This is the mail archive of the
systemtap@sources.redhat.com
mailing list for the systemtap project.
Re: Runtime probe example causes oops
- From: Martin Hunt <hunt at redhat dot com>
- To: Hien Nguyen <hien at us dot ibm dot com>
- Cc: "systemtap at sources dot redhat dot com" <systemtap at sources dot redhat dot com>, "Spirakis, Charles" <charles dot spirakis at intel dot com>
- Date: Wed, 13 Jul 2005 09:47:46 -0700
- Subject: Re: Runtime probe example causes oops
- Organization: Red Hat Inc.
- References: <42D404BC.2060306@us.ibm.com>
On Tue, 2005-07-12 at 10:58 -0700, Hien Nguyen wrote:
> Hi Martin,
>
> I just get the latest runtime from the CVS, and try it out on my
> machine. My machine is running 2.6.13-rc1.
>
> I did the following
> make
> cd probes/os_timer
> ./stp os_timers.ko
> Press Control-C to stop.
> Using channel with 4 sub-buffers of size 65536.
>
> This message came in the log
> Jul 12 10:41:53 elm3b99 kernel: os_timer: no version for
> "relay_subbufs_consumed" found: kernel tainted.
>
> Press Control-c I got an Oops
os_timer was written by Charles. I hadn't really looked at it. It
works fine on my Pentium M running RHEL4 with the latest patches. Well,
it doesn't crash but the data it collects is not right. I'm a bit
concerned that the oops shows it was in the runtime, but I cannot see
how it could have crashed at that particular point. Do you always get
the Oops? If so, could you try putting some printks in
_stp_symbol_sprint() in sym.c? The Oops says
_stp_symbol_print() called kallsyms_lookup() and that is where it
crashed.
Martin