linux-stable.git/net, branch v5.6.4

Bluetooth: RFCOMM: fix ODEBUG bug in rfcomm_dev_ioctl

2020-04-13T11:18:09+00:00

commit 71811cac8532b2387b3414f7cd8fe9e497482864 upstream.

Needn't call 'rfcomm_dlc_put' here, because 'rfcomm_dlc_exists' didn't
increase dlc->refcnt.

Reported-by: syzbot+4496e82090657320efc6@syzkaller.appspotmail.com
Signed-off-by: Qiujun Huang 
Suggested-by: Hillf Danton 
Signed-off-by: Marcel Holtmann 
Signed-off-by: Greg Kroah-Hartman

net_sched: fix a missing refcnt in tcindex_init()

2020-04-13T11:17:52+00:00

[ Upstream commit a8eab6d35e22f4f21471f16147be79529cd6aaf7 ]

The initial refcnt of struct tcindex_data should be 1,
it is clear that I forgot to set it to 1 in tcindex_init().
This leads to a dec-after-zero warning.

Reported-by: syzbot+8325e509a1bf83ec741d@syzkaller.appspotmail.com
Fixes: 304e024216a8 ("net_sched: add a temporary refcnt for struct tcindex_data")
Cc: Jamal Hadi Salim 
Cc: Jiri Pirko 
Cc: Paul E. McKenney 
Signed-off-by: Cong Wang 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

net_sched: add a temporary refcnt for struct tcindex_data

2020-04-13T11:17:51+00:00

[ Upstream commit 304e024216a802a7dc8ba75d36de82fa136bbf3e ]

Although we intentionally use an ordered workqueue for all tc
filter works, the ordering is not guaranteed by RCU work,
given that tcf_queue_work() is esstenially a call_rcu().

This problem is demostrated by Thomas:

  CPU 0:
    tcf_queue_work()
      tcf_queue_work(&r->rwork, tcindex_destroy_rexts_work);

  -> Migration to CPU 1

  CPU 1:
     tcf_queue_work(&p->rwork, tcindex_destroy_work);

so the 2nd work could be queued before the 1st one, which leads
to a free-after-free.

Enforcing this order in RCU work is hard as it requires to change
RCU code too. Fortunately we can workaround this problem in tcindex
filter by taking a temporary refcnt, we only refcnt it right before
we begin to destroy it. This simplifies the code a lot as a full
refcnt requires much more changes in tcindex_set_parms().

Reported-by: syzbot+46f513c3033d592409d2@syzkaller.appspotmail.com
Fixes: 3d210534cc93 ("net_sched: fix a race condition in tcindex_destroy()")
Cc: Thomas Gleixner 
Cc: Paul E. McKenney 
Cc: Jamal Hadi Salim 
Cc: Jiri Pirko 
Signed-off-by: Cong Wang 
Reviewed-by: Paul E. McKenney 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

ipv6: don't auto-add link-local address to lag ports

2020-04-13T11:17:48+00:00

[ Upstream commit 744fdc8233f6aa9582ce08a51ca06e59796a3196 ]

Bonding slave and team port devices should not have link-local addresses
automatically added to them, as it can interfere with openvswitch being
able to properly add tc ingress.

Basic reproducer, courtesy of Marcelo:

$ ip link add name bond0 type bond
$ ip link set dev ens2f0np0 master bond0
$ ip link set dev ens2f1np2 master bond0
$ ip link set dev bond0 up
$ ip a s
1: lo:  mtu 65536 qdisc noqueue state UNKNOWN
group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: ens2f0np0:  mtu 1500 qdisc
mq master bond0 state UP group default qlen 1000
    link/ether 00:0f:53:2f:ea:40 brd ff:ff:ff:ff:ff:ff
5: ens2f1np2:  mtu 1500 qdisc
mq master bond0 state DOWN group default qlen 1000
    link/ether 00:0f:53:2f:ea:40 brd ff:ff:ff:ff:ff:ff
11: bond0:  mtu 1500 qdisc
noqueue state UP group default qlen 1000
    link/ether 00:0f:53:2f:ea:40 brd ff:ff:ff:ff:ff:ff
    inet6 fe80::20f:53ff:fe2f:ea40/64 scope link
       valid_lft forever preferred_lft forever

(above trimmed to relevant entries, obviously)

$ sysctl net.ipv6.conf.ens2f0np0.addr_gen_mode=0
net.ipv6.conf.ens2f0np0.addr_gen_mode = 0
$ sysctl net.ipv6.conf.ens2f1np2.addr_gen_mode=0
net.ipv6.conf.ens2f1np2.addr_gen_mode = 0

$ ip a l ens2f0np0
2: ens2f0np0:  mtu 1500 qdisc
mq master bond0 state UP group default qlen 1000
    link/ether 00:0f:53:2f:ea:40 brd ff:ff:ff:ff:ff:ff
    inet6 fe80::20f:53ff:fe2f:ea40/64 scope link tentative
       valid_lft forever preferred_lft forever
$ ip a l ens2f1np2
5: ens2f1np2:  mtu 1500 qdisc
mq master bond0 state DOWN group default qlen 1000
    link/ether 00:0f:53:2f:ea:40 brd ff:ff:ff:ff:ff:ff
    inet6 fe80::20f:53ff:fe2f:ea40/64 scope link tentative
       valid_lft forever preferred_lft forever

Looks like addrconf_sysctl_addr_gen_mode() bypasses the original "is
this a slave interface?" check added by commit c2edacf80e15, and
results in an address getting added, while w/the proposed patch added,
no address gets added. This simply adds the same gating check to another
code path, and thus should prevent the same devices from erroneously
obtaining an ipv6 link-local address.

Fixes: d35a00b8e33d ("net/ipv6: allow sysctl to change link-local address generation mode")
Reported-by: Moshe Levi 
CC: Stephen Hemminger 
CC: Marcelo Ricardo Leitner 
CC: netdev@vger.kernel.org
Signed-off-by: Jarod Wilson 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

udp: initialize is_flist with 0 in udp_gro_receive

2020-04-08T07:11:04+00:00

[ Upstream commit bde1b56f898ca8112912d7b36e55e1543b3be0cf ]

Without NAPI_GRO_CB(skb)->is_flist initialized, when the dev doesn't
support NETIF_F_GRO_FRAGLIST, is_flist can still be set and fraglist
will be used in udp_gro_receive().

So fix it by initializing is_flist with 0 in udp_gro_receive.

Fixes: 9fd1ff5d2ac7 ("udp: Support UDP fraglist GRO/GSO.")
Signed-off-by: Xin Long 
Acked-by: Steffen Klassert 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

net: fix fraglist segmentation reference count leak

2020-04-08T07:11:03+00:00

[ Upstream commit cf673ed0e057a2dd68d930c6d7e30d53c70c5789 ]

Xin Long says:
 On udp rx path udp_rcv_segment() may do segment where the frag skbs
 will get the header copied from the head skb in skb_segment_list()
 by calling __copy_skb_header(), which could overwrite the frag skbs'
 extensions by __skb_ext_copy() and cause a leak.

 This issue was found after loading esp_offload where a sec path ext
 is set in the skb.

Fix this by discarding head state of the fraglist skb before replacing
its contents.

Fixes: 3a1296a38d0cf62 ("net: Support GRO/GSO fraglist chaining.")
Cc: Steffen Klassert 
Reported-by: Xiumei Mu 
Tested-by: Xin Long 
Signed-off-by: Florian Westphal 
Acked-by: Steffen Klassert 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

sctp: fix refcount bug in sctp_wfree

2020-04-08T07:11:03+00:00

[ Upstream commit 5c3e82fe159622e46e91458c1a6509c321a62820 ]

We should iterate over the datamsgs to move
all chunks(skbs) to newsk.

The following case cause the bug:
for the trouble SKB, it was in outq->transmitted list

sctp_outq_sack
        sctp_check_transmitted
                SKB was moved to outq->sacked list
        then throw away the sack queue
                SKB was deleted from outq->sacked
(but it was held by datamsg at sctp_datamsg_to_asoc
So, sctp_wfree was not called here)

then migrate happened

        sctp_for_each_tx_datachunk(
        sctp_clear_owner_w);
        sctp_assoc_migrate();
        sctp_for_each_tx_datachunk(
        sctp_set_owner_w);
SKB was not in the outq, and was not changed to newsk

finally

__sctp_outq_teardown
        sctp_chunk_put (for another skb)
                sctp_datamsg_put
                        __kfree_skb(msg->frag_list)
                                sctp_wfree (for SKB)
	SKB->sk was still oldsk (skb->sk != asoc->base.sk).

Reported-and-tested-by: syzbot+cea71eec5d6de256d54d@syzkaller.appspotmail.com
Signed-off-by: Qiujun Huang 
Acked-by: Marcelo Ricardo Leitner 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

sctp: fix possibly using a bad saddr with a given dst

2020-04-08T07:11:03+00:00

[ Upstream commit 582eea230536a6f104097dd46205822005d5fe3a ]

Under certain circumstances, depending on the order of addresses on the
interfaces, it could be that sctp_v[46]_get_dst() would return a dst
with a mismatched struct flowi.

For example, if when walking through the bind addresses and the first
one is not a match, it saves the dst as a fallback (added in
410f03831c07), but not the flowi. Then if the next one is also not a
match, the previous dst will be returned but with the flowi information
for the 2nd address, which is wrong.

The fix is to use a locally stored flowi that can be used for such
attempts, and copy it to the parameter only in case it is a possible
match, together with the corresponding dst entry.

The patch updates IPv6 code mostly just to be in sync. Even though the issue
is also present there, it fallback is not expected to work with IPv6.

Fixes: 410f03831c07 ("sctp: add routing output fallback")
Reported-by: Jin Meng 
Signed-off-by: Marcelo Ricardo Leitner 
Tested-by: Xin Long 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

net, ip_tunnel: fix interface lookup with no key

2020-04-08T07:11:03+00:00

[ Upstream commit 25629fdaff2ff509dd0b3f5ff93d70a75e79e0a1 ]

when creating a new ipip interface with no local/remote configuration,
the lookup is done with TUNNEL_NO_KEY flag, making it impossible to
match the new interface (only possible match being fallback or metada
case interface); e.g: `ip link add tunl1 type ipip dev eth0`

To fix this case, adding a flag check before the key comparison so we
permit to match an interface with no local/remote config; it also avoids
breaking possible userland tools relying on TUNNEL_NO_KEY flag and
uninitialised key.

context being on my side, I'm creating an extra ipip interface attached
to the physical one, and moving it to a dedicated namespace.

Fixes: c54419321455 ("GRE: Refactor GRE tunneling code.")
Signed-off-by: William Dauchy 
Signed-off-by: Nicolas Dichtel 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

ipv4: fix a RCU-list lock in fib_triestat_seq_show

2020-04-08T07:11:02+00:00

[ Upstream commit fbe4e0c1b298b4665ee6915266c9d6c5b934ef4a ]

fib_triestat_seq_show() calls hlist_for_each_entry_rcu(tb, head,
tb_hlist) without rcu_read_lock() will trigger a warning,

 net/ipv4/fib_trie.c:2579 RCU-list traversed in non-reader section!!

 other info that might help us debug this:

 rcu_scheduler_active = 2, debug_locks = 1
 1 lock held by proc01/115277:
  #0: c0000014507acf00 (&p->lock){+.+.}-{3:3}, at: seq_read+0x58/0x670

 Call Trace:
  dump_stack+0xf4/0x164 (unreliable)
  lockdep_rcu_suspicious+0x140/0x164
  fib_triestat_seq_show+0x750/0x880
  seq_read+0x1a0/0x670
  proc_reg_read+0x10c/0x1b0
  __vfs_read+0x3c/0x70
  vfs_read+0xac/0x170
  ksys_read+0x7c/0x140
  system_call+0x5c/0x68

Fix it by adding a pair of rcu_read_lock/unlock() and use
cond_resched_rcu() to avoid the situation where walking of a large
number of items  may prevent scheduling for a long time.

Signed-off-by: Qian Cai 
Reviewed-by: Eric Dumazet 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman