Yes, my sources were too old. The latest code has the problem you
predicted. I'll find and fix that and send a patch, then I'll test
Sasha's patch to avoid the unnecessary SubnSet(PortInfo).
Rolf
On Thu, 2006-03-30 at 13:38, Rolf Manderscheid wrote:
Also, is PortInfo:PartitionEnforcementInbound/Outbound set properly (0)
for these ports which don't support partition enforcement ? (That might
be another issue).
Yes, those components are set to zero.
How did you observe this ? What is your topology ?
Actually, it looks like they're
unconditionally set to zero.
Where ?
The only place I could find an osm_req_set(..., IB_MAD_ATTR_PORT_INFO,
...) call is in osm_link_mgr_set_physp_pi. Unless I'm missing
something, that function clears the payload and never assigns the
partition enforcement components.
What about osm_pkey_mgr.c::osm_pkey_mgr_enforce_partition which is
called from osm_pkey_mgr_update_peer_port ?
-- Hal
Rolf
_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
