OpenSolaris

Printable Version Enter a New Search
Bug ID 6610415
Synopsis fix to CR 6597119 failed (another panic in daplt)
State 10-Fix Delivered (Fix available in build)
Category:Subcategory ib_sw:ibcm
Keywords
Responsible Engineer Bill Taylor
Reported Against s10
Duplicate Of
Introduced In solaris_nevada
Commit to Fix snv_76
Fixed In snv_76
Release Fixed solaris_nevada(snv_76) , solaris_10u5(s10u5_07) (Bug ID:2153940)
Related Bugs 6434915 , 6597119 , 2152178
Submit Date 27-September-2007
Last Update Date 29-October-2007
Description 
Another occurrence of the panic reported in CR 6597119 has occurred.
It is so infrequent that it took many weeks of MPI testing by the
ClusterTools group (Don Kerr) to encounter it.  They use S10, but
the problem is in Nevada and OpenSolaris too.

The problem is a system panic with a stack trace like:

panic[cpu1]/thread=fffffe8001416c80: 
mutex_enter: bad mutex, lp=ffffffffac998da0 owner=ffffffff986d6ba0 thread=fffffe8001416c80
fffffe8001416860 unix:mutex_panic+6f ()
fffffe80014168c0 unix:mutex_vector_enter+292 ()
fffffe80014168e0 genunix:cv_wait+70 ()
fffffe8001416910 daplt:daplka_ep_get_state+71 ()
fffffe8001416940 daplt:daplka_cm_service_conn_closed+3b ()
fffffe8001416950 daplt:daplka_cm_service_handler+69 ()
fffffe8001416af0 ibcm:ibcm_process_dreq_msg+37d ()
fffffe8001416ba0 ibcm:ibcm_process_incoming_mad+273 ()
fffffe8001416bd0 ibcm:ibcm_recv_task+20 ()
fffffe8001416c60 genunix:taskq_d_thread+1ba ()
fffffe8001416c70 unix:thread_start+8 ()
Work Around 
N/A
Comments 
N/A