[jira] [Commented] (FELIX-4190) The framework should not hold any lock while calling ServiceFactory#unget
[ https://issues.apache.org/jira/browse/FELIX-4190?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13930818#comment-13930818 ] David Jencks commented on FELIX-4190: - I'm ok with the duplicated loop. > The framework should not hold any lock while calling ServiceFactory#unget > - > > Key: FELIX-4190 > URL: https://issues.apache.org/jira/browse/FELIX-4190 > Project: Felix > Issue Type: Bug > Components: Framework >Reporter: Guillaume Nodet >Assignee: Guillaume Nodet > Fix For: framework-4.4.0 > > -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (FELIX-4190) The framework should not hold any lock while calling ServiceFactory#unget
[ https://issues.apache.org/jira/browse/FELIX-4190?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13930291#comment-13930291 ] Richard S. Hall commented on FELIX-4190: I believe duplicating the loop would work. The first loop tries to be friendly, then the call to invalidate() makes it impossible to get the service object again, and then finally the second loop deals more harshly with clients who won't let go. > The framework should not hold any lock while calling ServiceFactory#unget > - > > Key: FELIX-4190 > URL: https://issues.apache.org/jira/browse/FELIX-4190 > Project: Felix > Issue Type: Bug > Components: Framework >Reporter: Guillaume Nodet >Assignee: Guillaume Nodet > Fix For: framework-4.4.0 > > -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (FELIX-4190) The framework should not hold any lock while calling ServiceFactory#unget
[
https://issues.apache.org/jira/browse/FELIX-4190?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13910322#comment-13910322
]
Guillaume Nodet commented on FELIX-4190:
So I come back to my earlier proposal which is to duplicate the loop at the end
after having invalidted the registration.
I.e.
{code}
// Now forcibly unget the service object for all stubborn clients.
Bundle[] clients = getUsingBundles(reg.getReference());
for (int i = 0; (clients != null) && (i < clients.length); i++)
{
while (ungetService(clients[i], reg.getReference()))
; // Keep removing until it is no longer possible
}
((ServiceRegistrationImpl) reg).invalidate();
for (int i = 0; (clients != null) && (i < clients.length); i++)
{
while (ungetService(clients[i], reg.getReference()))
; // Keep removing until it is no longer possible
}
{code}
The call to getUsingBundles does not need to be synchronized, as the method
itself synchronizes on this.
At this point, no new bundles can acquire a service reference, so we can keep
the list of bundles and iterator on it safely.
Then, we do a first pass on ungetService which should remove most of the
references, except if a bundle is re-acquiring the service.
We then invalidate the registration, which will prevent any further call to
getService.
We iterate another time on the bundles in case a bundle has acquired the
service during the first loop.
> The framework should not hold any lock while calling ServiceFactory#unget
> -
>
> Key: FELIX-4190
> URL: https://issues.apache.org/jira/browse/FELIX-4190
> Project: Felix
> Issue Type: Bug
> Components: Framework
>Reporter: Guillaume Nodet
>Assignee: Guillaume Nodet
> Fix For: framework-4.4.0
>
>
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
[jira] [Commented] (FELIX-4190) The framework should not hold any lock while calling ServiceFactory#unget
[ https://issues.apache.org/jira/browse/FELIX-4190?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13731347#comment-13731347 ] David Jencks commented on FELIX-4190: - I agree with Richard that this is not a satisfactory patch. This is a duplicate of FELIX-3687 where I proposed another unsatisfactory patch and determined that Richard's suggestion wouldn't work. I have a further proposal there that I still think might work. > The framework should not hold any lock while calling ServiceFactory#unget > - > > Key: FELIX-4190 > URL: https://issues.apache.org/jira/browse/FELIX-4190 > Project: Felix > Issue Type: Bug > Components: Framework >Reporter: Guillaume Nodet >Assignee: Guillaume Nodet > Fix For: framework-4.4.0 > > -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (FELIX-4190) The framework should not hold any lock while calling ServiceFactory#unget
[ https://issues.apache.org/jira/browse/FELIX-4190?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13728122#comment-13728122 ] Richard S. Hall commented on FELIX-4190: I won't revert it, since I don't expect it will cause any issues in normal usage. We can just think about how to fix it when you are back. It is a little tricky, though. > The framework should not hold any lock while calling ServiceFactory#unget > - > > Key: FELIX-4190 > URL: https://issues.apache.org/jira/browse/FELIX-4190 > Project: Felix > Issue Type: Bug > Components: Framework >Reporter: Guillaume Nodet >Assignee: Guillaume Nodet > Fix For: framework-4.4.0 > > -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (FELIX-4190) The framework should not hold any lock while calling ServiceFactory#unget
[ https://issues.apache.org/jira/browse/FELIX-4190?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13728091#comment-13728091 ] Guillaume Nodet commented on FELIX-4190: I suggest you revert the patch if you want until further discussed as I'll be in vacation for the next weeks and won't be able to actually do anything nor discuss that further. Another option might be to duplicate the loop after having invalidated the registration to get rid of possible new references acquired while processing the ServiceEvent.UNREGISTERING or just before the service registration has been invalidated. That would avoid grabbing the lock again I think. > The framework should not hold any lock while calling ServiceFactory#unget > - > > Key: FELIX-4190 > URL: https://issues.apache.org/jira/browse/FELIX-4190 > Project: Felix > Issue Type: Bug > Components: Framework >Reporter: Guillaume Nodet >Assignee: Guillaume Nodet > Fix For: framework-4.4.0 > > -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (FELIX-4190) The framework should not hold any lock while calling ServiceFactory#unget
[ https://issues.apache.org/jira/browse/FELIX-4190?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13727714#comment-13727714 ] Richard S. Hall commented on FELIX-4190: Reiterating a comment from email: Technically, I think this patch is not correct. We specifically allow for bundles that already hold the service reference to acquire the service object while a service is being unregistered (see comments in ServiceRegistry.unregisterService()). This means that a bundle that already holds the service reference could, technically, acquire the unregistering service just after you call getUsingBundles(), thus we wouldn't be forcing it to unget. The only way I think we can do what you want is to somehow acquire the registry lock (i.e., this) again and put us into some new state that doesn't allow any more bundles to acquire the unregistering service...or perhaps by adding something to the service registration to put it into a new state where it won't return anything from getService. Then we could release the registry lock and continue with what you have. > The framework should not hold any lock while calling ServiceFactory#unget > - > > Key: FELIX-4190 > URL: https://issues.apache.org/jira/browse/FELIX-4190 > Project: Felix > Issue Type: Bug > Components: Framework >Reporter: Guillaume Nodet >Assignee: Guillaume Nodet > Fix For: framework-4.4.0 > > -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
