On Fri, Jul 28, 2023 at 08:15:45AM +0000, Shinichiro Kawasaki wrote:
> On Jul 26, 2023 / 14:46, Daniel Wagner wrote:
> > Only allow to connect to our setup target with the correct hostnqn.
> >
> > Thus we have to explicitly add the test hostnqn to the test subsysnqn
> > allow_host list.
>
> [...]
>
> > --- a/tests/nvme/030
> > +++ b/tests/nvme/030
> > @@ -28,6 +28,7 @@ test() {
> >
> > _create_nvmet_subsystem "${subsys}1" "$(losetup -f)"
> > _add_nvmet_subsys_to_port "${port}" "${subsys}1"
> > + _create_nvmet_host "${subsys}1" "${def_hostnqn}"
> >
> > genctr=$(_discovery_genctr)
> >
> > @@ -36,13 +37,13 @@ test() {
> >
> > genctr=$(_check_genctr "${genctr}" "adding a subsystem to a port")
> >
> > - echo 0 > "${NVMET_CFS}/subsystems/${subsys}2/attr_allow_any_host"
> > + _add_nvmet_allow_hosts "${subsys}2" "${def_hostnqn}"
> >
> > - genctr=$(_check_genctr "${genctr}" "clearing attr_allow_any_host")
> > + genctr=$(_check_genctr "${genctr}" "adding host to allow_hosts")
> >
> > - echo 1 > "${NVMET_CFS}/subsystems/${subsys}2/attr_allow_any_host"
> > + _remove_nvmet_allow_hosts "${subsys}2" "${def_hostnqn}"
> >
> > - genctr=$(_check_genctr "${genctr}" "setting attr_allow_any_host")
> > + genctr=$(_check_genctr "${genctr}" "removing host from allow_hosts")
> >
> > _remove_nvmet_subsystem_from_port "${port}" "${subsys}2"
> > _remove_nvmet_subsystem "${subsys}2"
>
> The hunk above looks different from other changes. Is it changing test
> content slightly to meet request by Max? If so, it would be good to note
> in the commit message.
I should have explained this change in the commit message. When we use
allowed_hosts, the attr_allow_any_host feature is disabled and thus has no side
effects like updating the genctr. This makes the test fail. I opted to use
_[add|remove]_nvmet_allow_hosts to trigger the same side effect. While at it, I
also updated the logging info.
