Motivation:
OIO/NIO use a volatile variable to track if a read is pending. EPOLL does not use a volatile an executes a Runnable on the event loop thread to set readPending to false. These mechansims should be consistent, and not using a volatile variable is preferable because the variable is written to frequently in the event loop thread.
OIO also does not set readPending to false before each fireChannelRead operation and may result in reading more data than the user desires.
Modifications:
- OIO/NIO should not use a volatile variable for readPending
- OIO should set readPending to false before each fireChannelRead
Result:
OIO/NIO/EPOLL are more consistent w.r.t. readPending and volatile variable operations are reduced
Fixes https://github.com/netty/netty/issues/5069
Motivation:
We should better use Math utilities as these are intrinsics. This is a cleanup for ea3ffb8536.
Modifications:
Use Math utilities.
Result:
Cleaner code and use of intrinsics.
Motivation:
Fixes#5084. We (gRPC) encountered a bug that was triggered by
grpc/grpc-java@d927180. After that commit, event loop threads are
created per task by NioEventLoopGroup, and inherits the thread group of
the caller, which in our case is an application-provided request-scope
thread. Things go south when the application tries to manipulate (e.g.,
interrupt and join) all threads of the request-scope thread group, which
unexpectedly include the event loop threads.
Modifications:
DefaultThreadFactory will save the current thread group in constructor,
and apply it to all new threads.
Result:
Threads created by DefaultThreadFactory will be in the same thread group
as the thread where the factory is created.
Motivation:
When a PoolChunk needs to get moved to the previous PoolChunkList because of the minUsage / maxUsage constraints we always just moved it one level which is incorrect and so could lead to have PoolChunks in the wrong PoolChunkList (in respect to their minUsage / maxUsage settings). This then could have the effect that PoolChunks are not released / freed in a timely fashion and so.
Modifications:
- Correctly move PoolChunks between PoolChunkLists, which includes moving it multiple "levels".
- Add unit test
Result:
Correctlty move the PoolChunk to PoolChunkList when it is freed, even if its multiple layers.
Related: #3449
Motivation:
When a user shut down an EventExecutor/Loop prematurely, a Promise will
fail to execute its listeners. When it happens, DefaultPromise will log
a message at ERROR level, but there's no way to get notified about it
programmatically.
Modifications:
Do not catch and log the RejectedExecutionException unconditionally,
but only catch and log for non-late listener notifications, so that a
user gets notified on submission failure at least when the listener is
late.
Result:
Remedies #3449 to some extent, although we will need fundamental fix for
that, such as #3566
Motivation:
When always triggered fireChannelWritabilityChanged() directly when the update the pending bytes in the ChannelOutboundBuffer was made from within the EventLoop. This is problematic as this can cause some re-entrance issue if the user has a custom ChannelOutboundHandler that does multiple writes from within the write(...) method and also has a handler that will intercept the channelWritabilityChanged event and trigger another write when the Channel is writable. This can also easily happen if the user just use a MessageToMessageEncoder subclass and triggers a write from channelWritabilityChanged().
Beside this we also triggered fireChannelWritabilityChanged() too often when a user did a write from outside the EventLoop. In this case we increased the pending bytes of the outboundbuffer before scheduled the actual write and decreased again before the write then takes place. Both of this may trigger a fireChannelWritabilityChanged() event which then may be re-triggered once the actual write ends again in the ChannelOutboundBuffer.
The third gotcha was that a user may get multiple events even if the writability of the channel not changed.
Modification:
- Always invoke the fireChannelWritabilityChanged() later on the EventLoop.
- Only trigger the fireChannelWritabilityChanged() if the channel is still active and if the writability of the channel changed. No need to cause events that were already triggered without a real writability change.
- when write(...) is called from outside the EventLoop we only increase the pending bytes in the outbound buffer (so that Channel.isWritable() is updated directly) but not cause a fireChannelWritabilityChanged(). The fireChannelWritabilityChanged() is then triggered once the task is picked up by the EventLoop as usual.
Result:
No more re-entrance possible because of writes from within channelWritabilityChanged(...) method and no events without a real writability change.
Motivation:
The PoolChunkList.minUsage() and maxUsage() needs to take special action to translate Integer.MIN_VALUE / MAX_VALUE as these are used internal for tail and head of the linked-list structure.
Modifications:
- Correct the minUsage() and maxUsage() methods.
- Add unit test.
Result:
Correct metrics
Motivation:
fcbeebf6df introduced a unit test to verify ApplicationProtocolNegotiationHandler is compatible with SniHandler. However only the server attempts ALPN and verifies that it completes and the client doesn't verify the handshake is completed. This can lead to the client side SSL engine to prematurely close and throw an exception.
Modifications:
- The client should wait for the SSL handshake and ALPN to complete before the test exits.
Result:
SniHandlerTest.testSniWithApnHandler is more reliable.
Motivation:
441aa4c575 introduced a bug in transport-native-epoll where readPending is set to false before a read is attempted, but this should happen before fireChannelRead is called. The NIO transport also only sets the readPending variable to false on the first read in the event loop. This means that if the user only calls read() on the first channelRead(..) the select loop will still listen for read events even if the user does not call read() on subsequent channelRead() or channelReadComplete() in the same event loop run. If the user only needs 2 channelRead() calls then by default they will may get 14 more channelRead() calls in the current event loop, and then 16 more when the event loop is woken up for a read event. This will also read data off the TCP stack and allow the peer to queue more data in the local RECV buffers.
Modifications:
- readPending should be set to false before each call to channelRead()
- make NIO readPending set to false consistent with EPOLL
Result:
NIO and EPOLL transport set readPending to false at correct times which don't read more data than intended by the user.
Fixes https://github.com/netty/netty/issues/5082
Motivation:
When a promise is notified that was already added to the ChannelOutboundBuffer and we try to notify it later on we only see a warning that it was notified before. This is often not very useful as we have no idea where it was notified at all. We can do better in case it was failed before (which is most of the times the case) and just also log the cause that was used for it.
Modifications:
Add the cause that was used to notify the promise when we fail to notify it as part of the ChannelOutboundBuffer.
Result:
Easier to debug user errors.
Motivation:
There is a spelling error in FileRegion.transfered() as it should be transferred().
Modifications:
Deprecate old method and add a new one.
Result:
Fix typo and can remove the old method later.
Motivation:
See #3095
Modifications:
Add Log4J2LoggerFactory and Log4J2Logger which is an InternalLogger implementation based on log4j2.
Result:
The user can use log4j2 directly without a special slf4j binding.
Motivation:
DefaultChannelHandlerInvoker currently blindly cast to AbstractChannelHandlerContext without checking if the ChannelHandlerContext is really a sub-type of it. It should check it first and if not just use slow-path implementation.
Modifications:
Do instanceof check first and if it fails just create a new Runnable instance of used the cached.
Result:
DefaultChannelHandlerInvoker works with any ChannelHandlerContext implementations.
Motivation:
Sometimes it is useful to allow to disable the leak detection of buffers if the UnpooledByteBufAllocator is used. This is for example true if the app wants to leak buffers into user code but not want to put the burden on the user to always release the buffer.
Modifications:
Add another constructor to UnpooledByteBufAllocator that allows to completely disable leak-detection for all buffers that are allocator out of the UnpooledByteBufAllocator.
Result:
It's possible to disable leak-detection when the UnpooledByteBufAllocator is used.
Motivation:
We should only increment the metric for the huge / normal allocation after it is done. Also we should only decrement once deallocate.
Modifications:
- Move increment after the allocation.
- Fix deallocation metric and move it after deallocation
Result:
More correct metrics.
Motivation:
PoolThreadCache includes the wrong value when throwing a IllegalArgumentException because of freeSweepAllocationThreshold.
Modifications:
Use the correct value.
Result:
Correct exception message.
Motivation:
ApplicationProtocolNegotiationHandler attempts to get a reference to an SslHandler in handlerAdded, but when SNI is in use the actual SslHandler will be added to the pipeline dynamically at some later time. When the handshake completes ApplicationProtocolNegotiationHandler throws an IllegalStateException because its reference to SslHandler is null.
Modifications:
- Instead of saving a reference to SslHandler in handlerAdded just search the pipeline when the SslHandler is needed
Result:
ApplicationProtocolNegotiationHandler support SniHandler.
Fixes https://github.com/netty/netty/issues/5066
Motivation:
- The decoded name should always end with a dot (.), but we currently
strip it, which is incorrect.
- (O) 0 -> "."
- (X) 0 -> ""
- (O) 5 netty 2 io 0 -> "netty.io."
- (X) 5 netty 2 io 0 -> "netty.io"
- The encoded name should end with a null-label, which is a label whose
length is 0, but we currently append an extra NUL, causing FORMERR(1)
on a strict DNS server:
- (O) . -> 0
- (X) . -> 0 0
- (O) netty.io. -> 5 netty 2 io 0
- (X) netty.io. -> 5 netty 2 io 0 0
Modifications:
- Make sure to append '.' when decoding a name.
- Improve index checks so that the decoder can raise
CorruptFrameException instead of IIOBE
- Do not encode extra NUL
- Add more tests
Result:
Robustness and correctness
Motivation:
bfbef036a8 made EPOLL respect autoRead while in ET mode. However it is possible that we may miss data pending on the RECV queue if autoRead is off. This is because maybeMoreDataToRead is updated after fireChannelRead and if a user calls read() from here maybeMoreDataToRead will be false because it is updated after the fireChannelRead call. The way maybeMoreDataToRead was updated also causes a single channel to continuously read on the event loop and not relinquish and give other channels to try reading.
Modifications:
- Ensure maybeMoreDataToRead is always set after all user events, and is evaluated with readPending to execute a epollInReady on the EventLoop
- Combine the checkResetEpollIn and maybeMoreDataToRead logic to invoke a epollInReady later into the epollInFinally method due to similar responsibilities
- Update unit tests to reflect the user calling read() on the event loop from channelRead()
Result:
EPOLL ET with autoRead set to false will not leave data on the RECV queue.
Motivation:
Commit 2696778 changed MqttEncoder to be a singelton but missed to add @Sharable annotation. This broke the encoder as it can not be added to multiple pipelines.
Modifications:
Add @Sharable annotation
Result:
MqttEncoder can be used in multiple pipelines again.
Motivation:
Setting the WRITE_BUFFER_LOW_WATER_MARK before WRITE_BUFFER_HIGH_WATER_MARK results in an internal Exception (appears only in the logs) if the value is larger than the default high water mark value. The WRITE_BUFFER_HIGH_WATER_MARK call appears to have no effect in this context.
Setting the values in the reverse order works.
Modifications:
- deprecated ChannelOption.WRITE_BUFFER_HIGH_WATER_MARK and
ChannelOption.WRITE_BUFFER_LOW_WATER_MARK.
- add one new option called ChannelOption.WRITE_BUFFER_WATER_MARK.
Result:
The high/low water mark values limits caused by default values are removed.
Setting the WRITE_BUFFER_LOW_WATER_MARK before WRITE_BUFFER_HIGH_WATER_MARK results in an internal Exception (appears only in the logs) if the value is larger than the default high water mark value. The WRITE_BUFFER_HIGH_WATER_MARK call appears to have no effect in this context.
Setting the values in the reverse order works.
Motivation:
If a handler is added to the pipeline within ChannelInitializer::initChannel via
addFirst(...) then it will not receive the channelRegistered event. The same
handler added via addLast(...) will receive the event. This different behavior
is unlikely to be expected by users and can cause confusion.
Modifications:
Let ChannelInitializer::channelRegistered propagate the event by passing it to
the pipeline instead of firing it on the ChannelHandlerContext.
Result:
The channelRegistered event is propagated to handlers regardless of the method
used to add it to the pipeline (addFirst/addLast).
Motivation:
NIO now supports a pluggable select strategy, but EPOLL currently doesn't support this. We should strive for feature parity for EPOLL.
Modifications:
- Add SelectStrategy to EPOLL transport.
Result:
EPOLL transport supports SelectStategy.
Motivation:
Under high throughput/low latency workloads, selector wakeups are
degrading performance when the incoming operations are triggered
from outside of the event loop. This is a common scenario for
"client" applications where the originating input is coming from
application threads rather from the socket attached inside the
event loops.
As a result, it can be desirable to defer the blocking select
so that incoming tasks (write/flush) do not need to wakeup
the selector.
Modifications:
This changeset adds the notion of a generic SelectStrategy which,
based on its contract, allows the implementation to optionally
defer the blocking select based on some custom criteria.
The default implementation resembles the original behaviour, that
is if tasks are in the queue `selectNow()` and move on, and if no
tasks need to be processed go into the blocking select and wait
for wakeup.
The strategy can be customized per `NioEventLoopGroup` in the
constructor.
Result:
High performance client applications are now given the chance to
customize for how long the actual selector blocking should be
deferred by employing a custom select strategy.
Motivation:
There is no need to make DefaultChannelId package private as it may be useful for the user. For example EmbeddedChannel allows to inject a ChannelId when it is constructed. For this case the user can just use DefaultChannelId.
Modifications:
Change visibility of DefaultChannelId to public.
Result:
It's possible to create a new instance of DefaultChannelId by the user.
Motivation:
We need to ensure we run all pending tasks before doing any flush in writeOutbound(...) to ensure all pending tasks are run first. Also we should remove the assert of the future and just add a listener to it so it is processed later if needed. This is true as a user may schedule a write for later execution.
Modifications:
- Remove assert of future in writeOutbound(...)
- Correctly run pending tasks before doing the flush and also before doing the close of the channel.
- Add unit tests to proof the defect is fixed.
Result:
Correclty handle the situation of delayed writes.
Motivation:
We need to break out of the read loop for two reasons:
- If the input was shutdown in between (which may be the case when the user did it in the
fireChannelRead(...) method we should not try to read again to not produce any
miss-leading exceptions.
- If the user closes the channel we need to ensure we not try to read from it again as
the filedescriptor may be re-used already by the OS if the system is handling a lot of
concurrent connections and so needs a lot of filedescriptors. If not do this we risk
reading data from a filedescriptor that belongs to another socket then the socket that
was "wrapped" by this Channel implementation.
Modification:
Break the reading loop if the input was shutdown from within the channelRead(...) method.
Result:
No more meaningless exceptions and no risk to read data from wrong socket after the original was closed.
Motivation:
There are some use cases when a client may only be willing to read from a channel once
its previous write is finished (eg: serial dispatchers in Finagle). In this case, a
connection with SslHandler installed and ctx.channel().config().isAutoRead() == false
will stall in 100% of cases no matter what order of "channel active", "write", "flush"
events was.
The use case is following (how Finagle serial dispatchers work):
1. Client writeAndFlushes and waits on a write-promise to perform read() once it's satisfied.
2. A write-promise will only be satisfied once SslHandler finishes with handshaking and
sends the unencrypted queued message.
3. The handshaking process itself requires a number of read()s done by a client but the
SslHandler doesn't request them explicitly assuming that either auto-read is enabled
or client requested at least one read() already.
4. At this point a client will stall with NEED_UNWRAP status returned from underlying engine.
Modifiations:
Always request a read() on NEED_UNWRAP returned from engine if
a) it's handshaking and
b) auto read is disabled and
c) it wasn't requested already.
Result:
SslHandler is now completely tolerant of whether or not auto-read is enabled and client
is explicitly reading a channel.
Motivation:
We should throw a more helpful exception when a non PKCS#8 key is used by the user.
Modifications:
Change exception message to give a hint what is wrong.
Result:
Easier for user to understand whats wrong with their used key.
Motivation:
This allows using handlers for Streams in normal Netty-style. Frames are
read/written to the channel as messages, not directly as a
callback/method call. Handlers allow mixing and can ease HTTP/1 and
HTTP/2 interoperability by eventually supporting HTTP/1 handlers in
HTTP/2 and vise versa.
Modifications:
New handler Http2MultiplexCodec that converts from the current HTTP/2
API to a message-based API and child channels for streams.
Result:
The basics are done for server-side: new streams trigger creation of new
channels in much the same appearance to how new connections trigger new
channel creation. The basic frames HEADERS and DATA are handled, but
also GOAWAY and RST_STREAM.
Inbound flow control is implemented, but outbound is not. That will be
done later, along with not completing write promises on the child
channel until the write actually completes on the parent.
There is not yet support for outbound priority/weight, push promises,
and many other features.
There is a generic Object that may be set on stream frames. This also
paves the way for client-side support which needs a way to refer to
yet-to-be-created streams (due to how HEADERS allocates a stream id, and
the allocation order must be the same as transmission order).
Motivation:
8dbf5d02e5 modified the shutdown code for Socket but did not correctly calculate the change in shutdown state and only applying this change. This is significant because if sockets are being opening and closed quickly and the underlying FD happens to be reused we need to take care that we don't unintentionally change the state of the new FD by acting on an object which represents the old incarnation of that FD.
Modifications:
- Calculate the shutdown change, and only apply what has changed, or exit if no change.
Result:
Socket.shutdown can not inadvertently affect the state of another logical FD.
Motivation:
cf171ff525 introduced a change in behavior when dealing with closing channel in the read loop. This changed behavior may use stale state to determine if a channel should be shutdown and may be incorrect.
Modifications:
- Revert the usage of potentially stale state
Result:
Closing a channel in the read loop is based upon current state instead of potentially stale state.
Motivation:
Zero-length names needs to be "prefixed" by the length as well when encoded into a ByteBuf. Also some servers not correctly prefix these so we should ensure we can workaround this and even decode in such case.
Modifications:
- Always encode the length of the name into the ByteBuf even if its zero-length.
- If there are no readable bytes for the name just asume its an empty name to workaround dns servers that not fully respect the RFC.
Result:
Correctly encode zero-length names and be able to decode empty names even when the rfc is not strictly followed.
Motivation:
Often the user uses EmbeddedChannel within unit tests where the only "important" thing is to know if any pending messages were in the buffer and then release these.
We should provide methods for this so the user not need to manually loop through these and release.
Modifications:
Add methods to easily handle releasing of messages.
Result:
Less boiler-plate code for the user to write.
Motivation:
codec-mqtt had some typos and was not restrict enough in terms of making things final and private constructors.
Modifications:
- Fix typos
- Make most pojos final
- Remove redundant else blocks.
Result:
Cleaner and more restrict code.
Motivation:
Compile crash w/ JDK8:
```
[ERROR]
/Users/slandelle/Documents/dev/workspaces/workspace-ahc2/async-http-clie
nt-project/netty-bp/codec-dns/src/main/java/io/netty/handler/codec/dns/D
nsMessageUtil.java:[176,16] reference to append is ambiguous
both method append(java.lang.String) in java.lang.StringBuilder and
method append(java.lang.StringBuffer) in java.lang.StringBuilder match
```
Modification:
Force type explicitly
Result:
Class compile w/ JDK8
Motivation:
If a single Encoder object is promoted to the old generation then every object
reachable from the promoted object will eventually be promoted as well. A queue
illustrates the problem very well. Say a sequence of inserts and deletions
generate an object graph:
A -> B -> C -> D -> E -> F -> G -> H,
the head of the queue is E, the tail of the queue is H, and A, B, C, D are
dead. If all queue nodes are in the young generation, then a young gc will
clean up the object graph and leave us with:
E -> F -> G -> H
on the other hand, if B and C were previously promoted to the old generation,
then a young collection assumes the refernece from C to D is from a live object
(this is a key result of generational gc, no need to mark the old generation).
Hence the young collection assumes the refence to D is a gc root and leave us
with the object graph:
B-> C -> D -> E -> F -> G -> H.
Eventually D, E, F, G, H, and all queue nodes ever seen from this point on will
be promoted, regardless of their global live or dead status. It is generally
trivial to fix nepotism issues by simply breaking the reference chain after
dequeuing a node.
Currently Encoder objects do not null their references when removed from the
hash map. We have observed a 20X increase in promoted Encoder objects due to
nepotism.
Modifications:
Null before, after, and next fields when removing Encoder objects from maps.
Result:
Fewer promoted Encoder objects, fewer Encoder objects in the old generation,
shorter young collection times, old collections spaced further apart (nepotism
is just really bad). Enjoy.
Motivation:
We should upgrade to latest netty-tcnative version.
Modifications:
Upgrade to version 1.1.33.Fork15
Result:
Latest netty-tcnative version is used.
Motivation:
The current slow path of FastThreadLocal is much slower than JDK ThreadLocal. See #4418
Modifications:
- Add FastThreadLocalSlowPathBenchmark for the flow path of FastThreadLocal
- Add final to speed up the slow path of FastThreadLocal
Result:
The slow path of FastThreadLocal is improved.
Motivation:
The code of transport-native-epoll missed some things in terms of static keywords, @deprecated annotations and other minor things.
Modifications:
- Add missing @deprecated annotation
- Not using FQCN in javadocs
- Add static keyword where possible
- Use final fields when possible
- Remove throws IOException from method where it is not needed.
Result:
Cleaner code.
