Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1729956 > unrolled thread

Division by zero on UP (was: Re: netfilter: nat: use keyed locks)

Started byGeert Uytterhoeven <geert@linux-m68k.org>
First post2017-09-10 13:10 +0200
Last post2017-09-10 13:40 +0200
Articles 3 — 2 participants

Back to article view | Back to linux.kernel

Contents

  Division by zero on UP (was: Re: netfilter: nat: use keyed locks) Geert Uytterhoeven <geert@linux-m68k.org> - 2017-09-10 13:10 +0200
    Re: Division by zero on UP (was: Re: netfilter: nat: use keyed locks) Florian Westphal <fw@strlen.de> - 2017-09-10 13:30 +0200
      Re: Division by zero on UP (was: Re: netfilter: nat: use keyed locks) Geert Uytterhoeven <geert@linux-m68k.org> - 2017-09-10 13:40 +0200

#1729956 — Division by zero on UP (was: Re: netfilter: nat: use keyed locks)

FromGeert Uytterhoeven <geert@linux-m68k.org>
Date2017-09-10 13:10 +0200
SubjectDivision by zero on UP (was: Re: netfilter: nat: use keyed locks)
Message-ID<uo8hH-2QD-5@gated-at.bofh.it>
Hi Florian, Pablo,

On Sat, Sep 9, 2017 at 9:21 PM, Linux Kernel Mailing List
<linux-kernel@vger.kernel.org> wrote:
> Web:        https://git.kernel.org/torvalds/c/8073e960a03bf7b5d5ebfc5ff18ac475e1688f46
> Commit:     8073e960a03bf7b5d5ebfc5ff18ac475e1688f46
> Parent:     e1bf1687740ce1a3598a1c5e452b852ff2190682
> Refname:    refs/heads/master
> Author:     Florian Westphal <fw@strlen.de>
> AuthorDate: Wed Sep 6 14:39:52 2017 +0200
> Committer:  Pablo Neira Ayuso <pablo@netfilter.org>
> CommitDate: Fri Sep 8 18:55:52 2017 +0200
>
>     netfilter: nat: use keyed locks
>
>     no need to serialize on a single lock, we can partition the table and
>     add/delete in parallel to different slots.
>     This restores one of the advantages that got lost with the rhlist
>     revert.
>
>     Cc: Ivan Babrou <ibobrik@gmail.com>
>     Signed-off-by: Florian Westphal <fw@strlen.de>
>     Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
> ---
>  net/netfilter/nf_nat_core.c | 36 ++++++++++++++++++++++++------------
>  1 file changed, 24 insertions(+), 12 deletions(-)
>
> diff --git a/net/netfilter/nf_nat_core.c b/net/netfilter/nf_nat_core.c
> index f090419f5f97..f393a7086025 100644
> --- a/net/netfilter/nf_nat_core.c
> +++ b/net/netfilter/nf_nat_core.c
> @@ -30,7 +30,7 @@
>  #include <net/netfilter/nf_conntrack_zones.h>
>  #include <linux/netfilter/nf_nat.h>
>
> -static DEFINE_SPINLOCK(nf_nat_lock);
> +static spinlock_t nf_nat_locks[CONNTRACK_LOCKS];
>
>  static DEFINE_MUTEX(nf_nat_proto_mutex);
>  static const struct nf_nat_l3proto __rcu *nf_nat_l3protos[NFPROTO_NUMPROTO]
> @@ -425,13 +425,15 @@ nf_nat_setup_info(struct nf_conn *ct,
>
>         if (maniptype == NF_NAT_MANIP_SRC) {
>                 unsigned int srchash;
> +               spinlock_t *lock;
>
>                 srchash = hash_by_src(net,
>                                       &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple);
> -               spin_lock_bh(&nf_nat_lock);
> +               lock = &nf_nat_locks[srchash % ARRAY_SIZE(nf_nat_locks)];
> +               spin_lock_bh(lock);

If no spinlock debugging options (CONFIG_GENERIC_LOCKBREAK,
CONFIG_DEBUG_SPINLOCK, CONFIG_DEBUG_LOCK_ALLOC) are enabled on a UP platform
(e.g. m68k defconfig), arch_spinlock_t is an empty struct, and thus
ARRAY_SIZE(nf_nat_locks)] is zero, leading to:

    net/netfilter/nf_nat_core.c: In function ‘nf_nat_setup_info’:
    net/netfilter/nf_nat_core.c:432: warning: division by zero
    net/netfilter/nf_nat_core.c: In function ‘__nf_nat_cleanup_conntrack’:
    net/netfilter/nf_nat_core.c:535: warning: division by zero
    net/netfilter/nf_nat_core.c:537: warning: division by zero
    net/netfilter/nf_nat_core.c: In function ‘nf_nat_init’:
    net/netfilter/nf_nat_core.c:810: warning: division by zero
    net/netfilter/nf_nat_core.c:811: warning: division by zero
    net/netfilter/nf_nat_core.c:824: warning: division by zero

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

[toc] | [next] | [standalone]

#1729960

FromFlorian Westphal <fw@strlen.de>
Date2017-09-10 13:30 +0200
Message-ID<uo8B4-309-9@gated-at.bofh.it>
In reply to#1729956
Geert Uytterhoeven <geert@linux-m68k.org> wrote:
> >                 srchash = hash_by_src(net,
> >                                       &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple);
> > -               spin_lock_bh(&nf_nat_lock);
> > +               lock = &nf_nat_locks[srchash % ARRAY_SIZE(nf_nat_locks)];
> > +               spin_lock_bh(lock);
> 
> If no spinlock debugging options (CONFIG_GENERIC_LOCKBREAK,
> CONFIG_DEBUG_SPINLOCK, CONFIG_DEBUG_LOCK_ALLOC) are enabled on a UP platform
> (e.g. m68k defconfig), arch_spinlock_t is an empty struct, and thus
> ARRAY_SIZE(nf_nat_locks)] is zero, leading to:
> 
>     net/netfilter/nf_nat_core.c: In function ‘nf_nat_setup_info’:
>     net/netfilter/nf_nat_core.c:432: warning: division by zero

Gah. Sorry.  This is the 2nd time I add such a bug :(

Can you send a patch?  It should be enough to use CONNTRACK_LOCKS
instead of ARRAY_SIZE(nf_nat_locks) to avoid this problem.

Otherwise I will take care of this next thing monday morning.

[toc] | [prev] | [next] | [standalone]

#1729962

FromGeert Uytterhoeven <geert@linux-m68k.org>
Date2017-09-10 13:40 +0200
Message-ID<uo8KK-33V-3@gated-at.bofh.it>
In reply to#1729960
Hi Florian,

On Sun, Sep 10, 2017 at 1:21 PM, Florian Westphal <fw@strlen.de> wrote:
> Geert Uytterhoeven <geert@linux-m68k.org> wrote:
>> >                 srchash = hash_by_src(net,
>> >                                       &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple);
>> > -               spin_lock_bh(&nf_nat_lock);
>> > +               lock = &nf_nat_locks[srchash % ARRAY_SIZE(nf_nat_locks)];
>> > +               spin_lock_bh(lock);
>>
>> If no spinlock debugging options (CONFIG_GENERIC_LOCKBREAK,
>> CONFIG_DEBUG_SPINLOCK, CONFIG_DEBUG_LOCK_ALLOC) are enabled on a UP platform
>> (e.g. m68k defconfig), arch_spinlock_t is an empty struct, and thus
>> ARRAY_SIZE(nf_nat_locks)] is zero, leading to:
>>
>>     net/netfilter/nf_nat_core.c: In function ‘nf_nat_setup_info’:
>>     net/netfilter/nf_nat_core.c:432: warning: division by zero
>
> Gah. Sorry.  This is the 2nd time I add such a bug :(
>
> Can you send a patch?  It should be enough to use CONNTRACK_LOCKS
> instead of ARRAY_SIZE(nf_nat_locks) to avoid this problem.
>
> Otherwise I will take care of this next thing monday morning.

Thanks for the suggestion, will send a patch RSN.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web