[ofa-general] Re: [PATCH] opensm: disable switch ports only

Sasha Khapyorsky sashak at voltaire.com
Tue Nov 18 05:29:22 PST 2008


On 08:04 Tue 18 Nov     , Hal Rosenstock wrote:
> On Tue, Nov 18, 2008 at 7:53 AM, Sasha Khapyorsky <sashak at voltaire.com> wrote:
> >
> > When "babbling port" policy is on disable switch ports even when trap
> > source is endport.
> 
> So does disables the peer switch port to an endport which is babbling
> ?

Yes.

> That could be made clearer in the description.

Ok.

> What happens if the end port is switch port 0 ?

When it should work as usual (it doesn't have remote port).

> > This will allow to handle disable ports remotely
> > (with ibportstate, etc.).
> 
> I'm not following what you mean by the ibportstate comment here. What
> port can ibportstate now disable differently from before ?

It is the same, but when endport is disabled how could we reenable this
remotely via downed link?

Sasha

> 
> -- Hal
> 
> > Signed-off-by: Sasha Khapyorsky <sashak at voltaire.com>
> > ---
> >  opensm/opensm/osm_trap_rcv.c |    4 ++++
> >  1 files changed, 4 insertions(+), 0 deletions(-)
> >
> > diff --git a/opensm/opensm/osm_trap_rcv.c b/opensm/opensm/osm_trap_rcv.c
> > index 5de283b..07c5183 100644
> > --- a/opensm/opensm/osm_trap_rcv.c
> > +++ b/opensm/opensm/osm_trap_rcv.c
> > @@ -239,6 +239,10 @@ static int disable_port(osm_sm_t *sm, osm_physp_t *p)
> >        ib_port_info_t *pi = (ib_port_info_t *)payload;
> >        int ret;
> >
> > +       /* in case of endport - disable switch's peer port */
> > +       if (osm_node_get_type(p->p_node) != IB_NODE_TYPE_SWITCH)
> > +               p = p->p_remote_physp;
> > +
> >        /* If trap 131, might want to disable peer port if available */
> >        /* but peer port has been observed not to respond to SM requests */
> >
> > --
> > 1.6.0.3.517.g759a
> >
> >



More information about the general mailing list