Segmentation fault

  • Vladimir

    Vladimir - 2006-11-20

    Please note that this is about "Segmentation fault", not ORA-12514.

    Description on how to reproduce the bug:

    $ opr -a <database> <schemaname> <osuser>
    please enter the password :
    please re-enter the password :
    ORA-12514: TNS:listener could not resolve SERVICE_NAME given in connect descriptor
    entry not added.
    Segmentation fault


    OS version and hardware architecture:

    uname -m = i686
    uname -r = 2.4.21-47.ELsmp
    uname -s = Linux
    uname -v = #1 SMP Wed Jul 5 20:38:41 EDT 2006


    Vendor, name and version of the compiler:

    gcc version 2.96 20000731 (Red Hat Linux 7.3 2.96-128)


    Oracle Database 10g Enterprise Edition Release

    Client: Release


    Thank you,

    • Jan-Marten Spit

      Jan-Marten Spit - 2006-11-22

      oracle@quibus[ORCL] /opt/oracle $ opr -a ORSCL system oracle
      please enter the password :
      please re-enter the password :
      ORA-12154: TNS:could not resolve the connect identifier specified
      entry not added.
      oracle@quibus[ORCL] /opt/oracle $

      weird, cannot reproduce it on gentoo linux. it looks like it's related to the unloading of the oracle this could very well be an oracle bug. 9204 is not the most stable of releases :) Also a slim change that this is gcc 2.95 related.

      as you are on linux, could you run it from gdb:
      $ gdb opr
      (gdb) set args -a SID dbuser osuser
      (gdb) run

      and report what gdb says is the call that causes the segfault?

      you might have to recompile opr if you stripped the debug info from the executable with strip, e.g. when
      oracle@quibus[ORCL] /opt/oracle $ file `which opr`
      /usr/local/bin/opr: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), for GNU/Linux 2.4.1, dynamically linked (uses shared libs), for GNU/Linux 2.4.1, not stripped

      says "stripped" instead of "not stripped"



Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

No, thanks