Customer reported that one of their production machine crashed with SIGSEGV.
According to customer it is reoccuring and four hs_err files are available for investigation.
This has occured on only one machine consistently.

And Core file is available for one PID crash as below :

Location : /net/cores.central/cores/dir33/831750

Files :

1) hs_err_pid10228.log
2) hs_err_pid14944.log -->> hs_err for cored process
3) hs_err_pid9268.log
4) pkg_java_core.java.14944.1219945037.tar.Z -->> Core File



Java Version : 1.6.0_04-b12
O/S : Solaris 10
Platform : x86



Here is the stack trace info :

Instructions: (pc=0xfe91a624)
0xfe91a614: 84 4e ff ff ff 8b 4d e0 8b 41 04 83 c0 08 8b 10
0xfe91a624: 8b 92 d4 00 00 00 53 51 50 ff d2 83 c4 0c e9 30
fe91a614 84 4e ff ff ff
fe91a619 8b 4d e0 mov 0xffffffe0(%ebp),%ecx
fe91a61c 8b 41 04 mov 0x4(%ecx),%eax
fe91a61f 83 c0 08 add $0x8,%eax
fe91a622 8b 10 mov (%eax),%edx
---------------
fe91a624 8b 92 d4 00 00 00 mov 0xd4(%edx),%edx
fe91a62a 53 push %ebx
fe91a62b 51 push %ecx
fe91a62c 50 push %eax
fe91a62d ff d2 call *%edx
fe91a62f 83 c4 0c add $0xc,%esp
fe91a632 e9 30 ff ff ff jmp 0xfe91a567

Stack: [0x00000000,0x00000000], sp=0xfaf78e34, free space=-82461k
Native frames: (J=compiled Java code, j=interpreted, Vv=VM code, C=native code)
V [libjvm.so+0x31a624] __1cWPar_ConcMarkingClosureKtrim_queue6MI_v_+0xcc
V [libjvm.so+0x31a750] __1cSCMSConcMarkingTaskNdo_work_steal6Mi_v_+0x84
V [libjvm.so+0x31a066] __1cSCMSConcMarkingTaskEwork6Mi_v_+0x16a
V [libjvm.so+0x583d77] __1cbAYieldingFlexibleGangWorkerEloop6M_v_+0xc3
V [libjvm.so+0x58215b] __1cKGangWorkerDrun6M_v_+0x27
V [libjvm.so+0x4d6d62] java_start+0xee
C [libc.so.1+0xa3a05]
C [libc.so.1+0xa3cf0]

Need to investigate the root cause for this crash.