I put a new snapshot into coverity yesterday. There was nothing new found
and nothing old fixed. The table with component breakdown on the Coverity
website is still broken, not sure why.
Your request for analysis of Apache ActiveMQ Artemis has been completed
successfully.
The results are available at
https://scan.coverity.com/projects/apache-activemq-artemis
Analysis Summary:
New defects found: 0
Defects eliminated: 0
There is 753 outstanding finds now, more than half of it in tests and
examples, quite lot of it is caused by closing two JMS sessions in a
finally block, where closing the first may throw exception, which would
mean the second one does not get closed. Something that is not wroth
fixing, I'd say.
On Sat, Feb 25, 2017 at 9:23 AM, nigro_franz <[email protected]> wrote:
> It is awesome...the issue on the AbstractJDBCDriver seems correct,
> considering that the analysis is not aware of the lifecycle of the
> connection field, as intended by the developer!
>
It can probably be made aware of things like this [1], but it requires
adding hints to the sourcecode. I did it only once long time ago in C++, so
I don't recall the details. I have a feeling it is not worth it for Artemis.
[1] https://scan.coverity.com/models#java_models_annotations ; accessing
the link requires registration ;(
On Sat, Feb 25, 2017 at 4:18 PM, Clebert Suconic <[email protected]>
wrote:
> Where is this "// TODO BEFORE MERGE: (is null a good option here?)"
>
>
> This some mark I usually do, I meant to fix it before I sent the commit :)
>
in artemis-server/src/main/java/org/apache/activemq/artemis/
core/remoting/impl/invm/InVMConnection.java
<https://github.com/apache/activemq-artemis/blob/fbc77b44c2bf72cede87b7dd3b74ba2604ccb7b6/artemis-server/src/main/java/org/apache/activemq/artemis/core/remoting/impl/invm/InVMConnection.java>,
--
Jiří Daněk
