[
https://issues.apache.org/jira/browse/DIRMINA-672?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
James Talmage reopened DIRMINA-672:
-----------------------------------
Current patch will loop endlessly if it's a nonFragmented transport and
doDecode never fully consumes the buffer (i.e. it wants to discard the
remainder of the buffer).
Take the "CrLfTerminatedCommandLineDecoder" example from the JavaDoc. If you
send it "Line1\r\nLine2\r\nLine3" it will parse the first two lines and loop
endlessly. We should check the result of doDecode to make sure it wants us to
continue.
The fragmented transport code also checks to make sure that some of the buffer
is consumed if doDecode returns true (this would likely mean that the doDecode
method had a bug, but we should probably check for it and throw an exception)
> CumulativeProtocolDecoder/ DemuxingProtocolDecoder
> --------------------------------------------------
>
> Key: DIRMINA-672
> URL: https://issues.apache.org/jira/browse/DIRMINA-672
> Project: MINA
> Issue Type: Bug
> Components: Filter
> Affects Versions: 2.0.0-M1, 2.0.0-M2, 2.0.0-M3, 2.0.0-M4
> Environment: JDK / OS independent
> Reporter: James Talmage
> Assignee: Emmanuel Lecharny
> Fix For: 2.0.0-RC1
>
> Attachments: DemuxingProtocolDecoderBugTest.java
>
>
> Excerpt from forum discussion at:
> http://www.nabble.com/Potential-DemuxingProtocolDecoderBug-td22489974.html
> I'm using 2.0.0-M4. Upon inspection of the source code, I can tell it's
> going to be a JDK / OS independent issue. Also upon inspection, I've
> discovered the bug is actually in the CumulativeProtocolDecoder (starting @
> line 125):
> if (!session.getTransportMetadata().hasFragmentation()) {
> doDecode(session, in, out);
> return;
> }
> This breaks the contract with the doDecode method as it is only called once
> (the documentation says it should be called repeatedly until it returns
> false). The following changes makes my previous test case pass, but it's
> probably a little simplistic.
> if (!session.getTransportMetadata().hasFragmentation()) {
> while(in.hasRemaining() && doDecode(session, in, out)){
> //Do nothing
> }
> return;
> }
> The code should probably make sure that if doDecode returns true, some of the
> buffer was actually consumed (as the code for fragmented transports does).
> Also, it may make sense to provide another method (i.e.
> finishNonFragmentedDecode(...)), for handling the remainder of the buffer
> after doDecode returns false.
> I see where the author was headed with this code. Transports (such as UDP)
> that don't support fragmentation probably don't jive with the concept of an
> accumulating buffer (what do we do with the accumulation buffer if we drop a
> UDP packet?). It does however make perfect sense to use such transports with
> the concept of a DemuxingProtocolDecoder. Perhaps it would be better to
> refactor the DemuxingProtocolDecoder so that it's not a subclass of
> CumulativeProtocolDecoder. Create a helper class that handles the fragment
> accumulation aspect. CumulativeProtocolDecoder would always use said helper
> class (throwing an error if the protocol doesn't support fragmentation), but
> DemuxingProtocolDecoder could opt to use it depending on the protocol it
> encounters.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
