[openib-general] OpenSM doesn't start on p5 570

[Andrey Slepuhin]

Dear folks,

I have a problem starting opensm on a p5 570 machine. The following messages
appear in the opensm log file:

******************************************************************
******************** INITIATING HEAVY SWEEP **********************
******************************************************************


Jan 16 13:30:55 737114 [40018DC0] -> osm_req_get: [
Jan 16 13:30:55 737130 [40018DC0] -> osm_mad_pool_get: [
Jan 16 13:30:55 737147 [40018DC0] -> osm_vendor_get: [
Jan 16 13:30:55 737161 [40018DC0] -> osm_vendor_get: Acquiring UMAD for p_madw = 0x100747dc, size = 256
Jan 16 13:30:55 737176 [40018DC0] -> osm_vendor_get: Acquired UMAD 0x1008ee40, size = 256
Jan 16 13:30:55 737192 [40018DC0] -> osm_vendor_get: ]
Jan 16 13:30:55 737208 [40018DC0] -> osm_mad_pool_get: Acquired p_madw = 0x100747d0, p_mad = 0x1008ee78, size = 256
Jan 16 13:30:55 737223 [40018DC0] -> osm_mad_pool_get: ]
Jan 16 13:30:55 737238 [40018DC0] -> osm_req_get: Getting NodeInfo (0x11), modifier = 0x0, TID = 0x1234
Jan 16 13:30:55 737255 [40018DC0] -> osm_vl15_post: [
Jan 16 13:30:55 737269 [40018DC0] -> osm_vl15_post: Posting p_madw = 0x0x100747d0
Jan 16 13:30:55 737284 [40018DC0] -> osm_vl15_post: 0 QP0 MADs on wire, 1 QP0 MADs outstanding
Jan 16 13:30:55 737299 [40018DC0] -> osm_vl15_poll: [
Jan 16 13:30:55 737313 [40018DC0] -> osm_vl15_poll: Signalling poller thread
Jan 16 13:30:55 737334 [40018DC0] -> osm_vl15_poll: ]
Jan 16 13:30:55 737338 [42827B20] -> __osm_vl15_poller: Servicing p_madw = 0x100747d0
Jan 16 13:30:55 737352 [40018DC0] -> osm_vl15_post: ]
Jan 16 13:30:55 737388 [40018DC0] -> osm_req_get: ]
Jan 16 13:30:55 737404 [40018DC0] -> __osm_state_mgr_sweep_hop_0: ]
Jan 16 13:30:55 737420 [40018DC0] -> osm_state_mgr_process: ]
Jan 16 13:30:55 737436 [40018DC0] -> osm_sm_sweep: ]
Jan 16 13:30:55 737464 [42827B20] -> SMP dump:
                                base_ver................0x1
                                mgmt_class..............0x81
                                class_ver...............0x1
                                method..................0x1 (SubnGet)
                                D bit...................0x0
                                status..................0x0
                                hop_ptr.................0x0
                                hop_count...............0x0
                                trans_id................0x1234
                                attr_id.................0x11 (NodeInfo)
                                resv....................0x0
                                attr_mod................0x0
                                m_key...................0x0000000000000000
                                dr_slid.................0xFFFF
                                dr_dlid.................0xFFFF

                                Initial path: [0]
                                Return path:  [0]
                                Reserved:     [0][0][0][0][0][0][0]

                                00 00 00 00 00 00 00 00   00 00 00 00 00 00 00 00

                                00 00 00 00 00 00 00 00   00 00 00 00 00 00 00 00

                                00 00 00 00 00 00 00 00   00 00 00 00 00 00 00 00

                                00 00 00 00 00 00 00 00   00 00 00 00 00 00 00 00

Jan 16 13:30:55 737604 [42827B20] -> osm_vendor_send: [
Jan 16 13:30:55 737742 [42827B20] -> osm_vendor_send: Completed Sending Request p_madw = 0x100747d0
Jan 16 13:30:55 737761 [42827B20] -> osm_vendor_send: ]
Jan 16 13:30:55 737768 [43027B20] -> osm_mad_pool_get: [
Jan 16 13:30:55 737784 [42827B20] -> __osm_vl15_poller: 1 QP0 MADs on wire, 1 outstanding, 0 unicasts sent, 1 total sent
Jan 16 13:30:55 737812 [43027B20] -> osm_vendor_get: [
Jan 16 13:30:55 737848 [43027B20] -> osm_vendor_get: Acquiring UMAD for p_madw = 0x10074724, size = 256
Jan 16 13:30:55 737866 [43027B20] -> osm_vendor_get: Acquired UMAD 0x1008ef80, size = 256
Jan 16 13:30:55 737883 [43027B20] -> osm_vendor_get: ]
Jan 16 13:30:55 737897 [43027B20] -> osm_mad_pool_get: Acquired p_madw = 0x10074718, p_mad = 0x1008efb8, size = 256
Jan 16 13:30:55 737915 [43027B20] -> osm_mad_pool_get: ]
Jan 16 13:30:55 737939 [43027B20] -> umad_receiver: ERR 5413: Failed to obtain request madw for received MAD(method=0x81
attr=0x11) -- dropping
Jan 16 13:30:55 737960 [43027B20] -> osm_mad_pool_put: [
Jan 16 13:30:55 737975 [43027B20] -> osm_mad_pool_put: Releasing p_madw = 0x10074718, p_mad = 0x1008ed00
Jan 16 13:30:55 737993 [43027B20] -> osm_vendor_put: [
Jan 16 13:30:55 738008 [43027B20] -> osm_vendor_put: Retiring UMAD 0x1008ecc8
Jan 16 13:30:55 738026 [43027B20] -> osm_vendor_put: ]
Jan 16 13:30:55 738041 [43027B20] -> osm_mad_pool_put: ]


My configuration consists of two 23108 HCAs directly connected without a switch,
firmware is 3.3.3, kernel is 2.6.15-4 from OpenSUSE repository, userspace
revision is 4978.

Any help will be much appreciated

Best regards,
Andrey