On Tue, Aug 27, 2002 at 10:24:30AM +1000, Rusty Russell wrote:
> In message <20020827022239.C31269@in.ibm.com> you write:
> > +static struct rcu_data rcu_data[NR_CPUS] __cacheline_aligned;
>
> Not "static DEFINE_PER_CPU(struct rcu_data, rcu_data)"?
Yes, I can use per-cpu data areas here. Done.
>
> > +/* Fake initialization to work around compiler breakage */
> > +DEFINE_PER_CPU(long, cpu_quiescent) = 0L;
>
> static? And I assume you're talking about the tendency for gcc 2.95
> to put uninitialized static vars in the bss, even if they are marked
> as having a section attribute? If so, you should say so.
>
> > +#ifdef CONFIG_PREEMPT
> > +/* Fake initialization to work around compiler breakage */
> > +DEFINE_PER_CPU(atomic_t[2], rcu_preempt_cntr) =
> > + {ATOMIC_INIT(0), ATOMIC_INIT(0)};
> > +DEFINE_PER_CPU(atomic_t, *curr_preempt_cntr) = NULL;
> > +DEFINE_PER_CPU(atomic_t, *next_preempt_cntr) = NULL;
>
> Also static I assume?
So, only statics are broken by gcc 2.95, right ? If so, then
the fake initializers aren't required. I got bitten by this in
another piece of RCU code where there was a static per-cpu
tasklet and I got paranoic after that.
>
> Other than that, it looks good. You should probably cc: Ingo Molnar
> as it touches the scheduler...
Ok, I will do that from now on. Thanks for the review.
I will have the new bits up as soon as I finish testing.
Thanks
-- Dipankar Sarma <dipankar@in.ibm.com> http://lse.sourceforge.net Linux Technology Center, IBM Software Lab, Bangalore, India. - 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/