[openib-general] OpenSM -> 'open /dev/infiniband/umad1 failed'

[Sasha Khapyorsky]

Hi,

On 04:21 Wed 27 Sep     , chris_youb at yahoo.ca wrote:
> I'm trying to setup OpenSM on one of our boxes.  I've installed the RPMs from ofed-1.0-sles10-rpms_i686.tar.gz and updated the firmware on our Mellanox card.
> When I try to start opensm I get the following error message: 'umad_open_port: open /dev/infiniband/umad1 failed'.  Any suggestions of what I can try next?

Be sure that device node '/dev/infiniband/umad1' exists and you have
permission to access it for read/write.

Sasha

> 
> ******** Setup ********
> H/W: Dell 1550
> O/S: Suse 10.0 (linux 2.6.13-15.12-default)
> HBC: Mellanox MT23108 rev 3.5.000
> S/W: ofed-1.0-sles10-rpms_i686.tar.gz
> 
> ******** OpenSM ********
> linux:/usr/local/ofed/bin # ./opensm -V -d5
> -------------------------------------------------
> OpenSM Rev:openib-1.2.1
> Based on OpenIB svn Exported revision
> Command Line Arguments:
>  Big V selected
>  d level = 0x5
>  Log File: /var/log/osm.log
> -------------------------------------------------
> OpenSM Rev:openib-1.2.1 OpenIB svn Exported revision
> 
> ibwarn: [6860] umad_init:
> ibwarn: [6860] umad_get_cas_names: max 32
> ibwarn: [6860] umad_get_cas_names: return 1 cas
> ibwarn: [6860] umad_get_ca_portguids: ca name mthca0 max port guids 64
> ibwarn: [6860] umad_get_ca: ca_name mthca0
> ibwarn: [6860] umad_get_ca: opened mthca0
> ibwarn: [6860] umad_get_ca_portguids: mthca0: 3 ports
> ibwarn: [6860] umad_get_ca: ca_name mthca0
> ibwarn: [6860] umad_get_ca: opened mthca0
> ibwarn: [6860] umad_get_port: ca_name (null) portnum 0
> ibwarn: [6860] umad_get_cas_names: max 20
> ibwarn: [6860] umad_get_cas_names: return 1 cas
> ibwarn: [6860] resolve_ca_name: checking ca 'mthca0'
> ibwarn: [6860] resolve_ca_port: checking ca 'mthca0'
> ibwarn: [6860] umad_get_ca: ca_name mthca0
> ibwarn: [6860] umad_get_ca: opened mthca0
> ibwarn: [6860] resolve_ca_port: checking port 0
> ibwarn: [6860] resolve_ca_port: checking port 1
> ibwarn: [6860] resolve_ca_port: checking port 2
> ibwarn: [6860] resolve_ca_name: found ca mthca0 with port 2 type 0
> ibwarn: [6860] resolve_ca_name: phys found 0 on mthca0 port 2
> ibwarn: [6860] umad_release_port: port mthca0:2
> ibwarn: [6860] umad_release_port: releasing mthca0:2
> Using default GUID 0x2c90107fbfcf2
> ibwarn: [6860] umad_get_ca_portguids: ca name mthca0 max port guids 32
> ibwarn: [6860] umad_get_ca: ca_name mthca0
> ibwarn: [6860] umad_get_ca: opened mthca0
> ibwarn: [6860] umad_get_ca_portguids: mthca0: 3 ports
> ibwarn: [6860] umad_get_ca: ca_name mthca0
> ibwarn: [6860] umad_get_ca: opened mthca0
> ibwarn: [6860] umad_get_port: ca_name mthca0 portnum 2
> ibwarn: [6860] umad_open_port: ca mthca0 port 2
> ibwarn: [6860] umad_open_port: opening mthca0 port 2
> ibwarn: [6860] dev_to_umad_id: mapped mthca0 2 to 1
> ibwarn: [6860] umad_open_port: open /dev/infiniband/umad1 failed
> 
> Error from osm_opensm_bind (0x2A)
> Exiting SM
> 
> ibwarn: [6860] umad_done:
> 
> ******** Drivers ********
> ib_mthca               97692  0
> ib_mad                 34324  2 ib_umad,ib_mthca
> ib_core                39680  3 ib_umad,ib_mthca,ib_mad
> 
> ******** Logs ********
> linux:/usr/local/ofed/bin # tail -f /var/log/osm.log
> Jan 28 14:35:41 017194 [4018DFE0] -> osm_report_notice: Reporting Generic Notice type:3 num:66 from LID:0x0000 GID:0xfe80000000000000,0x0000000000000000
> Jan 28 14:35:41 017349 [4018DFE0] -> osm_report_notice: Reporting Generic Notice type:3 num:66 from LID:0x0000 GID:0xfe80000000000000,0x0000000000000000
> Jan 28 14:35:41 025501 [4018DFE0] -> osm_vendor_bind: Binding to port 0x2c90107fbfcf2
> Jan 28 14:35:41 030909 [4018DFE0] -> osm_vendor_open_port: ERR 542C: umad_open_port() failed
> Jan 28 14:35:41 030986 [4018DFE0] -> osm_vendor_bind: ERR 5424: Unable to Open Port 0x2c90107fbfcf2
> Jan 28 14:35:41 031015 [4018DFE0] -> osm_sm_mad_ctrl_bind: ERR 3118: Vendor specific bind failed
> Jan 28 14:35:41 031228 [4018DFE0] -> osm_sm_bind: ERR 2E10: SM MAD Controller bind failed (IB_ERROR)
> Jan 28 14:35:41 031742 [4018DFE0] -> osm_sa_mad_ctrl_unbind: ERR 1A11: No previous bind
> Jan 28 14:35:41 032313 [0000] -> Exiting SM
> 
> 
> --
> This message was sent on behalf of chris_youb at yahoo.ca at openSubscriber.com
> http://www.opensubscriber.com/messages/openib-general@openib.org/topic.html
> 
> _______________________________________________
> openib-general mailing list
> openib-general at openib.org
> http://openib.org/mailman/listinfo/openib-general
> 
> To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
>