlerContext
- Extracted some handler methods from ChannelInboundHandler into
  ChannelStateHandler
- Extracted some handler methods from ChannelOutboundHandler into
  ChannelOperationHandler
- Moved exceptionCaught and userEventTriggered are now in
  ChannelHandler
  
- Channel(Inbound|Outbound)HandlerContext is merged into
  ChannelHandlerContext
- ChannelHandlerContext adds direct access methods for inboud and
  outbound buffers
  - The use of ChannelBufferHolder is minimal now.
    - Before: inbound().byteBuffer()
    - After: inboundByteBuffer()
    - Simpler and better performance
    
- Bypass buffer types were removed because it just does not work at all
  with the thread model.
  - All handlers that uses a bypass buffer are broken.  Will fix soon.

- CombinedHandlerAdapter does not make sense anymore either because
  there are four handler interfaces to consider and often the two
  handlers will implement the same handler interface such as
  ChannelStateHandler.  Thinking of better ways to provide this feature
2012-06-07 14:52:33 +09:00
Trustin Lee
4831eb49da Fix a test failure in LocalChannelRegistryTest 2012-06-07 08:54:37 +09:00
Trustin Lee
843a94b989 Fix compilation errors (#384)
- ECJ worked fine with the old code, but the official compiler does not
2012-06-06 23:02:47 +09:00
norman
caa35c9772 Rethrow the Throwable which accoured while execution the modification in the EventExecutor 2012-06-05 11:34:47 +02:00
norman
4eb42125a7 Make sure we can't deadlock even if the ChannelPipeline modification is executed by the EventExecutor 2012-06-05 11:21:44 +02:00
Trustin Lee
c2e65016fd Fixed some checkstyle errors 2012-06-04 13:43:02 -07:00
Trustin Lee
1eced1e9e3 Update license headers 2012-06-04 13:31:44 -07:00
Trustin Lee
36dbbc4867 Fix a test failure 2012-06-04 12:14:42 -07:00
Trustin Lee
4630905b92 Do not wrap exceptions if possible 2012-06-04 12:10:13 -07:00
Trustin Lee
b9fc67c9b8 Get the executor of the context earlier if possible. 2012-06-04 11:59:31 -07:00
Trustin Lee
b322e98712 Remove unnecessary parameters and fields / Clean-up 2012-06-04 11:56:00 -07:00
Norman Maurer
6b637ab22f No need for a custom Runnable implementation, just use EventExecutor.submit(..).get(). Thanks @trusting for review 2012-06-04 20:49:31 +02:00
Norman Maurer
ecea558e03 Add missing synchronized 2012-06-04 20:40:17 +02:00
Norman Maurer
0e74b089c2 Merge branch 'master' of github.com:netty/netty 2012-06-04 20:34:16 +02:00
Norman Maurer
c31fbc67c2 Let DefaultChannelPipeline call lifecycle methods of ChannelHandler with respect to the new ThreadModel 2012-06-04 20:34:09 +02:00
Norman Maurer
aac5586428 Fix NPE in test 2012-06-04 20:32:12 +02:00
Trustin Lee
1a9c0fd569 Make use of the existance of bridge references to find a buffer 2012-06-04 11:18:48 -07:00
Trustin Lee
c8c577efc5 Remove an unused field 2012-06-04 09:25:39 -07:00
Trustin Lee
0aa99606d9 Fix a race where 2 handlers in different threads access the same buffer
- DefaultChannelPipeline detects such cases and creates an object called
  'bridge' that works as a man-in-the-middle to deal with a race
  condition
- Slight performance drop is observed but still faster than v3. 
  Couldn't find much from a profiler yet.
2012-06-04 00:24:34 -07:00
Trustin Lee
c1afe3d8c3 Exchanging messages between two handlers is now thread safe
- (not byte buffers yet)
2012-06-03 19:39:35 -07:00
Trustin Lee
f3734e1eb9 Simplified DefaultChannelPipeline by making its list head final
- Previously, head was a volatile field which is null at the beginning.
  While iterating over the pipeline, if the loop hits null, it called
  Channel.Unsafe explicitly.
- Instead, I created an outbound handler that redirects all requests
  to the unsafe and made it a final field of the pipeline.
- As a result, DefaultChannelPipeline code became much simpler.
2012-06-03 18:51:42 -07:00
Trustin Lee
f6e14b636f Improve concurrent message buffer access test to reproduce known issue 2012-06-03 13:21:57 -07:00
Trustin Lee
4d8e2198fa Fix a bug where LocalServerChannel's localAddress is unregistered 2012-06-03 13:20:56 -07:00
Trustin Lee
1cd3156eba Make test case more robust 2012-06-03 13:12:29 -07:00
Trustin Lee
234c4c70db Ensure LocalChannel fire channelActive after peers's channelRegistered
- Also:
  - Made the test case more robust
  - Added a simple concurrent buffer modification test (needs more work)
2012-06-03 12:54:26 -07:00
Trustin Lee
d48973b0ff Add ChannelHandlerContext.hasNext(In|Out)bound(Byte|Message)Buffer 2012-06-03 04:35:38 -07:00
Trustin Lee
bbed0602c1 Use NoSuchBufferException instead of IllegalArgumentException 2012-06-03 04:26:11 -07:00
Trustin Lee
13d7ee1b2f Optimized DefaultChannelPipeline.write(...)
- Also replaced unnecessary function calls with field accesses
2012-06-03 04:25:03 -07:00
Trustin Lee
f991a8c7d4 Fix a bug where DefaultChannelPipeline.write() doesn't find the buffer
- Also fixed failures in SpdySessionHandlerTest
2012-06-03 04:10:32 -07:00
Trustin Lee
ada61d4985 Ported multicast test / Fixed bugs in NioDatagramChannelConfig 2012-06-03 02:57:52 -07:00
Trustin Lee
e6ceb91a85 Add AbstractDatagramTest / Port unicast test / Ignore 'Socket closed' 2012-06-03 02:40:58 -07:00
Trustin Lee
955c89fcf1 Add SocketEchoTest that will simplify a lot of socket testing
- SocketTestCombination generates all possible test combinations of
  socket transports.
- SocketEchoTest iterates over the combinations and runs all tests
  using reflection.
2012-06-03 01:52:39 -07:00
Trustin Lee
3b2c25e8ed Rename (Server)Bootstrap.(child)initializer to (child)handler
- The handler you specify with initializer() is actually simply added
  to the pipeline and that's all.  It's ChannelInitializer which does
  additional work.  For example, a user can specify just a single
  handler with initializer() and it will still work.  This is especially
  common for Bootstrap, so I renamed initializer to handler, which makes
  more sense.
2012-06-03 01:00:16 -07:00
Trustin Lee
19dcb81727 Add comments for easier understanding 2012-06-02 02:43:26 -07:00
Trustin Lee
01a5bd41f0 Add Channel.type() which tells if stream-oriented or message-oriented
- DefaultChannelPipeline uses this information to reject invalid buffer
  access in inbound(Message|Byte)Buffer.  Otherwise, a user can access
  a message buffer when the channel is stream-oriented.
- Because ChannelType cannot be both STREAM and MESSAGE, catch-all
  buffer has been removed to avoid confusion and unexpected behavior
  (it's already causing headache.)
- As a result, codec embedder needs rework.
2012-06-02 01:58:15 -07:00
Trustin Lee
cc4f705029 Replace ChannelBuffer.toByteBuffer() with hasNioBuffer() and nioBuffer()
... just like we do with byte arrays.  toByteBuffer() and
toByteBuffers() had an indeterministic behavior and thus it could not
tell when the returned NIO buffer is shared or not.  nioBuffer() always
returns a view buffer of the Netty buffer.  The only case where
hasNioBuffer() returns false and nioBuffer() fails is the
CompositeChannelBuffer, which is not very commonly used and *slow*.
2012-06-02 01:30:55 -07:00
Trustin Lee
04cf1c8199 Use custom thread factory by default to better recognize the threads
... from their names
2012-06-01 22:57:54 -07:00
Trustin Lee
61e169e53a Remove EventExecutor.parent(), which is of no use 2012-06-01 22:33:53 -07:00
Trustin Lee
4440386494 Little bit of optimization 2012-06-01 18:34:19 -07:00
Trustin Lee
141a05c831 Strict thread model / Allow assign an executor to a handler
- Add EventExecutor and make EventLoop extend it
- Add SingleThreadEventExecutor and MultithreadEventExecutor
- Add EventExecutor's default implementation
- Fixed an API design problem where there is no way to get non-bypass
  buffer of desired type
2012-06-01 17:51:19 -07:00
Trustin Lee
ab5043b3c7 Fixed LocalAddressTest 2012-05-31 21:22:03 -07:00