netdev - Re: [PATCH net] ipv6: kill sk_dst

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <1448935312.25582.6.camel@edumazet-glaptop2.roam.corp.google.com>
Date:	Mon, 30 Nov 2015 18:01:52 -0800
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	YOSHIFUJI Hideaki <hideaki.yoshifuji@...aclelinux.com>
Cc:	Dmitry Vyukov <dvyukov@...gle.com>,
	"David S. Miller" <davem@...emloft.net>,
	Vlad Yasevich <vyasevich@...il.com>,
	Hideaki YOSHIFUJI <yoshfuji@...ux-ipv6.org>,
	netdev <netdev@...r.kernel.org>,
	Eric Dumazet <edumazet@...gle.com>
Subject: Re: [PATCH net] ipv6: kill sk_dst_lock

On Tue, 2015-12-01 at 10:44 +0900, YOSHIFUJI Hideaki wrote:
> Hi,
> 
> Eric Dumazet wrote:
> > diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
> > index 2bfb2ad2fab1..877f682989b8 100644
> > --- a/include/net/ip6_route.h
> > +++ b/include/net/ip6_route.h
> > @@ -133,27 +133,18 @@ void rt6_clean_tohost(struct net *net, struct in6_addr *gateway);
> >  /*
> >   *	Store a destination cache entry in a socket
> >   */
> > -static inline void __ip6_dst_store(struct sock *sk, struct dst_entry *dst,
> > -				   const struct in6_addr *daddr,
> > -				   const struct in6_addr *saddr)
> > +static inline void ip6_dst_store(struct sock *sk, struct dst_entry *dst,
> > +				 const struct in6_addr *daddr,
> > +				 const struct in6_addr *saddr)
> >  {
> >  	struct ipv6_pinfo *np = inet6_sk(sk);
> > -	struct rt6_info *rt = (struct rt6_info *) dst;
> >  
> > +	np->dst_cookie = rt6_get_cookie((struct rt6_info *)dst);
> >  	sk_setup_caps(sk, dst);
> >  	np->daddr_cache = daddr;
> >  #ifdef CONFIG_IPV6_SUBTREES
> >  	np->saddr_cache = saddr;
> >  #endif
> > -	np->dst_cookie = rt6_get_cookie(rt);
> > -}
> > -
> 
> I believe you do not have to change function inside, right?

I knew I forgot something in my changelog :

ip6_dst_store() can be called from process context.

As soon as the dst is installed in sk->sk_dst_cache, dst can be freed
from another cpu doing a concurrent ip6_dst_store()

Doing the dst dereference before doing the install is safer.

Otherwise, we need to add rcu_read_lock() extra sections.

I believe we have other bugs like this one (deref dst after
sk_setup_caps() calls) that need an audit.

But I prefer making smaller patches addressing one problem at a time...

Thanks !


--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html