linux-stable.git/include/net, branch v5.4.192

tcp: fix potential xmit stalls caused by TCP_NOTSENT_LOWAT

2022-05-09T07:03:25+00:00

[ Upstream commit 4bfe744ff1644fbc0a991a2677dc874475dd6776 ]

I had this bug sitting for too long in my pile, it is time to fix it.

Thanks to Doug Porter for reminding me of it!

We had various attempts in the past, including commit
0cbe6a8f089e ("tcp: remove SOCK_QUEUE_SHRUNK"),
but the issue is that TCP stack currently only generates
EPOLLOUT from input path, when tp->snd_una has advanced
and skb(s) cleaned from rtx queue.

If a flow has a big RTT, and/or receives SACKs, it is possible
that the notsent part (tp->write_seq - tp->snd_nxt) reaches 0
and no more data can be sent until tp->snd_una finally advances.

What is needed is to also check if POLLOUT needs to be generated
whenever tp->snd_nxt is advanced, from output path.

This bug triggers more often after an idle period, as
we do not receive ACK for at least one RTT. tcp_notsent_lowat
could be a fraction of what CWND and pacing rate would allow to
send during this RTT.

In a followup patch, I will remove the bogus call
to tcp_chrono_stop(sk, TCP_CHRONO_SNDBUF_LIMITED)
from tcp_check_space(). Fact that we have decided to generate
an EPOLLOUT does not mean the application has immediately
refilled the transmit queue. This optimistic call
might have been the reason the bug seemed not too serious.

Tested:

200 ms rtt, 1% packet loss, 32 MB tcp_rmem[2] and tcp_wmem[2]

$ echo 500000 >/proc/sys/net/ipv4/tcp_notsent_lowat
$ cat bench_rr.sh
SUM=0
for i in {1..10}
do
 V=`netperf -H remote_host -l30 -t TCP_RR -- -r 10000000,10000 -o LOCAL_BYTES_SENT | egrep -v "MIGRATED|Bytes"`
 echo $V
 SUM=$(($SUM + $V))
done
echo SUM=$SUM

Before patch:
$ bench_rr.sh
130000000
80000000
140000000
140000000
140000000
140000000
130000000
40000000
90000000
110000000
SUM=1140000000

After patch:
$ bench_rr.sh
430000000
590000000
530000000
450000000
450000000
350000000
450000000
490000000
480000000
460000000
SUM=4680000000  # This is 410 % of the value before patch.

Fixes: c9bee3b7fdec ("tcp: TCP_NOTSENT_LOWAT socket option")
Signed-off-by: Eric Dumazet 
Reported-by: Doug Porter 
Cc: Soheil Hassas Yeganeh 
Cc: Neal Cardwell 
Acked-by: Soheil Hassas Yeganeh 
Signed-off-by: David S. Miller 
Signed-off-by: Sasha Levin

tcp: ensure to use the most recently sent skb when filling the rate sample

2022-05-09T07:03:24+00:00

[ Upstream commit b253a0680ceadc5d7b4acca7aa2d870326cad8ad ]

If an ACK (s)acks multiple skbs, we favor the information
from the most recently sent skb by choosing the skb with
the highest prior_delivered count. But in the interval
between receiving ACKs, we send multiple skbs with the same
prior_delivered, because the tp->delivered only changes
when we receive an ACK.

We used RACK's solution, copying tcp_rack_sent_after() as
tcp_skb_sent_after() helper to determine "which packet was
sent last?". Later, we will use tcp_skb_sent_after() instead
in RACK.

Fixes: b9f64820fb22 ("tcp: track data delivery rate for a TCP connection")
Signed-off-by: Pengcheng Yang 
Cc: Paolo Abeni 
Acked-by: Neal Cardwell 
Tested-by: Neal Cardwell 
Reviewed-by: Eric Dumazet 
Link: https://lore.kernel.org/r/1650422081-22153-1-git-send-email-yangpc@wangsu.com
Signed-off-by: Jakub Kicinski 
Signed-off-by: Sasha Levin

tcp: fix race condition when creating child sockets from syncookies

2022-04-27T11:50:45+00:00

[ Upstream commit 01770a166165738a6e05c3d911fb4609cc4eb416 ]

When the TCP stack is in SYN flood mode, the server child socket is
created from the SYN cookie received in a TCP packet with the ACK flag
set.

The child socket is created when the server receives the first TCP
packet with a valid SYN cookie from the client. Usually, this packet
corresponds to the final step of the TCP 3-way handshake, the ACK
packet. But is also possible to receive a valid SYN cookie from the
first TCP data packet sent by the client, and thus create a child socket
from that SYN cookie.

Since a client socket is ready to send data as soon as it receives the
SYN+ACK packet from the server, the client can send the ACK packet (sent
by the TCP stack code), and the first data packet (sent by the userspace
program) almost at the same time, and thus the server will equally
receive the two TCP packets with valid SYN cookies almost at the same
instant.

When such event happens, the TCP stack code has a race condition that
occurs between the momement a lookup is done to the established
connections hashtable to check for the existence of a connection for the
same client, and the moment that the child socket is added to the
established connections hashtable. As a consequence, this race condition
can lead to a situation where we add two child sockets to the
established connections hashtable and deliver two sockets to the
userspace program to the same client.

This patch fixes the race condition by checking if an existing child
socket exists for the same client when we are adding the second child
socket to the established connections socket. If an existing child
socket exists, we drop the packet and discard the second child socket
to the same client.

Signed-off-by: Ricardo Dias 
Signed-off-by: Eric Dumazet 
Link: https://lore.kernel.org/r/20201120111133.GA67501@rdias-suse-pc.lan
Signed-off-by: Jakub Kicinski 
Signed-off-by: Sasha Levin

ax25: fix reference count leaks of ax25_dev

2022-04-20T07:19:39+00:00

commit 87563a043cef044fed5db7967a75741cc16ad2b1 upstream.

The previous commit d01ffb9eee4a ("ax25: add refcount in ax25_dev
to avoid UAF bugs") introduces refcount into ax25_dev, but there
are reference leak paths in ax25_ctl_ioctl(), ax25_fwd_ioctl(),
ax25_rt_add(), ax25_rt_del() and ax25_rt_opt().

This patch uses ax25_dev_put() and adjusts the position of
ax25_addr_ax25dev() to fix reference cout leaks of ax25_dev.

Fixes: d01ffb9eee4a ("ax25: add refcount in ax25_dev to avoid UAF bugs")
Signed-off-by: Duoming Zhou 
Reviewed-by: Dan Carpenter 
Link: https://lore.kernel.org/r/20220203150811.42256-1-duoming@zju.edu.cn
Signed-off-by: Jakub Kicinski 
[OP: backport to 5.4: adjust context]
Signed-off-by: Ovidiu Panait 
Signed-off-by: Greg Kroah-Hartman

ax25: add refcount in ax25_dev to avoid UAF bugs

2022-04-20T07:19:39+00:00

commit d01ffb9eee4af165d83b08dd73ebdf9fe94a519b upstream.

If we dereference ax25_dev after we call kfree(ax25_dev) in
ax25_dev_device_down(), it will lead to concurrency UAF bugs.
There are eight syscall functions suffer from UAF bugs, include
ax25_bind(), ax25_release(), ax25_connect(), ax25_ioctl(),
ax25_getname(), ax25_sendmsg(), ax25_getsockopt() and
ax25_info_show().

One of the concurrency UAF can be shown as below:

  (USE)                       |    (FREE)
                              |  ax25_device_event
                              |    ax25_dev_device_down
ax25_bind                     |    ...
  ...                         |      kfree(ax25_dev)
  ax25_fillin_cb()            |    ...
    ax25_fillin_cb_from_dev() |
  ...                         |

The root cause of UAF bugs is that kfree(ax25_dev) in
ax25_dev_device_down() is not protected by any locks.
When ax25_dev, which there are still pointers point to,
is released, the concurrency UAF bug will happen.

This patch introduces refcount into ax25_dev in order to
guarantee that there are no pointers point to it when ax25_dev
is released.

Signed-off-by: Duoming Zhou 
Signed-off-by: David S. Miller 
[OP: backport to 5.4: adjusted context]
Signed-off-by: Ovidiu Panait 
Signed-off-by: Greg Kroah-Hartman

net/sched: flower: fix parsing of ethertype following VLAN header

2022-04-20T07:19:34+00:00

[ Upstream commit 2105f700b53c24aa48b65c15652acc386044d26a ]

A tc flower filter matching TCA_FLOWER_KEY_VLAN_ETH_TYPE is expected to
match the L2 ethertype following the first VLAN header, as confirmed by
linked discussion with the maintainer. However, such rule also matches
packets that have additional second VLAN header, even though filter has
both eth_type and vlan_ethtype set to "ipv4". Looking at the code this
seems to be mostly an artifact of the way flower uses flow dissector.
First, even though looking at the uAPI eth_type and vlan_ethtype appear
like a distinct fields, in flower they are all mapped to the same
key->basic.n_proto. Second, flow dissector skips following VLAN header as
no keys for FLOW_DISSECTOR_KEY_CVLAN are set and eventually assigns the
value of n_proto to last parsed header. With these, such filters ignore any
headers present between first VLAN header and first "non magic"
header (ipv4 in this case) that doesn't result
FLOW_DISSECT_RET_PROTO_AGAIN.

Fix the issue by extending flow dissector VLAN key structure with new
'vlan_eth_type' field that matches first ethertype following previously
parsed VLAN header. Modify flower classifier to set the new
flow_dissector_key_vlan->vlan_eth_type with value obtained from
TCA_FLOWER_KEY_VLAN_ETH_TYPE/TCA_FLOWER_KEY_CVLAN_ETH_TYPE uAPIs.

Link: https://lore.kernel.org/all/Yjhgi48BpTGh6dig@nanopsycho/
Fixes: 9399ae9a6cb2 ("net_sched: flower: Add vlan support")
Fixes: d64efd0926ba ("net/sched: flower: Add supprt for matching on QinQ vlan headers")
Signed-off-by: Vlad Buslov 
Reviewed-by: Jiri Pirko 
Signed-off-by: David S. Miller 
Signed-off-by: Sasha Levin

ipv4: Invalidate neighbour for broadcast address upon address addition

2022-04-15T12:18:33+00:00

[ Upstream commit 0c51e12e218f20b7d976158fdc18019627326f7a ]

In case user space sends a packet destined to a broadcast address when a
matching broadcast route is not configured, the kernel will create a
unicast neighbour entry that will never be resolved [1].

When the broadcast route is configured, the unicast neighbour entry will
not be invalidated and continue to linger, resulting in packets being
dropped.

Solve this by invalidating unresolved neighbour entries for broadcast
addresses after routes for these addresses are internally configured by
the kernel. This allows the kernel to create a broadcast neighbour entry
following the next route lookup.

Another possible solution that is more generic but also more complex is
to have the ARP code register a listener to the FIB notification chain
and invalidate matching neighbour entries upon the addition of broadcast
routes.

It is also possible to wave off the issue as a user space problem, but
it seems a bit excessive to expect user space to be that intimately
familiar with the inner workings of the FIB/neighbour kernel code.

[1] https://lore.kernel.org/netdev/55a04a8f-56f3-f73c-2aea-2195923f09d1@huawei.com/

Reported-by: Wang Hai 
Signed-off-by: Ido Schimmel 
Tested-by: Wang Hai 
Signed-off-by: David S. Miller 
Signed-off-by: Sasha Levin

udp: call udp_encap_enable for v6 sockets when enabling encap

2022-04-15T12:18:01+00:00

commit a4a600dd301ccde6ea239804ec1f19364a39d643 upstream.

When enabling encap for a ipv6 socket without udp_encap_needed_key
increased, UDP GRO won't work for v4 mapped v6 address packets as
sk will be NULL in udp4_gro_receive().

This patch is to enable it by increasing udp_encap_needed_key for
v6 sockets in udp_tunnel_encap_enable(), and correspondingly
decrease udp_encap_needed_key in udpv6_destroy_sock().

v1->v2:
  - add udp_encap_disable() and export it.
v2->v3:
  - add the change for rxrpc and bareudp into one patch, as Alex
    suggested.
v3->v4:
  - move rxrpc part to another patch.

Acked-by: Willem de Bruijn 
Signed-off-by: Xin Long 
Signed-off-by: Jakub Kicinski 
Tested-by: Antonio Quartulli 
Signed-off-by: Greg Kroah-Hartman

esp: Fix possible buffer overflow in ESP transformation

2022-03-28T06:46:44+00:00

commit ebe48d368e97d007bfeb76fcb065d6cfc4c96645 upstream.

The maximum message size that can be send is bigger than
the  maximum site that skb_page_frag_refill can allocate.
So it is possible to write beyond the allocated buffer.

Fix this by doing a fallback to COW in that case.

v2:

Avoid get get_order() costs as suggested by Linus Torvalds.

Fixes: cac2661c53f3 ("esp4: Avoid skb_cow_data whenever possible")
Fixes: 03e2a30f6a27 ("esp6: Avoid skb_cow_data whenever possible")
Reported-by: valis 
Signed-off-by: Steffen Klassert 
Signed-off-by: Vaibhav Rustagi 
Signed-off-by: Greg Kroah-Hartman

xfrm: Check if_id in xfrm_migrate

2022-03-19T12:40:16+00:00

[ Upstream commit c1aca3080e382886e2e58e809787441984a2f89b ]

This patch enables distinguishing SAs and SPs based on if_id during
the xfrm_migrate flow. This ensures support for xfrm interfaces
throughout the SA/SP lifecycle.

When there are multiple existing SPs with the same direction,
the same xfrm_selector and different endpoint addresses,
xfrm_migrate might fail with ENODATA.

Specifically, the code path for performing xfrm_migrate is:
  Stage 1: find policy to migrate with
    xfrm_migrate_policy_find(sel, dir, type, net)
  Stage 2: find and update state(s) with
    xfrm_migrate_state_find(mp, net)
  Stage 3: update endpoint address(es) of template(s) with
    xfrm_policy_migrate(pol, m, num_migrate)

Currently "Stage 1" always returns the first xfrm_policy that
matches, and "Stage 3" looks for the xfrm_tmpl that matches the
old endpoint address. Thus if there are multiple xfrm_policy
with same selector, direction, type and net, "Stage 1" might
rertun a wrong xfrm_policy and "Stage 3" will fail with ENODATA
because it cannot find a xfrm_tmpl with the matching endpoint
address.

The fix is to allow userspace to pass an if_id and add if_id
to the matching rule in Stage 1 and Stage 2 since if_id is a
unique ID for xfrm_policy and xfrm_state. For compatibility,
if_id will only be checked if the attribute is set.

Tested with additions to Android's kernel unit test suite:
https://android-review.googlesource.com/c/kernel/tests/+/1668886

Signed-off-by: Yan Yan 
Signed-off-by: Steffen Klassert 
Signed-off-by: Sasha Levin