Re: rcu_sched stall detected, but no state dump

From: Paul E. McKenney
Date: Mon Dec 22 2014 - 14:40:36 EST


On Mon, Dec 22, 2014 at 01:42:14PM +0100, Miroslav Benes wrote:
> On Fri, 19 Dec 2014, Paul E. McKenney wrote:
> > On Fri, Dec 19, 2014 at 12:09:48PM +0100, Miroslav Benes wrote:

[ . . . ]

> > But what do you do if your test module happens to start spinning when
> > there is no callback?
> >
> > One approach is to do a call_rcu() just after starting your loop.
> > Alternatively, do several just before starting your loop.
>
> Yes, that would be the way if I understand it correctly. Fortunately there
> is always some pending RCU callback during my test. I wonder where it
> comes from (probably module handling. I'll try to find out just to be
> sure.).
>
> So, is the following patch correct? I tested it and it seems ok. And sorry
> for the delay, I took the weekend off. Thank you for your patience.

Good! I forward ported it to -rcu, merged the two "if" statements,
queued it, and have started testing. Thank you!!!

Thanx, Paul

> Miroslav
>
> -- >8 --
> >From 0dccc8e322e1a38016422be8f4af110eb7a405fc Mon Sep 17 00:00:00 2001
> From: Miroslav Benes <mbenes@xxxxxxx>
> Date: Mon, 22 Dec 2014 13:24:44 +0100
> Subject: [PATCH] rcu: Fix RCU CPU stall detection in tiny implementation
>
> The tiny RCU CPU stall detection depends on *rcp->curtail not being
> NULL. It is however a tail pointer and thus NULL by definition. Instead we
> should check rcp->rcucblist for the presence of pending callbacks which need to
> be processed. With this fix INFO about the stall is printed and jiffies_stall
> (jiffies at next stall) correctly updated.
>
> Note that the check for pending callback is necessary to avoid spurious warnings
> if there are no pendings callbacks.
>
> Signed-off-by: Miroslav Benes <mbenes@xxxxxxx>
> ---
> kernel/rcu/tiny_plugin.h | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/kernel/rcu/tiny_plugin.h b/kernel/rcu/tiny_plugin.h
> index 858c565..868e132 100644
> --- a/kernel/rcu/tiny_plugin.h
> +++ b/kernel/rcu/tiny_plugin.h
> @@ -145,13 +145,13 @@ static void check_cpu_stall(struct rcu_ctrlblk *rcp)
> rcp->ticks_this_gp++;
> j = jiffies;
> js = ACCESS_ONCE(rcp->jiffies_stall);
> - if (*rcp->curtail && ULONG_CMP_GE(j, js)) {
> + if (rcp->rcucblist && ULONG_CMP_GE(j, js)) {
> pr_err("INFO: %s stall on CPU (%lu ticks this GP) idle=%llx (t=%lu jiffies q=%ld)\n",
> rcp->name, rcp->ticks_this_gp, rcu_dynticks_nesting,
> jiffies - rcp->gp_start, rcp->qlen);
> dump_stack();
> }
> - if (*rcp->curtail && ULONG_CMP_GE(j, js))
> + if (rcp->rcucblist && ULONG_CMP_GE(j, js))
> ACCESS_ONCE(rcp->jiffies_stall) = jiffies +
> 3 * rcu_jiffies_till_stall_check() + 3;
> else if (ULONG_CMP_GE(j, js))
> --
> 2.1.2
>

--
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/