[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <1465423725.7945.25.camel@edumazet-glaptop3.roam.corp.google.com>
Date: Wed, 08 Jun 2016 15:08:45 -0700
From: Eric Dumazet <eric.dumazet@...il.com>
To: Florian Westphal <fw@...len.de>
Cc: netdev@...r.kernel.org
Subject: Re: [PATCH -next 3/4] sched: remove qdisc_rehape_fail
On Wed, 2016-06-08 at 23:01 +0200, Florian Westphal wrote:
>
> Would you mind an annotation rather than covering the hole?
>
> --- a/include/net/sch_generic.h
> +++ b/include/net/sch_generic.h
> @@ -71,11 +71,11 @@ struct Qdisc {
> struct gnet_stats_basic_cpu __percpu *cpu_bstats;
> struct gnet_stats_queue __percpu *cpu_qstats;
>
> - struct Qdisc *next_sched;
> - struct sk_buff *gso_skb;
> /*
> * For performance sake on SMP, we put highly modified fields at the end
> */
> + struct Qdisc *next_sched ____cacheline_aligned_in_smp;
> + struct sk_buff *gso_skb;
>
>
> ... it creates 16 byte hole after cpu_qstats and keeps the rest as-is
> (i.e. next_sched is at beginning of 2nd cacheline, as before the removal).
>
> I could also cover the hole by moving rcu_head there but it seems fragile
> and doesn't reduce total struct size anyway (we get larger hole at end).
>
> If you have no objection I'd resubmit the series as-is but with this patch.
>
> Let me know, thanks Eric!
This all looks fine to me.
Powered by blists - more mailing lists