I am getting following crash in our 32 bit C++ application built on RHEL5.4 with Oracle version 11.2.0.2.
Program terminated with signal 6, Aborted.
[New process 22157]
#0 0xffffe410 in __kernel_vsyscall ()
(gdb) bt
#0 0xffffe410 in __kernel_vsyscall ()
#1 0xf7dc5c81 in raise () from /lib/libpthread.so.0
#2 0xf73d4d43 in skgesigOSCrash () from /opt/oracle/11.2.0.2/lib/libclntsh.so.11.1
#3 0xf7643d61 in kpeDbgSignalHandler () from /opt/oracle/11.2.0.2/lib/libclntsh.so.11.1
#4 0xf73d5003 in skgesig_sigactionHandler () from /opt/oracle/11.2.0.2/lib/libclntsh.so.11.1
#5 <signal handler called>
#6 0xffffe410 in __kernel_vsyscall ()
#7 0x00be7df0 in raise () from /lib/libc.so.6
#8 0x00be9701 in abort () from /lib/libc.so.6
#9 0x0804b716 in main (argc=2, argv=0xffabe244) at app.cc:371
From the stack trace, it seems that the crash is occurring at oracle client library libclntsh.so.11.1 . When looked at Oracle forums, it says that Pro *C application consistently produces core dumps with above stack trace in Oracle client 11.2.0.2 . Such traces are resulting from the function sqlnst (). But in this case, the stack is called from the main function of app.
Is there a possibility that this crash is happening while making system calls as indicated by __kernel_vsyscall ().
Can you please advice what is the most effective way to handle such crashes and what could be the reason for these crashes.