Commit Graph

9944 Commits

Author SHA1 Message Date
Aayush Atharva
cf589a8f31
Add DoT and TCP DNS Client Example (#10256)
Motivation:

[DNS-over-TLS (DoT)](https://tools.ietf.org/html/rfc7858.html) encrypts DNS queries and sends it over TLS connection to make sure queries are secure in transit.

[TCP DNS](https://tools.ietf.org/html/rfc7766) sends DNS queries over TCP connection (unencrypted).

Modification:

Add DNS-over-TLS (DoT) Client Example which uses TLSv1.2 and TLSv1.3.
Add TCP DNS Client Example

Result:

DNS-over-TLS (DoT) Client Example
TCP DNS Client Example
2020-05-07 15:26:14 +02:00
louxiu
909e7c9c29
Add option to configure recycler delayed queue drop ratio (#10251) (#10255)
Motivation

- Recycler stack and delayed queue drop ratio can only be configured with
the same value. The overall drop ratio is ratio^2.

- #10251 shows that enable drop in `WeakOrderQueue` may introduce
performance degradation. Though the final reason is not clear now,
it would be better to add option to configure delayed queue drop ratio separately.

Modification

- "io.netty.recycler.delayedQueue.ratio" as the drop ratio of delayed queue
- default "delayedQueue.ratio" is same as "ratio"

Results

Able to configure recycler delayed queue drop ratio separately
2020-05-07 15:08:35 +02:00
feijermu
731d33070f
Add a DNS client example. (#10237)
Motivation:

It seems that there is no DNS client example in Netty project so far.

Modification:

Add a Netty DNS client example.

Result:

More examples
2020-05-07 10:46:41 +02:00
louxiu
4c9a30d5f9
Use io.netty.recycler.ratio directly (#10253)
Motivation

1. It's inable to collect all object because RATIO is always >=1 after
`safeFindNextPositivePowerOfTwo`

2. Enable drop object in `WeakOrderQueue`(commit:
71860e5b94) enlarge the drop ratio. We
can subtly control the overall drop ratio by using `io.netty.recycler.ratio` directly,

Modification

- Remove `safeFindNextPositivePowerOfTwo` before set the ratio

Results

Able to disable drop when recycle object
2020-05-07 10:29:05 +02:00
Norman Maurer
f00160bca3
Don't reuse ChannelPromise in WebSocketProtocolHandler (#10248)
Motivation:

We cant reuse the ChannelPromise as it will cause an error when trying to ful-fill it multiple times.

Modifications:

- Use a new promise and chain it with the old one
- Add unit test

Result:

Fixes https://github.com/netty/netty/issues/10240
2020-05-07 08:11:33 +02:00
feijermu
40448db5bb
Remove a unused import in DefaultHttp2ConnectionEncoder.java (#10249)
Motivation:

`io.netty.handler.codec.http2.Http2CodecUtil.DEFAULT_PRIORITY_WEIGHT` is never used in DefaultHttp2ConnectionEncoder.java.

Modification:

Just remove this unused import.

Result:

Make the DefaultHttp2ConnectionEncoder.java's imports clean.
2020-05-07 08:10:36 +02:00
Nick Hill
c3db0391af
Correctly handle WrappedCompositeByteBufs in addFlattenedComponents() (#10247)
Motivation

An NPE was reported in #10245, caused by a regression introduced in
#8939. This in particular affects ByteToMessageDecoders that use the
COMPOSITE_CUMULATOR.

Modification

- Unwrap WrappedCompositeByteBufs passed to
CompositeByteBuf#addFlattenedComponents(...) method before accessing
internal components field
- Extend unit test to cover this case and ensure more of the
CompositeByteBuf tests are also run on the wrapped variant

Results

Fixes #10245
2020-05-05 13:56:40 +02:00
Norman Maurer
cfcd7a4fde
Rename testmethods to make these more clear (#10231)
Motivation:

The currently used method names don't make a lot of sense.

Modifications:

Rename to cleanup

Result:

Cleanup
2020-04-29 17:52:43 +02:00
Aayush Atharva
9427255ffb
Specify algorithm for key pair in self signed certificate to generate EC or RSA based certificate. (#10223)
Motivation:

EC is better than RSA because of the small key size, efficient and secure which makes it perfect for testing purposes.

Modification:

Added support to specify an algorithm (EC or RSA) in constructors for key pair generation. The default key size is 256-bits as promulgated by NSA Suite B.

Result:
Able to generate a self-signed certificate of EC or RSA.
2020-04-29 16:52:07 +02:00
Norman Maurer
8f7ca2b4ef
Reuse the same allocator as used by the ByteBuf that is used during… (#10226)
Motivation:

We should not use Unpooled to allocate buffers if possible to ensure we can make use of pooling etc.

Modifications:

- Only allocate a buffer if really needed
- Use the ByteBufAllocator of the offered ByteBuf
- Ensure we not use buffer.copy() but explicitly allocate a buffer and then copy into it to not hit the limit of maxCapacity()

Result:

Improve allocations
2020-04-29 14:39:14 +02:00
Norman Maurer
ff36f2826c
Remove some debugging cruft (#10229)
Motivation:

RtspDecoderTest did include a println(...) call which was a left over from debugging.

Modifications:

Remove println(...)

Result:

Cleanup
2020-04-29 11:35:22 +02:00
Norman Maurer
987a68eb02
Fix memory leak in HttpPostMultipartRequestDecoder (#10227)
Motivation:

We need to release all ByteBufs that we allocate to prevent leaks. We missed to release the ByteBufs that are used to aggregate in two cases

Modifications:

Add release() calls

Result:

No more memory leak in HttpPostMultipartRequestDecoder
2020-04-29 08:23:41 +02:00
Norman Maurer
6cd193e83f
Don't log with warn level in the DnsNameResolver in most cases (#10225)
Motivation:

We should only log with warn level if something really critical happens as otherwise we may spam logs and confuse the user.

Modifications:

- Change log level to debug for most cases

Result:

Less noisy logging
2020-04-29 08:00:14 +02:00
Norman Maurer
387e451c82
Detect CNAME loops in the CNAME cache while trying to resolve (#10221)
Motivation:

We need to detect CNAME loops during lookup the DnsCnameCache as otherwise we may try to follow cnames forever.

Modifications:

- Correctly detect CNAME loops in the cache
- Add unit test

Result:

Fixes https://github.com/netty/netty/issues/10220
2020-04-28 10:47:10 +02:00
Norman Maurer
83012a038b
Update to latest Conscrypt release and add workarounds for bugs (#10211)
Motivation:

We are far behind with the version of Conscrypt we are using during testing. We should ensure we use the latest.

Modifications:

- Update conscrypt dependency
- Ensure we use conscrypt provider in tests
- Add workarounds for conscrypt bugs in testsuite

Result:

Use latest Conscrypt release
2020-04-28 09:50:05 +02:00
Fabien Renaud
c354fa48e1
HttpPostRequestDecoder: retain instead of copy when first buf is last (#10209)
Motivations
-----------
There is no need to copy the "offered" ByteBuf in HttpPostRequestDecoder
when the first HttpContent ByteBuf is also the last (LastHttpContent) as
the full content can immediately be decoded. No extra bookeeping needed.

Modifications
-------------
HttpPostMultipartRequestDecoder
 - Retain the first ByteBuf when it is both the first HttpContent offered
to the decoder and is also LastHttpContent.
 - Retain slices of the final buffers values

Results
-------
ByteBufs of FullHttpMessage decoded by HttpPostRequestDecoder are no longer
unnecessarily copied. Attributes are extracted as retained slices when
the content is multi-part. Non-multi-part content continues to return
Unpooled buffers.

Partially addresses issue #10200
2020-04-28 09:43:05 +02:00
feijermu
9751bb3ebc
Move up the size check in AbstractDiskHttpData.setContent. (#10222)
Motivation:

`AbstractHttpData.checkSize` may throw an IOException if we set the max size limit via `AbstractHttpData.setMaxSize`. However, if this exception happens, the `AbstractDiskHttpData.file` and the `AbstractHttpData.size` are still be modified. In other words, it may break the failure atomicity here.

Modification:

Just move up the size check.

Result:

Keep the failure atomicity even if `AbstractHttpData.checkSize` fails.
2020-04-28 09:34:33 +02:00
Fabien Renaud
4f72cdf233
Dns resolver: honor resolv.conf timeout, rotate and attempts options (#10207)
Motivations
-----------
DnsNameResolverBuilder and DnsNameResolver do not auto-configure
themselves uing default options define in /etc/resolv.conf.
In particular, rotate, timeout and attempts options are ignored.

Modifications
-------------
 - Modified UnixResolverDnsServerAddressStreamProvider to parse ndots,
attempts and timeout options all at once and use these defaults to
configure DnsNameResolver when values are not provided by the
DnsNameResolverBuilder.
 - When rotate option is specified, the DnsServerAddresses returned by
UnixResolverDnsServerAddressStreamProvider is rotational.
 - Amend resolv.conf options with the RES_OPTIONS environment variable
when present.

Result: 

Fixes https://github.com/netty/netty/issues/10202
2020-04-28 09:28:05 +02:00
Saranya Krishnakumar
5fa5ce34e1
Add check for DefaultFileRegion to calculate size of msg in AbstractTrafficShapingHandler (#10215)
Motivation:

Currently calculateSize method in AbstractTrafficShapingHandler calculates size for object of type ByteBuf or ByteBufHolder. Adding a check for FileRegion, makes it possible to do traffic shaping for FileRegion objects as well

Modification:

Check if object to be sent is of type FileRegion, if yes calculate the size using its count() method.


Co-authored-by: Dinesh Joshi <dinesh.joshi@apple.com>
2020-04-27 13:59:01 +02:00
feijermu
be28b266d5
Remove unused import in JsonObjectDecoder.java (#10213)
Motivation:

`io.netty.channel.ChannelHandler` is never used in JsonObjectDecoder.java.

Modification:

Just remove this unused import.

Result:

Make the JsonObjectDecoder.java's imports simple and clean.
2020-04-27 13:58:02 +02:00
feijermu
eb3721b971
Fix a potential fd leak in AbstractDiskHttpData.delete (#10212)
Motivation:

An unexpected IOException may be thrown from `FileChannel.force`. If it happens, the `FileChannel.close` may not be invoked.

Modification:

Place the `FileChannel.close` in a finally block.

Result:

Avoid fd leak.
2020-04-27 07:03:45 +02:00
Norman Maurer
9778f05e14
Update testsuite / pom.xml to be able to build with Java15 (#10210)
Motivation:

We need to make some slightly changes to be able to build on Java15 as some previous deprecated methods now throw UnsupportedOperationException

Modifications:

- Add code to handle UnsupportedOperationException
- Revert previous applied workaround for bug that was fixed in Java15
- Add maven profile

Result:

Be able to build with latest Java15 EA release
2020-04-27 06:27:54 +02:00
Norman Maurer
5a08dc0d9a
Add fastpath implementation for Unpooled.copiedBuffer(CharSequence, Charset) when UTF-8 or US-ASCII is used (#10206)
Motivation:

We can make use of our optimized implementations for UTF-8 and US-ASCII if the user request a copy of a sequence for these charsets

Modifications:

- Add fastpath implementation
- Add unit tests

Result:

Fixes https://github.com/netty/netty/issues/10205
2020-04-23 17:47:16 +02:00
wangxiyuan
660611c450
Add epoll aarch64 maven config and Dockerfile (#9804)
Motivation:

`transport-native-epoll` doesn't have ARM release package. 

Modification:

This PR added cross compile profile for epoll. Then we can easily build aarch64 package on X86 machine. 

Result:
Fixes #8279
2020-04-23 13:58:08 +02:00
Norman Maurer
0cf919c389
HTTP2: Guard against multiple ctx.close(...) calls with the same ChannelPromise (#10201)
Motivation:

Http2ConnectionHandler may call ctx.close(...) with the same promise instance multiple times if the timeout for gracefulShutdown elapse and the listener itself is notified. This can cause problems as other handlers in the pipeline may queue these promises and try to notify these later via setSuccess() or setFailure(...) which will then throw an IllegalStateException if the promise was notified already

Modification:

- Add boolean flag to ensure doClose() will only try to call ctx.close(...) one time

Result:

Don't call ctx.close(...) with the same promise multiple times when gradefulShutdown timeout elapses.
2020-04-23 11:25:39 +02:00
root
9c5008b109 [maven-release-plugin] prepare for next development iteration 2020-04-22 09:57:54 +00:00
root
d0ec961cce [maven-release-plugin] prepare release netty-4.1.49.Final 2020-04-22 09:57:26 +00:00
Norman Maurer
d0c2f6e8b3
Clarify exception message when ALPN is not supported (#10155)
Motivation:

We should provide more informations when ALPN is not supported and a user tries to use it.

Modifications:

- Use UnsupportedOperationException

Result:

Easier to debug ALPN problems
2020-04-22 08:07:27 +02:00
Norman Maurer
5cd00d22d9
Ensure we support ALPN when using java 8u251 (#10196)
Motivation:

ALPN support was backported to java 8 lately. Ensure we support it if the user uses the latest java 8 release

Modifications:

- Update logic to be able to detect if ALPN is supported out of the box when using Java8
- Update jetty alpn version

Result:

Be able to use ALPN out of the box when using java 8u251
2020-04-21 15:01:59 +02:00
Norman Maurer
232c669fa4
Guard against overflow when calling CompositeByteBuf.addComponent(...) (#10197)
Motivation:

We need to ensure we not overflow when adding buffers to a CompositeByteBuf

Modifications:

- Correctly validate overflow before adding to the internal storage
- Add testcase

Result:

Fixes https://github.com/netty/netty/issues/10194
2020-04-21 12:04:25 +02:00
feijermu
32e32fe98f
Fix a potential fd leak in AbstractDiskHttpData.setContent (#10198)
Motivation:

`RandomAccessFile.setLength` may throw an IOException. We must deal with this in case of the occurrence of `I/O` error.

Modification:

Place the `RandomAccessFile.setLength` method call in the `try-finally` block.

Result:

Avoid fd leak.
2020-04-21 10:54:09 +02:00
jrhee17
37948bc9de
Add support for HAProxyMessageEncoder (#10175)
Motivation:

Add support for HAProxyMessageEncoder.
This should help java based HAProxy server implementations propagate proxy information.

Modification:

Add public constructors for `HAProxyMessage`, `HAProxyTLV`, `HAProxySSLTLV`.
Add additional argument checks for `HAProxyMessage` and modify exceptions thrown when creating via public constructors directly.
Introduce a `@Sharable` `HAProxyMessageEncoder` which encodes a `HAProxyMessage` into a byte array.
Add an example `HAProxyServer` and `HAProxyClient` to `io.netty.example`


Result:

Fixes #10164
2020-04-16 09:35:06 +02:00
Norman Maurer
2c78b4c84f
Correctly propagate exceptions from inbound operations in all cases (#10176)
Motivation:

In AbstractChannelHandlerContext we had some code where we tried to guard against endless loops caused by exceptions thrown by exceptionCaught(...) that would trigger exceptionCaught again. This code was proplematic for two reasons:
 - It is quite expensive as we need to compare all the stacks
 - We may end up not notify another handlers exceptionCaught(...) if in our exeuction stack we triggered actions that will cause an exceptionCaught somewhere else in the pipeline

Modifications:

- Just remove the detection code as we already handle everything correctly when we invoke exceptionCaught(...)
- Add unit tests

Result:

Ensure we always notify correctly and also fixes performance issue reported as https://github.com/netty/netty/issues/10165
2020-04-16 08:56:45 +02:00
Norman Maurer
6dad12defa
Add workaround for possible classloader deadlock when trying to load JNI code (#10190)
Motivation:

netty_epoll_linuxsocket_JNI_OnLoad(...) may produce a deadlock with another thread that will load IOUtil in a static block. This seems to be a JDK bug which is not yet fixed. To workaround this we force IOUtil to be loaded from without java code before init the JNI code

Modifications:

Use Selector.open() as a workaround to load IOUtil

Result:

Fixes https://github.com/netty/netty/issues/10187
2020-04-16 08:40:59 +02:00
Linas Medžiūnas
fb5e2cd3aa
Efficient BytBuf search algorithms (#9914) (#9955)
Motivation:

We have found out that ByteBufUtil.indexOf can be inefficient for substring search on
ByteBuf, both in terms of algorithm complexity (worst case O(needle.readableBytes *
haystack.readableBytes)), and in constant factor (esp. on Composite buffers).
With implementation of more performant search algorithms we have seen improvements on
the order of magnitude.

Modifications:

This change introduces three search algorithms:
1. Knuth Morris Pratt - classical textbook algorithm, a good default choice.
2. Bit mask based algorithm - stable performance on any input, but limited to maximum
search substring (the needle) length of 64 bytes.
3. Aho–Corasick - worse performance and higher memory consumption than [1] and [2], but
it supports multiple substring (the needles) search simultaneously, by inspecting every
byte of the haystack only once.

Each algorithm processes every byte of underlying buffer only once, they are implemented
as ByteProcessor.

Result:

Efficient search algorithms with linear time complexity available in Netty (I will share
benchmark results in a comment on a PR).
2020-04-15 10:21:24 +02:00
时无两丶
ff6d2e4949
Remove duplicate code in PoolArena. (#10174)
Motivation:

Remove duplicate code in PoolArena.

Modification:

Replace duplicate code with `tinyIdx` and `smallIdx`.

Result:

Clean code.
2020-04-15 09:29:23 +02:00
feijermu
17f205fab3
Close the FileChannel in case of an IOException in AbstractDiskHttpData.addContent. (#10188)
Motivation:

`FileChannel.force` may throw an IOException. A fd leak may happen here.

Modification:

Close the fileChannel in a finally block.

Result:

Avoid fd leak.
2020-04-15 09:24:26 +02:00
Norman Maurer
4c9d7492fb
CodecOutputList should be recycled in a finally block (#10186)
Motivation:

To ensure we always recycle the CodecOutputList we should better do it in a finally block

Modifications:

Call CodecOutputList.recycle() in finally

Result:

Less chances of non-recycled lists. Related to https://github.com/netty/netty/issues/10183
2020-04-15 09:08:19 +02:00
feijermu
1b3e1985b4
Release the channel attribute--REOPEN_TASK after removing the TrafficShapingHandler from channel pipeline. (#10177)
Motivation:

The `AbstractTrafficShapingHandler` caches the `ReopenReadTimerTask` instance in the channel attribute. However, if this handler is removed from the channel pipeline, this `ReopenReadTimerTask` instance may not be released.

Modification:

Release the channel attribute `REOPEN_TASK` in `handlerRemoved` method.

Result:

Avoid a channel attribute leak.
2020-04-15 08:52:08 +02:00
Cenjie Ho
2d4d64dc42
Add the unit test of ChannelOutboundBuffer (#10180)
Motivation:

Parameter maxCount needs the unit test.

Modifications:

1. Change the conditional statement to avoid the ineffective maxCount (enhance the robustness of the code merely).
2. Add the unit test for maxCount.

Result:

Enable this parameter to be tested.
2020-04-15 08:22:51 +02:00
Norman Maurer
b996bf2151
Update Bouncycastle dependency (#10185)
Motivation:

We should update our optional bouncycastle dependency to ensure we use the latest which has all the security fixes

Modifications:

Update bouncycastle version

Result:

Fixes https://github.com/netty/netty/issues/10184
2020-04-14 14:23:45 +02:00
Stephane Landelle
d5e42f8553
Log protocol version during handshake (#10178)
Motivation:

Only cipher suite is logged during handshake. Picked protocol is interesting too.

Modification:

Log protocol as well.

Result:

More interesting information when debugging TLS handshakes.
2020-04-14 10:48:43 +02:00
Sérgio Santos
79ef0c4706
Allow Conscrypt to be used on Android (#10182)
**Motivation:**

Following up on https://github.com/netty/netty/issues/10181

I was able to successfully setup a Netty server with TLS on Android, using Conscrypt, with the change proposed here. Conscrypt publishes an [Android specific version](https://github.com/google/conscrypt#android). But the current availability check prevents Netty from using it (the call `PlatformDependent.javaVersion()` returns `6` on Android).

**Modification:**

Check whether the java version is above 8, or if it's running on an Android device, to flag Conscrypt as available (besides the remaining checks).

**Result:**

Netty with TLS runs fine on Android 👍
2020-04-14 10:42:51 +02:00
Andrey Mizurov
2b14775446
Stomp over WebSocket Chat example (#10152)
Motivation:

Often people want to use `stomp-codec` with WebSocket transport or other but cannot figure out how can do this staff on Netty.

Modification:

Create example for demonstrating integration between STOMP and WebSocket.
Inspired by https://github.com/jmesnil/stomp-websocket

Result:

Fixes #9383
2020-04-08 12:04:48 +02:00
minux
7d7e1f5ed7
Stop sending DNS queries if promise is cancelled (#10171)
Motivation:
A user might want to cancel DNS resolution when it takes too long.
Currently, there's no way to cancel the internal DNS queries especially when there's a lot of search domains.

Modification:
- Stop sending a DNS query if the original `Promise`, which was passed calling `resolve()`, is canceled.

Result:
- You can now stop sending DNS queries by cancelling the `Promise`.
2020-04-07 12:25:02 +02:00
minux
a4c96483d1
Fix a bug where making IPv6 DnsQuestion when it's not supported (#10170)
Motivation:
Related https://github.com/line/armeria/issues/2463
Here is an example that an NIC has only link local address for IPv6.
```
$ ipaddr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
3: eth0@if18692: <BROADCAST,MULTICAST,UP,LOWER_UP,M-DOWN> mtu 1460 qdisc noqueue
    link/ether 1a:5e:xx:xx:xx:xx brd ff:ff:ff:ff:ff:ff
    inet 10.xxx.xxx.xxx/24 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::xxxx:xxxx:xxxx:xxxx/64 scope link
       valid_lft forever preferred_lft forever
```
If the NICs have only local or link local addresses, We should not send IPv6 DNS queris.

Modification:
- Ignore link-local IPv6 addresses which may exist even on a machine without IPv6 network.

Result:
- `DnsNameResolver` does not send DNS queries for AAAA when IPv6 is not available.
2020-04-07 12:08:04 +02:00
feijermu
3ebf3d7705
Close the file in case of an IOException in AbstractMemoryHttpData.renameTo. (#10163)
Motivation:

An `IOException` may be thrown from `FileChannel.write` or `FileChannel.force`, and cause the fd leak.

Modification:

Close the file in a finally block.

Result:

Avoid fd leak.
2020-04-06 10:12:46 +02:00
Cenjie Ho
363c4ecc7f
Add the missing comment (#10169)
Motivation:

The comment of SingleThreadEventExecutor#run() is missing

Modification:

Just fill it in.

Result:

Method comment is not blank now.
2020-04-06 09:05:29 +02:00
Dmitry Konstantinov
05b7af7ae4
Propagate ref to pool thread cache down in the allocation stack to avoid extra thread local lookup (#10166)
Motivation:
PoolChunk requires a link to a PoolThreadCache to init ByteBuf. Currently the link is retrieved from a thread local: arena.parent.threadCache().
It has some performance cost. At the beginning of the allocation call the PoolThreadCache is already retrieved from the thread local. The reference can be propagated through the calls and used.

Modifications:
Replace second lookup of PoolThreadCache during ByteBuf init by propagation of a reference to PoolThreadCache down in the allocation stack explicitly

Result:
Improve performance of ByteBuf allocation
--Before--
Benchmark                                            (size)  (tokens)  (useThreadCache)  Mode  Cnt    Score   Error  Units
SimpleByteBufPooledAllocatorBenchmark.getAndRelease     123         0              true  avgt   20   57.112 ± 1.004  ns/op
SimpleByteBufPooledAllocatorBenchmark.getAndRelease     123       100              true  avgt   20  222.827 ± 1.307  ns/op

--After--
Benchmark                                            (size)  (tokens)  (useThreadCache)  Mode  Cnt    Score   Error  Units
SimpleByteBufPooledAllocatorBenchmark.getAndRelease     123         0              true  avgt   20   50.732 ± 1.321  ns/op
SimpleByteBufPooledAllocatorBenchmark.getAndRelease     123       100              true  avgt   20  216.892 ± 3.806  ns/op
2020-04-06 08:01:01 +02:00
feijermu
bdaa935756
Release the temporary ByteBuf in case of unexpected exception in WebSocketUtil.base64. (#10160)
Motivation:

An exception may occur between ByteBuf's allocation and release. So I think it's a good idea to place the release operation in a finally block.

Modification:

Release the temporary ByteBuf in finally blocks.

Result:

Avoid ByteBuf leak.
2020-04-03 14:49:40 +02:00