Re: [RFC] [PATCH 2.6.37-rc5-tip 4/20] 4: uprobes: Adding andremove a uprobe in a rb tree.

[Peter Zijlstra]

On Wed, 2011-01-26 at 14:11 +0530, Srikar Dronamraju wrote:
> * Peter Zijlstra <peterz@xxxxxxxxxxxxx> [2011-01-25 13:15:42]:
> 
> > On Thu, 2010-12-16 at 15:28 +0530, Srikar Dronamraju wrote:
> > > +/* Should be called lock-less */
> > > +static void put_uprobe(struct uprobe *uprobe)
> > > +{
> > > +       if (atomic_dec_and_test(&uprobe->ref))
> > > +               kfree(uprobe);
> > > +} 
> > 
> > Since this instantly frees the uprobe once ref hits 0, the
> > atomic_inc_not_zero() in find_uprobe() doesn't really make sense does
> > it?
> 
> Okay, I can move the atomic_inc_not_zero() in find_uprobe() to
> atomic_inc().
> 
> Do you see any side-effects of using atomic_inc_not_zero?

No, its just slower, once you want to start doing RCU lookups in the
probe tree you'll need it through.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/