[dpdk-dev] [PATCH v4] test/ipsec: measure libipsec performance

Thomas Monjalon thomas at monjalon.net
Tue Apr 21 14:58:31 CEST 2020


21/04/2020 14:04, Ananyev, Konstantin:
> > 21/04/2020 13:07, Ananyev, Konstantin:
> > > > 21/04/2020 12:21, Ananyev, Konstantin:
> > > > > > 21/04/2020 04:29, Thomas Monjalon:
> > > > > > > > --- a/MAINTAINERS
> > > > > > > > +++ b/MAINTAINERS
> > > > > > > > @@ -1259,6 +1259,8 @@ F: lib/librte_ipsec/
> > > > > > > >  M: Bernard Iremonger <bernard.iremonger at intel.com>
> > > > > > > >  F: app/test/test_ipsec.c
> > > > > > > >  F: doc/guides/prog_guide/ipsec_lib.rst
> > > > > > > > +M: Savinay Dharmappa <savinay.dharmappa at intel.com>
> > > > > > > > +F: app/test/test_ipsec_perf.c
> > > > > > > >  M: Vladimir Medvedkin <vladimir.medvedkin at intel.com>
> > > > > > > >  F: app/test/test_ipsec_sad.c
> > > > > > > >  F: app/test-sad/
> > > > > > >
> > > > > > > Repeating what I said on v3:
> > > > > > > Having one different maintainer per test file is quite ridiculous.
> > > > > > > The maintainers of a lib are expected to maintain the related tests.
> > > > [...]
> > > > > About having separate MAINTAINER for the test -
> > > > > honestly I don't understand why it is a problem for you.
> > > > > Obviously we would like to spread the load - what's wrong with it?
> > > >
> > > > This is a problem of ownership.
> > > > Maintaining a library means you take care of every aspect, including tests.
> > > > That's why I would like to see you as a global maintainer of IPsec.
> > > >
> > > > It doesn't prevent you to delegate workload, of course.
> > > > But at the end it is more convenient to know there is a limited number
> > > > of persons responsible for the global quality of a component,
> > > > a person which is accountable and answering questions on the topic,
> > > > no matter which exact file we are talking about.
> > >
> > > Just talked with Bernard, he kindly agreed to be a maintainer for all ipsec UT:
> > > app/test/test_ipsec*
> > > Hope that will fulfil your concern?
> > 
> > My concern was to have the library maintainer maintaining also the related tests.
> > I don't understand why you don't want to take this responsibility,
> > but I cannot force you.
> > Having only one maintainer for IPsec tests is better than the current situation.
> 
> Ok, are you going to drop v4, so we can submit v5 with the fixes?
> Or should we submit a patch with fixes on top of v4?

v4 is dropped, you can send a v5.






More information about the dev mailing list