* Avoid creating listener list for rarely used requestAttributeListener
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #3964
Keep a list of EventListeners in the AbstractConnector to make it
more efficient to add and iterate over them.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #2578 EventListener
Use addEventListener rather than bespoke listener methods.
Support getEventListenerBeans at Container level for fast lookup
improve javadoc
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #2578 EventListener
fixed test
more javadoc
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #2578 EventListener
fixed tests
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #3964
Don't use null for empty lists of listeners
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* fix merge
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #2578 EventListener
Resolve differences between eventListeners added as beans and beans
added as EventListeners. The behaviour should now be the same
regardless of how they listener is added and all listeners are now
beans.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #2578 EventListener
Add only SelectorManager listeners to manager from connector
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #2578 EventListener
Fixed javadoc
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #2578 EventListener
removed old TODO
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #2578 EventListener
connector cannot be null
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #2578 EventListener
javadoc
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #3964
AbstractConnector keeps a specific list of HttpChannel.Listeners
to avoid Connection.Listeners and MBean listeners being added to
the HttpChannel listener list.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #2578 EventListener
fixed merge
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* fixed javadoc
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #2578 EventListener
removed the ability to set/clear context listeners
Instead just remove non-durable ones.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #3964 Listeners
Simplified listener handling by avoiding null connector, previously
only needed for testing.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* fixed bad merge
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #3964 Listeners
Fixed test that assumed HttpChannel listeners were not cleared by a recycle
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #3964 Listeners
Separated out durable vs cyclic HttpChannel.Listeners, so as to
simplify handling.
Deprecated cyclic HttpChannel.Listeners, as I'm not sure the channel is
the right place for them.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* removed deprecated cyclic HttpChannel listeners
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* removed deprecated cyclic HttpChannel listeners - import
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #4003 Cleanup quickstart
* Fixed tests that scan for "Started" on console
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* updates from review
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #3863 - Enforce use of SNI.
Introduced SslContextFactory.rejectUnmatchedSNIHost (default false)
so that if no SNI is sent, or SNI does not match a certificate,
then the TLS handshake is aborted.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
* Issue #3863 - Enforce use of SNI.
Updates after review.
Introduced SslContextFactory.SNISelector to allow application to write
their custom logic to select a certificate based on SNI information.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
* Issue #3863 Enforce SNI
Added two sniRequired fields - one at SslContextLevel and the other at the SecureRequestCustomizer. This allows rejection either at TLS handshake or by 400 response.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #3863 Enforce SNI
cleanups from review
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #3863 Enforce SNI
improved comments
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #3863 Enforce SNI
syntax sugar
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #3863 SNI
Updates from review. Extra test for sniSelector function
Signed-off-by: Greg Wilkins <gregw@webtide.com>
Made method reportDifferences(...) private since it was exposing
package private class TimeNSize and no code outside of jetty-util
could have used it.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
* issue exclude/include con name InetAccesHandler - add better unit test
this logic:
String name =
baseRequest.getHttpChannel().getConnector().getName();
return _names.test(name) && _addrs.test(addr);
Is not correct. it's treating the connector name exactly like the
filter. But that's not what it's intended to do. It's supposed to tell
what connectors are applicable to this filter. And what connectors are
not affected.
For example in the unit test there exists 2 connectors:
http
tls
We want to restrict the http connector, but we want to leave tls
connector alone.
So we would specify:
include = 192.168.1.1-192.168.1.254
includeConnector = http
The way the logic is above, it is treating the connector name as if it's
the filter itself. Which is not what I intended.
What i need in psuedo-code is this:
if (there are no "include connectors" OR if this connector is
included) AND (if this connector is not in the excluded list)
---> Then apply the IP filter.
Signed-off-by: Nicholas DiPiazza <nicholas.dipiazza@lucidworks.com>
* exclude should take precedence over include
Signed-off-by: Nicholas DiPiazza <nicholas.dipiazza@lucidworks.com>
* Issue #4193 InetAccessHandler
reverted changes to IncludeExcludeSet
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #4193 InetAccessHandler
updates from review
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #4188 Spin in close of GzipHandler
Cleanup and simplify code
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #4188 Spin in close of GzipHandler
Fix slice code. Added unit test for it.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #4188 Spin in close of GzipHandler
Fixed last slice.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* cleanup from review
Signed-off-by: Greg Wilkins <gregw@webtide.com>
There is a race between the doStop clearing the key map and the watching thread
checking isRunning before iterating over the key map.
While more sophisticated approaches could be used, I think that is best to defer
until this class is reworked entirely. For now just using a ConcurrentHashMap will
avoid the exception and the closing of the pathwatcher will prevent watching forever.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
Fixed setting of host/port in AbstractConnectorHttpClientTransport
and HttpProxy so that the creation of SSLEngine can use the proper
host/port pair, and can be subsequently used in TLS components.
Introduced SslContextFactory X509ExtendedKeyManagerWrapper and
X509ExtendedTrustManagerWrapper as utility classes used internally
and in tests.
The test case for this issue required 3 keystores, so other test
classes have been refactored to use the new keystores.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Several QTP fixes:
* #4105 Threads without jobs now check if they should idle die before waiting rather than before, this allows idling under steady load. 3ad6780
* #4121 ThreadFactory behaviour supported by doing thread config within newThread call. 7b306d7
* #4122 Always clear the interrupted status. c37a4ff
task = queue.poll(timeout);
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #4105 - starting a thread in QTP now increments idle count
Signed-off-by: Lachlan Roberts <lachlan@webtide.com>
* Issue #4105 - improve comments in test
Signed-off-by: Lachlan Roberts <lachlan@webtide.com>
Added test to reproduce issue
Fixed bug from #2772 where output was shutdown on DONE without checking for END.
Fixed aggregation logic to aggregate last write if aggregation already started
Improved comments and clarify conditions
Signed-off-by: Greg Wilkins <gregw@webtide.com>
Now using the _size AtomicInteger to resolve the race between stopping
and adding to the stack.
A _size of -1 now means no more threads can be added to the stack, and
the loop in doStop() will now only exit if it can set the size to -1.
Signed-off-by: Lachlan Roberts <lachlan@webtide.com>
Introduced a Response.DemandedContentListener to explicitly separate
the will to request more content from the notification that the content
has been consumed.
Updated all transports to follow the new semantic: rather than waiting
for the callback to complete before delivering more content, now they
wait for the demand to be positive to deliver more content.
Since now the content may be unconsumed but there can be more demand,
all transport implementation had to be changed to use RetainableByteBuffer
to retain content buffers that were not consumed.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
* Issue #4003 Fix quickstart.
Modes:
AUTO: use quickstart xml if present, start normally otherwise
GENERATE: re/generate quickstart xml
QUICKSTART: use a pregenerated quickstart xml, fail otherwise
Signed-off-by: Jan Bartel <janb@webtide.com>
* Issue #4003 Cleanup quickstart
Removed TerminateException in favour of Server.setDryRun(boolean)
and AbstractLifeCycle.StopException
Signed-off-by: Greg Wilkins <gregw@webtide.com>
Removes the Locker class, replaced by AutoLock.
Removed usages of Locker.isLocked() from the session code
since it was not necessary.
Took the chance to do a little code cleanup.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Code cleanups and reformatting.
Fixed logic for SETTINGS frame replies: they are not subject to rate control.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Issue #3964
* Avoid creating listener list for rarely used requestAttributeListener
* AbstractConnector keeps a specific list of HttpChannel.Listeners
to avoid Connection.Listeners and MBean listeners being added to
the HttpChannel listener list.
* Simplified listener handling by avoiding null connector, previously
only needed for testing.
* Fixed test that assumed HttpChannel listeners were not cleared by a recycle
* Separated out durable vs cyclic HttpChannel.Listeners, so as to
simplify handling.
* Deprecated cyclic HttpChannel.Listeners, as I'm not sure the channel is
the right place for them.
* Added improved method to combine multiple HttpChannel Listeners
into a single Listener.
* Fixed MockConnector
* Added benchmark
* Improved benchmark
* Updates from review
* Removed benchmark and alternate implementations.
* Updated javadoc
* Updates from review
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #3806 async sendError
Avoid using isHandled as a test withing sendError as this can be
called asynchronously and is in a race with the normal dispatch of the
request, which could also be setting handled status.
The ErrorHandler was dispatching directly to a context from within
sendError. This meant that an async thread can call sendError and be
dispatched to within the servlet container at the same time that the
original thread was still dispatched to the container.
This commit fixes that problem by using an async dispatch for error
pages within the ErrorHandler. However, this introduces a new problem
that a well behaved async app will call complete after calling
sendError. Thus we have ignore complete ISEs for the remainder of
the current async cycle.
Fixed the closing of the output after calling sendError. Do not
close if the request was async (and thus might be dispatched to an
async error) or if it is now async because the error page itself is
async.
* updates from review
* better tests
* revert ignore complete
* added some TODOs
* more TODOs
* fixed rename
* cleanup ISE and more TODOs
* refactored to call sendError for uncaught exceptions rather than onError
* more of the refactor
* extra tests for sendError from completing state
Reworked HttpChannelState and sendError so that sendError is now
just a change of state. All the work is done in the ErrorDispatch
action, including calling the ErrorHandler. Async not yet working.
Additional tests
Converted ERRORED state to a separate boolean so it can be used for
both Sync and Async dispatches.
Removed ASYNC_IO state as it was just the same as DISPATCHED
The async onError listener handling is now most likely broken.
WIP making sendError simpler and more tests pass
WIP handling async and thrown exceptions
WIP passing tests
Improved thread handling
removed bad test
Implemented error dispatch on complete properly
more fixed tests
sendError state looks committed
- Added resetContent method to leave more non-content headers during sendError
- Fixed security tests
- simplified the non dispatch error page writing. Moved towards being able to write async
* fixed gzipHandlerTest
* Updated handling of timeout errors. According to servlet spec,
exceptions thrown from onTimeout should not be passed to onError, but
just logged and ignored:
If an exception is thrown while invoking methods in an AsyncListener,
it is logged and will not affect the invocation of any other AsyncListeners.
* This changes several tests.
* Dispatcher/ContextHandler changes for new ERROR dispatch handling. Feels a bit fragile!
* Fixed tests in jetty-servlets
* Fixed tests in jetty-proxy
* more test fixes
* Fixed head handling
reverted unnecessary changes
Improved reason handling
WIP on fully async error handling.
Simplified HttpChannelState state machines to allow for async actions
during completing
more WIP on fully async error handling.
sendError and completion are not both non-blocking, without using
a startAsync operation. However we are lacking unit tests that actually
exercise those code paths.
* Simplified name of states
Added test for async completion
* Cleanups and javadoc
* Cleanups and javadoc
* remove snake case
* feedback from review
* Write error page into fixed pooled buffer
Use the response to get/release a pooled buffer into which the error
page can be written. Make it a fixed sized buffer and if it overflows
then no error page is generated (first overflow turns off showstacks
to save space).
The ErrorHandler badly needs to be refactored, but we cannot change
API in jetty-9
* More test fixes for different error page format
* minor cleanups
* Cleanup from Review
* Fixed javadoc
* cleanups and simplifications
* Cleanup from Review
* renaming and some TODOs
* Cleanup from Review
* Checkstyle fixes
* Cleanup from Review
* Code cleanups and simplifications
* fixed debug
* Cleanup from Review
* Ensure response sent before server shutdown
* removed unnecessary optimisation
* fixed duplicate from merge
* Updates from review
Signed-off-by: Greg Wilkins <gregw@webtide.com>
+ Adding SameFileAliasChecker to help with environments
where the Alias and the Path point to the same file,
by relying on the FileSystem and Path implementation
to make the determination if the two Path references
truly point to the same file.
+ Minor cleanup of FileSystemResourceTest
+ Additional DefaultServletTest for this UTF-8 differences
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
+ OSX File is `swedish-å.txt`
+ OSX has NFD form UTF-8 characters. `swedish-a%CC%8A.txt`
+ HTTP uses normal form UTF-8. `swedish-%C3%A5.txt`
+ A HTTP GET request should work against the resource
being requested, regardless of UTF-8 style used.
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Updated code after review.
Now a managed failed bean is restarted if its container is restarted.
Added more test cases.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Removed usage of StringUtil.
Moved fields together at top of the class to avoid
class initialization ordering issues.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Previously beans were started if they were not running, which
included them being in a STOPPING state.
Now beans are only started if they are in STOPPED state.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Updated code to reflect reviews.
Now lookup of system properties and server attributes is done in
ContextHandler.doStart(), so that the getter always return the
actual value (and this is good for JMX too).
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
+ Construction of PathResource now tests if path belongs
to the Default FileSystem or not. This important info
for later actions against the PathResource that would
need to know the File object for the Path object.
Non-Default FileSystem == null
Default FileSystem == Path.toFile()
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
+ SPARSE hint only applies to real os file systems or
default file systems, not for all file systems.
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
+ Reverting toFile().getInputStream() on PathResource
+ Adding RangeWriter concept for managing open resource
across multiple range writes
+ RangeWriter implementation delegates to HttpContent behaviors
Lookup is :
- Direct Buffer
- Indirect Buffer
- ReadableByteChannel (as SeekableByteChannel)
- InputStream
+ Adding unit tests for all RangeWriter implementation to ensure
that they behave the same way everywhere.
+ Making ResourceService use new RangeWriter implementation
+ Existing DefaultServletRangeTest still works as-is
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Made TRUST_ALL_CERTS implement X509ExtendedTrustManager,
so that it does not get wrapped in sun.security.ssl.AbstractTrustManagerWrapper,
which performs additional trust checks.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
+ Allowing CachedHttpContent._contentLengthValue actually hold
the `long` resource size (for the 4G variant on test)
+ Allowing BufferUtil to not throw Exception if resource length
is a positive value, but exceeds Integer.MAX_VALUE, opting instead
to return a null to prevent excessive memory usage. (fixes the 10G
variant of test)
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Issue #3700 if a location is null then consider if there are inclusions and/or exclusions when matching a class.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Removing Legacy Method Separators
* Restyling branch `jetty-9.4.x`
* Applying changes highlighted by checkstyle
* Applying XML restyling
* Fixing XML codestyle for IntelliJ
* Fixing XML style mistakes
* Revert "Applying XML restyling"
* Updating checkstyle for XML codestyle
* Reformatting pom.xml files
* Fixed empty string from line wraps
* Update intellij style to not do expression relative formatting. Reformatted code based on that.
* Increasing line split on Eclipse IDE Formatter to 512
* Restoring setting on internal default value.
+ IntelliJ will not export settings on things that set to their
internal default values.
We want to keep those values as a hedge against future default
value changes in future releases of IntelliJ.
* Fixing intellij codestyle
* do not allow single line simple methods
* misc checkstyle fixes
* re-exported with correct name and all values
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Signed-off-by: Greg Wilkins <gregw@webtide.com>
+ Reverting commit against IncludeExcludeSet
+ TypeUtil.getLocationOfClass is now MethodHandle based
+ new ModuleLocation class provides Module behaviors for Java 9+
Runtimes, and is used by new MethodHandle based getLocationOfClass
+ Jetty 10 doesn't need reflect complexity of ModuleLocation class
and can be inlined as real code in TypeUtil
+ TypeUtil.getLocationOfClass is only guaranteed to return the
URI to the container holding the location of the class, or null.
- The JAR file - most common
- The Directory - 2nd most common - a classpath dir entry
- The jrt:// module reference - if a module reference
+ URLResource support for results of getLocationOfClass is deprecated
(nothing was using it anyway)
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
+ TypeUtil.getLocationOfClass() respects Class.getClassLoader() of null
which means a class belonging to Boot ClassLoader, a Primitive, Void,
or a dynamic in-memory class. Using system classloader is incorrect
and invalid in Java 9+
+ Fixing IncludeExcludeSet.test() to always return TRUE or FALSE
never null.
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
* Issue #3755 Annotation example cleanup
+ Created JettyDistribution class as common utility to locate a jetty distribution for examples.
+ Fixed ServerWithAnnotations to correctly use the test-spec-webapp
+ Added AttributeContainerMap as a better way to treat attribute values as beans. This avoids them appearing twice in a dump and always associates them with their key.
+ Added NamingDump and use it in EnvConfiguration and jetty-plus.xml so that a server dump will contain dumps of the server local tree and each contexts java:comp/env tree
+ Improved the dump format of NamingContext and WebAppContext
+ Improved the toString format of several associated classes
Signed-off-by: Greg Wilkins <gregw@webtide.com>
+ StringUtil.replace()
+ StringUtil.replaceFirst()
+ StringUtil.sanitizeFileSystemPath()
Change existing usages of String.replace() to either
use new StringUtil.replace() or other methods elsewhere
that better suit that specific need.
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
* upgrade log4j2 version and fix log4j2.xml provided sample, add disruptor if users wants to use async logging
Signed-off-by: olivier lamy <oliver.lamy@gmail.com>
+ Removed regex usage
+ More unit tests
+ Adding jmh test for improvements (see PR #3716 for published results
of old vs new implementation)
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Added warnings for invocation of deprecated constructors and methods.
Moved methods exclusively used by XmlConfiguration from TypeUtil.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
+ Deprecated B64Code
+ All code that isn't B64CodeTest is now using java.util.Base64
+ B64CodeTest is updated to confirm change to java.util.Base64
is possible without change in behavior. Just have to make
sure you use the appropriate Encoder / Decoder for the task
at hand (default vs mime vs url)
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Added support for the X-Forwarded-Port header.
Reimplemented header scanning using more efficient Trie and MethodHandles
Signed-off-by: Greg Wilkins <gregw@webtide.com>
previously if there were no idle threads, QueuedThreadPool.execute()
would just queue the job and not start a new thread to run it
Signed-off-by: lachan-roberts <lachlan@webtide.com>
previously if there were no idle threads, QueuedThreadPool.execute()
would just queue the job and not start a new thread to run it
Signed-off-by: lachan-roberts <lachlan@webtide.com>
+ Even though this was reported against Windows, the solution
implemented should be sane for all OS or FileSystem combinations.
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
+ Even though this was reported against Windows, the solution
implemented should be sane for all OS or FileSystem combinations.
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
+ If optional Security Provider is configured, then use it consistently
for all security objects that take a Provider argument.
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
- Fixes deflater/inflater pool validation
- Removes validation tests of Flators for pool
- Eroding pool usage; fixes cleanup of Deflaters
Signed-off-by: bkmz <ilya.cherkasov@gmail.com>
as the QueuedThreadPool was being stopped, jobs could be taken out of
the QueuedThreadPool job queue but then not run
also cleaned up QueuedThreadPool _runnable code to simplify it
Signed-off-by: Lachlan Roberts <lachlan@webtide.com>
Introduced SslContextFactory subclasses Client and Server.
Replaced all usages of SslContextFactory with either Client or Server
as required.
Refactored configuration checking so that warnings are not emitted
when non necessary.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
pass the cause from the AbnormalCloseStatus to closeConnection in
WSChannel
If onOpen throws we now throw a runtime exception in addition to
failing the callback, this is to fail the CompletableFuture in the
ClientUpgradeRequest.
Signed-off-by: lachan-roberts <lachlan@webtide.com>
+ Changing usages of JVM deprecated classes / methods as well
+ Cleaning up test cases and javadoc
+ Removing test methods that rely on deprecated and now removed concepts
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
* add jdk12 build stage
* test last release of infinispan for jdk12 build
* fix upperBound dependency for jboss logging
* upgrade juniper to 5.4.0 and disable infinispan test if JRE.JAVA_12
* Fixed annotations to exclude any JDK > 11 for 3 tests.
Signed-off-by: olivier lamy <oliver.lamy@gmail.com>
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
After review, keeping "requires static" only if the
module tree can be built without the static module.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Reviewed module-info.java files.
For those that have a "requires static" added a comment about
why the dependency is optional.
Rearranged directives in alphabetical order.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Added a SslHandshakeListener to SslConnection that performs
the host name verification (only on the client) if the
HostnameVerifier has been configured in SslContextFactory.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Changes after review:
+ removed Adaptor from FrameHandler, so all non test usages of
FrameHandler now use async API.
+ Fixed sequencing of multiple async operation so callback is notified
after completion (created more Callback.from utilities for this)
+ fixed import order
Signed-off-by: Greg Wilkins <gregw@webtide.com>
Made default endpointIdentificationAlgorithm=HTTPS to avoid warnings
when SslContextFactory is used on a server.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Improve ws error handling by splitting processError into handling for
errors from the network and errors from the application.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
+ Do not rely on Object.toString() for representing self (aka:this)
in the dump output. Using Object.toString() can result in lots
of duplicated information in the dump (and super long lines)
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Improve on #3167 with major refactor of the context initialization of
websocket:
+ The Javax and Jetty sides are more symmetric - both use shared
filter and mapping.
+ Regularised usage of beans rather than attributes for ws components
+ Customization is now part of the mapping, so ws are configured by
how they were mapped and not by who does the upgrade.
+ Filter still can be configured to customize defaults
+ Servlet can be configured to customize any ws mappings added via the
servlet
There is still some strangeness as the WebSocketServlet is mostly
generic, yet can only map Jetty API websockets.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Updated module-info.java to reference the "jetty.servlet.api" module.
* Updated POMs to reference the o.e.j.toolchain:jetty-servlet-api artifact.
* Removed references to jetty-schemas.jar.
* Updated attribute "org.eclipse.jetty.server.webapp.ContainerIncludeJarPattern"
to match the new Jetty Servlet API jar.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Fixes#2978 - Add module-info to relevant Jetty modules.
Added module-info.java for Jetty modules that are not test modules.
Moved jetty-http test utility classes to new module "jetty-http-tools".
Removed generation of test-jar from websocket-core.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Fixes Issue #3103 - HttpClient buffer leak found by dump improvements
The leak problem was an additional, unnecessary, call retain() in ResponseNotifier.notifyContent() that was leaking `ByteBuffer`s.
* Adding more todos
* More websocket test updates
* Adding more information to exception about mismatched hash
* AcceptHash improved testing
* Prevent duplicate request headers
* Removing ability to use LocalFuzzer entirely (until its fixed)
* JSR356 HandshakeResponse header modifications are now allowed during Configurator
* Correcting MessageSink creation
* Using proper sub-protocol
* Metadata remembers OnMessage.maxMessageSize now
* Minor test updates
* Issue #2172 - Fixing JSR-356 Decoder discovery and OnMessage wiring
* More websocket test updates
* Update for null/empty payload
* Partial String Message Sink fix
* More websocket fixes around InputStream handling
* Moving proposed Abstract FrameHandler hierarchy into core main
* More testing fixes, along with a change to use the new core Abstract*FrameHandler
* Adding WebSocket URI validation
* Updating copy found in other API with improvements made in first API.
* More websocket test updates
* Fixing compilation
* WebSocket test fixes
* Fixing ContinuationTest assertions
* fixed package
* Added some better examples of QuotedCSV
I think this test is trying to use QuotedCSV in the wrong way, as it often is offering values when it should be testing parameters.
I have added some examples to show that QuotedCSV is capable of handling the quotes, spaces and embedded commas of the tests, but only
for parameters and not for values. The question is, is this sufficient for websocket?
* Improve close handling
The handling of a Close Frame may close the endpoint, so parsing should not continue
* moved old AbstractFrameHandler to test
* implement ingoing and outgoing frame sequence checking
* More work on javax.websocket parameterized server
* Bumping up pom plugin versions to jetty-9.4.x standards
* Merging up Jenkinsfile to attempt to fix CI build issues
* JSR356 @PathParam testing
* More WebSocket testing updates
* More WebSocket testing updates
* WebSocket Idle Timeout from FrameHandler specific policy
* WebSocket/JSR Pong support
* WebSocket/JSR Correcting MessageHandler.Whole<PongMessage> handling
* URL Decode UriTemplatePathSpec values
* Fixing test expectations
* Fixing AbstractWholeMessageHandler buffer flip
* More websocket test fixes
* Updated ALPN version for JDK 8u181.
* Removing QuotedUtil from WebSocket-core
+ Moved to QuotedCSV for most usages
+ Introduced HeaderUtil to satisfy remaining use cases
* Review of HttpClient names and Thread names.
+ Javax WebSocket Server = "Javax-WebSocketServer@####"
+ Javax WebSocket Client = "Javax-WebSocketClient@####"
* added OpCodeTest.java
* Minor Frame cleanup
* MessageOS cleanup
* Made the OpCode of WebSocket Frame immutable.
Frames can no longer be set as a continuation and should now be copied using the ContinuationFrame constructor
DataFrames can no longer be created without one of the subclasses as the public constructor was removed in favour of newDataFrame method
* Chat test
* Work on Chat Test
Moved implementation of assertValid to WebSocketFrame
Removed getFinRsvOp
* removed constructors for DataFrame using a basedOn Frame
instead use constructors with opCode, payload and fin bit
* refactor of WebSocket Frames WIP
* WIP: resolved failing tests
* CloseStatus rework, now use CloseStatus.toFrame to create Close Frames
new DelegateFrame class which ReadOnlyFrame extends
trim utf8 byte array in CloseStatus to utf8 character boundaries only
* working on test to test socket io erros in websocket-core
* renaming for clarity
Gave up on maintaining old names, as this is a different package anyway so some adaption will be needed regardless.
* Moved AbstractFrameHandlers to common
Only have a AbstractTestFrameHandler in core
* Close uses CloseStatus
Reduced the parsing and generation of CloseStatus when closing
* merge
* work on WebSocketTest
* cleanup log
* wip
* files missing comment header
jetty-jmh was using 9.4.12-SNAPSHOT now using 10.0.0-SNAPSHOT
* thread names
* remove sequence check from parser
* minor cleanups
* Call frameHandler externally to parser
* OpCode.check now throws ProtocolException instead of returning boolean
changed ProtocolException messages in OpCode.check
* removed Frame.Type and use OpCode instead
separated out FrameValidation for parser into method in WebSocketChannel
* moved remaining frame validation checks into WebSocketChannel
* referenced counted buffer
* cleanups
* wip on demand side
* First attempt at demandable
* demand control
* clean up reference counting
* clean ups
* WebSocketServerTest
* wip
* fixed auto fragmentation
* simple demand test
* test demand and retain
* Added validation back into the Parser
split up validation in WSChannel into methods for incoming and outgoing
fixed various tests in websocket-core
* more tests
* fixed remaining tests failing due to the changes to frame validation
* cleanups
* more autobahn refixes
* adjust mask for auto fragments
* clear mask
* fully close on unrecoverable error and not wait for close response
* Removed the CoreFuzzer and ContinuationTest
These tests are now covered elsewhere.
* remove declared exception onClosed
* Test delay close
* Test handler close
* fixed test dependency
* moved the frame sequence check from OutgoingState from after the ExtensionStack to channel.sendFrame
changed error handling to fail the callback of invalid/outOfSequence frames
added WebSocket Close tests revealing some problems with demand in state ICLOSE
* cleanup
* timing issues with WebSocketCloseTest
* test cleanups
* Cancel demand on close
* removed frame validation from generator
* finished close testing
* Issue #2901 - Moving Http Upgrader to Http Conversation
+ Solves for Authentication and Redirect
* changes from #2902 to jetty-10.0.x
* Validation Extension
introduced a ValidationExtension to do frame validation in the ExtensionStack
added the WebSocketChannel to ExtensionStack.connect to set it on the Extensions
* added parameters for ValidationExtension to choose what to validate
* ValidationExtension Cleanup
ValidationExtension now uses the Validation methods from its WebSocketChannel
simplified the parameter passing to the ValidationExtension
* core cleanup
* Fixed intermittent test failures
* clean up duplicates and imports
* simplified package structure
* many cleanups, simplifications and duplication removal
* cleanup
* added utf8 validation for text frames into ValidationExtension
added tests for utf8 validation
* create ValidationExtensionTest
* cleanup
* merge with jetty-10
update websocket to junit5
* fix to autobahn client report directory
* Fixes and cleanup
Fixed initial buffer too large for websocket buffer
less verbose debug
better debug
* Removed behavior from policy
* cleanup after merge from hell
* bug fixes and cleanups
* refactor packages with JPMS in mind
* unit test for TextMessageHandler
* cleanups
* more cleanups
* provided client/server dependencies
* fixed bad refactor
* use lambdas for TextMessageHandler
* cleanup
* cleanup
* cleanups
* cleanup of websocket-common framehandlers
* reverting to jetty websocket api from 9.4
* better chat
* renamed websocket common to util
* save wip on jetty-websocket-tests
* giving up on jetty-websocket-tests as mostly duplicated tests
* renamed websocket-util back to websocket-common
* moved InvalidWebSocketException to javax common
* reworked TextMessageHandler to MessageHandler
now handles binary messages as well
* implemented getByteBufferPool in remaining DummyChannel classes
* Moved the abstract Frame and Message handlers out of websocket-common
* replaced usages of AbstractWholeMessageHandler with CoreMessageHandler
renamed MessageHandler to CoreMessageHandler
* removed AbstractFrameTypeHandler
* cleanup
* jetty-websocket-common passing tests without websocket-common dependency
* fixed compile errors
* renamed MessageHandler
* Removing reference to websocket-server (no longer exists)
* Removing, Extensions now exist in websocket-core
* Do not append an empty ByteBuffer
* NPE Fix for empty/unused UTF-8 String Builder
* NPE Fix for SendPartialBinaryFrameHandler test
* Punching holes in WebAppContext needed for JSR356 use
* Updating "9.x.x" to "10.x.x"
* Removing Bad URI test case that is ignored
* Fixing test expectation now that we moved to QuotedCSV
* Adding TODO about onOpen failure requiring onError + onClose.
* Throwing JSR356 DeploymentException on addEndpoint() failure
* Test cleanup (for better error messages)
* Adding some comments to explain wrapNonVoidReturnType()
* cleanup methodHandle return type filtering
* test fixes
* Fix change return type to Void
* Fixed test by wiring up MessageHandler
* test cleanups
removed changeReturnType for MethodHandle as filterReturnValue already changes to return type of the filter
* fixed MessageReceivingTest.testPartialBinaryFrameHandler
* Core BatchMode replaced by boolean
* fixed LargeContainerTest.testEcho by allowing the policy to be set on the WSServer as an attribute
* acquire batchBuffer before calculating batchSpace in FrameFlusher
* Deprecated WebSocketPolicy
* Fixing MessageReceivingTest
* Refactoring out websocket-servlet
+ WebSocketPolicy removed from websocket-servlet
+ websocket-common is removed
+ Migrating back up to APIs
* WIP removing core policy
* added missing copyright headers
* customize client session
* customize cleanup
* wip
* wip
* Removing WebSocketPolicy usage within websocket-core
* Allowing DummyCoreSession to have a Behavior
* Removing core WebSocketPolicy.clonePolicy() - no longer used
* Jetty Native WebSocket now compiles / tests
* Remember WebSocketServletFactoryImpl as Context Attribute
* JSR356 codebase now compiles
* updated websocket servlets
* fix handling for no max message size
* Fixed LargeContainerTest.testEcho
Changed WebSocketServletFactoryImpl.defaultAutFragment to true
only check the incoming frame size if autoFragment is false
* Added WebSocketConstants to avoid duplicate default locations
* applications extensions discovered via headers
* fixed ConfiguratorTest.testNoExtensionsConfigurator
* cleaned up extension negotiation
* cleaned up extension negotiation
* cleanups
* Convert exceptions to API versions
* cleanup imports
* convert requestURI to a WebsocketURI in Negotiated
* fixed SessionTrackingTest
* moved session tracking management to be done by the JavaxWebSocketFrameHandler
* Cleanup dump for jdk11 classloaders
* fix to ConfiguratorTest
added the EchoSocket endpoint on the ServerContainer instead of using server.registerWebSocket
* Simplify websocket-servlet
fixed jetty websocket chat example
fixes for javax websocket chat example
fixed Upgrade response header issues
Set default configurations
improved dump
depends on javax server impl
do not expose core classes
fix javadoc
removed FrameHandlerFactory self reference
* Removed -impl from javax client and server
* Fixed common tests for onClosed event
* Improve creation of JavaxWebSocketServerContainer
Better handling of executor and httpClient as managed beans
* Fixed MessageOutputStream using empty buffers and looping forever
no longer enforce maxTextMessageBufferSize on partial messages
* Updates after review
reintroduced WebSocketServletFactory
reverted other API changes
* Ignored TextStreamTest tests incorrectly assuming no frame fragmentation
When message fragmentation is implemented in PartialStringMessageSink
then update these tests to check on the server side for no buffers
larger than the maxTextMessageBufferSize.
* fixed check for previous WebsocketUpgrade filter
Disable the AltFilterTest until problems of the WebSocketUpgradeFilter
attributes on the ServletContext are fixed.
* fixed flaky SessionTrackingTest by waiting for Sessions to open
* remove payload length check for outgoing frames
fix bug sending frames after they have failed validation checks
* reverted maven coords for javax client and server
* updated to latest jetty-10 dump changes
* Fix to WebSocketUpgradeFilter Attributes on the ServletContext
Removed deprecated method
WebSocketUpgradeFilter.configureContext(ServletContext)
Changed return type of
WebSocketUpgradeFilter.configureContext(ServletContext) to void.
WebSocketUpgradeFilter.configureContext was checking the ServletContext
for a WebSocketUpgradeFilter set as an attribute before creating a new
one. The FilterHolder from web.xml is being set as an attribute in init
which is called which happens after configureContext.
So instead of checking the attribute in
WebSocketUpgradeFilter.configureContext, we are now searching through
the filters on the ServletHandler to see if a WebSocketUpgradeFilter
has been registered with the CONTEXT_ATTRIBUTE_KEY.
* cleanup and identifying problems with FrameHandlerMetadata
* properly cache Metadata and copy MessageMetadata
properly cache Metadata in the FrameHandlerFactory using the metadataMap
copy the MessageMetadata from Metadata to prevent mutating the MethodHandle
* added new WebSocketServerContainerExecutorTest
removed DelayedStartClientOnServerTest which doesnt seem to be a valid
test when executors are shared
removed last check for executor==null in JavaxWebSocketServerContainerInitializer
as the threadPool is always created in the server
* reformatted websocket code
* reformat code
* Rervert unnecessary changes from 10.0.x
* Rervert unnecessary changes from 10.0.x
* fix build issues testing Hazelcast sessions
* QuotesDecoderTest sessions not being closed
* Fixed CookiesTest
CookiesTest.testCookiesAreSentToClient was failing due to the
Assertion being caught and logged by the upgrade request
the response is now captured by a FuturePromise and the assertions are
done in the main test
the reason the assertion was failing was due to lowercase "set-cookie"
instead of "Set-Cookie"
* fixed maven coords
* javadoc fixes
* fixed javadoc
* revert maven artifact names
* cleanup after review
Signed-off-by: Lachlan Roberts <lachlan@webtide.com>
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #3049 Warn on common SslContext vulnerable configurations
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Adding documentation notes for weak cipher warnings
Signed-off-by: WalkerWatch <ctwalker@gmail.com>
* Issue #3049 - SslContextFactory warnings on known bad config
+ Changes warnings from being a boolean on SslContextFactory
to being a logger named
"org.eclipse.jetty.util.ssl.SslContextFactory.config"
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
* Issue #3049 - SslContextFactory warnings on known bad config
+ Cleanup based on review
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
* Issue #3049 - SslContextFactory warnings on known bad config
+ Cleanup based on review
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Fixed dump() in QueuedThreadPool.
When the dump was not detailed, it was printing jobs=0 even if there
were jobs in the queue.
Given that it was adding no information (actually misleading information)
and that the queue size is already reported by QueuedThreadPool.toString()
the jobs are not dumped if the dump is not detailed.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
+ Moving MultiPartInputStreamParser from jetty-util to jetty-http/internal
+ Moving related classes ReadLineInputStream to jetty-http/internal
+ Moving related tests to jetty-http as well
+ Moved MultiPartFormDataCompliance to jetty-http
+ Renaming test to have same name as implementation
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
* Cleanup the dump implementation
* improved the clarity of utility methods for dump and updated most dump methods
* fixed upgrade filter dump
* Improved dump after review
* Moved dumpObjects to Dumpable
* implemented dumpBeans with dumpObjects
* less verbose dump
* Dump streams
* fixed dump test
Signed-off-by: Greg Wilkins <gregw@webtide.com>
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Introduced --jpms option in jetty-start to run Jetty from the module-path.
Introduced [jpms] sections in *.mod files, to specify JPMS command line
options that needs to be added to the command line generated by jetty-start.
Bumped java.transaction-api to 1.3 because it has Automatic-Module-Name.
Fixed ASM version lookup using ManifestUtils.
Fixed WebInfConfiguration.findAndFilterContainerPaths() to properly
scan the module-path, which may contain both files and directories.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Turns out that Package.getImplementationVersion() and similar return
null when running from the module path, so the logic in StartArgs
broke and prevented Jetty to start.
Introduced ManifestUtils to retrieve the manifest file from JarFile,
so that attribute entries can be retrieved independently from the JDK
version or module-path vs class-path.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
+ Changes needed for new Junit 5
+ Migrating from Vintage junit API to Jupiter junit API
+ Relies on SNAPSHOT jetty-test-helper
- this will be a formal release once this issue has been
resolved satisfactory
+ Have jenkins always pull latest SNAPSHOT for each build
+ Adding jetty.snapshots repository
+ Using surefire 2.22.0 per advice from junit
+ Ensuring <reuseForks>true</reuseForks> to work around issue junit-team/junit5#801
+ Disabling <forkMode>always</forkMode> in maven-surefire-plugin
due to bug https://github.com/junit-team/junit5/issues/801
+ OSGi tests must remain at vintage due to PaxExam
+ Moving from vintage TestingDir to jupiter WorkDir
+ Fixing imports to use jupiter, not vintage
+ Migrating vintage ExpectedException to jupiter assertThrows
+ Migrating vintage TestName to jupiter TestInfo
+ Migrating @RunWith(Parameterized.class)
to @ParameterizedTest with Argument Sources
+ Migrating assertTrue(val.contains(needle))
to assertThat(val, containsString(needle))
+ Aligning junit versions per recommendations from @sormuras
+ Adjusting parameter order change for assertEquals()
+ Test LifeCycle Annotation Migration
junit 4 | junit 5 / jupiter
------------ | -----------
@Before | @BeforeEach
@After | @AfterEach
@BeforeClass | @BeforeAll
@AfterClass | @AfterAll
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Signed-off-by: olivier lamy <oliver.lamy@gmail.com>
+ Changes needed for new Junit 5
+ Migrating from Vintage junit API to Jupiter junit API
+ Relies on SNAPSHOT jetty-test-helper
- this will be a formal release once this issue has been
resolved satisfactory
+ Have jenkins always pull latest SNAPSHOT for each build
+ Adding jetty.snapshots repository
+ Using surefire 2.22.0 per advice from junit
+ Ensuring <reuseForks>true</reuseForks> to work around issue junit-team/junit5#801
+ Disabling <forkMode>always</forkMode> in maven-surefire-plugin
due to bug https://github.com/junit-team/junit5/issues/801
+ OSGi tests must remain at vintage due to PaxExam
+ Moving from vintage TestingDir to jupiter WorkDir
+ Fixing imports to use jupiter, not vintage
+ Migrating vintage ExpectedException to jupiter assertThrows
+ Migrating vintage TestName to jupiter TestInfo
+ Migrating @RunWith(Parameterized.class)
to @ParameterizedTest with Argument Sources
+ Migrating assertTrue(val.contains(needle))
to assertThat(val, containsString(needle))
+ Aligning junit versions per recommendations from @sormuras
+ Adjusting parameter order change for assertEquals()
+ Test LifeCycle Annotation Migration
junit 4 | junit 5 / jupiter
------------ | -----------
@Before | @BeforeEach
@After | @AfterEach
@BeforeClass | @BeforeAll
@AfterClass | @AfterAll
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Signed-off-by: olivier lamy <oliver.lamy@gmail.com>
* Issue #2798 thread budget warning default to available cores
* Issue #2798 - Testcase example for ExecutorThreadPool
* Issue #2798 - Better TestCases for ThreadPoolBudget
* Simplified
* Issue #2798 - ThreadPoolBudget improvements.
* Removed unused imports.
* Made fields private.
* Removed field info.
* Changed the logic to print the info on leases only when warning or throwing.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Cleanup of Graceful shutdown, plus ensure Connection:close if connector is shutdown for #2749
* WIP close connection when shutting down
* WIP use HttpChannel.Listener
* cleanups
* support graceful stop of a context
* only close connections if the connector is shutdown
* minor cleanups
* fixed toString and test
* fixed imports
* Move close logic to HttpConnection
* fixed generator to not override persistent
* Issue #2749 - Close connections on graceful shutdown.
* Small fix after review.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
+ ensuring SLOTH fix still works with updated testcase
+ Dump detection of JreDisabled:java.security only checked
default SSLEngine.enabledCipherSuites
New implementation changes meaning to be more general
JVM:disabled (as in not selected by default), and
only reports it as disabled in the dump if not specifically
included from the SSLEngine.supportedCipherSuites via
the SslContextFactory.includedCipherSuites configuration
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Improved defaulting values for the truststore, avoiding to default
the password, which is often missing for a truststore.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
* Issue #2597 - Improving UnixSocketTest behavior.
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
jnr doesn't like filename/path with strange character so use a new property for jenkins build with docker
Signed-off-by: olivier lamy <oliver.lamy@gmail.com>
* Adding test case to verify existing PathResource behavior
* Skipping Resource class usage in OSGi entirely
* Only using the URL class, and removing of conversions to
other more correct forms (File and URI)
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Introduced ProcessorUtils.setAvailableProcessor(int) to allow
usage in embedded code without resorting to system properties.
Refactored the static initializer to be testable.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Now newly created threads set _lastShrink to avoid they are immediately
removed from the pool if they happen to meet the shrink conditions.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
1. The virtual method to stop fillInStackTrace() from being called by the
constructor has been removed.
2. The public no argument constructor for MultiException puts the class
into a builder mode where it just lets you add exceptions.
MultiException created from the constructor should not be thrown because
fillInStackTrace() is not called.
3. In the `ifExceptionThrow*()` methods a new `MultiException` is
created, with the stack trace filled, before it is thrown.
Signed-off-by: Luke Butters <lbutters@funnelback.com>
This doesn't actually stop creating the unnecessary exception instead it
stop running the expensive `fillInStackTrace()` unless we actually need
it.
This does result in the stack trace being set to where we throw the
exception rather than when we create it.
Signed-off-by: Luke Butters <lbutters@funnelback.com>
* Issue #2439 - Remove HTTP/2 data copy.
Implemented reference counting for the network buffer, with the
semantic that calling succeeded() on callbacks decrements the
reference count.
Introduced interface Retainable, used by the client when notifying
multiple application content listeners.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
+ Adding removeThread(Thread) to allow instrumentation
libraries to track removal of threads from Pool.
Signed-off-by: Joakim Erdfelt <joakim.erdfelt@gmail.com>
Fixes#2468 - EWYK concurrent production.
Fixed the case when a task throws an exception, which causes
the thread that ran the task to continue production, even
when it should not.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Now explicitly using a _mappedBuffer field in
CachedContentFactory.CachedHttpContent.
Deprecated BufferUtil.isMappedBuffer().
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
## SearchPattern
New class which does a fast search for patterns within strings and arrays of bytes using an implementation of the Boyer–Moore–Horspool algorithm. This was written to be used in the new MultiPartParser class to search for delimeter boundaries.
## MultiPartParser
New class which uses the SearchPattern to parse a MultiPart Mime given a ByteBuffer. Written in a non-blocking style so can be used asynchronously (although not currently be being used this way).
## MultiPartFormInputStream
New class which uses the MultiPartParser to parse a MultiPart Mime input stream into a Collection of Parts. This class is in org/eclipse/jetty/http and is designed to replace org/eclipse/jetty/util/MultiPartInputStreamParser.
## MultiPartInputStreamParser and Non Compliances
This class has been deprecated and replaced by org.eclipse.jetty.http.MultiPartFormInputStream. It accepts formats non compliant with the RFC that the new MultiPartFormInputStream does not accept. When this occurs violations are recorded by the method getNonComplianceWarnings().
## MultiParts
New interface to allow switching between the different implementations. This allows MultiParts to function in two different modes. The LEGACY implementation using the UTIL parser which may parse forms containing non compliances with the RFC, and the RFC7578 implementation using the new and faster HTTP parser. This file contains the implementations of MultiParts for HTTP and UTIL parsers as nested classes which are used by Request.
## Request
Changed to use the new MultiParts interface instead of the MultiPartInputStreamParser class. with a method called newMultiParts which will construct a MultiPart using one of the HTTP or UTIL implementations depending on what compliance mode is set.
## Jetty Test Webapp Dump Servlet
Code added to display parts while running the dump test webapp if MuliPart form is submitted.
## MultiPartBenchMark
JMH Benchmark of the HTTP multipart parser vs the UTIL multipart parser.
testLargeGenerated parses a 10MB file of random binary data.
testParser parses a series of small multipart forms captured by a browser.
```
# Run complete. Total time: 00:02:09
Benchmark (parserType) Mode Cnt Score Error Units
MultiPartBenchmark.testLargeGenerated UTIL avgt 10 0.252 ± 0.025 s/op
MultiPartBenchmark.testLargeGenerated HTTP avgt 10 0.035 ± 0.004 s/op
MultiPartBenchmark.testParser UTIL avgt 10 0.028 ± 0.005 s/op
MultiPartBenchmark.testParser HTTP avgt 10 0.015 ± 0.006 s/op
```
Changed Request.MultiPartInputStream to an interface called MultiParts where there is an implementation for both the HTTP and UTIL parsers.
Resolved some issues with default charsets in regards to request.setCharacterEncoding and the _charset_ part for issue #2398.
Changed HTTP parser to operate the same as UTIL parser in situtions with parts not of type form-data or without name field. HTTP parser was ignoring these parts, UTIL parser was throwing exceptions.
Replaced the context attribute with a field in MultiParts.
Signed-off-by: Lachlan Roberts <lachlan@webtide.com>
This allows Eclipse to continue to make use of jetty-* bundles
without requiring the presence of Apache Aries.
This fixes a problem introduced by the resolution to #2164 and
the same solution is used as in 10cdf16
See also the discussion at Eclipse:
https://bugs.eclipse.org/bugs/show_bug.cgi?id=532294
Signed-off-by: Mat Booth <mat.booth@redhat.com>
Fixed method, added Javadocs and cleaned up code with a few renamings
to better comply with AtomicLong naming.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Modified tests in new MultiPartInputStreamTest which were failing because they didn't comply with RFC.
Signed-off-by: Lachlan Roberts <lachlan@webtide.com>
the --task was not protected with a memory barrier, so different producing threads could contend on the field.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
For #2284 no longer parse the optional detail of the java version.
Parse failures will result in a guessed JVM platform 8
Signed-off-by: Greg Wilkins <gregw@webtide.com>
* Issue #2131 - Introduce a monitored thread pool.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
* Issue #2131 - Introduce a monitored thread pool.
Updated to use SampleStatistic and CounterStatistics.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
* Issue #2131 - Introduce a monitored thread pool.
Adding statistics fields as beans.
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Issue #2293 Pending Multiplexed Connections
Added a AtomicBiInteger to allow both total connections and pending connections to be encoded in the
same atomic int.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
Objects which inherit or implement an `equals()` method should not be compared with == or !=
When the comparison of references is intentional `@SuppressWarnings("ReferenceEquality")` can be used
Signed-off-by: Lachlan Roberts <lachlan@webtide.com>
Merging this PR for #2199 as we currently think it is not worse than previous and the code base is definitely simpler. While JMH has shown significant benefits for this approach, we have yet been able to demonstrate them in full scale integration tests - however that is likely due to GC and thread pool issues dominating.
A `org.eclipse.jetty.http2.PEC_MODE` System property has been added to allow the EWYK scheduling to be disabled for HTTP/2 if need be.
* Implementation of #1803 proposal 2 - EITHER dispatch type for EWYK
* made code more readable
* increase small threadpools in tests for extra reserved thread
* clean up
* minor code simplifications
* Work in progress to simplify reserved thread pool
* use a single ReservedThreadPool built into the QueuedThreadPool
* fixed javadoc
* removed more old reserved thread references
* disable EWYK for h2
* fixes from review
* use EWYK for HTTP2
* Fixed javadocs, imports and QueuedThreadPool constructors.
* fix javadoc
* EWYK avoid unnecessary pendings
* after review
* fixed merge with jetty-threadpool.xml
* alternate EWYK implementations
* added jetty copyright headers
* Simplified EWYK code after review
* fixed bad merge
* Code cleanups.
* Improved Javadocs for deprecated property "reservedThreads".
* Improved Javadocs for deprecated property "reservedThreads".
* added a system property to enable only PEC for HTTP/2
Signed-off-by: Greg Wilkins <gregw@webtide.com>
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Issue #1803 - Review usage of scheduling strategies
Use a single ReservedThreadExecutor built into the QueuedThreadPool
via new interface TryExecutor.
Signed-off-by: Greg Wilkins <gregw@webtide.com>
Signed-off-by: Simone Bordet <simone.bordet@gmail.com>
Joakim Erdfelt
Greg Wilkins
Simone Bordet
Jan Bartel
contextshuffling
Lachlan Roberts
Chris Walker
康智冬
olivier lamy
olivier lamy
Michael Hausegger
bkmz
Joakim Erdfelt
WalkerWatch
Yanming Zhou
Jesse McConnell
Kazuhiro Sera
Matthias Perktold
Luke Butters
Mat Booth
olivier lamy
David Ha