Hide Forgot
While testing MCE injection, got this vmcore porcessing failure: --- Running analyze_VMcore --- Installing kernel debuginfo(s) Coredump references 0 debuginfo files, 1 of them are not installed Setting up yum repositories Looking for needed packages in repositories Can't find packages for 1 debuginfo files All debuginfo files are available Generating backtrace abrt-dump-oops: Found oopses: 0 There may be more than one problem here. First one is this clearly broken behavior of debuginfo installer: # /usr/libexec/abrt-action-install-debuginfo-to-abrt-cache --size_mb=4096 --exact=/usr/lib/debug/lib/modules/2.6.32-430.el6.i686/vmlinux Coredump references 0 debuginfo files, 1 of them are not installed Setting up yum repositories Looking for needed packages in repositories Can't find packages for 1 debuginfo files All debuginfo files are available [root@guinea ~]# /usr/libexec/abrt-action-install-debuginfo-to-abrt-cache --size_mb=4096 --exact=/usr/lib/debug/lib/modules/2.6.32-430.el6.i686/vmlinux; echo $? Coredump references 0 debuginfo files, 1 of them are not installed Setting up yum repositories Looking for needed packages in repositories Can't find packages for 1 debuginfo files All debuginfo files are available 0 # ls /usr/lib/debug/lib/modules/2.6.32-430.el6.i686/vmlinux ls: cannot access /usr/lib/debug/lib/modules/2.6.32-430.el6.i686/vmlinux: No such file or directory Debuginfo was NOT installed. Exit code is wrong as well.
Without double-pasted mess: # /usr/libexec/abrt-action-install-debuginfo-to-abrt-cache --size_mb=4096 --exact=/usr/lib/debug/lib/modules/2.6.32-430.el6.i686/vmlinux; echo $? Coredump references 0 debuginfo files, 1 of them are not installed Setting up yum repositories Looking for needed packages in repositories Can't find packages for 1 debuginfo files All debuginfo files are available 0
Since debuginfo installer lied to us, abrt-action-analyze-vmcore thinks that everything is fine, and runs "crash -s vmcore <path-to-missing-debuginfo", storing result in "kernel_log" element... which ends up as: kernel_log ===== crash: /var/cache/abrt-di/usr/lib/debug/lib/modules/2.6.32-430.el6.i686/vmlinux: No such file or directory Usage: crash [OPTION]... NAMELIST MEMORY-IMAGE (dumpfile form) crash [OPTION]... [NAMELIST] (live system form) Enter "crash -h" for details. ===== abrt-action-analyze-vmcore does not check "crash -s" exit code (this is a second bug). Then abrt-dump-oops "Found oopses: 0" in kernel_log. Not really a surprise...
This request was not resolved in time for the current release. Red Hat invites you to ask your support representative to propose this request, if still desired, for consideration in the next release of Red Hat Enterprise Linux.
Upstream already has a bit different code, so the fix below may be unnucessary, but on RHEL6.5 machine the folloving hack in vmcore_event.conf works around the problem: EVENT=analyze_VMcore analyzer=vmcore # hack if test -f vmcore-dmesg.txt; then exec abrt-dump-oops -o vmcore-dmesg.txt >backtrace fi # end hack abrt-action-analyze-vmcore It uses the fact that kdump in RHEL6 already saves dmesg output as a separate file.
This is an artificial issue. It rarely can happen. The upstream code is different now. It does not make sense to me fixing this.