Commit Graph

1312 Commits

Author SHA1 Message Date
Norman Maurer
dfd6b9009c [#2144] Fix NPE in Local transport caused by a race
Motivation:
At the moment it is possible to see a NPE when the LocalSocketChannels doRegister() method is called and the LocalSocketChannels doClose() method is called before the registration was completed.

Modifications:
Make sure we delay the actual close until the registration task was executed.

Result:
No more NPE
2014-04-17 15:19:23 +02:00
Norman Maurer
541abb8515 [#2375] [#2404] Fix bug in respecting ChannelConfig.setAutoRead(false) and also fix Channel.read() for OIO
Motivation:
At the moment ChanneConfig.setAutoRead(false) only is guaranteer to not have an extra channelRead(...) triggered when used from within the channelRead(...) or channelReadComplete(...) method. This is not the correct behaviour as it should also work from other methods that are triggered from within the EventLoop. For example a valid use case is to have it called from within a ChannelFutureListener, which currently not work as expected.

Beside this there is another bug which is kind of related. Currently Channel.read() will not work as expected for OIO as we will stop try to read even if nothing could be read there after one read operation on the socket (when the SO_TIMEOUT kicks in).

Modifications:
Implement the logic the right way for the NIO/OIO/SCTP and native transport, specific to the transport implementation. Also correctly handle Channel.read() for OIO transport by trigger a new read if SO_TIMEOUT was catched.

Result:
It is now also possible to use ChannelConfig.setAutoRead(false) from other methods that are called from within the EventLoop and have direct effect.
2014-04-17 08:19:20 +02:00
Norman Maurer
238af8cbd1 [#2390] Minimize memory usage of NioDatagramChannel
Motivation:
At the moment we create a HashMap that holds the MembershipKeys for multicast with every NioDatagramChannel even when most people not need it at al

Modifications:
Lazy create the HashMap when needed.

Result:
Less memory usage and less object creation
2014-04-15 06:56:33 +02:00
Norman Maurer
31a36e09ad [#2353] Use a privileged block to get ClassLoader and System property if needed
Motivation:
When using System.getProperty(...) and various methods to get a ClassLoader it will fail when a SecurityManager is in place.

Modifications:
Use a priveled block if needed. This work is based in the PR #2353 done by @anilsaldhana .

Result:
Code works also when SecurityManager is present
2014-04-08 14:13:49 +02:00
Norman Maurer
cd579f75d2 [#2363] SelectedSelectionKeySet may hold strong reference to SelectionKey after Channel is closed
Motivation:
Because we not null out the array entry in the SelectionKey[] which is produced by SelectedSelectionKeySet.flip() we may end up with a few SelectionKeyreferences still hanging around here even after the Channel was closed. As these entries may be present at the end of the SelectionKey[] which is never updated for a long time as not enough SelectionKeys are ready.

Modifications:
Once we access the SelectionKey out of the SelectionKey[] we directly null it out.

Result:
Reference can be GC'ed right away once the Channel was closed.
2014-04-05 19:27:09 +02:00
Norman Maurer
db5285950b [#2362] AbstractChannel.AbstractUnsafe.write(...) is slow
Motivation:
At the moment we do a Channel.isActive() check in every AbstractChannel.AbstractUnsafe.write(...) call which gives quite some overhead as shown in the profiler when you write fast enough. We can eliminate the check and do something more smart here.

Modifications:
Remove the isActive() check and just check if the ChannelOutboundBuffer was set to null before, which means the Channel was closed. The rest will be handled in flush0() anyway.

Result:
Less overhead when doing many write calls
2014-04-04 09:55:01 +02:00
Norman Maurer
ce52b04d4f [#2349] Correctly handle cancelled ChannelPromise in DefaultChannelHandlerContext
Motivation:
At the moment an IllegalArgumentException will be thrown if a ChannelPromise is cancelled while propagate through the ChannelPipeline. This is not correct, we should just stop to propagate it as it is valid to cancel at any time.

Modifications:
Stop propagate the operation through the ChannelPipeline once a ChannelPromise is cancelled.

Result:
No more IllegalArgumentException when cancel a ChannelPromise while moving through the ChannelPipeline.
2014-03-31 08:46:51 +02:00
Daniel Bevenius
5d141242c9 Remove overridden next() method in EventLoop interface.
Motivation:
EventLoop overrides the next() method with the same signature of its
super class EventLoopGroup.

Modifications:
Remove the duplicate method declaration from the EventLoop interface.

Result:
Perhaps makes the API slightly more readable as there is one less
method.
2014-03-24 12:12:48 +09:00
CoNDoRip
d56e55d51d Allow specifying SelectorProvider when constructing an NIO channel #2311
Motivation:

At the moment we use the system-wide default selector provider for this invocation of the Java virtual machine when constructing a new NIO channel, which makes using an alternative SelectorProvider practically useless.
This change allows user specify his/her preferred SelectorProvider.

Modifications:

Add SelectorProvider as a param for current `private static *Channel newSocket` method of NioSocketChannel, NioServerSocketChannel and NioDatagramChannel.
Change default constructors of NioSocketChannel, NioServerSocketChannel and NioDatagramChannel to use DEFAULT_SELECTOR_PROVIDER when calling newSocket(SelectorProvider).
Add new constructors for NioSocketChannel, NioServerSocketChannel and NioDatagramChannel which allow user specify his/her preferred SelectorProvider.

Result:

Now users can specify his/her preferred SelectorProvider when constructing an NIO channel.
2014-03-23 16:20:16 +01:00
Trustin Lee
d641b6a97a Use the length of MAC address as the last property to compare to get the best MAC address
Motivation:

Some operating systems like Windows 7 uses a valid globally unique EUI-64 MAC address for a virtual device (e.g. 00:00:00:00:00:00:00:E0), and because it's usually longer than the legit MAC-48 address, we should not use the length of MAC address when two MAC addresses are of the same quality.  Instead, we should compare the INET address of the NICs before comparing the length of the MAC addresses.

Modification:

Compare the length of MAC addresses as a last resort.

Result:

Correct MAC address detection in Windows with IPv6 enabled.
2014-03-21 13:01:55 +09:00
Trustin Lee
6cb238a673 Prefer the NIC with global IP address when getting the default machine ID
Motivation:

When there are two MAC addresses which are good enough, we can choose the one with better IP address rather than just choosing the first appeared one.

Modification:

Replace isBetterAddress() with compareAddresses() to make it return if both addresses are in the same preference level.
Add compareAddresses() which compare InetAddresses and use it when compareAddress(byte[], byte[]) returns 0 (same preference)

Result:

More correct primary MAC address detection
2014-03-21 13:01:55 +09:00
Trustin Lee
3621a0169c Reduce the time taken by NetUtil and DefaultChannelId class initialization
Motivation:

As reported in #2331, some query operations in NetworkInterface takes much longer time than we expected.  For example, specifying -Djava.net.preferIPv4Stack=true option in Window increases the execution time by more than 4 times.  Some Windows systems have more than 20 network interfaces, and this problem gets bigger as the number of unused (virtual) NICs increases.

Modification:

Use NetworkInterface.getInetAddresses() wherever possible.
Before iterating over all NICs reported by NetworkInterface, filter the NICs without proper InetAddresses.  This reduces the number of candidates quite a lot.
NetUtil does not query hardware address of NIC in the first place but uses InetAddress.isLoopbackAddress().
Do not call unnecessary query operations on NetworkInterface.  Just get hardware address and compare.

Result:

Significantly reduced class initialization time, which should fix #2331.
2014-03-21 13:01:55 +09:00
Daniel Bevenius
278fc05e28 Fixing javadoc ascii table for ChannelFuture
Motivation:
The current ascii table [1] showing the various states that a ChannelFuture
can be in, but the table is slightly "off" for the 'cause'.

Modifications:
- Updated the ascii table to display nicely.

Result:
Easier to read the states of a ChannelFuture.

[1] http://netty.io/5.0/api/io/netty/channel/ChannelFuture.html
2014-03-20 08:32:40 +01:00
Norman Maurer
4e2bc95ef9 [#2326] Add constructor to NioServerSocketChannel which accepts a ServerSocketChannel
Motivation:
Allow the user to create a NioServerSocketChannel from an existing ServerSocketChannel.

Modifications:
Add an extra constructor

Result:
Now the user is be able to create a NioServerSocketChannel from an existing ServerSocketChannel, like he can do with all the other Nio*Channel implemntations.
2014-03-16 07:05:39 -07:00
Norman Maurer
ba201f61ea [#2323] Make it clear a Channel must be closed to release all resources
Motivation:
Ensure the user know the Channel must be closed to release resources like filehandles.

Modifications:
Add some extra javadoc.

Result:
More clear documentation
2014-03-16 07:03:44 -07:00
Norman Maurer
77295c1230 [#2308] Use SelectorProvider.open*() to open NIO channels and so remove condition when create new NIO channels.
Motivation:
At the moment we use SocketChannel.open(), ServerSocketChannel.open() and DatagramSocketChannel.open(...) within the constructor of our
NIO channels. This introduces a bottleneck if you create a lot of connections as these calls delegate to SelectorProvider.provider() which
uses synchronized internal. This change removed the bottleneck.

Modifications:
Obtain a static instance of the SelectorProvider and use SelectorProvider.openSocketChannel(), SelectorProvider.openServerSocketChannel() and
SelectorProvider.openDatagramChannel(). This eliminates the bottleneck as SelectorProvider.provider() is not called on every channel creation.

Result:
Less conditions when create new channels.
2014-03-13 07:03:20 +01:00
Norman Maurer
a402b80ad0 Remove condition in ChannelHandlerAdapter.isSharable() by caching the result of the annotation lookup.
Motivation:
Remove the synchronization bottleneck and so speed up things

Modifications:
Introduce a ThreadLocal cache that holds mappings between classes of ChannelHandlerAdapater implementations and the result of checking if the @Sharable annotation is present.
This way we only will need to do the real check one time and server the other calls via the cache. A ThreadLocal and WeakHashMap combo is used to implement the cache
as this way we can minimize the conditions while still be sure we not leak class instances in containers.

Result:
Less conditions during adding ChannelHandlerAdapter to the ChannelPipeline
2014-03-12 13:43:39 +01:00
Jakob Buchgraber
aad4082d0f Corrected inconsistencies in the Javadoc. 2014-03-04 06:25:30 +01:00
Jakob Buchgraber
7d04136734 Added asserts to make sure ChannelHandlers are removed from the pipeline 2014-03-03 06:43:19 +01:00
Norman Maurer
6efac6179e [#1259] Add optimized queue for SCMP pattern and use it in NIO and native transport
This queue also produces less GC then CLQ when make use of OneTimeTask
2014-02-27 13:56:15 +01:00
Norman Maurer
01cd5929da Fix check to clear READ_OP and EPOLLIN. Part of [#2254] 2014-02-22 20:06:21 +01:00
Norman Maurer
52535a12f8 [#2254] Correctly handle Channel.read() and ChannelHandlerContext.read()
This includes also when it is called from channelRead(...) and channelReadComplete(...) methods.
2014-02-22 18:54:03 +01:00
Norman Maurer
60b830ba89 Directly use memory addresses for gathering writes to reduce gc pressure. Part of [#2239]
This also does factor out some logic of ChannelOutboundBuffer. Mainly we not need nioBuffers() for many
transports and also not need to copy from heap to direct buffer. So this functionality was moved to
NioSocketChannelOutboundBuffer. Also introduce a EpollChannelOutboundBuffer which makes use of
memory addresses for all the writes to reduce GC pressure
2014-02-21 14:16:37 +01:00
Norman Maurer
bea810707c [#2254] Fix regression in handling autoRead and Channel.read()
This regression was introduced by e0b39159657c9eb711047bc32367537c4870d467
2014-02-21 09:46:28 +01:00
Norman Maurer
b49490e239 Move marking ChannelPromise for writes uncancellable to addFlush for keep things simple 2014-02-17 16:15:49 +01:00
Trustin Lee
d5e897bba2 Determine the default allocator from system property
- Add ByteBufAllocator.DEFAULT
- The default allocator is now 'pooled'
2014-02-14 13:04:48 -08:00
Norman Maurer
657af70576 Prettify exception message 2014-02-13 06:47:40 +01:00
Norman Maurer
0ad029bf96 Allow to set IoRatio to 100% 2014-02-12 19:30:57 +01:00
Norman Maurer
4dc337a717 Correctly respect isAutoRead() and make it consistent across OIO/NIO 2014-02-11 20:46:58 +01:00
Norman Maurer
a96eb96646 Allow to cancel non-flushed writes 2014-02-11 20:00:56 +01:00
Trustin Lee
6c094237e5 Do not warn if failed to mark a void promise as success
- it's always supposed to fail.
2014-02-10 15:04:03 -08:00
Trustin Lee
fee1d9e75c Make most outbound operations cancellable / More robust promise update
- Inspired by #2214 by @normanmaurer
- Call setUncancellable() before performing an outbound operation
- Add safeSetSuccess/Failure() and use them wherever
2014-02-10 14:52:24 -08:00
Trustin Lee
a34d5baad2 Make most outbound operations are cancellable
- Inspired by #2214
- It actually reduces the chance of trying to marking a cancelled promise as success again, which raises an IllegalStateException.
2014-02-10 14:05:29 -08:00
Norman Maurer
bd20f80ca0 [#2215] DefaultChannelHandlerContext tasks needs to be volatile to ensure every thread only see full initialized instances 2014-02-08 10:35:54 +01:00
Norman Maurer
7a1a30f0ad Provide an optimized AtomicIntegerFieldUpdater, AtomicLongFieldUpdater and AtomicReferenceFieldUpdater 2014-02-06 21:07:31 +01:00
Norman Maurer
1d8a200849 Not wakeup the EventLoop for writes as they will not cause a flush anyway 2014-02-01 15:00:14 +01:00
Norman Maurer
f7f808c7e0 Better exception message to tell the user why it is not supported 2014-01-30 07:00:53 +01:00
Norman Maurer
96f94cfcb1 [#2164] Only reregister SelectionKeys that are still valid 2014-01-29 07:19:41 +01:00
Trustin Lee
0f1b1be0aa Enable a user specify an arbitrary information with ReferenceCounted.touch()
- Related: #2163
- Add ResourceLeakHint to allow a user to provide a meaningful information about the leak when touching it
- DefaultChannelHandlerContext now implements ResourceLeakHint to tell where the message is going.
- Cleaner resource leak report by excluding noisy stack trace elements
2014-01-29 11:44:59 +09:00
Trustin Lee
65ebecb85f Touch a ReferenceCounted while it traverses across a pipeline 2014-01-28 20:10:19 +09:00
Trustin Lee
b887e35ac2 Add ReferenceCounted.touch() / Add missing retain() overrides
- Fixes #2163
- Inspector warnings
2014-01-28 20:06:55 +09:00
Trustin Lee
7d501db7f8 Fix API documentation on the usage of AttributeKey 2014-01-28 13:57:08 +09:00
Norman Maurer
cac449a5e4 Add testcase to try to reproduce #2144 2014-01-23 07:10:08 +01:00
Trustin Lee
13f508154a Merge the attribute map of ChannelHandlerContext into Channel
- Fixes #2136
2014-01-20 15:16:12 +09:00
William Kemper
cab2a03760 fix grouping for isActive - socket.isBound is almost always true and should not override 'isOpen' 2014-01-17 07:23:03 +01:00
Trustin Lee
b165134044 Better exception message 2014-01-13 23:30:32 +09:00
Trustin Lee
17c0722862 Add missing validation which results in NPE later 2014-01-13 23:29:41 +09:00
Trustin Lee
9c1b9492d5 Fix a bug where DefaultChannelPipelineTest.testFireChannelRegistered() triggers channelRegistered() twice 2014-01-13 23:27:56 +09:00
Trustin Lee
999b51b026 Get the PID properly on Android
- Related: #2109
2014-01-13 22:28:28 +09:00
Trustin Lee
6df3bb5a79 Fix a problem where DefaultChannelId prevents Netty 5 from running on Android
- Fixes #2109
- Use reflection to find the current PID
2014-01-13 17:33:55 +09:00