On 01/10/2013 07:28 PM, Eric Blake wrote:
On 01/10/2013 05:14 PM, John Ferlan wrote:
>
> Still get a Coverity error, but I think it has something to do with the
> VIR_FREE(*priv); done in the esxFreePrivate(). I'm beginning to believe
> it's a Coverity "issue". I have no idea why what I had done
previously
> was able to bypass the error. I'm trying to figure out the right
> mechanism to ask in the Coverity world.
Can you paste the actual Coverity analysis on this code? Knowing what
call stack it is tracing through will help others look at this setup to
see why it is still complaining.
The full analysis is a bit large so I cut it down a bit. The attached
esx.html has the trace in question - the links within won't work, but it
hopefully gives you a picture of the trace. The key is that the
esxFreePrivate(&priv); frees both the data in the structure and the
structure itself which is what I believe is confusing coverity. If I
add a VIR_FREE(priv) after this call - then the error goes away even
though it's redundant with the VIR_FREE(*priv) in esxFreePrivate().