| Message ID | 20210330064551.545964-1-eric.dumazet@gmail.com |
|---|---|
| State | New |
| Headers | show |
| Series | [net-next] ip6_tunnel: sit: proper dev_{hold|put} in ndo_[un]init methods | expand |
On Tue, Mar 30, 2021 at 8:45 AM Eric Dumazet <eric.dumazet@gmail.com> wrote: > > From: Eric Dumazet <edumazet@google.com> > > Same reasons than for the previous commits : > 6289a98f0817 ("sit: proper dev_{hold|put} in ndo_[un]init methods") > 40cb881b5aaa ("ip6_vti: proper dev_{hold|put} in ndo_[un]init methods") > 7f700334be9a ("ip6_gre: proper dev_{hold|put} in ndo_[un]init methods") > > After adopting CONFIG_PCPU_DEV_REFCNT=n option, syzbot was able to trigger > a warning [1] > > Issue here is that: > > - all dev_put() should be paired with a corresponding prior dev_hold(). > > - A driver doing a dev_put() in its ndo_uninit() MUST also > do a dev_hold() in its ndo_init(), only when ndo_init() > is returning 0. > > Otherwise, register_netdevice() would call ndo_uninit() > in its error path and release a refcount too soon. > > Note to David & Jakub Can you merge this patch so that I can send my global fix for fallback tunnels, with a correct Fixes: tag for this patch ? Thanks !
On Wed, Mar 31, 2021 at 8:00 AM Eric Dumazet <edumazet@google.com> wrote: > > On Tue, Mar 30, 2021 at 8:45 AM Eric Dumazet <eric.dumazet@gmail.com> wrote: > > > > From: Eric Dumazet <edumazet@google.com> > > > > Same reasons than for the previous commits : > > 6289a98f0817 ("sit: proper dev_{hold|put} in ndo_[un]init methods") > > 40cb881b5aaa ("ip6_vti: proper dev_{hold|put} in ndo_[un]init methods") > > 7f700334be9a ("ip6_gre: proper dev_{hold|put} in ndo_[un]init methods") > > > > After adopting CONFIG_PCPU_DEV_REFCNT=n option, syzbot was able to trigger > > a warning [1] > > > > Issue here is that: > > > > - all dev_put() should be paired with a corresponding prior dev_hold(). > > > > - A driver doing a dev_put() in its ndo_uninit() MUST also > > do a dev_hold() in its ndo_init(), only when ndo_init() > > is returning 0. > > > > Otherwise, register_netdevice() would call ndo_uninit() > > in its error path and release a refcount too soon. > > > > > > Note to David & Jakub > > Can you merge this patch so that I can send my global fix for fallback > tunnels, with a correct Fixes: tag for this patch ? > > Thanks ! Forgot to attach what the global fix would look like : From 2c9cce067a9a1aec32b323ef80179b1856c796af Mon Sep 17 00:00:00 2001 From: Eric Dumazet <edumazet@google.com> Date: Tue, 30 Mar 2021 13:29:08 -0700 Subject: [PATCH net-next] ipv6: remove extra dev_hold() for fallback tunnels My previous commits added a dev_hold() in tunnels ndo_init(), but forgot to remove it from special functions setting up fallback tunnels. Fallback tunnels do call their respective ndo_init() This leads to various reports like : unregister_netdevice: waiting for ip6gre0 to become free. Usage count = 2 Fixes: f21eb0195662 ("ip6_tunnel: sit: proper dev_{hold|put} in ndo_[un]init methods") Fixes: 6289a98f0817 ("sit: proper dev_{hold|put} in ndo_[un]init methods") Fixes: 40cb881b5aaa ("ip6_vti: proper dev_{hold|put} in ndo_[un]init methods") Fixes: 7f700334be9a ("ip6_gre: proper dev_{hold|put} in ndo_[un]init methods") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> --- net/ipv6/ip6_gre.c | 3 --- net/ipv6/ip6_tunnel.c | 1 - net/ipv6/ip6_vti.c | 1 - net/ipv6/sit.c | 1 - 4 files changed, 6 deletions(-) diff --git a/net/ipv6/ip6_gre.c b/net/ipv6/ip6_gre.c index 9689bf9f46f347562330a4d8630c0b0b13a411fc..bc224f917bbd53beb9b8af5bdef3fb9794b8ee44 100644 --- a/net/ipv6/ip6_gre.c +++ b/net/ipv6/ip6_gre.c @@ -387,7 +387,6 @@ static struct ip6_tnl *ip6gre_tunnel_locate(struct net *net, if (!(nt->parms.o_flags & TUNNEL_SEQ)) dev->features |= NETIF_F_LLTX; - dev_hold(dev); ip6gre_tunnel_link(ign, nt); return nt; @@ -1539,8 +1538,6 @@ static void ip6gre_fb_tunnel_init(struct net_device *dev) strcpy(tunnel->parms.name, dev->name); tunnel->hlen = sizeof(struct ipv6hdr) + 4; - - dev_hold(dev); } static struct inet6_protocol ip6gre_protocol __read_mostly = { diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index 67ee9d58ec5efcc81e8b27406bd4f57a0caea70b..07a0a06a9b52bc9974e2f36b1477c341c952f94a 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -1925,7 +1925,6 @@ static int __net_init ip6_fb_tnl_dev_init(struct net_device *dev) struct ip6_tnl_net *ip6n = net_generic(net, ip6_tnl_net_id); t->parms.proto = IPPROTO_IPV6; - dev_hold(dev); rcu_assign_pointer(ip6n->tnls_wc[0], t); return 0; diff --git a/net/ipv6/ip6_vti.c b/net/ipv6/ip6_vti.c index a018afdb3e062c9e664d4ca424176a859f0a332c..856e46ad0895b47b58896852afee3d4a398b139e 100644 --- a/net/ipv6/ip6_vti.c +++ b/net/ipv6/ip6_vti.c @@ -963,7 +963,6 @@ static int __net_init vti6_fb_tnl_dev_init(struct net_device *dev) struct vti6_net *ip6n = net_generic(net, vti6_net_id); t->parms.proto = IPPROTO_IPV6; - dev_hold(dev); rcu_assign_pointer(ip6n->tnls_wc[0], t); return 0; diff --git a/net/ipv6/sit.c b/net/ipv6/sit.c index 488d3181aec3a5558dbefb6145400627535df761..ff2ca2e7c7f5045663069ea572560d58abee2970 100644 --- a/net/ipv6/sit.c +++ b/net/ipv6/sit.c @@ -1470,7 +1470,6 @@ static void __net_init ipip6_fb_tunnel_init(struct net_device *dev) iph->ihl = 5; iph->ttl = 64; - dev_hold(dev); rcu_assign_pointer(sitn->tunnels_wc[0], tunnel); }
Hello: This patch was applied to netdev/net-next.git (refs/heads/master): On Mon, 29 Mar 2021 23:45:51 -0700 you wrote: > From: Eric Dumazet <edumazet@google.com> > > Same reasons than for the previous commits : > 6289a98f0817 ("sit: proper dev_{hold|put} in ndo_[un]init methods") > 40cb881b5aaa ("ip6_vti: proper dev_{hold|put} in ndo_[un]init methods") > 7f700334be9a ("ip6_gre: proper dev_{hold|put} in ndo_[un]init methods") > > [...] Here is the summary with links: - [net-next] ip6_tunnel: sit: proper dev_{hold|put} in ndo_[un]init methods https://git.kernel.org/netdev/net-next/c/48bb5697269a You are awesome, thank you! -- Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html
From: Eric Dumazet <edumazet@google.com> Date: Wed, 31 Mar 2021 08:00:24 +0200 > Can you merge this patch so that I can send my global fix for fallback > tunnels, with a correct Fixes: tag for this patch ? Done.
On Wed, Mar 31, 2021 at 11:39 PM David Miller <davem@davemloft.net> wrote: > > From: Eric Dumazet <edumazet@google.com> > Date: Wed, 31 Mar 2021 08:00:24 +0200 > > > Can you merge this patch so that I can send my global fix for fallback > > tunnels, with a correct Fixes: tag for this patch ? > > Done. Thanks a lot David, I have sent the followup patch, with updated Fixes: tags.
diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index cd78f5b2cd75e20a79fd6c4cbfafd548204c2b7c..67ee9d58ec5efcc81e8b27406bd4f57a0caea70b 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -266,7 +266,6 @@ static int ip6_tnl_create2(struct net_device *dev) strcpy(t->parms.name, dev->name); - dev_hold(dev); ip6_tnl_link(ip6n, t); return 0; @@ -1882,6 +1881,7 @@ ip6_tnl_dev_init_gen(struct net_device *dev) dev->min_mtu = ETH_MIN_MTU; dev->max_mtu = IP6_MAX_MTU - dev->hard_header_len; + dev_hold(dev); return 0; destroy_dst: