Re: Processed: reassign 633535 to systemtap

["Frank Ch. Eigler" <fche@redhat.com>]

Hi, Ben -

> > > > reassign 633535 systemtap
> > > Bug reassigned from package 'linux-2.6' to 'systemtap'.
> > 
> > Can you elaborate why you believe systemtap is related?  According to
> > the dmesg, the last systemtap invocation was several hours before the
> > kernel error.
>  
> The last module unloaded was generated by SystemTap.  

Is that all?  That will often be the case on a machine that runs
systemtap occasionally, since it is the only widget that regularly
removes (its own) kernel modules.  In this case, the recorded time lag
between the last systemtap startup and the crash was just under six
days (!).  In our experience, a few minutes after exit would
constitute a reasonable statute of limitations.  (Inconveniently, stap
exits are not normally recorded in dmesg.)


> If you don't think the bug has anything to do with SystemTap then we
> can ask the submitter to try to reproduce the bug without it.

There is much more information needed from the submitter, regardless.
For example:
- What was that ptrace-related workload being run?
  (On debian, ptrace has no relations to systemtap.)
- How long before the kernel error was the last systemtap script completed?
- What are those itrace / itrace2 dmesg's?
- Can the problem be reproduced?  Reliably or intermittently? 
  On a freshly booted machine?


- FChE