Do net put an ack to the queue if it has already been shutdown. Some
servers, like ejabberd, like to request an ack even after we have send
a stream close (and hance the queue was shutdown). If we would not
check here, then the ack would dangle around in the queue, and be send
on the next re-connection attempt even before the stream open.
See the following trace of the MUC bookmarks integration test. The
fact that it is a MUC test does not matter, but this test does
disconnect the connection and reconnect it. Not how the server,
ejabberd in this case, requests an SM ack by sending an <r/> even
though we already send the </stream:stream>:
22:22:05 SENT (4):
<iq id='MD4UC-61' type='set'>
<query xmlns='jabber:iq:private'>
<storage xmlns='storage:bookmarks'>
<conference name='Smack Inttest: 7in7j' autojoin='true' jid='y9jcn5@conference.salem.geekplace.eu'>
<nick>
Nick-P2VXD7
</nick>
</conference>
</storage>
</query>
</iq>
22:22:05 RECV (4):
<r xmlns='urn:xmpp:sm:3'/>
22:22:05 SENT (4):
<a xmlns='urn:xmpp:sm:3' h='29'/>
22:22:05 RECV (4):
<message to='sinttest-7in7j-4@salem.geekplace.eu' from='sinttest-7in7j-4@salem.geekplace.eu' type='headline'>
<event xmlns='http://jabber.org/protocol/pubsub#event'>
<items node='storage:bookmarks'>
<item id='current'>
<storage xmlns='storage:bookmarks'>
<conference name='Smack Inttest: 7in7j' autojoin='true' jid='y9jcn5@conference.salem.geekplace.eu'>
<nick>
Nick-P2VXD7
</nick>
</conference>
</storage>
</item>
</items>
</event>
<addresses xmlns='http://jabber.org/protocol/address'>
<address jid='sinttest-7in7j-4@salem.geekplace.eu/1415073683806426185213090' type='replyto'/>
</addresses>
</message>
22:22:05 RECV (4):
<iq xml:lang='en-US' to='sinttest-7in7j-4@salem.geekplace.eu/1415073683806426185213090' from='sinttest-7in7j-4@salem.geekplace.eu' type='result' id='MD4UC-61'/>
22:22:05 SENT (4):
<presence id='6MS6J-20' type='unavailable'/>
<a xmlns='urn:xmpp:sm:3' h='31'/>
<!-- We have closed the stream -->
</stream:stream>
<!-- But the server still requests an SM ack -->
22:22:05 RECV (4):
<r xmlns='urn:xmpp:sm:3'/>
22:22:05 RECV (4):
</stream:stream>
22:22:05 XMPPConnection closed (XMPPTCPConnection[sinttest-7in7j-4@salem.geekplace.eu/1415073683806426185213090] (4))
22:22:05 SENT (4):
<a xmlns='urn:xmpp:sm:3' h='31'/>
22:22:05 SENT (4):
<stream:stream xmlns='jabber:client' to='salem.geekplace.eu' xmlns:stream='http://etherx.jabber.org/streams' version='1.0' from='sinttest-7in7j-4@salem.geekplace.eu' xml:lang='en-US'>
22:22:05 RECV (4): ?xml version='1.0'?>
<stream:stream id='3379123514446782311' ver
22:22:05 RECV (4): sion='1.0' xml:lang='en' xmlns:stream='http://etherx.jabber.org/streams' xmlns='jabber:client'>
<stream:error>
<invalid-xml xmlns='urn:ietf:params:xml:ns:xmpp-streams'/>
</stream:error>
</stream:stream>
22:22:05 XMPPConnection closed due to an exception (XMPPTCPConnection[sinttest-7in7j-4@salem.geekplace.eu/1415073683806426185213090] (4))
org.jivesoftware.smack.XMPPException$StreamErrorException: invalid-xml You can read more about the meaning of this stream error at http://xmpp.org/rfcs/rfc6120.html#streams-error-conditions
<stream:error><invalid-xml xmlns='urn:ietf:params:xml:ns:xmpp-streams'/></stream:error>
at org.jivesoftware.smack.tcp.XMPPTCPConnection$PacketReader.parsePackets(XMPPTCPConnection.java:981)
at org.jivesoftware.smack.tcp.XMPPTCPConnection$PacketReader.access$700(XMPPTCPConnection.java:913)
at org.jivesoftware.smack.tcp.XMPPTCPConnection$PacketReader$1.run(XMPPTCPConnection.java:936)
at java.base/java.lang.Thread.run(Thread.java:834)
Instead of breaking in case the SSLEngine signals NEED_WRAP, which
leads to an endless loop while holding the
channelSelectedCallbackLock, we have to return, so that the
asynchronously invoked callback can aquire it, and do its work.
We previously only set the SM session ID to zero, but that is not
enough. On a clean shutdown, i.e. where we send a </stream> close tag,
we also have to nullify the unacknowledgedStanzas queue.
We previously only set 'connected' after connectInternal()
returned. This could lead to notifyConnectionError() ignoring stream
error exceptions, e.g. when establishing TLS which happens also in
connectInternal(), because 'connected' was still 'false'.
2020-08-06 13:08:06.265 19830-20423/org.atalk.android D/SMACK: SENT (0):
<stream:stream xmlns='jabber:client' to='atalk.sytes.net' xmlns:stream='http://etherx.jabber.org/streams' version='1.0' xml:lang='en'>
2020-08-06 13:08:06.333 19830-20424/org.atalk.android D/SMACK: RECV (0): ?xml version='1.0'?>
<stream:stream id='16420577292739412012' version='1.0' xml:lang='en' xmlns:stream='http://etherx.jabber.org/streams' from='atalk.sytes.net' xmlns='jabber:client'>
<stream:error>
<policy-violation xmlns='urn:ietf:params:xml:ns:xmpp-streams'/>
<text xml:lang='en' xmlns='urn:ietf:params:xml:ns:xmpp-streams'>
Too many (20) failed authentications from this IP address (::ffff:42.60.7.13). The address will be unblocked at 05:15:34 06.08.2020 UTC
</text>
</stream:error>
</stream:stream>
2020-08-06 13:08:06.346 19830-20424/org.atalk.android I/aTalk: [241896] org.jivesoftware.smack.AbstractXMPPConnection.notifyConnectionError() Connection was already disconnected when attempting to handle org.jivesoftware.smack.XMPPException$StreamErrorException: policy-violation You can read more about the meaning of this stream error at http://xmpp.org/rfcs/rfc6120.html#streams-error-conditions
<stream:error><policy-violation xmlns='urn:ietf:params:xml:ns:xmpp-streams'/><text xml:lang='en'>Too many (20) failed authentications from this IP address (::ffff:42.60.7.13). The address will be unblocked at 05:15:34 06.08.2020 UTC</text></stream:error>
org.jivesoftware.smack.XMPPException$StreamErrorException: policy-violation You can read more about the meaning of this stream error at http://xmpp.org/rfcs/rfc6120.html#streams-error-conditions
<stream:error><policy-violation xmlns='urn:ietf:params:xml:ns:xmpp-streams'/><text xml:lang='en'>Too many (20) failed authentications from this IP address (::ffff:42.60.7.13). The address will be unblocked at 05:15:34 06.08.2020 UTC</text></stream:error>
at org.jivesoftware.smack.tcp.XMPPTCPConnection$PacketReader.parsePackets(XMPPTCPConnection.java:966)
at org.jivesoftware.smack.tcp.XMPPTCPConnection$PacketReader.access$700(XMPPTCPConnection.java:898)
at org.jivesoftware.smack.tcp.XMPPTCPConnection$PacketReader$1.run(XMPPTCPConnection.java:921)
at java.lang.Thread.run(Thread.java:919)
Which eventually leads to a NoResponseException
org.jivesoftware.smack.SmackException$NoResponseException: No response
received within reply timeout. Timeout was 30000ms (~30s). While
waiting for establishing TLS
[XMPPTCPConnection[not-authenticated] (4)]
We now set 'connected' to 'true' as soon as the transport (e.g. TCP,
BOSH, …) is connected. While this is in other ways also sensible, it
also allows notifyConnectionError() to handle exceptions in the early
connection stage.
Thanks to Eng Chong Meng for reporting this.
The current code would work just fine for a connection having
multiple endpoints. However, when there is only one endpoint
ConnectionAttemptState.nextAddress() would return null, since
connectionEndpointIterator has already iterated over the only
possible value in the contructor leading to a NullPointerException.
This means that during establishment of a connection having multiple
endpoints, the first value inside connectionEndpointIterator would
always be overlooked.
To ensure the thread starting the reader/writer threads sees them
running and eventually waits until the 'running' boolean is reset to
'false' upon connection termination.
Since the current variant of notifyConnectionError() does not execute
most of its work in a new thread, especially since instantShutdown()
is called in the invoking thread, we have to mark the connections
reader or writer threads as no longer running prior them invoking
notifyConnectionError(). Otherwise they will end up waiting for
themselves to terminate.
To make it clear that this will either return if the condition is
true *or* if a connection exception happened.
Also introduce waitFor(), which is deliberately not named
waitForCondition() because it carries a different semantic.
This continues the design started with e98d42790 ("SmackReactor/NIO,
Java8/Android19, Pretty print XML, FSM connections"), where the
exceptions that caused an operation to fail, are not recorded within
SynchronizationPoint but within the connection instance itself.
This moves the logic in AbstractXMPPConnection.getSmackTlsContext()
into the ConnectionConfiguration constructor.
Also introduce SslContextFactory and use it in
ConnectionConfiguration.
This also removes the powermock dependency. Although powermock is a
fine library, it currently prevents dropping Junit4. And since we only
use the Whitebox API of powermock, this simply replaced powermock's
Whitebox with our own.
This is a complete redesign of what was previously
XmppNioTcpConnection. The new architecture allows to extend an XMPP
client to server (c2s) connection with new transport bindings and
other extensions.
Besides the way the transport handles the stream after SASL
<success/>, the SASL logic is independend from the underlying
transport (BOSH, TCP, …). Hence move it up into
AbstractXMPPConnection.
This also has the benefit that we can make some more methods private
or package-private.
Also introduce XmlStringBuilder.optTextChild(), which causes some
associated changes.
Since using a pipe, as we did previously would not error the target if
the first command in the pipe fails.
It is still far from ideal, since the dot file is also generated if
the gradle command fails. At some point, this should probably become
part of gradle build step instead of shelling out to a Makefile.
Florian Schmaus
Aditya Borikar