Commit Graph

7138 Commits

Author SHA1 Message Date
Brendt Lucas
41d0a81691 Use ByteBufAllocator to allocate ByteBuf for FullHttpMessage Motivation: When converting SPDY or HTTP/2 frames to HTTP/1.x, netty always used an unpooled heap ByteBuf.
Modifications:
When constructing the FullHttpMessage pass in the ByteBuf to use via the ByteBufAllocator assigned via the context.

Result:
The ByteBuf assigned to the FullHttpMessage can now be configured as a pooled/unpooled, direct/heap based ByteBuf via the ByteBufAllocator used.
2016-02-17 19:55:52 -08:00
Scott Mitchell
839e2ca508 Revert JDK GCM direct buffer crash workaround
Motivation:
Commit 108dc23cab introduced a workaround due to a JDK crash when GCM cipher was used during an unwrap operation. Attempting to reproduce this issue with the latest JDK (1.8.0_72-b15) demonstrate that this issue no longer exists while it can be reliably reproduced on earlier JDKs (1.8.0_25-b17 and earlier)

Modifications:
- Remove the copy-to-heap-buffer workaround for JDK engine

Result:
Fixes https://github.com/netty/netty/issues/3256
2016-02-17 19:54:02 -08:00
Norman Maurer
0b1c82b254 Ensure read components can be discarded when COMPOSITE_CUMULATOR is used.
Motivation:

ByteToMessageDecoder must ensure that read components of the CompositeByteBuf can be discard by default when discardSomeReadBytes() is called. This may not be the case before as because of the default maxNumComponents that will cause consolidation.

Modifications:

Ensure we not do any consolidation to actually be abel to discard read components

Result:

Less memory usage and allocations.
2016-02-17 19:51:08 -08:00
Norman Maurer
74495fd27f Add advanced leakdetection for methods introduced by [#4842]
Motivation:

[#4842] introduced 4 new methods but missed to implement advanced leak detection for these.

Modifications:

Correctly implement advanced leak detection for these methods.

Result:

Advanced leak detection works for all methods as expected.
2016-02-17 09:40:42 -08:00
Moses Nakamura
f0f0b69d90 fixed "sensative" typo to read "sensitive" 2016-02-17 08:18:11 -08:00
Norman Maurer
94f2748f1b Upgrade to netty-tcnative-1.1.33.Fork13
Motivation:

netty-tcnative-1.1.33.Fork was released, we should upgrade. Also we should skip renegotiate tests if boringssl is used because boringssl does not support renegotiation.

Modifications:

- Upgrade to netty-tcnative-1.1.33.Fork13
- Skip renegotiate tests if boringssl is used.

Result:

Use newest version of netty-tcnative and be able to build if boringssl is used.
2016-02-17 08:16:35 -08:00
Xiaoyan Lin
333f55e9ce Add unescapeCsvFields to parse a CSV line and implement CombinedHttpHeaders.getAll
Motivation:

See #4855

Modifications:

Unfortunately, unescapeCsv cannot be used here because the input could be a CSV line like `"a,b",c`. Hence this patch adds unescapeCsvFields to parse a CSV line and split it into multiple fields and unescaped them. The unit tests should define the behavior of unescapeCsvFields.

Then this patch just uses unescapeCsvFields to implement `CombinedHttpHeaders.getAll`.

Result:

`CombinedHttpHeaders.getAll` will return the unescaped values of a header.
2016-02-15 15:26:15 -08:00
Xiaoyan Lin
ccb0870600 Add methods with position independent FileChannel calls to ByteBuf
Motivation

See ##3229

Modifications:

Add methods with position independent FileChannel calls to ByteBuf and its subclasses.

Results:

The user can use these new methods to read/write ByteBuff without updating FileChannel's position.
2016-02-14 20:37:37 -08:00
Scott Mitchell
b112673554 ByteToMessageDecoder ChannelInputShutdownEvent support
Motivation:
b714297a44 introduced ChannelInputShutdownEvent support for HttpObjectDecoder. However this should have been added to the super class ByteToMessageDecoder, and ByteToMessageDecoder should not propegate a channelInactive event through the pipeline in this case.

Modifications:
- Move the ChannelInputShutdownEvent handling from HttpObjectDecoder to ByteToMessageDecoder
- ByteToMessageDecoder doesn't call ctx.fireChannelInactive() on ChannelInputShutdownEvent

Result:
Half closed events are treated more generically, and don't get translated into a channelInactive pipeline event.
2016-02-12 16:15:17 -08:00
Scott Mitchell
06e29e0d1b HTTP/2 codec may not always call Http2Connection.onStreamRemoved
Motivation:
Http2Connection.onStreamRemoved is not always called if Http2Connection.onStreamAdded is called. This is problematic as users may rely on the onStreamRemoved method to be called to release ByteBuf objects and do other cleanup.

Modifications:
- Http2Connection.close will remove all streams existing streams and prevent new ones from being created
- Http2ConnectionHandler will call the new close method in channelInactive

Result:
Http2Connection.onStreamRemoved is always called when Http2Connection.onStreamRemoved is called to preserve the Http2Connection guarantees.
Fixes https://github.com/netty/netty/issues/4838
2016-02-12 16:01:37 -08:00
Norman Maurer
c6a3729e4c Ensure handlerAdded(...) and handlerRemoved(...) is always called from the right thread
Motiviation:

We should ensure that handlerAdded(...) and handlerRemoved(...) is always called from the EventExecutor that also invokes the other methods of the ChannelHandler. Also we need to ensure we always call handlerAdded(...) before any other method can be calld to ensure correct ordering.

Motifications:

- Ensure that the right thread is used to call the methods
- Ensure correct ordering
- Add tests

Result:

Respect the thread-model for handlerAdded(...) and handlerRemoved(...) and preserve correct ordering in all cases.
2016-02-12 11:18:40 -08:00
Scott Mitchell
fdc6a5e87f Leak Detector disclosing when records dropped
Motivation:
ResourceLeakDetector enforces a limit as to how large the queue is allowed to grow for stack traces in order to keep memory from growing too large. However it is not always clear when records are dropped, or how many have been dropped. This can make interpreting leak reports more difficult if you assume all information is present when it may not be. Also we should increase the limit (currently 4) when running tests on the CI servers.

Modifications:
- Increase leak detector record limit on CI servers from 4 to 32.
- Track how many records have been discarded and disclose this in the leak report.

Result:
Leak reports clarify how many records were dropped, and how to increase the limit.
2016-02-12 09:57:43 -08:00
Scott Mitchell
691bc1690e ByteBufUtil.writeUtf8 not consistent with String.getBytes(Charset)
Motivation:
f750d6e36c added support for surrogates in the writeUtf8 conversion. However exceptions are thrown if invalid input is detected, but the JDK (and slow path of writeUtf8) uses a replacement character and does not throw. We should behave the same way.

Modificiations:
- Don't throw in ByteBufUtil.writeUtf8, and instead use a replacement character consistent with the JDK

Result:
ByteBufUtil.writeUtf8 behavior is consistent with the JDK UTF_8 conversion.
2016-02-11 08:24:26 -08:00
Norman Maurer
cd56f87ca1 Remove invalid return
Motivation:

JNI_OnUnload(...) does not return anything (has void in its signature) so we should not try to return something.

Modifications:

Remove return.

Result:

Fix incorrect but harmless code.
2016-02-10 16:48:39 -08:00
Scott Mitchell
56e6e07b25 HTTP/2 RST_STREAM Regression f990f99
Motivation:
Commit f990f99 introduced a bug into the RST_STREAM processing that would prevent a RST_STREAM from being sent when it should have been. The promise would be marked as successful even though the RST_STREAM frame would never be sent.

Modifications:
- Fix conditional in Http2ConnectionHandler.resetStream to allow reset streams to be sent in all stream states besides IDLE.

Result:
RST_STREAM frames are now sent when they are supposed to be sent
Fixes https://github.com/netty/netty/issues/4856
2016-02-10 13:47:53 -08:00
Xiaoyan Lin
36aa11937d ByteBuf for Key instead of String for codec-memcache
Motivation:

The key can be ByteBuf to avoid converting between ByteBuf and String. See #3689.

Modifications:

Replace the type of key with ByteBuf.

Result:

The type of key becomes ByteBuf.
2016-02-08 15:21:24 -08:00
Xiaoyan Lin
f43dc7d551 Add unescapeCsv to StringUtil
Motivation:

See #3435

Modifications:

Add unescapeCsv to StringUtil

Result:

StringUtil has the counter part of escapeCsv: unescapeCsv
2016-02-08 14:35:30 -08:00
Scott Mitchell
a15ff32608 HttpObjectDecoder configurable initial buffer size
Motivation:
The initial buffer size used to decode HTTP objects is currently fixed at 128. This may be too small for some use cases and create a high amount of overhead associated with resizing/copying. The user should be able to configure the initial size as they please.

Modifications:
- Make HttpObjectDecoder's AppendableCharSequence initial size configurable

Result:
Users can more finely tune initial buffer size for increased performance or to save memory.
Fixes https://github.com/netty/netty/issues/4807
2016-02-07 21:23:29 -08:00
Xiaoyan Lin
f59392d9f5 Make "CorsConfigBuilder.allowNullOrigin()" public
Motivation:

"CorsConfigBuilder.allowNullOrigin()" should be public otherwise people can not set it. See #4835

Modifications:

Make "CorsConfigBuilder.allowNullOrigin()" public.

Result:

The user can call "CorsConfigBuilder.allowNullOrigin()" now.
2016-02-07 10:23:24 -08:00
Scott Mitchell
b9682a26b1 EPOLL dladdr unexpected return value
Motivation:
netty_epoll_native.c uses dladdr in attempt to get the name of the library that the code is running in. However the address passed to this funciton (JNI_OnLoad) may not be unique in the context of the application which loaded it. For example if another JNI library is loaded this address may first resolve to the other JNI library and cause the path name parsing to fail, which will cause the library to fail.

Modifications:
- Pass an addresses which is local to the current library to dladdr

Result:
EPOLL JNI library can be loaded in an environment where multiple JNI libraries are loaded.
Fixes https://github.com/netty/netty/issues/4840
2016-02-06 19:43:57 +01:00
Norman Maurer
f10d66b45e Epoll.isAvailable() must return false if sun.misc.Unsafe is not present.
Motivation:

Currently our epoll native transport requires sun.misc.Unsafe and so we need to take this into account for Epoll.isAvailable().

Modifications:

Take into account if sun.misc.Unsafe is present.

Result:

Only return true for Epoll.isAvailable() if sun.misc.Unsafe is present.
2016-02-06 09:59:12 +01:00
Norman Maurer
a88c89a426 Add maven profile to run tests with -Dio.netty.noUnsafe=true
Motivation:

We had reports of failures before when sun.misc.Unsafe was not present. We should run our tests also with it disable to ensure everything works even if sun.misc.Unsafe is not present on the system.

Modifications:

Add a new profile which allows to run tests without Unsafe (using -PnoUnsafe)

Result:

Better testing of netty for systems where sun.misc.Unsafe is not present.
2016-02-06 09:58:02 +01:00
Norman Maurer
65b3470456 [#4793] Correctly add newlines when encode base64
Motivation:

We not correctly added newlines if the src data needed to be padded. This regression was introduced by '63426fc3ed083513c07a58b45381f5c10dd47061'

Modifications:

- Correctly handling newlines
- Add unit test that proves the fix.

Result:

No more invalid base64 encoded data.
2016-02-06 09:56:21 +01:00
Scott Mitchell
acbf1b9e7e Update jetty-alpn-agent version
Motivation:
Builds fail with java 1.8.0_72 because jetty-alpn-boot has absorbed new code from openjdk and older version are now incompatible.

Modifications:
- Updated jetty-alpn-agent version

Result:
We can now build/develop using java 1.8.0_72
2016-02-06 09:55:31 +01:00
Jon Chambers
61f812ea2a Allow InputStreams for key/trust managers in SslContextBuilder
Motivation:

Sometimes it's easier to get keys/certificates as `InputStream`s than it is to
get an actual `File`. This is especially true when operating in a container
environment and `getResourceAsInputStream` is the best way to load resources
packaged with an application.

Modifications:

- Add read-from-`InputStream` methods to `PemReader`
- Allow `SslContext` to get keys/certificates from `InputStreams`
- Add `InputStream`-based setters for key/trust managers to `SslContextBuilder`

Result:

Callers may pass an `InputStream` instead of a `File` to `SslContextBuilder`.
2016-02-05 14:39:55 -08:00
Norman Maurer
0f91ad841d Fix possible testfailure due not waiting on Channel.close() (introduced by e220c56823) 2016-02-05 12:28:11 +01:00
Norman Maurer
0f3d47199b [#4834] Fix race in AddressResolverGroup
Motivation:

We miss to use synchronized when remove the executor from the map.

Modifications:

Add synchronized(...) keyword

Result:

No more race.
2016-02-05 09:29:13 +01:00
Fabian Lange
a51e2c8769 Expose Helper to obtain the "best" mac address.
Motivation:

The implementation of obtaining the best possible mac address is very good. There are many sub-par implementations proposed on stackoverflow.
While not strictly a netty concern, it would be nice to offer this util also to netty users.

Modifications:

extract DefaultChannelId#defaultMachineId code obtaining the "best" mac into a new helper called MacAddress, keep the random bytes fallback in DefaultChannelID.

Result:

New helper available.
2016-02-05 09:27:43 +01:00
Norman Maurer
d9f938ca03 [#4828] OpenSslContext throws UnsupportedOperationException when Unsafe not available
Motivation:

OpenSslContext constructor fails with a UnsupportedOperationException if Unsafe is not present on the system.

Modifications:

Make OpenSslContext work also when Unsafe is not present by fallback to using JNI to get the memory address.

Result:

Using OpenSslContext also works on systems without Unsafe.
2016-02-05 09:25:18 +01:00
Norman Maurer
19907030d1 [#4841] Fix segfault if UnpooledUnsafeHeapByteBuf.getShort(..) is used and UNALGINED access is not possible.
Motivation:

We missed to take the byte[] into account when try to access the bytes and so produce a segfault.

Modifications:

Correctly pass the byte[] in.

Result:

No more segfault.
2016-02-05 09:24:32 +01:00
Norman Maurer
75a2ddd61c [maven-release-plugin] prepare for next development iteration 2016-02-04 16:51:44 +01:00
Norman Maurer
7eb3a60dba [maven-release-plugin] prepare release netty-4.1.0.CR2 2016-02-04 16:37:06 +01:00
Norman Maurer
465a190c3f [#4805] Respect isAutoRead() once re-register Channel
Motivation:

When a channel was registered before and is re-registered we need to respect ChannelConfig.isAutoRead() and so start reading one the registration task completes. This was done "by luck" before 15162202fb.

Modifications:

Explicit start reading once a Channel was re-registered if isAutoRead() is true.

Result:

Correctly receive data after re-registration completes.
2016-02-04 15:34:24 +01:00
Norman Maurer
08a7ca3747 Correctly pass ChannelPromise on to the next ChannelOutboundHandler when use CombinedChannelDuplexHandler.
Motivation:

Due a regression introduced by e969b6917c we missed to pass the original ChannelPromise to the next ChannelOutboundHandler and so
may never notify the origin ChannelPromise. This is related to #4805.

Modifications:

- Correctly pass the ChannelPromise
- Add unit test.

Result:

Correctly pass the ChannelPromise on deregister(...)
2016-02-04 15:28:46 +01:00
Norman Maurer
7ef6db3ffd [#4754] Correctly detect websocket upgrade
Motivation:

If the Connection header contains multiple values (which is valid) we fail to detect a websocket upgrade

Modification:

- Add new method which allows to check if a header field contains a specific value (and also respect multiple header values)
- Use this method to detect handshake

Result:

Correct detect handshake if Connection header contains multiple values (seperated by ',').
2016-02-04 14:03:08 +01:00
Norman Maurer
a0758e7e60 [#4794] Support window size flag by default if ZlibCodecFactory supports it.
Motivation:

If the ZlibCodecFactory can support using a custom window size we should support it by default in the websocket extensions as well.

Modifications:

Detect if a custom window size can be handled by the ZlibCodecFactory and if so enable it by default for PerMessageDeflate*ExtensionHandshaker.

Result:

Support window size flag by default in most installations.
2016-02-04 14:01:40 +01:00
Norman Maurer
7a562943ad [#4533] Ensure replacement of decoder is delayed after finishHandshake() is called
Motivation:

If the user calls handshake.finishHandshake() we need to ensure that the user has the chance to setup the pipeline before any WebSocketFrames are read. Because of this we need
to delay the removal of the HttpRequestDecoder.

Modifications:

- Remove the HttpRequestDecoder via the EventLoop and so delay it which gives the user a chance to setup the pipeline after finishHandshake() completes
- Add unit test for this.

Result:

Less surpising and correct behaviour even if the http response and websocket frame are received in one read operation.
2016-02-04 13:57:35 +01:00
Trustin Lee
ef0e053202 Preserve the host name of address when parsing /etc/hosts file
Motivation:

When an InetNameResolver resolves a name, it is expected to reserve the
requested host name in the resolved InetAddress.

DefaultHostsFileEntriesResolver does not preserve the host name. For
example, resolving 'localhost' will return an InetAddress whose address
is '127.0.0.1', but its getHostString() will not return 'localhost' but
just '127.0.0.1'.

Modifications:

Fix the construction of parsed InetAddresses in HostsFileParser

Result:

Host name is preserved in the resolved InetAddress
2016-02-04 13:45:01 +01:00
Scott Mitchell
075a54af3e Native EPOLL Library Allows Shading
Motivation:
If Netty's class files are renamed and the type references are updated (shaded) the native libraries will not function. The native epoll module uses implicit JNI bindings which requires the fully qualified java type names to match the method signatures of the native methods. This means EPOLL cannot be used with a shaded Netty.

Modifications:
- Make the JNI method registration dynamic
- support a system property io.netty.packagePrefix which must be prepended to the name of the native library (to ensure the correct library is loaded) and all class names (to allow classes to be correctly referenced)
- remove system property io.netty.native.epoll.nettyPackagePrefix which was recently added and the code to support it was incomplete

Result:
transport-native-epoll can be used when Netty has been shaded.
Fixes https://github.com/netty/netty/issues/4800
2016-02-03 14:40:28 -08:00
Norman Maurer
eb1d9da76c Enable SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER when using OpenSslContext
Motivation:

We need to enable SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER when using OpenSslContext as the memory address of the buffer that is passed to OpenSslEngine.wrap(...) may change during calls and retries. This is the case as
if the buffer is a heap-buffer we will need to copy it to a direct buffer to hand it over to the JNI layer. When not enable this mode we may see errors like: 'error:1409F07F:SSL routines:SSL3_WRITE_PENDING: bad write retry'.
Related to https://github.com/netty/netty-tcnative/issues/100.

Modifications:

Explitict set mode to SSL.SSL_MODE_RELEASE_BUFFERS | SSL.SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER . (SSL.SSL_MODE_RELEASE_BUFFERS was used before implicitly).

Result:

No more 'error:1409F07F:SSL routines:SSL3_WRITE_PENDING: bad write retry' possible when writing heap buffers.
2016-02-03 11:29:07 +01:00
Norman Maurer
e220c56823 [#4746] Support SNI when using OpenSSL
Motivation:

When using SslProvider.OPENSSL we currently not handle SNI on the client side.

Modifications:

Correctly enable SNI when using clientMode and peerHost != null.

Result:

SNI works even with SslProvider.OPENSSL.
2016-02-03 10:46:10 +01:00
fu.jian
a06708f81b fix the issue netty#2944 in 4.1
Motivation:

fix the issue netty#2944

Modifications:

use - instead of =>, use ! instead of :> due to the connection is bidirectional. What's more, toString() method don't know the direction or there is no need to know the direction when only log channel information.
add L: before local address and R: before remote address.

Result:

after the fix, log won't confuse the user
2016-02-02 21:43:38 +01:00
Scott Mitchell
6312c2f00b CompositeByteBuf.addComponent always assume reference count ownership
Motivation:
The current interface for CompositeByteBuf.addComponent is not clear under what conditions ownership is transferred when addComponent is called. There should be a well defined behavior so that users can ensure that no leaks occur.

Modifications:
- CompositeByteBuf.addComponent should always assume reference count ownership

Result:
Users that call CompositeByteBuf.addComponent do not have to independently check if the buffer's ownership has been transferred and if not independently release the buffer.
Fixes https://github.com/netty/netty/issues/4760
2016-02-02 11:38:11 -08:00
Scott Mitchell
7a7160f176 HTTP/2 Buffer Leak if UTF8 Conversion Fails
Motivation:
Http2CodecUtil uses ByteBufUtil.writeUtf8 but does not account for it
throwing an exception. If an exception is thrown because the format is
not valid UTF16 encoded UTF8 then the buffer will leak.

Modifications:
- Make sure the buffer is released if an exception is thrown
- Ensure call sites of the Http2CodecUtil.toByteBuf can tolerate and
  exception being thrown

Result:
No leak if exception data can not be converted to UTF8.
2016-02-02 11:22:17 -08:00
Travis Haagen
a75dcb2756 Made it easier to use custom ChannelId instances with Channel implementations that rely on the AbstractChannel(Channel parent) constructor.
Motivation:

The AbstractChannel(Channel parent) constructor was previously hard-coded to always
call DefaultChannelId.newInstance(), and this made it difficult to use a custom
ChannelId implementation with some commonly used Channel implementations.

Modifications:

Introduced newId() method in AbstractChannel, which by default returns
DefaultChannelId.newInstance() but can be overridden by subclasses. Added
ensureDefaultChannelId() test to AbstractChannelTest, to ensure the prior
behavior of calling DefaultChannelId.newInstance() still holds with the
AbstractChannel(Channel parent) constructor.

Result:

AbstractChannel now has the protected newId() method, but there is no functional
difference.
2016-02-02 08:33:42 +01:00
Luke Daley
d97f17060f Support non chunked HTTP request bodies larger than Integer.MAX_VALUE.
Motivation:

Request bodies can easily be larger than Integer.MAX_VALUE in practice.
There's no reason, or intention, for Netty to impose this artificial constraint.

Worse, it currently does not fail if the body is larger than this value;
it just silently only reads the first Integer.MAX_VALUE bytes and discards the rest.

This restriction doesn't effect chunked transfers, with no Content-Length header.

Modifications:

Force the use of `long HttpUtil.getContentLength(HttpMessage, long)` instead of
`long HttpUtil.getContentLength(HttpMessage, long)`.

Result:

Netty will support HTTP request bodies of up to Long.MAX_VALUE length.
2016-02-02 08:28:27 +01:00
Scott Mitchell
f990f9983d HTTP/2 Don't Flow Control Iniital Headers
Motivation:
Currently the initial headers for every stream is queued in the flow controller. Since the initial header frame may create streams the peer must receive these frames in the order in which they were created, or else this will be a protocol error and the connection will be closed. Tolerating the initial headers being queued would increase the complexity of the WeightedFairQueueByteDistributor and there is benefit of doing so is not clear.

Modifications:
- The initial headers will no longer be queued in the flow controllers

Result:
Fixes https://github.com/netty/netty/issues/4758
2016-02-01 13:37:43 -08:00
Scott Mitchell
5fb18e3415 InboundHttp2ToHttpAdapter leak and logic improvements
Motivation:
In HttpConversionUtil's toHttpRequest and toHttpResponse methods can
allocate FullHttpMessage objects, and if an exeception is thrown during
the header conversion then this object will not be released. If a
FullHttpMessage is not fired up the pipeline, and the stream is closed
then we remove from the map, but do not release the object. This leads
to a ByteBuf leak. Some of the logic related to stream lifetime management
and FullHttpMessage also predates the RFC being finalized and is not correct.

Modifications:
- Fix leaks in HttpConversionUtil
- Ensure the objects are released when they are removed from the map.
- Correct logic and unit tests where they are found to be incorrect.

Result:
Fixes https://github.com/netty/netty/issues/4780
Fixes https://github.com/netty/netty/issues/3619
2016-02-01 12:38:40 -08:00
Trustin Lee
4d6ab1d30d Fix missing trailing data on HTTP client upgrade
Motivation:

When HttpClientUpgradeHandler upgrades from HTTP/1 to another protocol,
it performs a two-step opertion:

1. Remove the SourceCodec (HttpClientCodec)
2. Add the UpgradeCodec

When HttpClientCodec is removed from the pipeline, the decoder being
removed triggers channelRead() event with the data left in its
cumulation buffer. However, this is not received by the UpgradeCodec
becuase it's not added yet. e.g. HTTP/2 SETTINGS frame sent by the
server can be missed out.

To fix the problem, we need to reverse the steps:

1. Add the UpgradeCodec
2. Remove the SourceCodec

However, this does not work as expected either, because UpgradeCodec can
send a greeting message such as HTTP/2 Preface. Such a greeting message
will be handled by the SourceCodec and will trigger an 'unsupported
message type' exception.

To fix the problem really, we need to make the upgrade process 3-step:

1. Remove/disable the encoder of SourceCodec
2. Add the UpgradeCodec
3. Remove the SourceCodec

Modifications:

- Add SourceCodec.prepareUpgradeFrom() so that SourceCodec can remove or
  disable its encoder
- Implement HttpClientCodec.prepareUpgradeFrom() properly
- Miscellaneous:
  - Log the related channel as well When logging the failure to send a
    GOAWAY

Result:

Cleartext HTTP/1-to-HTTP/2 upgrade works again.
2016-02-01 15:52:37 +01:00
Xiaoyan Lin
17df8171b3 Use AbstractReferenceCounted to clean up the codes for codec-memcache
Motivation:

Some duplicated methods in message types of codec-memcache can be cleaned using AbstractReferenceCounted.

Modifications:

Use AbstractReferenceCounted to avoid duplicated methods.

Result:

Duplicated methods are cleaned.
2016-02-01 15:51:08 +01:00