[ 
https://issues.apache.org/jira/browse/MINIFICPP-9?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16188246#comment-16188246
 ] 

ASF subversion and git services commented on MINIFICPP-9:
---------------------------------------------------------

Commit 9f161a27e5fa0ea9aac4d59d1c23edbad3d77858 in nifi-minifi-cpp's branch 
refs/heads/master from Marc Parisi
[ https://git-wip-us.apache.org/repos/asf?p=nifi-minifi-cpp.git;h=9f161a2 ]

MINIFI-339: Add C2 base allowing for 1 protocol and n heartbeat reporters
MINIFI-339: Add GetTCP Processor
MINIFI-339: Add listener server
MINIFI-339: Update to listener
MINIFI-339: Resolve Issue with stack based processor nodes losing scope
MINIFI-369: Update ListenHTTP processor to allow transfer encoding
MINIFI-339: Update rest receiver instantiation and fix issue found in GetFile
MINIFI-339: Rename content to operational arguments
MINIFI-371: remove virtual destructors when not needed
MINIFI-339: Fixing issues with GetTCP
MINIFI-378: Resolve issues with shutdown. Took the approach to call notifyStop 
at the destructor to avoid larger changes
MINIFI-339: Allow C2 to be disabled

This closes #134.

Signed-off-by: Aldrin Piri <ald...@apache.org>


> Shutdown can cause ProcessSessionFactory pointer to be deleted during use, 
> causing page fault at shutdown
> ---------------------------------------------------------------------------------------------------------
>
>                 Key: MINIFICPP-9
>                 URL: https://issues.apache.org/jira/browse/MINIFICPP-9
>             Project: NiFi MiNiFi C++
>          Issue Type: Bug
>            Reporter: marco polo
>            Assignee: marco polo
>            Priority: Blocker
>
> If a thread within a processor uses the process session factory to create a 
> session, the pointer will result in an invalid memory access, causing a seg 
> fault. This is because the raw pointer is passed around and deleted. We 
> should either convert this to a shared pointer or be more mindful of the 
> variable's lifecycle given the complexity of processors will only ever 
> increase. Further, we should not be deleting infrastructure components 
> without ensuring all processors are finished. 



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to