On 11:35 Tue 15 Jan , Ira Weiny wrote: > Perhaps the variable name should be pkey_index?
Yes, I thought about this too. There are some mess in this area this other old ones osmvendor libraries, I'm reviewing this yet. Sasha > > Ira > > On Tue, 15 Jan 2008 18:21:42 +0000 > Sasha Khapyorsky <[EMAIL PROTECTED]> wrote: > > > > > Use pkey at index 0 of port's pkey table, now this value is passed to > > user_mad. For some reason the old code (where 0xffff was passed) worked > > too. > > > > Signed-off-by: Sasha Khapyorsky <[EMAIL PROTECTED]> > > --- > > opensm/opensm/osm_perfmgr.c | 2 +- > > 1 files changed, 1 insertions(+), 1 deletions(-) > > > > diff --git a/opensm/opensm/osm_perfmgr.c b/opensm/opensm/osm_perfmgr.c > > index 76ef080..66d919d 100644 > > --- a/opensm/opensm/osm_perfmgr.c > > +++ b/opensm/opensm/osm_perfmgr.c > > @@ -399,7 +399,7 @@ osm_perfmgr_send_pc_mad(osm_perfmgr_t * perfmgr, > > ib_net16_t dest_lid, > > p_madw->mad_addr.addr_type.gsi.remote_qkey = > > cl_hton32(IB_QP1_WELL_KNOWN_Q_KEY); > > /* FIXME what about other partitions */ > > - p_madw->mad_addr.addr_type.gsi.pkey = cl_hton16(0xFFFF); > > + p_madw->mad_addr.addr_type.gsi.pkey = 0; > > p_madw->mad_addr.addr_type.gsi.service_level = 0; > > p_madw->mad_addr.addr_type.gsi.global_route = FALSE; > > p_madw->resp_expected = TRUE; > > -- > > 1.5.4.rc2.60.gb2e62 _______________________________________________ general mailing list [email protected] http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
