[dpdk-dev] unchecked return value in enic driver

John Daley (johndale) johndale at cisco.com
Mon Jun 20 20:44:33 CEST 2016



> -----Original Message-----
> From: Kerlin, MarcinX [mailto:marcinx.kerlin at intel.com]
> Sent: Monday, June 20, 2016 4:12 AM
> To: John Daley (johndale) <johndale at cisco.com>; Nelson Escobar
> (neescoba) <neescoba at cisco.com>
> Cc: 'dev at dpdk.org' <dev at dpdk.org>
> Subject: RE: unchecked return value in enic driver
> 
> Hi John and Nelson,
> 
> > -----Original Message-----
> > From: Kerlin, MarcinX
> > Sent: Monday, June 13, 2016 1:18 PM
> > To: johndale at cisco.com; neescoba at cisco.com
> > Cc: dev at dpdk.org
> > Subject: unchecked return value in enic driver
> >
> > Hi John and Nelson,
> >
> > I have a question regarding Coverity defects:
> >
> > File: /drivers/net/enic/enic_ethdev.c
> > Line: 379
> >
> > CID 13197: Unchecked return value
> > (CHECKED_RETURN)1. check_return: Calling rte_atomic64_cmpset without
> > checking return value (as is done elsewhere 15 out of 17 times)
> >
> > Can I mark this error as "False Positive" in Coverity Classification ? reason:
> > 1. Function returns a void type so change the return type to int
> > requires changes all drivers 2. rte_atomic64_cmpset is at the end of
> > function so nonsense added a return
> >
> > What is your opinion?

I agree with marking it false positive for the reason you mention. 
Thanks!
John

> 
> I marked this Coverity as false-positive with an explanation. If in your opinion
> it is not ok, you can reopen/change/fix it.
> 
> >
> > Regards,
> > Marcin


More information about the dev mailing list