[jira] [Resolved] (MYFACES-3877) Add @ListenerFor(systemEventClass = PostRestoreStateEvent.class) causes StackOverflowException on MyFaces 2.2

["Leonardo Uribe (JIRA)" <de...@myfaces.apache.org>]

     [ https://issues.apache.org/jira/browse/MYFACES-3877?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Leonardo Uribe resolved MYFACES-3877.
-------------------------------------

       Resolution: Fixed
    Fix Version/s: 2.2.3

> Add @ListenerFor(systemEventClass = PostRestoreStateEvent.class) causes StackOverflowException on MyFaces 2.2
> -------------------------------------------------------------------------------------------------------------
>
>                 Key: MYFACES-3877
>                 URL: https://issues.apache.org/jira/browse/MYFACES-3877
>             Project: MyFaces Core
>          Issue Type: Bug
>          Components: JSR-344
>    Affects Versions: 2.2.2
>            Reporter: Leonardo Uribe
>            Assignee: Leonardo Uribe
>             Fix For: 2.2.3
>
>
> From MyFaces users list Oleg Varaksin:
> Hello MyFaces team,
> We get a StackOverflowError since MyFaces 2.x. The stack trace is shown here http://forum.primefaces.org/viewtopic.php?f=14&t=36999
> The problem: Wenn we call super.processEvent(event) in the processEvent() of a custom component, we get a recursion which ends in StackOverflowError.
> @Override
> public void processEvent(ComponentSystemEvent event) throws AbortProcessingException {
>     super.processEvent(event);
>     ...
> }
> The call super.processEvent(event) is necessary because e.g. Mojarra executes there some important code. But if we look at processEvent() in the MyFaces' UIComponent, it iterates over all listeners for processEvent() and invokes them. That means, processEvent() of the custom component is invoked again. Does it work as designed or a is it a bug?
> It was working before MyFaces 2.x and it works for all Mojarra versions.
> Thanks in advance.
> The problem was caused by this line:
> @ListenerFor(systemEventClass = PostRestoreStateEvent.class)
> The line is not necessary, because all components are already subscribed to the event by default.
> The code we have in MyFaces is correct from spec perspective. See:
> https://java.net/jira/browse/JAVASERVERFACES_SPEC_PUBLIC-1061
> But a simple check in this part can help to avoid the exception and keep running code that comes from 2.0/2.1. In fact, in 2.0/2.1 the offending line has no effect, but the problem was caused because MyFaces handles the case. It is recommended to remove the offending line from the users code anyway.



--
This message was sent by Atlassian JIRA
(v6.2#6252)