November 2017 - Linux-stable-mirror

[Linux-stable-mirror] Patch "tcp/dccp: fix lockdep splat in inet_csk_route_req()" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tcp/dccp: fix lockdep splat in inet_csk_route_req() to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tcp-dccp-fix-lockdep-splat-in-inet_csk_route_req.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Wed Nov 15 17:24:03 CET 2017 From: Eric Dumazet <edumazet(a)google.com> Date: Sun, 22 Oct 2017 12:33:57 -0700 Subject: tcp/dccp: fix lockdep splat in inet_csk_route_req() From: Eric Dumazet <edumazet(a)google.com> [ Upstream commit a6ca7abe53633d08eea1c6756cb49c9b2d4c90bf ] This patch fixes the following lockdep splat in inet_csk_route_req() lockdep_rcu_suspicious inet_csk_route_req tcp_v4_send_synack tcp_rtx_synack inet_rtx_syn_ack tcp_fastopen_synack_time tcp_retransmit_timer tcp_write_timer_handler tcp_write_timer call_timer_fn Thread running inet_csk_route_req() owns a reference on the request socket, so we have the guarantee ireq->ireq_opt wont be changed or freed. lockdep can enforce this invariant for us. Fixes: c92e8c02fe66 ("tcp/dccp: fix ireq->opt races") Signed-off-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/inet_connection_sock.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -410,7 +410,8 @@ struct dst_entry *inet_csk_route_req(con struct ip_options_rcu *opt; struct rtable *rt; - opt = rcu_dereference(ireq->ireq_opt); + opt = rcu_dereference_protected(ireq->ireq_opt, + atomic_read(&req->rsk_refcnt) > 0); flowi4_init_output(fl4, ireq->ir_iif, ireq->ir_mark, RT_CONN_FLAGS(sk), RT_SCOPE_UNIVERSE, sk->sk_protocol, inet_sk_flowi_flags(sk), Patches currently in stable-queue which might be from edumazet(a)google.com are queue-4.9/net-call-cgroup_sk_alloc-earlier-in-sk_clone_lock.patch queue-4.9/tcp-dccp-fix-ireq-opt-races.patch queue-4.9/tcp-fix-tcp_mtu_probe-vs-highest_sack.patch queue-4.9/ipv6-addrconf-increment-ifp-refcount-before-ipv6_del_addr.patch queue-4.9/ipv6-flowlabel-do-not-leave-opt-tot_len-with-garbage.patch queue-4.9/packet-avoid-panic-in-packet_getsockopt.patch queue-4.9/sctp-add-the-missing-sock_owned_by_user-check-in-sctp_icmp_redirect.patch queue-4.9/net_sched-avoid-matching-qdisc-with-zero-handle.patch queue-4.9/tun-tap-sanitize-tunsetsndbuf-input.patch queue-4.9/tcp-dccp-fix-lockdep-splat-in-inet_csk_route_req.patch queue-4.9/tcp-dccp-fix-other-lockdep-splats-accessing-ireq_opt.patch

7 years, 1 month

1
0
0 0

[Linux-stable-mirror] Patch "soreuseport: fix initialization race" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled soreuseport: fix initialization race to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: soreuseport-fix-initialization-race.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Wed Nov 15 17:24:03 CET 2017 From: Craig Gallek <kraig(a)google.com> Date: Thu, 19 Oct 2017 15:00:29 -0400 Subject: soreuseport: fix initialization race From: Craig Gallek <kraig(a)google.com> [ Upstream commit 1b5f962e71bfad6284574655c406597535c3ea7a ] Syzkaller stumbled upon a way to trigger WARNING: CPU: 1 PID: 13881 at net/core/sock_reuseport.c:41 reuseport_alloc+0x306/0x3b0 net/core/sock_reuseport.c:39 There are two initialization paths for the sock_reuseport structure in a socket: Through the udp/tcp bind paths of SO_REUSEPORT sockets or through SO_ATTACH_REUSEPORT_[CE]BPF before bind. The existing implementation assumedthat the socket lock protected both of these paths when it actually only protects the SO_ATTACH_REUSEPORT path. Syzkaller triggered this double allocation by running these paths concurrently. This patch moves the check for double allocation into the reuseport_alloc function which is protected by a global spin lock. Fixes: e32ea7e74727 ("soreuseport: fast reuseport UDP socket selection") Fixes: c125e80b8868 ("soreuseport: fast reuseport TCP socket selection") Signed-off-by: Craig Gallek <kraig(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/core/sock_reuseport.c | 12 +++++++++--- net/ipv4/inet_hashtables.c | 5 +---- net/ipv4/udp.c | 5 +---- 3 files changed, 11 insertions(+), 11 deletions(-) --- a/net/core/sock_reuseport.c +++ b/net/core/sock_reuseport.c @@ -36,9 +36,14 @@ int reuseport_alloc(struct sock *sk) * soft irq of receive path or setsockopt from process context */ spin_lock_bh(&reuseport_lock); - WARN_ONCE(rcu_dereference_protected(sk->sk_reuseport_cb, - lockdep_is_held(&reuseport_lock)), - "multiple allocations for the same socket"); + + /* Allocation attempts can occur concurrently via the setsockopt path + * and the bind/hash path. Nothing to do when we lose the race. + */ + if (rcu_dereference_protected(sk->sk_reuseport_cb, + lockdep_is_held(&reuseport_lock))) + goto out; + reuse = __reuseport_alloc(INIT_SOCKS); if (!reuse) { spin_unlock_bh(&reuseport_lock); @@ -49,6 +54,7 @@ int reuseport_alloc(struct sock *sk) reuse->num_socks = 1; rcu_assign_pointer(sk->sk_reuseport_cb, reuse); +out: spin_unlock_bh(&reuseport_lock); return 0; --- a/net/ipv4/inet_hashtables.c +++ b/net/ipv4/inet_hashtables.c @@ -455,10 +455,7 @@ static int inet_reuseport_add_sock(struc return reuseport_add_sock(sk, sk2); } - /* Initial allocation may have already happened via setsockopt */ - if (!rcu_access_pointer(sk->sk_reuseport_cb)) - return reuseport_alloc(sk); - return 0; + return reuseport_alloc(sk); } int __inet_hash(struct sock *sk, struct sock *osk, --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -222,10 +222,7 @@ static int udp_reuseport_add_sock(struct } } - /* Initial allocation may have already happened via setsockopt */ - if (!rcu_access_pointer(sk->sk_reuseport_cb)) - return reuseport_alloc(sk); - return 0; + return reuseport_alloc(sk); } /** Patches currently in stable-queue which might be from kraig(a)google.com are queue-4.9/soreuseport-fix-initialization-race.patch queue-4.9/tun-tap-sanitize-tunsetsndbuf-input.patch

7 years, 1 month

1
0
0 0

[Linux-stable-mirror] Patch "tap: double-free in error path in tap_open()" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tap: double-free in error path in tap_open() to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tap-double-free-in-error-path-in-tap_open.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Wed Nov 15 17:24:03 CET 2017 From: Girish Moodalbail <girish.moodalbail(a)oracle.com> Date: Wed, 25 Oct 2017 00:23:04 -0700 Subject: tap: double-free in error path in tap_open() From: Girish Moodalbail <girish.moodalbail(a)oracle.com> [ Upstream commit 78e0ea6791d7baafb8a0ca82b1bd0c7b3453c919 ] Double free of skb_array in tap module is causing kernel panic. When tap_set_queue() fails we free skb_array right away by calling skb_array_cleanup(). However, later on skb_array_cleanup() is called again by tap_sock_destruct through sock_put(). This patch fixes that issue. Fixes: 362899b8725b35e3 (macvtap: switch to use skb array) Signed-off-by: Girish Moodalbail <girish.moodalbail(a)oracle.com> Acked-by: Jason Wang <jasowang(a)redhat.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/macvtap.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) --- a/drivers/net/macvtap.c +++ b/drivers/net/macvtap.c @@ -559,6 +559,10 @@ static int macvtap_open(struct inode *in &macvtap_proto, 0); if (!q) goto err; + if (skb_array_init(&q->skb_array, dev->tx_queue_len, GFP_KERNEL)) { + sk_free(&q->sk); + goto err; + } RCU_INIT_POINTER(q->sock.wq, &q->wq); init_waitqueue_head(&q->wq.wait); @@ -582,22 +586,18 @@ static int macvtap_open(struct inode *in if ((dev->features & NETIF_F_HIGHDMA) && (dev->features & NETIF_F_SG)) sock_set_flag(&q->sk, SOCK_ZEROCOPY); - err = -ENOMEM; - if (skb_array_init(&q->skb_array, dev->tx_queue_len, GFP_KERNEL)) - goto err_array; - err = macvtap_set_queue(dev, file, q); - if (err) - goto err_queue; + if (err) { + /* macvtap_sock_destruct() will take care of freeing skb_array */ + goto err_put; + } dev_put(dev); rtnl_unlock(); return err; -err_queue: - skb_array_cleanup(&q->skb_array); -err_array: +err_put: sock_put(&q->sk); err: if (dev) Patches currently in stable-queue which might be from girish.moodalbail(a)oracle.com are queue-4.9/tap-double-free-in-error-path-in-tap_open.patch

7 years, 1 month

1
0
0 0

[Linux-stable-mirror] Patch "sctp: reset owner sk for data chunks on out queues when migrating a sock" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sctp: reset owner sk for data chunks on out queues when migrating a sock to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sctp-reset-owner-sk-for-data-chunks-on-out-queues-when-migrating-a-sock.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Wed Nov 15 17:24:03 CET 2017 From: Xin Long <lucien.xin(a)gmail.com> Date: Sat, 28 Oct 2017 02:13:29 +0800 Subject: sctp: reset owner sk for data chunks on out queues when migrating a sock From: Xin Long <lucien.xin(a)gmail.com> [ Upstream commit d04adf1b355181e737b6b1e23d801b07f0b7c4c0 ] Now when migrating sock to another one in sctp_sock_migrate(), it only resets owner sk for the data in receive queues, not the chunks on out queues. It would cause that data chunks length on the sock is not consistent with sk sk_wmem_alloc. When closing the sock or freeing these chunks, the old sk would never be freed, and the new sock may crash due to the overflow sk_wmem_alloc. syzbot found this issue with this series: r0 = socket$inet_sctp() sendto$inet(r0) listen(r0) accept4(r0) close(r0) Although listen() should have returned error when one TCP-style socket is in connecting (I may fix this one in another patch), it could also be reproduced by peeling off an assoc. This issue is there since very beginning. This patch is to reset owner sk for the chunks on out queues so that sk sk_wmem_alloc has correct value after accept one sock or peeloff an assoc to one sock. Note that when resetting owner sk for chunks on outqueue, it has to sctp_clear_owner_w/skb_orphan chunks before changing assoc->base.sk first and then sctp_set_owner_w them after changing assoc->base.sk, due to that sctp_wfree and it's callees are using assoc->base.sk. Reported-by: Dmitry Vyukov <dvyukov(a)google.com> Signed-off-by: Xin Long <lucien.xin(a)gmail.com> Acked-by: Marcelo Ricardo Leitner <marcelo.leitner(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/sctp/socket.c | 32 ++++++++++++++++++++++++++++++++ 1 file changed, 32 insertions(+) --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -168,6 +168,36 @@ static inline void sctp_set_owner_w(stru sk_mem_charge(sk, chunk->skb->truesize); } +static void sctp_clear_owner_w(struct sctp_chunk *chunk) +{ + skb_orphan(chunk->skb); +} + +static void sctp_for_each_tx_datachunk(struct sctp_association *asoc, + void (*cb)(struct sctp_chunk *)) + +{ + struct sctp_outq *q = &asoc->outqueue; + struct sctp_transport *t; + struct sctp_chunk *chunk; + + list_for_each_entry(t, &asoc->peer.transport_addr_list, transports) + list_for_each_entry(chunk, &t->transmitted, transmitted_list) + cb(chunk); + + list_for_each_entry(chunk, &q->retransmit, list) + cb(chunk); + + list_for_each_entry(chunk, &q->sacked, list) + cb(chunk); + + list_for_each_entry(chunk, &q->abandoned, list) + cb(chunk); + + list_for_each_entry(chunk, &q->out_chunk_list, list) + cb(chunk); +} + /* Verify that this is a valid address. */ static inline int sctp_verify_addr(struct sock *sk, union sctp_addr *addr, int len) @@ -7826,7 +7856,9 @@ static void sctp_sock_migrate(struct soc * paths won't try to lock it and then oldsk. */ lock_sock_nested(newsk, SINGLE_DEPTH_NESTING); + sctp_for_each_tx_datachunk(assoc, sctp_clear_owner_w); sctp_assoc_migrate(assoc, newsk); + sctp_for_each_tx_datachunk(assoc, sctp_set_owner_w); /* If the association on the newsk is already closed before accept() * is called, set RCV_SHUTDOWN flag. Patches currently in stable-queue which might be from lucien.xin(a)gmail.com are queue-4.9/sctp-reset-owner-sk-for-data-chunks-on-out-queues-when-migrating-a-sock.patch queue-4.9/ipip-only-increase-err_count-for-some-certain-type-icmp-in-ipip_err.patch queue-4.9/sctp-full-support-for-ipv6-ip_nonlocal_bind-ip_freebind.patch queue-4.9/sctp-add-the-missing-sock_owned_by_user-check-in-sctp_icmp_redirect.patch queue-4.9/ip6_gre-only-increase-err_count-for-some-certain-type-icmpv6-in-ip6gre_err.patch queue-4.9/ip6_gre-update-dst-pmtu-if-dev-mtu-has-been-updated-by-toobig-in-__gre6_xmit.patch

7 years, 1 month

1
0
0 0

[Linux-stable-mirror] Patch "sctp: full support for ipv6 ip_nonlocal_bind & IP_FREEBIND" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sctp: full support for ipv6 ip_nonlocal_bind & IP_FREEBIND to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sctp-full-support-for-ipv6-ip_nonlocal_bind-ip_freebind.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Wed Nov 15 17:24:03 CET 2017 From: Laszlo Toth <laszlth(a)gmail.com> Date: Mon, 23 Oct 2017 19:19:33 +0200 Subject: sctp: full support for ipv6 ip_nonlocal_bind & IP_FREEBIND From: Laszlo Toth <laszlth(a)gmail.com> [ Upstream commit b71d21c274eff20a9db8158882b545b141b73ab8 ] Commit 9b9742022888 ("sctp: support ipv6 nonlocal bind") introduced support for the above options as v4 sctp did, so patched sctp_v6_available(). In the v4 implementation it's enough, because sctp_inet_bind_verify() just returns with sctp_v4_available(). However sctp_inet6_bind_verify() has an extra check before that for link-local scope_id, which won't respect the above options. Added the checks before calling ipv6_chk_addr(), but not before the validation of scope_id. before (w/ both options): ./v6test fe80::10 sctp bind failed, errno: 99 (Cannot assign requested address) ./v6test fe80::10 tcp bind success, errno: 0 (Success) after (w/ both options): ./v6test fe80::10 sctp bind success, errno: 0 (Success) Signed-off-by: Laszlo Toth <laszlth(a)gmail.com> Reviewed-by: Xin Long <lucien.xin(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/sctp/ipv6.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) --- a/net/sctp/ipv6.c +++ b/net/sctp/ipv6.c @@ -881,8 +881,10 @@ static int sctp_inet6_bind_verify(struct net = sock_net(&opt->inet.sk); rcu_read_lock(); dev = dev_get_by_index_rcu(net, addr->v6.sin6_scope_id); - if (!dev || - !ipv6_chk_addr(net, &addr->v6.sin6_addr, dev, 0)) { + if (!dev || !(opt->inet.freebind || + net->ipv6.sysctl.ip_nonlocal_bind || + ipv6_chk_addr(net, &addr->v6.sin6_addr, + dev, 0))) { rcu_read_unlock(); return 0; } Patches currently in stable-queue which might be from laszlth(a)gmail.com are queue-4.9/sctp-full-support-for-ipv6-ip_nonlocal_bind-ip_freebind.patch

7 years, 1 month

1
0
0 0

[Linux-stable-mirror] Patch "sctp: add the missing sock_owned_by_user check in sctp_icmp_redirect" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sctp: add the missing sock_owned_by_user check in sctp_icmp_redirect to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sctp-add-the-missing-sock_owned_by_user-check-in-sctp_icmp_redirect.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Wed Nov 15 17:24:03 CET 2017 From: Xin Long <lucien.xin(a)gmail.com> Date: Wed, 18 Oct 2017 21:37:49 +0800 Subject: sctp: add the missing sock_owned_by_user check in sctp_icmp_redirect From: Xin Long <lucien.xin(a)gmail.com> [ Upstream commit 1cc276cec9ec574d41cf47dfc0f51406b6f26ab4 ] Now sctp processes icmp redirect packet in sctp_icmp_redirect where it calls sctp_transport_dst_check in which tp->dst can be released. The problem is before calling sctp_transport_dst_check, it doesn't check sock_owned_by_user, which means tp->dst could be freed while a process is accessing it with owning the socket. An use-after-free issue could be triggered by this. This patch is to fix it by checking sock_owned_by_user before calling sctp_transport_dst_check in sctp_icmp_redirect, so that it would not release tp->dst if users still hold sock lock. Besides, the same issue fixed in commit 45caeaa5ac0b ("dccp/tcp: fix routing redirect race") on sctp also needs this check. Fixes: 55be7a9c6074 ("ipv4: Add redirect support to all protocol icmp error handlers") Reported-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Xin Long <lucien.xin(a)gmail.com> Acked-by: Marcelo Ricardo Leitner <marcelo.leitner(a)gmail.com> Acked-by: Neil Horman <nhorman(a)tuxdriver.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/sctp/input.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/net/sctp/input.c +++ b/net/sctp/input.c @@ -421,7 +421,7 @@ void sctp_icmp_redirect(struct sock *sk, { struct dst_entry *dst; - if (!t) + if (sock_owned_by_user(sk) || !t) return; dst = sctp_transport_dst_check(t); if (dst) Patches currently in stable-queue which might be from lucien.xin(a)gmail.com are queue-4.9/sctp-reset-owner-sk-for-data-chunks-on-out-queues-when-migrating-a-sock.patch queue-4.9/ipip-only-increase-err_count-for-some-certain-type-icmp-in-ipip_err.patch queue-4.9/sctp-full-support-for-ipv6-ip_nonlocal_bind-ip_freebind.patch queue-4.9/sctp-add-the-missing-sock_owned_by_user-check-in-sctp_icmp_redirect.patch queue-4.9/ip6_gre-only-increase-err_count-for-some-certain-type-icmpv6-in-ip6gre_err.patch queue-4.9/ip6_gre-update-dst-pmtu-if-dev-mtu-has-been-updated-by-toobig-in-__gre6_xmit.patch

7 years, 1 month

1
0
0 0

[Linux-stable-mirror] Patch "ppp: fix race in ppp device destruction" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ppp: fix race in ppp device destruction to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ppp-fix-race-in-ppp-device-destruction.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Wed Nov 15 17:24:03 CET 2017 From: Guillaume Nault <g.nault(a)alphalink.fr> Date: Fri, 6 Oct 2017 17:05:49 +0200 Subject: ppp: fix race in ppp device destruction From: Guillaume Nault <g.nault(a)alphalink.fr> [ Upstream commit 6151b8b37b119e8e3a8401b080d532520c95faf4 ] ppp_release() tries to ensure that netdevices are unregistered before decrementing the unit refcount and running ppp_destroy_interface(). This is all fine as long as the the device is unregistered by ppp_release(): the unregister_netdevice() call, followed by rtnl_unlock(), guarantee that the unregistration process completes before rtnl_unlock() returns. However, the device may be unregistered by other means (like ppp_nl_dellink()). If this happens right before ppp_release() calling rtnl_lock(), then ppp_release() has to wait for the concurrent unregistration code to release the lock. But rtnl_unlock() releases the lock before completing the device unregistration process. This allows ppp_release() to proceed and eventually call ppp_destroy_interface() before the unregistration process completes. Calling free_netdev() on this partially unregistered device will BUG(): ------------[ cut here ]------------ kernel BUG at net/core/dev.c:8141! invalid opcode: 0000 [#1] SMP CPU: 1 PID: 1557 Comm: pppd Not tainted 4.14.0-rc2+ #4 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1.fc26 04/01/2014 Call Trace: ppp_destroy_interface+0xd8/0xe0 [ppp_generic] ppp_disconnect_channel+0xda/0x110 [ppp_generic] ppp_unregister_channel+0x5e/0x110 [ppp_generic] pppox_unbind_sock+0x23/0x30 [pppox] pppoe_connect+0x130/0x440 [pppoe] SYSC_connect+0x98/0x110 ? do_fcntl+0x2c0/0x5d0 SyS_connect+0xe/0x10 entry_SYSCALL_64_fastpath+0x1a/0xa5 RIP: free_netdev+0x107/0x110 RSP: ffffc28a40573d88 ---[ end trace ed294ff0cc40eeff ]--- We could set the ->needs_free_netdev flag on PPP devices and move the ppp_destroy_interface() logic in the ->priv_destructor() callback. But that'd be quite intrusive as we'd first need to unlink from the other channels and units that depend on the device (the ones that used the PPPIOCCONNECT and PPPIOCATTACH ioctls). Instead, we can just let the netdevice hold a reference on its ppp_file. This reference is dropped in ->priv_destructor(), at the very end of the unregistration process, so that neither ppp_release() nor ppp_disconnect_channel() can call ppp_destroy_interface() in the interim. Reported-by: Beniamino Galvani <bgalvani(a)redhat.com> Fixes: 8cb775bc0a34 ("ppp: fix device unregistration upon netns deletion") Signed-off-by: Guillaume Nault <g.nault(a)alphalink.fr> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ppp/ppp_generic.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) --- a/drivers/net/ppp/ppp_generic.c +++ b/drivers/net/ppp/ppp_generic.c @@ -1338,7 +1338,17 @@ ppp_get_stats64(struct net_device *dev, static int ppp_dev_init(struct net_device *dev) { + struct ppp *ppp; + netdev_lockdep_set_classes(dev); + + ppp = netdev_priv(dev); + /* Let the netdevice take a reference on the ppp file. This ensures + * that ppp_destroy_interface() won't run before the device gets + * unregistered. + */ + atomic_inc(&ppp->file.refcnt); + return 0; } @@ -1361,6 +1371,15 @@ static void ppp_dev_uninit(struct net_de wake_up_interruptible(&ppp->file.rwait); } +static void ppp_dev_priv_destructor(struct net_device *dev) +{ + struct ppp *ppp; + + ppp = netdev_priv(dev); + if (atomic_dec_and_test(&ppp->file.refcnt)) + ppp_destroy_interface(ppp); +} + static const struct net_device_ops ppp_netdev_ops = { .ndo_init = ppp_dev_init, .ndo_uninit = ppp_dev_uninit, @@ -1386,6 +1405,7 @@ static void ppp_setup(struct net_device dev->tx_queue_len = 3; dev->type = ARPHRD_PPP; dev->flags = IFF_POINTOPOINT | IFF_NOARP | IFF_MULTICAST; + dev->destructor = ppp_dev_priv_destructor; netif_keep_dst(dev); } Patches currently in stable-queue which might be from g.nault(a)alphalink.fr are queue-4.9/ppp-fix-race-in-ppp-device-destruction.patch queue-4.9/l2tp-check-ps-sock-before-running-pppol2tp_session_ioctl.patch

7 years, 1 month

1
0
0 0

[Linux-stable-mirror] Patch "netlink: do not set cb_running if dump's start() errs" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled netlink: do not set cb_running if dump's start() errs to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: netlink-do-not-set-cb_running-if-dump-s-start-errs.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Wed Nov 15 17:24:03 CET 2017 From: "Jason A. Donenfeld" <Jason(a)zx2c4.com> Date: Mon, 9 Oct 2017 14:14:51 +0200 Subject: netlink: do not set cb_running if dump's start() errs From: "Jason A. Donenfeld" <Jason(a)zx2c4.com> [ Upstream commit 41c87425a1ac9b633e0fcc78eb1f19640c8fb5a0 ] It turns out that multiple places can call netlink_dump(), which means it's still possible to dereference partially initialized values in dump() that were the result of a faulty returned start(). This fixes the issue by calling start() _before_ setting cb_running to true, so that there's no chance at all of hitting the dump() function through any indirect paths. It also moves the call to start() to be when the mutex is held. This has the nice side effect of serializing invocations to start(), which is likely desirable anyway. It also prevents any possible other races that might come out of this logic. In testing this with several different pieces of tricky code to trigger these issues, this commit fixes all avenues that I'm aware of. Signed-off-by: Jason A. Donenfeld <Jason(a)zx2c4.com> Cc: Johannes Berg <johannes(a)sipsolutions.net> Reviewed-by: Johannes Berg <johannes(a)sipsolutions.net> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/netlink/af_netlink.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -2207,16 +2207,17 @@ int __netlink_dump_start(struct sock *ss cb->min_dump_alloc = control->min_dump_alloc; cb->skb = skb; + if (cb->start) { + ret = cb->start(cb); + if (ret) + goto error_unlock; + } + nlk->cb_running = true; mutex_unlock(nlk->cb_mutex); - ret = 0; - if (cb->start) - ret = cb->start(cb); - - if (!ret) - ret = netlink_dump(sk); + ret = netlink_dump(sk); sock_put(sk); Patches currently in stable-queue which might be from Jason(a)zx2c4.com are queue-4.9/netlink-do-not-set-cb_running-if-dump-s-start-errs.patch

7 years, 1 month

1
0
0 0

[Linux-stable-mirror] Patch "packet: avoid panic in packet_getsockopt()" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled packet: avoid panic in packet_getsockopt() to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: packet-avoid-panic-in-packet_getsockopt.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Wed Nov 15 17:24:03 CET 2017 From: Eric Dumazet <edumazet(a)google.com> Date: Wed, 18 Oct 2017 16:14:52 -0700 Subject: packet: avoid panic in packet_getsockopt() From: Eric Dumazet <edumazet(a)google.com> [ Upstream commit 509c7a1ecc8601f94ffba8a00889fefb239c00c6 ] syzkaller got crashes in packet_getsockopt() processing PACKET_ROLLOVER_STATS command while another thread was managing to change po->rollover Using RCU will fix this bug. We might later add proper RCU annotations for sparse sake. In v2: I replaced kfree(rollover) in fanout_add() to kfree_rcu() variant, as spotted by John. Fixes: a9b6391814d5 ("packet: rollover statistics") Signed-off-by: Eric Dumazet <edumazet(a)google.com> Cc: Willem de Bruijn <willemb(a)google.com> Cc: John Sperbeck <jsperbeck(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/packet/af_packet.c | 24 ++++++++++++++++-------- 1 file changed, 16 insertions(+), 8 deletions(-) --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -1720,7 +1720,7 @@ static int fanout_add(struct sock *sk, u out: if (err && rollover) { - kfree(rollover); + kfree_rcu(rollover, rcu); po->rollover = NULL; } mutex_unlock(&fanout_mutex); @@ -1747,8 +1747,10 @@ static struct packet_fanout *fanout_rele else f = NULL; - if (po->rollover) + if (po->rollover) { kfree_rcu(po->rollover, rcu); + po->rollover = NULL; + } } mutex_unlock(&fanout_mutex); @@ -3851,6 +3853,7 @@ static int packet_getsockopt(struct sock void *data = &val; union tpacket_stats_u st; struct tpacket_rollover_stats rstats; + struct packet_rollover *rollover; if (level != SOL_PACKET) return -ENOPROTOOPT; @@ -3929,13 +3932,18 @@ static int packet_getsockopt(struct sock 0); break; case PACKET_ROLLOVER_STATS: - if (!po->rollover) + rcu_read_lock(); + rollover = rcu_dereference(po->rollover); + if (rollover) { + rstats.tp_all = atomic_long_read(&rollover->num); + rstats.tp_huge = atomic_long_read(&rollover->num_huge); + rstats.tp_failed = atomic_long_read(&rollover->num_failed); + data = &rstats; + lv = sizeof(rstats); + } + rcu_read_unlock(); + if (!rollover) return -EINVAL; - rstats.tp_all = atomic_long_read(&po->rollover->num); - rstats.tp_huge = atomic_long_read(&po->rollover->num_huge); - rstats.tp_failed = atomic_long_read(&po->rollover->num_failed); - data = &rstats; - lv = sizeof(rstats); break; case PACKET_TX_HAS_OFF: val = po->tp_tx_has_off; Patches currently in stable-queue which might be from edumazet(a)google.com are queue-4.9/net-call-cgroup_sk_alloc-earlier-in-sk_clone_lock.patch queue-4.9/tcp-dccp-fix-ireq-opt-races.patch queue-4.9/tcp-fix-tcp_mtu_probe-vs-highest_sack.patch queue-4.9/ipv6-addrconf-increment-ifp-refcount-before-ipv6_del_addr.patch queue-4.9/ipv6-flowlabel-do-not-leave-opt-tot_len-with-garbage.patch queue-4.9/packet-avoid-panic-in-packet_getsockopt.patch queue-4.9/sctp-add-the-missing-sock_owned_by_user-check-in-sctp_icmp_redirect.patch queue-4.9/net_sched-avoid-matching-qdisc-with-zero-handle.patch queue-4.9/tun-tap-sanitize-tunsetsndbuf-input.patch queue-4.9/tcp-dccp-fix-lockdep-splat-in-inet_csk_route_req.patch queue-4.9/tcp-dccp-fix-other-lockdep-splats-accessing-ireq_opt.patch

7 years, 1 month

1
0
0 0

[Linux-stable-mirror] Patch "net_sched: avoid matching qdisc with zero handle" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net_sched: avoid matching qdisc with zero handle to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net_sched-avoid-matching-qdisc-with-zero-handle.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Wed Nov 15 17:24:03 CET 2017 From: Cong Wang <xiyou.wangcong(a)gmail.com> Date: Fri, 27 Oct 2017 22:08:56 -0700 Subject: net_sched: avoid matching qdisc with zero handle From: Cong Wang <xiyou.wangcong(a)gmail.com> [ Upstream commit 50317fce2cc70a2bbbc4b42c31bbad510382a53c ] Davide found the following script triggers a NULL pointer dereference: ip l a name eth0 type dummy tc q a dev eth0 parent :1 handle 1: htb This is because for a freshly created netdevice noop_qdisc is attached and when passing 'parent :1', kernel actually tries to match the major handle which is 0 and noop_qdisc has handle 0 so is matched by mistake. Commit 69012ae425d7 tries to fix a similar bug but still misses this case. Handle 0 is not a valid one, should be just skipped. In fact, kernel uses it as TC_H_UNSPEC. Fixes: 69012ae425d7 ("net: sched: fix handling of singleton qdiscs with qdisc_hash") Fixes: 59cc1f61f09c ("net: sched:convert qdisc linked list to hashtable") Reported-by: Davide Caratti <dcaratti(a)redhat.com> Cc: Jiri Kosina <jkosina(a)suse.cz> Cc: Eric Dumazet <edumazet(a)google.com> Cc: Jamal Hadi Salim <jhs(a)mojatatu.com> Signed-off-by: Cong Wang <xiyou.wangcong(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/sched/sch_api.c | 2 ++ 1 file changed, 2 insertions(+) --- a/net/sched/sch_api.c +++ b/net/sched/sch_api.c @@ -296,6 +296,8 @@ struct Qdisc *qdisc_lookup(struct net_de { struct Qdisc *q; + if (!handle) + return NULL; q = qdisc_match_from_root(dev->qdisc, handle); if (q) goto out; Patches currently in stable-queue which might be from xiyou.wangcong(a)gmail.com are queue-4.9/tun-call-dev_get_valid_name-before-register_netdevice.patch queue-4.9/net_sched-avoid-matching-qdisc-with-zero-handle.patch queue-4.9/tun-allow-positive-return-values-on-dev_get_valid_name-call.patch

7 years, 1 month

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror November 2017