[ofa-general] Standby OpenSM reporting "Bad LinearFDBTop" value

[Hal Rosenstock]

On Fri, 2008-03-14 at 00:24 +0000, Sasha Khapyorsky wrote:
> On 14:35 Thu 13 Mar     , Hal Rosenstock wrote:
> > On Thu, 2008-03-13 at 21:36 +0000, Sasha Khapyorsky wrote:
> > > On 07:10 Thu 13 Mar     , Hal Rosenstock wrote:
> > > > 
> > > > Isn't message which states "Forcing correction to 0x0" misleading
> > > 
> > > It is in sense of OpenSM internal SwitchInfo representation.
> > 
> > The message is unclear and makes it seem like it is updating SwitchInfo.
> 
> Which message would be better? Would care about the patch?

How about the following ?

opensm/osm_sw_info_rcv.c: Clarify LinearFDBTop correction log message

Signed-off-by: Hal Rosenstock <hal at xsigo.com>

diff --git a/opensm/opensm/osm_sw_info_rcv.c b/opensm/opensm/osm_sw_info_rcv.c
index d056c1d..46229dc 100644
--- a/opensm/opensm/osm_sw_info_rcv.c
+++ b/opensm/opensm/osm_sw_info_rcv.c
@@ -517,7 +517,7 @@ void osm_si_rcv_process(IN void *context, IN void *data)
 			OSM_LOG(sm->p_log, OSM_LOG_ERROR, "ERR 3610: "
 				"\n\t\t\t\tBad LinearFDBTop value = 0x%X "
 				"on switch 0x%" PRIx64
-				"\n\t\t\t\tForcing correction to 0x%X\n",
+				"\n\t\t\t\tForcing internal correction to 0x%X\n",
 				cl_ntoh16(p_si->lin_top),
 				cl_ntoh64(osm_node_get_node_guid(p_node)), 0);
 

> 
> > > > on
> > > > standby and should be conditionalized for master only ?
> > > 
> > > The message? Why? It should fix this value anyway.
> > 
> > on the standby ? For later, if it becomes master ?
> 
> Yes (although SwitchInfo should be refetched on new heavy sweep), and
> in general there is no reason to keep invalid data in OpenSM, so I think
> the warning is valid. The text itself can mislead...
> 
> Sasha
> _______________________________________________
> general mailing list
> general at lists.openfabrics.org
> http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
> 
> To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general