Re: CVS commit: src/sys

[Ryota Ozaki <ozaki-r%netbsd.org@localhost>]

On Fri, Jul 15, 2016 at 4:20 AM, Martin Husemann <martin%duskware.de@localhost> wrote:
>> > >> Modified Files:
>> > >>  src/sys/net: route.c
>> > >>  src/sys/netinet: ip_flow.c
>> > >>  src/sys/netinet6: ip6_flow.c nd6.c
>
> It is specifically the route.c change, backing that one out avoids the
> problem for me.
>
> I would suggest this patch:
>
> Index: route.c
> ===================================================================
> RCS file: /cvsroot/src/sys/net/route.c,v
> retrieving revision 1.171
> diff -u -p -r1.171 route.c
> --- route.c     13 Jul 2016 09:56:20 -0000      1.171
> +++ route.c     14 Jul 2016 19:12:41 -0000
> @@ -1160,12 +1160,12 @@ rt_timer_init(void)
>         assert(rt_init_done == 0);
>
>         LIST_INIT(&rttimer_queue_head);
> -       callout_init(&rt_timer_ch, 0);
> -       callout_reset(&rt_timer_ch, hz, rt_timer_timer, NULL);
> +       callout_init(&rt_timer_ch, CALLOUT_MPSAFE);
>         error = workqueue_create(&rt_timer_wq, "rt_timer",
>             rt_timer_work, NULL, PRI_SOFTNET, IPL_SOFTNET, WQ_MPSAFE);
>         if (error)
>                 panic("%s: workqueue_create failed (%d)\n", __func__, error);
> +       callout_reset(&rt_timer_ch, hz, rt_timer_timer, NULL);
>         rt_init_done = 1;
>  }

Thanks. LGTM.

>
>
> but it does not fix the issue. I don't see anything obviously wrong here
> nor understand why it seems to be mips specific.

Does NFS root matter?

  ozaki-r