This is the mail archive of the
systemtap@sourceware.org
mailing list for the systemtap project.
Re: 3 bugs found.
- From: Vara Prasad <prasadav at us dot ibm dot com>
- To: "Keshavamurthy, Anil S" <anil dot s dot keshavamurthy at intel dot com>, Prasanna S Panchamukhi <prasanna at in dot ibm dot com>, Ananth N Mavinakayanahalli <ananth at in dot ibm dot com>, Masami Hiramatsu <hiramatu at sdl dot hitachi dot co dot jp>
- Cc: David Smith <dsmith at redhat dot com>, James Dickens <jamesd dot wi at gmail dot com>, SystemTAP <systemtap at sources dot redhat dot com>
- Date: Wed, 20 Sep 2006 10:04:13 -0700
- Subject: Re: 3 bugs found.
- References: <cd09bdd10609151235m7ca053eek2636595a8b3f0565@mail.gmail.com> <450FF333.60701@redhat.com> <451166A2.5050009@redhat.com>
David Smith wrote:
[...]
I've found it. If I add '_raw_spin_unlock' to the blacklist (along
with 'atomic_notifier_call_chain' and '_spin_unlock_irqrestore'), then
probing kernel.function("*") works fine for me on x86.
David, thanks a bunch for narrowing it down to this small list.
Note that once again I'm not sure that is the correct fix (adding it
to the blacklist), I just wanted to get past it.
I think our kprobes experts can now write a simple kprobes module to
reproduce the problem and narrow it down further to see what is the
fundamental problem. If it turns out something we can't change these
functions or some call these functions make to be safe to probe we may
be able to put magic __kprobes macro to prevent anyone stumbling into
these functions via probes.
Do I hear any volunteers from the kprobe folks in the To list?