Re: [patch] kprobes for 2.5.73 with single-stepping out-of-line

Vamsi Krishna S . (vamsi@in.ibm.com)
Wed, 25 Jun 2003 18:16:33 +0530


On Wed, Jun 25, 2003 at 01:52:02PM +0200, Andi Kleen wrote:
> On Wed, Jun 25, 2003 at 04:11:13PM +0530, Vamsi Krishna S . wrote:
> > On Tue, Jun 24, 2003 at 06:01:09PM +0200, Andi Kleen wrote:
> > > "Vamsi Krishna S ." <vamsi@in.ibm.com> writes:
> > >
> > >
> > > > +static struct kprobe *current_kprobe;
> > >
> > > This global variable is quite unclean. It looks like it is for passing
> > > function arguments around. Why is it needed?
> > >
> > This is used for keeping track of the probe that is currently being
> > handled. This information is needed to be kept across a
> > trap 3 - singlestep - trap 1. So, we set store the current probe in
> > this variable while handling trap 3, for use while handling the
> > subsequent trap 1.
>
> But how can this be SMP safe? Do you hold a lock during all this?
>
Yes.. in the interests of keeping the locking to a minimum, only one
probe can be handled at a time, even on SMP.

> -Andi
Thanks,
Vamsi.

-- 
Vamsi Krishna S.
IBM Software Lab, Bangalore.
Ph: +91 80 5044959
Internet: vamsi@in.ibm.com
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/