[ Upstream commit c5d2177a72 ]
A socket in a sockmap may have different combinations of programs attached
depending on configuration. There can be no programs in which case the socket
acts as a sink only. There can be a TX program in this case a BPF program is
attached to sending side, but no RX program is attached. There can be an RX
program only where sends have no BPF program attached, but receives are hooked
with BPF. And finally, both TX and RX programs may be attached. Giving us the
permutations:
None, Tx, Rx, and TxRx
To date most of our use cases have been TX case being used as a fast datapath
to directly copy between local application and a userspace proxy. Or Rx cases
and TxRX applications that are operating an in kernel based proxy. The traffic
in the first case where we hook applications into a userspace application looks
like this:
AppA redirect AppB
Tx <-----------> Rx
| |
+ +
TCP <--> lo <--> TCP
In this case all traffic from AppA (after 3whs) is copied into the AppB
ingress queue and no traffic is ever on the TCP recieive_queue.
In the second case the application never receives, except in some rare error
cases, traffic on the actual user space socket. Instead the send happens in
the kernel.
AppProxy socket pool
sk0 ------------->{sk1,sk2, skn}
^ |
| |
| v
ingress lb egress
TCP TCP
Here because traffic is never read off the socket with userspace recv() APIs
there is only ever one reader on the sk receive_queue. Namely the BPF programs.
However, we've started to introduce a third configuration where the BPF program
on receive should process the data, but then the normal case is to push the
data into the receive queue of AppB.
AppB
recv() (userspace)
-----------------------
tcp_bpf_recvmsg() (kernel)
| |
| |
| |
ingress_msgQ |
| |
RX_BPF |
| |
v v
sk->receive_queue
This is different from the App{A,B} redirect because traffic is first received
on the sk->receive_queue.
Now for the issue. The tcp_bpf_recvmsg() handler first checks the ingress_msg
queue for any data handled by the BPF rx program and returned with PASS code
so that it was enqueued on the ingress msg queue. Then if no data exists on
that queue it checks the socket receive queue. Unfortunately, this is the same
receive_queue the BPF program is reading data off of. So we get a race. Its
possible for the recvmsg() hook to pull data off the receive_queue before the
BPF hook has a chance to read it. It typically happens when an application is
banging on recv() and getting EAGAINs. Until they manage to race with the RX
BPF program.
To fix this we note that before this patch at attach time when the socket is
loaded into the map we check if it needs a TX program or just the base set of
proto bpf hooks. Then it uses the above general RX hook regardless of if we
have a BPF program attached at rx or not. This patch now extends this check to
handle all cases enumerated above, TX, RX, TXRX, and none. And to fix above
race when an RX program is attached we use a new hook that is nearly identical
to the old one except now we do not let the recv() call skip the RX BPF program.
Now only the BPF program pulls data from sk->receive_queue and recv() only
pulls data from the ingress msgQ post BPF program handling.
With this resolved our AppB from above has been up and running for many hours
without detecting any errors. We do this by correlating counters in RX BPF
events and the AppB to ensure data is never skipping the BPF program. Selftests,
was not able to detect this because we only run them for a short period of time
on well ordered send/recvs so we don't get any of the noise we see in real
application environments.
Fixes: 51199405f9 ("bpf: skb_verdict, support SK_PASS on RX BPF path")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Tested-by: Jussi Maki <joamaki@gmail.com>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20211103204736.248403-4-john.fastabend@gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b8b8315e39 ]
We do not need to handle unhash from BPF side we can simply wait for the
close to happen. The original concern was a socket could transition from
ESTABLISHED state to a new state while the BPF hook was still attached.
But, we convinced ourself this is no longer possible and we also improved
BPF sockmap to handle listen sockets so this is no longer a problem.
More importantly though there are cases where unhash is called when data is
in the receive queue. The BPF unhash logic will flush this data which is
wrong. To be correct it should keep the data in the receive queue and allow
a receiving application to continue reading the data. This may happen when
tcp_abort() is received for example. Instead of complicating the logic in
unhash simply moving all this to tcp_close() hook solves this.
Fixes: 51199405f9 ("bpf: skb_verdict, support SK_PASS on RX BPF path")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Tested-by: Jussi Maki <joamaki@gmail.com>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20211103204736.248403-3-john.fastabend@gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit cf12e6f912 ]
v1: Implement a more general statement as recommended by Eric Dumazet. The
sequence number will be advanced, so this check will fix the FIN case and
other cases.
A customer reported sockets stuck in the CLOSING state. A Vmcore revealed that
the write_queue was not empty as determined by tcp_write_queue_empty() but the
sk_buff containing the FIN flag had been freed and the socket was zombied in
that state. Corresponding pcaps show no FIN from the Linux kernel on the wire.
Some instrumentation was added to the kernel and it was found that there is a
timing window where tcp_sendmsg() can run after tcp_send_fin().
tcp_sendmsg() will hit an error, for example:
1269 ▹ if (sk->sk_err || (sk->sk_shutdown & SEND_SHUTDOWN))↩
1270 ▹ ▹ goto do_error;↩
tcp_remove_empty_skb() will then free the FIN sk_buff as "skb->len == 0". The
TCP socket is now wedged in the FIN-WAIT-1 state because the FIN is never sent.
If the other side sends a FIN packet the socket will transition to CLOSING and
remain that way until the system is rebooted.
Fix this by checking for the FIN flag in the sk_buff and don't free it if that
is the case. Testing confirmed that fixed the issue.
Fixes: fdfc5c8594 ("tcp: remove empty skb from write queue in error cases")
Signed-off-by: Jon Maxwell <jmaxwell37@gmail.com>
Reported-by: Monir Zouaoui <Monir.Zouaoui@mail.schwarz>
Reported-by: Simon Stier <simon.stier@mail.schwarz>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 19757cebf0 ]
Use of percpu_counter structure to track count of orphaned
sockets is causing problems on modern hosts with 256 cpus
or more.
Stefan Bach reported a serious spinlock contention in real workloads,
that I was able to reproduce with a netfilter rule dropping
incoming FIN packets.
53.56% server [kernel.kallsyms] [k] queued_spin_lock_slowpath
|
---queued_spin_lock_slowpath
|
--53.51%--_raw_spin_lock_irqsave
|
--53.51%--__percpu_counter_sum
tcp_check_oom
|
|--39.03%--__tcp_close
| tcp_close
| inet_release
| inet6_release
| sock_close
| __fput
| ____fput
| task_work_run
| exit_to_usermode_loop
| do_syscall_64
| entry_SYSCALL_64_after_hwframe
| __GI___libc_close
|
--14.48%--tcp_out_of_resources
tcp_write_timeout
tcp_retransmit_timer
tcp_write_timer_handler
tcp_write_timer
call_timer_fn
expire_timers
__run_timers
run_timer_softirq
__softirqentry_text_start
As explained in commit cf86a086a1 ("net/dst: use a smaller percpu_counter
batch for dst entries accounting"), default batch size is too big
for the default value of tcp_max_orphans (262144).
But even if we reduce batch sizes, there would still be cases
where the estimated count of orphans is beyond the limit,
and where tcp_too_many_orphans() has to call the expensive
percpu_counter_sum_positive().
One solution is to use plain per-cpu counters, and have
a timer to periodically refresh this cache.
Updating this cache every 100ms seems about right, tcp pressure
state is not radically changing over shorter periods.
percpu_counter was nice 15 years ago while hosts had less
than 16 cpus, not anymore by current standards.
v2: Fix the build issue for CONFIG_CRYPTO_DEV_CHELSIO_TLS=m,
reported by kernel test robot <lkp@intel.com>
Remove unused socket argument from tcp_too_many_orphans()
Fixes: dd24c00191 ("net: Use a percpu_counter for orphan_count")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Stefan Bach <sfb@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit cd9733f5d7 upstream.
With two Msgs, msgA and msgB and a user doing nonblocking sendmsg calls (or
multiple cores) on a single socket 'sk' we could get the following flow.
msgA, sk msgB, sk
----------- ---------------
tcp_bpf_sendmsg()
lock(sk)
psock = sk->psock
tcp_bpf_sendmsg()
lock(sk) ... blocking
tcp_bpf_send_verdict
if (psock->eval == NONE)
psock->eval = sk_psock_msg_verdict
..
< handle SK_REDIRECT case >
release_sock(sk) < lock dropped so grab here >
ret = tcp_bpf_sendmsg_redir
psock = sk->psock
tcp_bpf_send_verdict
lock_sock(sk) ... blocking on B
if (psock->eval == NONE) <- boom.
psock->eval will have msgA state
The problem here is we dropped the lock on msgA and grabbed it with msgB.
Now we have old state in psock and importantly psock->eval has not been
cleared. So msgB will run whatever action was done on A and the verdict
program may never see it.
Fixes: 604326b41a ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: Liu Jian <liujian56@huawei.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20211012052019.184398-1-liujian56@huawei.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 86f1e3a848 ]
With net.ipv4.tcp_l3mdev_accept=1 it is possible for a listen socket to
accept connection from the same client address in different VRFs. It is
also possible to set different MD5 keys for these clients which differ
only in the tcpm_l3index field.
This appears to work when distinguishing between different VRFs but not
between non-VRF and VRF connections. In particular:
* tcp_md5_do_lookup_exact will match a non-vrf key against a vrf key.
This means that adding a key with l3index != 0 after a key with l3index
== 0 will cause the earlier key to be deleted. Both keys can be present
if the non-vrf key is added later.
* _tcp_md5_do_lookup can match a non-vrf key before a vrf key. This
casues failures if the passwords differ.
Fix this by making tcp_md5_do_lookup_exact perform an actual exact
comparison on l3index and by making __tcp_md5_do_lookup perfer
vrf-bound keys above other considerations like prefixlen.
Fixes: dea53bb80e ("tcp: Add l3index to tcp_md5sig_key and md5 functions")
Signed-off-by: Leonard Crestez <cdleonard@gmail.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8d6c414cd2 ]
The commit 6da5b0f027 ("net: ensure unbound datagram socket to be
chosen when not in a VRF") modified compute_score() so that a device
match is always made, not just in the case of an l3mdev skb, then
increments the score also for unbound sockets. This ensures that
sockets bound to an l3mdev are never selected when not in a VRF.
But as unbound and bound sockets are now scored equally, this results
in the last opened socket being selected if there are matches in the
default VRF for an unbound socket and a socket bound to a dev that is
not an l3mdev. However, handling prior to this commit was to always
select the bound socket in this case. Reinstate this handling by
incrementing the score only for bound sockets. The required isolation
due to choosing between an unbound socket and a socket bound to an
l3mdev remains in place due to the device match always being made.
The same approach is taken for compute_score() for stream sockets.
Fixes: 6da5b0f027 ("net: ensure unbound datagram socket to be chosen when not in a VRF")
Fixes: e78190581a ("net: ensure unbound stream socket to be chosen when not in a VRF")
Signed-off-by: Mike Manning <mmanning@vyatta.att-mail.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/cf0a8523-b362-1edf-ee78-eef63cbbb428@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 339031bafe ]
This is a revert of
7b1957b049 ("netfilter: nf_defrag_ipv4: use net_generic infra")
and a partial revert of
8b0adbe3e3 ("netfilter: nf_defrag_ipv6: use net_generic infra").
If conntrack is builtin and kernel is booted with:
nf_conntrack.enable_hooks=1
.... kernel will fail to boot due to a NULL deref in
nf_defrag_ipv4_enable(): Its called before the ipv4 defrag initcall is
made, so net_generic() returns NULL.
To resolve this, move the user refcount back to struct net so calls
to those functions are possible even before their initcalls have run.
Fixes: 7b1957b049 ("netfilter: nf_defrag_ipv4: use net_generic infra")
Fixes: 8b0adbe3e3 ("netfilter: nf_defrag_ipv6: use net_generic infra").
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit a9f5970767 upstream.
up->corkflag field can be read or written without any lock.
Annotate accesses to avoid possible syzbot/KCSAN reports.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 597aa16c78 ]
Multipath RTA_FLOW is embedded in nexthop. Dump it in fib_add_nexthop()
to get the length of rtnexthop correct.
Fixes: b0f6019363 ("ipv4: Refactor nexthop attributes in fib_dump_info")
Signed-off-by: Xiao Liang <shaw.leon@gmail.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3106a08475 ]
syzkaller discovered memory leaks [1] that can be reduced to the
following commands:
# ip nexthop add id 1 blackhole
# devlink dev reload pci/0000:06:00.0
As part of the reload flow, mlxsw will unregister its netdevs and then
unregister from the nexthop notification chain. Before unregistering
from the notification chain, mlxsw will receive delete notifications for
nexthop objects using netdevs registered by mlxsw or their uppers. mlxsw
will not receive notifications for nexthops using netdevs that are not
dismantled as part of the reload flow. For example, the blackhole
nexthop above that internally uses the loopback netdev as its nexthop
device.
One way to fix this problem is to have listeners flush their nexthop
tables after unregistering from the notification chain. This is
error-prone as evident by this patch and also not symmetric with the
registration path where a listener receives a dump of all the existing
nexthops.
Therefore, fix this problem by replaying delete notifications for the
listener being unregistered. This is symmetric to the registration path
and also consistent with the netdev notification chain.
The above means that unregister_nexthop_notifier(), like
register_nexthop_notifier(), will have to take RTNL in order to iterate
over the existing nexthops and that any callers of the function cannot
hold RTNL. This is true for mlxsw and netdevsim, but not for the VXLAN
driver. To avoid a deadlock, change the latter to unregister its nexthop
listener without holding RTNL, making it symmetric to the registration
path.
[1]
unreferenced object 0xffff88806173d600 (size 512):
comm "syz-executor.0", pid 1290, jiffies 4295583142 (age 143.507s)
hex dump (first 32 bytes):
41 9d 1e 60 80 88 ff ff 08 d6 73 61 80 88 ff ff A..`......sa....
08 d6 73 61 80 88 ff ff 01 00 00 00 00 00 00 00 ..sa............
backtrace:
[<ffffffff81a6b576>] kmemleak_alloc_recursive include/linux/kmemleak.h:43 [inline]
[<ffffffff81a6b576>] slab_post_alloc_hook+0x96/0x490 mm/slab.h:522
[<ffffffff81a716d3>] slab_alloc_node mm/slub.c:3206 [inline]
[<ffffffff81a716d3>] slab_alloc mm/slub.c:3214 [inline]
[<ffffffff81a716d3>] kmem_cache_alloc_trace+0x163/0x370 mm/slub.c:3231
[<ffffffff82e8681a>] kmalloc include/linux/slab.h:591 [inline]
[<ffffffff82e8681a>] kzalloc include/linux/slab.h:721 [inline]
[<ffffffff82e8681a>] mlxsw_sp_nexthop_obj_group_create drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c:4918 [inline]
[<ffffffff82e8681a>] mlxsw_sp_nexthop_obj_new drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c:5054 [inline]
[<ffffffff82e8681a>] mlxsw_sp_nexthop_obj_event+0x59a/0x2910 drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c:5239
[<ffffffff813ef67d>] notifier_call_chain+0xbd/0x210 kernel/notifier.c:83
[<ffffffff813f0662>] blocking_notifier_call_chain kernel/notifier.c:318 [inline]
[<ffffffff813f0662>] blocking_notifier_call_chain+0x72/0xa0 kernel/notifier.c:306
[<ffffffff8384b9c6>] call_nexthop_notifiers+0x156/0x310 net/ipv4/nexthop.c:244
[<ffffffff83852bd8>] insert_nexthop net/ipv4/nexthop.c:2336 [inline]
[<ffffffff83852bd8>] nexthop_add net/ipv4/nexthop.c:2644 [inline]
[<ffffffff83852bd8>] rtm_new_nexthop+0x14e8/0x4d10 net/ipv4/nexthop.c:2913
[<ffffffff833e9a78>] rtnetlink_rcv_msg+0x448/0xbf0 net/core/rtnetlink.c:5572
[<ffffffff83608703>] netlink_rcv_skb+0x173/0x480 net/netlink/af_netlink.c:2504
[<ffffffff833de032>] rtnetlink_rcv+0x22/0x30 net/core/rtnetlink.c:5590
[<ffffffff836069de>] netlink_unicast_kernel net/netlink/af_netlink.c:1314 [inline]
[<ffffffff836069de>] netlink_unicast+0x5ae/0x7f0 net/netlink/af_netlink.c:1340
[<ffffffff83607501>] netlink_sendmsg+0x8e1/0xe30 net/netlink/af_netlink.c:1929
[<ffffffff832fde84>] sock_sendmsg_nosec net/socket.c:704 [inline]
[<ffffffff832fde84>] sock_sendmsg net/socket.c:724 [inline]
[<ffffffff832fde84>] ____sys_sendmsg+0x874/0x9f0 net/socket.c:2409
[<ffffffff83304a44>] ___sys_sendmsg+0x104/0x170 net/socket.c:2463
[<ffffffff83304c01>] __sys_sendmsg+0x111/0x1f0 net/socket.c:2492
[<ffffffff83304d5d>] __do_sys_sendmsg net/socket.c:2501 [inline]
[<ffffffff83304d5d>] __se_sys_sendmsg net/socket.c:2499 [inline]
[<ffffffff83304d5d>] __x64_sys_sendmsg+0x7d/0xc0 net/socket.c:2499
Fixes: 2a014b200b ("mlxsw: spectrum_router: Add support for nexthop objects")
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 563f23b002 upstream.
The resilient nexthop group torture tests in fib_nexthop.sh exposed a
possible division by zero while replacing a resilient group [1]. The
division by zero occurs when the data path sees a resilient nexthop
group with zero buckets.
The tests replace a resilient nexthop group in a loop while traffic is
forwarded through it. The tests do not specify the number of buckets
while performing the replacement, resulting in the kernel allocating a
stub resilient table (i.e, 'struct nh_res_table') with zero buckets.
This table should never be visible to the data path, but the old nexthop
group (i.e., 'oldg') might still be used by the data path when the stub
table is assigned to it.
Fix this by only assigning the stub table to the old nexthop group after
making sure the group is no longer used by the data path.
Tested with fib_nexthops.sh:
Tests passed: 222
Tests failed: 0
[1]
divide error: 0000 [#1] PREEMPT SMP KASAN
CPU: 0 PID: 1850 Comm: ping Not tainted 5.14.0-custom-10271-ga86eb53057fe #1107
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-4.fc34 04/01/2014
RIP: 0010:nexthop_select_path+0x2d2/0x1a80
[...]
Call Trace:
fib_select_multipath+0x79b/0x1530
fib_select_path+0x8fb/0x1c10
ip_route_output_key_hash_rcu+0x1198/0x2da0
ip_route_output_key_hash+0x190/0x340
ip_route_output_flow+0x21/0x120
raw_sendmsg+0x91d/0x2e10
inet_sendmsg+0x9e/0xe0
__sys_sendto+0x23d/0x360
__x64_sys_sendto+0xe1/0x1b0
do_syscall_64+0x35/0x80
entry_SYSCALL_64_after_hwframe+0x44/0xae
Cc: stable@vger.kernel.org
Fixes: 283a72a559 ("nexthop: Add implementation of resilient next-hop groups")
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 8a0ed250f9 ]
The GRE tunnel device can pull existing outer headers in ipge_xmit.
This is a rare path, apparently unique to this device. The below
commit ensured that pulling does not move skb->data beyond csum_start.
But it has a false positive if ip_summed is not CHECKSUM_PARTIAL and
thus csum_start is irrelevant.
Refine to exclude this. At the same time simplify and strengthen the
test.
Simplify, by moving the check next to the offending pull, making it
more self documenting and removing an unnecessary branch from other
code paths.
Strengthen, by also ensuring that the transport header is correct and
therefore the inner headers will be after skb_reset_inner_headers.
The transport header is set to csum_start in skb_partial_csum_set.
Link: https://lore.kernel.org/netdev/YS+h%2FtqCJJiQei+W@shredder/
Fixes: 1d011c4803 ("ip_gre: add validation for csum_start")
Reported-by: Ido Schimmel <idosch@idosch.org>
Suggested-by: Alexander Duyck <alexander.duyck@gmail.com>
Signed-off-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit e50e711351 upstream.
Turn udp_tunnel_nic work-queue to an ordered work-queue. This queue
holds the UDP-tunnel configuration commands of the different netdevs.
When the netdevs are functions of the same NIC the order of
execution may be crucial.
Problem example:
NIC with 2 PFs, both PFs declare offload quota of up to 3 UDP-ports.
$ifconfig eth2 1.1.1.1/16 up
$ip link add eth2_19503 type vxlan id 5049 remote 1.1.1.2 dev eth2 dstport 19053
$ip link set dev eth2_19503 up
$ip link add eth2_19504 type vxlan id 5049 remote 1.1.1.3 dev eth2 dstport 19054
$ip link set dev eth2_19504 up
$ip link add eth2_19505 type vxlan id 5049 remote 1.1.1.4 dev eth2 dstport 19055
$ip link set dev eth2_19505 up
$ip link add eth2_19506 type vxlan id 5049 remote 1.1.1.5 dev eth2 dstport 19056
$ip link set dev eth2_19506 up
NIC RX port offload infrastructure offloads the first 3 UDP-ports (on
all devices which sets NETIF_F_RX_UDP_TUNNEL_PORT feature) and not
UDP-port 19056. So both PFs gets this offload configuration.
$ip link set dev eth2_19504 down
This triggers udp-tunnel-core to remove the UDP-port 19504 from
offload-ports-list and offload UDP-port 19056 instead.
In this scenario it is important that the UDP-port of 19504 will be
removed from both PFs before trying to add UDP-port 19056. The NIC can
stop offloading a UDP-port only when all references are removed.
Otherwise the NIC may report exceeding of the offload quota.
Fixes: cc4e3835ef ("udp_tunnel: add central NIC RX port offload infrastructure")
Signed-off-by: Aya Levin <ayal@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 4f884f3962 upstream.
Commit 10d3be5692 ("tcp-tso: do not split TSO packets at retransmit
time") may directly retrans a multiple segments TSO/GSO packet without
split, Since this commit, we can no longer assume that a retransmitted
packet is a single segment.
This patch fixes the tp->undo_retrans accounting in tcp_sacktag_one()
that use the actual segments(pcount) of the retransmitted packet.
Before that commit (10d3be5692), the assumption underlying the
tp->undo_retrans-- seems correct.
Fixes: 10d3be5692 ("tcp-tso: do not split TSO packets at retransmit time")
Signed-off-by: zhenggy <zhenggy@chinatelecom.cn>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 9756e44fd4 upstream.
The commit 733c99ee8b ("net: fix NULL pointer reference in
cipso_v4_doi_free") was merged by a mistake, this patch try
to cleanup the mess.
And we already have the commit e842cb60e8 ("net: fix NULL
pointer reference in cipso_v4_doi_free") which fixed the root
cause of the issue mentioned in it's description.
Suggested-by: Paul Moore <paul@paul-moore.com>
Signed-off-by: Michael Wang <yun.wang@linux.alibaba.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit e3faa49bce ]
Since the original TFO server code was implemented in commit
168a8f5805 ("tcp: TCP Fast Open Server -
main code path") the TFO server code has supported the sysctl bit flag
TFO_SERVER_COOKIE_NOT_REQD. Currently, when the TFO_SERVER_ENABLE and
TFO_SERVER_COOKIE_NOT_REQD sysctl bit flags are set, a server connection
will accept a SYN with N bytes of data (N > 0) that has no TFO cookie,
create a new fast open connection, process the incoming data in the SYN,
and make the connection ready for accepting. After accepting, the
connection is ready for read()/recvmsg() to read the N bytes of data in
the SYN, ready for write()/sendmsg() calls and data transmissions to
transmit data.
This commit changes an edge case in this feature by changing this
behavior to apply to (N >= 0) bytes of data in the SYN rather than only
(N > 0) bytes of data in the SYN. Now, a server will accept a data-less
SYN without a TFO cookie if TFO_SERVER_COOKIE_NOT_REQD is set.
Caveat! While this enables a new kind of TFO (data-less empty-cookie
SYN), some firewall rules setup may not work if they assume such packets
are not legit TFOs and will filter them.
Signed-off-by: Luke Hsiao <lukehsiao@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20210816205105.2533289-1-luke.w.hsiao@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 6321c7acb8 ]
Fix the following out-of-bounds warning:
In function 'ip_copy_addrs',
inlined from '__ip_queue_xmit' at net/ipv4/ip_output.c:517:2:
net/ipv4/ip_output.c:449:2: warning: 'memcpy' offset [40, 43] from the object at 'fl' is out of the bounds of referenced subobject 'saddr' with type 'unsigned int' at offset 36 [-Warray-bounds]
449 | memcpy(&iph->saddr, &fl4->saddr,
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
450 | sizeof(fl4->saddr) + sizeof(fl4->daddr));
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The problem is that the original code is trying to copy data into a
couple of struct members adjacent to each other in a single call to
memcpy(). This causes a legitimate compiler warning because memcpy()
overruns the length of &iph->saddr and &fl4->saddr. As these are just
a couple of struct members, fix this by using direct assignments,
instead of memcpy().
This helps with the ongoing efforts to globally enable -Warray-bounds
and get us closer to being able to tighten the FORTIFY_SOURCE routines
on memcpy().
Link: https://github.com/KSPP/linux/issues/109
Reported-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/d5ae2e65-1f18-2577-246f-bada7eee6ccd@intel.com/
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 92548b0ee2 ]
The UDP length field should be in network order.
This removes the following sparse error:
net/ipv4/route.c:3173:27: warning: incorrect type in assignment (different base types)
net/ipv4/route.c:3173:27: expected restricted __be16 [usertype] len
net/ipv4/route.c:3173:27: got unsigned long
Fixes: 404eb77ea7 ("ipv4: support sport, dport and ip_proto in RTM_GETROUTE")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Roopa Prabhu <roopa@nvidia.com>
Cc: David Ahern <dsahern@kernel.org>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 67d6d681e1 ]
Even after commit 6457378fe7 ("ipv4: use siphash instead of Jenkins in
fnhe_hashfun()"), an attacker can still use brute force to learn
some secrets from a victim linux host.
One way to defeat these attacks is to make the max depth of the hash
table bucket a random value.
Before this patch, each bucket of the hash table used to store exceptions
could contain 6 items under attack.
After the patch, each bucket would contains a random number of items,
between 6 and 10. The attacker can no longer infer secrets.
This is slightly increasing memory size used by the hash table,
by 50% in average, we do not expect this to be a problem.
This patch is more complex than the prior one (IPv6 equivalent),
because IPv4 was reusing the oldest entry.
Since we need to be able to evict more than one entry per
update_or_create_fnhe() call, I had to replace
fnhe_oldest() with fnhe_remove_oldest().
Also note that we will queue extra kfree_rcu() calls under stress,
which hopefully wont be a too big issue.
Fixes: 4895c771c7 ("ipv4: Add FIB nexthop exceptions.")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Keyu Man <kman001@ucr.edu>
Cc: Willy Tarreau <w@1wt.eu>
Signed-off-by: David S. Miller <davem@davemloft.net>
Reviewed-by: David Ahern <dsahern@kernel.org>
Tested-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 525e2f9fd0 ]
st->bucket stores the current bucket number.
st->offset stores the offset within this bucket that is the sk to be
seq_show(). Thus, st->offset only makes sense within the same
st->bucket.
These two variables are an optimization for the common no-lseek case.
When resuming the seq_file iteration (i.e. seq_start()),
tcp_seek_last_pos() tries to continue from the st->offset
at bucket st->bucket.
However, it is possible that the bucket pointed by st->bucket
has changed and st->offset may end up skipping the whole st->bucket
without finding a sk. In this case, tcp_seek_last_pos() currently
continues to satisfy the offset condition in the next (and incorrect)
bucket. Instead, regardless of the offset value, the first sk of the
next bucket should be returned. Thus, "bucket == st->bucket" check is
added to tcp_seek_last_pos().
The chance of hitting this is small and the issue is a decade old,
so targeting for the next tree.
Fixes: a8b690f98b ("tcp: Fix slowness in read /proc/net/tcp")
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Acked-by: Kuniyuki Iwashima <kuniyu@amazon.co.jp>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20210701200541.1033917-1-kafai@fb.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
In netlbl_cipsov4_add_std() when 'doi_def->map.std' alloc
failed, we sometime observe panic:
BUG: kernel NULL pointer dereference, address:
...
RIP: 0010:cipso_v4_doi_free+0x3a/0x80
...
Call Trace:
netlbl_cipsov4_add_std+0xf4/0x8c0
netlbl_cipsov4_add+0x13f/0x1b0
genl_family_rcv_msg_doit.isra.15+0x132/0x170
genl_rcv_msg+0x125/0x240
This is because in cipso_v4_doi_free() there is no check
on 'doi_def->map.std' when 'doi_def->type' equal 1, which
is possibe, since netlbl_cipsov4_add_std() haven't initialize
it before alloc 'doi_def->map.std'.
This patch just add the check to prevent panic happen for similar
cases.
Reported-by: Abaci <abaci@linux.alibaba.com>
Signed-off-by: Michael Wang <yun.wang@linux.alibaba.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A group of security researchers brought to our attention
the weakness of hash function used in fnhe_hashfun().
Lets use siphash instead of Jenkins Hash, to considerably
reduce security risks.
Also remove the inline keyword, this really is distracting.
Fixes: d546c62154 ("ipv4: harden fnhe_hashfun()")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Keyu Man <kman001@ucr.edu>
Cc: Willy Tarreau <w@1wt.eu>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently if BBR congestion control is initialized after more than 2B
packets have been delivered, depending on the phase of the
tp->delivered counter the tracking of BBR round trips can get stuck.
The bug arises because if tp->delivered is between 2^31 and 2^32 at
the time the BBR congestion control module is initialized, then the
initialization of bbr->next_rtt_delivered to 0 will cause the logic to
believe that the end of the round trip is still billions of packets in
the future. More specifically, the following check will fail
repeatedly:
!before(rs->prior_delivered, bbr->next_rtt_delivered)
and thus the connection will take up to 2B packets delivered before
that check will pass and the connection will set:
bbr->round_start = 1;
This could cause many mechanisms in BBR to fail to trigger, for
example bbr_check_full_bw_reached() would likely never exit STARTUP.
This bug is 5 years old and has not been observed, and as a practical
matter this would likely rarely trigger, since it would require
transferring at least 2B packets, or likely more than 3 terabytes of
data, before switching congestion control algorithms to BBR.
This patch is a stable candidate for kernels as far back as v4.9,
when tcp_bbr.c was added.
Fixes: 0f8782ea14 ("tcp_bbr: add BBR congestion control")
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Kevin Yang <yyd@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20210811024056.235161-1-ncardwell@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
GSO expects inner transport header offset to be valid when
skb->encapsulation flag is set. GSO uses this value to calculate the length
of an individual segment of a GSO packet in skb_gso_transport_seglen().
However, tcp/udp gro_complete callbacks don't update the
skb->inner_transport_header when processing an encapsulated TCP/UDP
segment. As a result a GRO skb has ->inner_transport_header set to a value
carried over from earlier skb processing.
This can have mild to tragic consequences. From miscalculating the GSO
segment length to triggering a page fault [1], when trying to read TCP/UDP
header at an address past the skb->data page.
The latter scenario leads to an oops report like so:
BUG: unable to handle page fault for address: ffff9fa7ec00d008
#PF: supervisor read access in kernel mode
#PF: error_code(0x0000) - not-present page
PGD 123f201067 P4D 123f201067 PUD 123f209067 PMD 0
Oops: 0000 [#1] SMP NOPTI
CPU: 44 PID: 0 Comm: swapper/44 Not tainted 5.4.53-cloudflare-2020.7.21 #1
Hardware name: HYVE EDGE-METAL-GEN10/HS-1811DLite1, BIOS V2.15 02/21/2020
RIP: 0010:skb_gso_transport_seglen+0x44/0xa0
Code: c0 41 83 e0 11 f6 87 81 00 00 00 20 74 30 0f b7 87 aa 00 00 00 0f [...]
RSP: 0018:ffffad8640bacbb8 EFLAGS: 00010202
RAX: 000000000000feda RBX: ffff9fcc8d31bc00 RCX: ffff9fa7ec00cffc
RDX: ffff9fa7ebffdec0 RSI: 000000000000feda RDI: 0000000000000122
RBP: 00000000000005c4 R08: 0000000000000001 R09: 0000000000000000
R10: ffff9fe588ae3800 R11: ffff9fe011fc92f0 R12: ffff9fcc8d31bc00
R13: ffff9fe0119d4300 R14: 00000000000005c4 R15: ffff9fba57d70900
FS: 0000000000000000(0000) GS:ffff9fe68df00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: ffff9fa7ec00d008 CR3: 0000003e99b1c000 CR4: 0000000000340ee0
Call Trace:
<IRQ>
skb_gso_validate_network_len+0x11/0x70
__ip_finish_output+0x109/0x1c0
ip_sublist_rcv_finish+0x57/0x70
ip_sublist_rcv+0x2aa/0x2d0
? ip_rcv_finish_core.constprop.0+0x390/0x390
ip_list_rcv+0x12b/0x14f
__netif_receive_skb_list_core+0x2a9/0x2d0
netif_receive_skb_list_internal+0x1b5/0x2e0
napi_complete_done+0x93/0x140
veth_poll+0xc0/0x19f [veth]
? mlx5e_napi_poll+0x221/0x610 [mlx5_core]
net_rx_action+0x1f8/0x790
__do_softirq+0xe1/0x2bf
irq_exit+0x8e/0xc0
do_IRQ+0x58/0xe0
common_interrupt+0xf/0xf
</IRQ>
The bug can be observed in a simple setup where we send IP/GRE/IP/TCP
packets into a netns over a veth pair. Inside the netns, packets are
forwarded to dummy device:
trafgen -> [veth A]--[veth B] -forward-> [dummy]
For veth B to GRO aggregate packets on receive, it needs to have an XDP
program attached (for example, a trivial XDP_PASS). Additionally, for UDP,
we need to enable GSO_UDP_L4 feature on the device:
ip netns exec A ethtool -K AB rx-udp-gro-forwarding on
The last component is an artificial delay to increase the chances of GRO
batching happening:
ip netns exec A tc qdisc add dev AB root \
netem delay 200us slot 5ms 10ms packets 2 bytes 64k
With such a setup in place, the bug can be observed by tracing the skb
outer and inner offsets when GSO skb is transmitted from the dummy device:
tcp:
FUNC DEV SKB_LEN NH TH ENC INH ITH GSO_SIZE GSO_TYPE
ip_finish_output dumB 2830 270 290 1 294 254 1383 (tcpv4,gre,)
^^^
udp:
FUNC DEV SKB_LEN NH TH ENC INH ITH GSO_SIZE GSO_TYPE
ip_finish_output dumB 2818 270 290 1 294 254 1383 (gre,udp_l4,)
^^^
Fix it by updating the inner transport header offset in tcp/udp
gro_complete callbacks, similar to how {inet,ipv6}_gro_complete callbacks
update the inner network header offset, when skb->encapsulation flag is
set.
[1] https://lore.kernel.org/netdev/CAKxSbF01cLpZem2GFaUaifh0S-5WYViZemTicAg7FCHOnh6kug@mail.gmail.com/
Fixes: bf296b125b ("tcp: Add GRO support")
Fixes: f993bc25e5 ("net: core: handle encapsulation offloads when computing segment lengths")
Fixes: e20cf8d3f1 ("udp: implement GRO for plain UDP sockets.")
Reported-by: Alex Forster <aforster@cloudflare.com>
Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In cases where the header straight after the tunnel header was
another ethernet header (TEB), instead of the network header,
the ECN decapsulation code would treat the ethernet header as if
it was an IP header, resulting in mishandling and possible
wrong drops or corruption of the IP header.
In this case, ECT(1) is sent, so IP_ECN_decapsulate tries to copy it to the
inner IPv4 header, and correct its checksum.
The offset of the ECT bits in an IPv4 header corresponds to the
lower 2 bits of the second octet of the destination MAC address
in the ethernet header.
The IPv4 checksum corresponds to end of the source address.
In order to reproduce:
$ ip netns add A
$ ip netns add B
$ ip -n A link add _v0 type veth peer name _v1 netns B
$ ip -n A link set _v0 up
$ ip -n A addr add dev _v0 10.254.3.1/24
$ ip -n A route add default dev _v0 scope global
$ ip -n B link set _v1 up
$ ip -n B addr add dev _v1 10.254.1.6/24
$ ip -n B route add default dev _v1 scope global
$ ip -n B link add gre1 type gretap local 10.254.1.6 remote 10.254.3.1 key 0x49000000
$ ip -n B link set gre1 up
# Now send an IPv4/GRE/Eth/IPv4 frame where the outer header has ECT(1),
# and the inner header has no ECT bits set:
$ cat send_pkt.py
#!/usr/bin/env python3
from scapy.all import *
pkt = IP(b'E\x01\x00\xa7\x00\x00\x00\x00@/`%\n\xfe\x03\x01\n\xfe\x01\x06 \x00eXI\x00'
b'\x00\x00\x18\xbe\x92\xa0\xee&\x18\xb0\x92\xa0l&\x08\x00E\x00\x00}\x8b\x85'
b'@\x00\x01\x01\xe4\xf2\x82\x82\x82\x01\x82\x82\x82\x02\x08\x00d\x11\xa6\xeb'
b'3\x1e\x1e\\xf3\\xf7`\x00\x00\x00\x00ZN\x00\x00\x00\x00\x00\x00\x10\x11\x12'
b'\x13\x14\x15\x16\x17\x18\x19\x1a\x1b\x1c\x1d\x1e\x1f !"#$%&\'()*+,-./01234'
b'56789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ')
send(pkt)
$ sudo ip netns exec B tcpdump -neqlllvi gre1 icmp & ; sleep 1
$ sudo ip netns exec A python3 send_pkt.py
In the original packet, the source/destinatio MAC addresses are
dst=18:be:92:a0:ee:26 src=18:b0:92:a0:6c:26
In the received packet, they are
dst=18:bd:92:a0:ee:26 src=18:b0:92:a0:6c:27
Thanks to Lahav Schlesinger <lschlesinger@drivenets.com> and Isaac Garzon <isaac@speed.io>
for helping me pinpoint the origin.
Fixes: b723748750 ("tunnel: Propagate ECT(1) when decapsulating as recommended by RFC6040")
Cc: David S. Miller <davem@davemloft.net>
Cc: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>
Cc: David Ahern <dsahern@kernel.org>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Gilad Naaman <gnaaman@drivenets.com>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Multiple complaints have been raised from the TFO users on the internet
stating that the TFO blackhole logic is too aggressive and gets falsely
triggered too often.
(e.g. https://blog.apnic.net/2021/07/05/tcp-fast-open-not-so-fast/)
Considering that most middleboxes no longer drop TFO packets, we decide
to disable the blackhole logic by setting
/proc/sys/net/ipv4/tcp_fastopen_blackhole_timeout_set to 0 by default.
Fixes: cf1ef3f071 ("net/tcp_fastopen: Disable active side TFO in certain scenarios")
Signed-off-by: Wei Wang <weiwan@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit d26796ae58 ("udp: check udp sock encap_type in __udp_lib_err")
added checks for encapsulated sockets but it broke cases when there is
no implementation of encap_err_lookup for encapsulation, i.e. ESP in
UDP encapsulation. Fix it by calling encap_err_lookup only if socket
implements this method otherwise treat it as legal socket.
Fixes: d26796ae58 ("udp: check udp sock encap_type in __udp_lib_err")
Signed-off-by: Vadim Fedorenko <vfedorenko@novek.ru>
Reviewed-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
tfo_active_disable_stamp is read and written locklessly.
We need to annotate these accesses appropriately.
Then, we need to perform the atomic_inc(tfo_active_disable_times)
after the timestamp has been updated, and thus add barriers
to make sure tcp_fastopen_active_should_disable() wont read
a stale timestamp.
Fixes: cf1ef3f071 ("net/tcp_fastopen: Disable active side TFO in certain scenarios")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Wei Wang <weiwan@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Acked-by: Wei Wang <weiwan@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Andrii Nakryiko says:
====================
pull-request: bpf 2021-07-15
The following pull-request contains BPF updates for your *net* tree.
We've added 9 non-merge commits during the last 5 day(s) which contain
a total of 9 files changed, 37 insertions(+), 15 deletions(-).
The main changes are:
1) Fix NULL pointer dereference in BPF_TEST_RUN for BPF_XDP_DEVMAP and
BPF_XDP_CPUMAP programs, from Xuan Zhuo.
2) Fix use-after-free of net_device in XDP bpf_link, from Xuan Zhuo.
3) Follow-up fix to subprog poke descriptor use-after-free problem, from
Daniel Borkmann and John Fastabend.
4) Fix out-of-range array access in s390 BPF JIT backend, from Colin Ian King.
5) Fix memory leak in BPF sockmap, from John Fastabend.
6) Fix for sockmap to prevent proc stats reporting bug, from John Fastabend
and Jakub Sitnicki.
7) Fix NULL pointer dereference in bpftool, from Tobias Klauser.
8) AF_XDP documentation fixes, from Baruch Siach.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The proc socket stats use sk_prot->inuse_idx value to record inuse sock
stats. We currently do not set this correctly from sockmap side. The
result is reading sock stats '/proc/net/sockstat' gives incorrect values.
The socket counter is incremented correctly, but because we don't set the
counter correctly when we replace sk_prot we may omit the decrement.
To get the correct inuse_idx value move the core_initcall that initializes
the UDP proto handlers to late_initcall. This way it is initialized after
UDP has the chance to assign the inuse_idx value from the register protocol
handler.
Fixes: edc6741cc6 ("bpf: Add sockmap hooks for UDP sockets")
Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Cong Wang <cong.wang@bytedance.com>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20210714154750.528206-1-jakub@cloudflare.com
The proc socket stats use sk_prot->inuse_idx value to record inuse sock
stats. We currently do not set this correctly from sockmap side. The
result is reading sock stats '/proc/net/sockstat' gives incorrect values.
The socket counter is incremented correctly, but because we don't set the
counter correctly when we replace sk_prot we may omit the decrement.
To get the correct inuse_idx value move the core_initcall that initializes
the TCP proto handlers to late_initcall. This way it is initialized after
TCP has the chance to assign the inuse_idx value from the register protocol
handler.
Fixes: 604326b41a ("bpf, sockmap: convert to generic sk_msg interface")
Suggested-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Cong Wang <cong.wang@bytedance.com>
Link: https://lore.kernel.org/bpf/20210712195546.423990-3-john.fastabend@gmail.com
It has 'if (err >0 )' statement in nlmsg_unicast(), so use nlmsg_unicast()
instead of netlink_unicast(), this looks more concise.
v2: remove the change in netfilter.
Signed-off-by: Yajun Deng <yajun.deng@linux.dev>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
If check_fully_established() causes a subflow reset, it should not
continue to process the packet in tcp_data_queue().
Add a return value to mptcp_incoming_options(), and return false if a
subflow has been reset, else return true. Then drop the packet in
tcp_data_queue()/tcp_rcv_state_process() if mptcp_incoming_options()
return false.
Fixes: d582484726 ("mptcp: fix fallback for MP_JOIN subflows")
Signed-off-by: Jianguo Wu <wujianguo@chinatelecom.cn>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 28e104d002 ("net: ip_tunnel: fix mtu calculation") removed
dev->hard_header_len subtraction when calculate MTU for tunnel devices
as there is an overhead for device that has header_ops.
But there are ETHER tunnel devices, like gre_tap or erspan, which don't
have header_ops but set dev->hard_header_len during setup. This makes
pkts greater than (MTU - ETH_HLEN) could not be xmited. Fix it by
subtracting the ETHER tunnel devices' dev->hard_header_len for MTU
calculation.
Fixes: 28e104d002 ("net: ip_tunnel: fix mtu calculation")
Reported-by: Jianlin Shi <jishi@redhat.com>
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
sk_wmem_schedule makes sure that sk_forward_alloc has enough
bytes for charging that is going to be done by sk_mem_charge.
In the transmit zerocopy path, there is sk_mem_charge but there was
no call to sk_wmem_schedule. This change adds that call.
Without this call to sk_wmem_schedule, sk_forward_alloc can go
negetive which is a bug because sk_forward_alloc is a per-socket
space that has been forward charged so this can't be negative.
Fixes: f214f915e7 ("tcp: enable MSG_ZEROCOPY")
Signed-off-by: Talal Ahmad <talalahmad@google.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Wei Wang <weiwan@google.com>
Reviewed-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While TCP stack scales reasonably well, there is still one part that
can be used to DDOS it.
IPv6 Packet too big messages have to lookup/insert a new route,
and if abused by attackers, can easily put hosts under high stress,
with many cpus contending on a spinlock while one is stuck in fib6_run_gc()
ip6_protocol_deliver_rcu()
icmpv6_rcv()
icmpv6_notify()
tcp_v6_err()
tcp_v6_mtu_reduced()
inet6_csk_update_pmtu()
ip6_rt_update_pmtu()
__ip6_rt_update_pmtu()
ip6_rt_cache_alloc()
ip6_dst_alloc()
dst_alloc()
ip6_dst_gc()
fib6_run_gc()
spin_lock_bh() ...
Some of our servers have been hit by malicious ICMPv6 packets
trying to _increase_ the MTU/MSS of TCP flows.
We believe these ICMPv6 packets are a result of a bug in one ISP stack,
since they were blindly sent back for _every_ (small) packet sent to them.
These packets are for one TCP flow:
09:24:36.266491 IP6 Addr1 > Victim ICMP6, packet too big, mtu 1460, length 1240
09:24:36.266509 IP6 Addr1 > Victim ICMP6, packet too big, mtu 1460, length 1240
09:24:36.316688 IP6 Addr1 > Victim ICMP6, packet too big, mtu 1460, length 1240
09:24:36.316704 IP6 Addr1 > Victim ICMP6, packet too big, mtu 1460, length 1240
09:24:36.608151 IP6 Addr1 > Victim ICMP6, packet too big, mtu 1460, length 1240
TCP stack can filter some silly requests :
1) MTU below IPV6_MIN_MTU can be filtered early in tcp_v6_err()
2) tcp_v6_mtu_reduced() can drop requests trying to increase current MSS.
This tests happen before the IPv6 routing stack is entered, thus
removing the potential contention and route exhaustion.
Note that IPv6 stack was performing these checks, but too late
(ie : after the route has been added, and after the potential
garbage collect war)
v2: fix typo caught by Martin, thanks !
v3: exports tcp_mtu_to_mss(), caught by David, thanks !
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Maciej Żenczykowski <maze@google.com>
Cc: Martin KaFai Lau <kafai@fb.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit fixes a bug (found by syzkaller) that could cause spurious
double-initializations for congestion control modules, which could cause
memory leaks or other problems for congestion control modules (like CDG)
that allocate memory in their init functions.
The buggy scenario constructed by syzkaller was something like:
(1) create a TCP socket
(2) initiate a TFO connect via sendto()
(3) while socket is in TCP_SYN_SENT, call setsockopt(TCP_CONGESTION),
which calls:
tcp_set_congestion_control() ->
tcp_reinit_congestion_control() ->
tcp_init_congestion_control()
(4) receive ACK, connection is established, call tcp_init_transfer(),
set icsk_ca_initialized=0 (without first calling cc->release()),
call tcp_init_congestion_control() again.
Note that in this sequence tcp_init_congestion_control() is called
twice without a cc->release() call in between. Thus, for CC modules
that allocate memory in their init() function, e.g, CDG, a memory leak
may occur. The syzkaller tool managed to find a reproducer that
triggered such a leak in CDG.
The bug was introduced when that commit 8919a9b31e ("tcp: Only init
congestion control if not initialized already")
introduced icsk_ca_initialized and set icsk_ca_initialized to 0 in
tcp_init_transfer(), missing the possibility for a sequence like the
one above, where a process could call setsockopt(TCP_CONGESTION) in
state TCP_SYN_SENT (i.e. after the connect() or TFO open sendmsg()),
which would call tcp_init_congestion_control(). It did not intend to
reset any initialization that the user had already explicitly made;
it just missed the possibility of that particular sequence (which
syzkaller managed to find).
Fixes: 8919a9b31e ("tcp: Only init congestion control if not initialized already")
Reported-by: syzbot+f1e24a0594d4e3a895d3@syzkaller.appspotmail.com
Signed-off-by: Nguyen Dinh Phi <phind.uet@gmail.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Tested-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If an UDP packet enters the GRO engine but is not eligible
for aggregation and is not targeting an UDP tunnel,
udp_gro_receive() will not set the flush bit, and packet
could delayed till the next napi flush.
Fix the issue ensuring non GROed packets traverse
skb_gro_flush_final().
Reported-and-tested-by: Matthias Treydte <mt@waldheinz.de>
Fixes: 18f25dc399 ("udp: skip L4 aggregation for UDP tunnel packets")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While tp->mtu_info is read while socket is owned, the write
sides happen from err handlers (tcp_v[46]_mtu_reduced)
which only own the socket spinlock.
Fixes: 563d34d057 ("tcp: dont drop MTU reduction indications")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull networking updates from Jakub Kicinski:
"Core:
- BPF:
- add syscall program type and libbpf support for generating
instructions and bindings for in-kernel BPF loaders (BPF loaders
for BPF), this is a stepping stone for signed BPF programs
- infrastructure to migrate TCP child sockets from one listener to
another in the same reuseport group/map to improve flexibility
of service hand-off/restart
- add broadcast support to XDP redirect
- allow bypass of the lockless qdisc to improving performance (for
pktgen: +23% with one thread, +44% with 2 threads)
- add a simpler version of "DO_ONCE()" which does not require jump
labels, intended for slow-path usage
- virtio/vsock: introduce SOCK_SEQPACKET support
- add getsocketopt to retrieve netns cookie
- ip: treat lowest address of a IPv4 subnet as ordinary unicast
address allowing reclaiming of precious IPv4 addresses
- ipv6: use prandom_u32() for ID generation
- ip: add support for more flexible field selection for hashing
across multi-path routes (w/ offload to mlxsw)
- icmp: add support for extended RFC 8335 PROBE (ping)
- seg6: add support for SRv6 End.DT46 behavior
- mptcp:
- DSS checksum support (RFC 8684) to detect middlebox meddling
- support Connection-time 'C' flag
- time stamping support
- sctp: packetization Layer Path MTU Discovery (RFC 8899)
- xfrm: speed up state addition with seq set
- WiFi:
- hidden AP discovery on 6 GHz and other HE 6 GHz improvements
- aggregation handling improvements for some drivers
- minstrel improvements for no-ack frames
- deferred rate control for TXQs to improve reaction times
- switch from round robin to virtual time-based airtime scheduler
- add trace points:
- tcp checksum errors
- openvswitch - action execution, upcalls
- socket errors via sk_error_report
Device APIs:
- devlink: add rate API for hierarchical control of max egress rate
of virtual devices (VFs, SFs etc.)
- don't require RCU read lock to be held around BPF hooks in NAPI
context
- page_pool: generic buffer recycling
New hardware/drivers:
- mobile:
- iosm: PCIe Driver for Intel M.2 Modem
- support for Qualcomm MSM8998 (ipa)
- WiFi: Qualcomm QCN9074 and WCN6855 PCI devices
- sparx5: Microchip SparX-5 family of Enterprise Ethernet switches
- Mellanox BlueField Gigabit Ethernet (control NIC of the DPU)
- NXP SJA1110 Automotive Ethernet 10-port switch
- Qualcomm QCA8327 switch support (qca8k)
- Mikrotik 10/25G NIC (atl1c)
Driver changes:
- ACPI support for some MDIO, MAC and PHY devices from Marvell and
NXP (our first foray into MAC/PHY description via ACPI)
- HW timestamping (PTP) support: bnxt_en, ice, sja1105, hns3, tja11xx
- Mellanox/Nvidia NIC (mlx5)
- NIC VF offload of L2 bridging
- support IRQ distribution to Sub-functions
- Marvell (prestera):
- add flower and match all
- devlink trap
- link aggregation
- Netronome (nfp): connection tracking offload
- Intel 1GE (igc): add AF_XDP support
- Marvell DPU (octeontx2): ingress ratelimit offload
- Google vNIC (gve): new ring/descriptor format support
- Qualcomm mobile (rmnet & ipa): inline checksum offload support
- MediaTek WiFi (mt76)
- mt7915 MSI support
- mt7915 Tx status reporting
- mt7915 thermal sensors support
- mt7921 decapsulation offload
- mt7921 enable runtime pm and deep sleep
- Realtek WiFi (rtw88)
- beacon filter support
- Tx antenna path diversity support
- firmware crash information via devcoredump
- Qualcomm WiFi (wcn36xx)
- Wake-on-WLAN support with magic packets and GTK rekeying
- Micrel PHY (ksz886x/ksz8081): add cable test support"
* tag 'net-next-5.14' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (2168 commits)
tcp: change ICSK_CA_PRIV_SIZE definition
tcp_yeah: check struct yeah size at compile time
gve: DQO: Fix off by one in gve_rx_dqo()
stmmac: intel: set PCI_D3hot in suspend
stmmac: intel: Enable PHY WOL option in EHL
net: stmmac: option to enable PHY WOL with PMT enabled
net: say "local" instead of "static" addresses in ndo_dflt_fdb_{add,del}
net: use netdev_info in ndo_dflt_fdb_{add,del}
ptp: Set lookup cookie when creating a PTP PPS source.
net: sock: add trace for socket errors
net: sock: introduce sk_error_report
net: dsa: replay the local bridge FDB entries pointing to the bridge dev too
net: dsa: ensure during dsa_fdb_offload_notify that dev_hold and dev_put are on the same dev
net: dsa: include fdb entries pointing to bridge in the host fdb list
net: dsa: include bridge addresses which are local in the host fdb list
net: dsa: sync static FDB entries on foreign interfaces to hardware
net: dsa: install the host MDB and FDB entries in the master's RX filter
net: dsa: reference count the FDB addresses at the cross-chip notifier level
net: dsa: introduce a separate cross-chip notifier type for host FDBs
net: dsa: reference count the MDB entries at the cross-chip notifier level
...
Merge misc updates from Andrew Morton:
"191 patches.
Subsystems affected by this patch series: kthread, ia64, scripts,
ntfs, squashfs, ocfs2, kernel/watchdog, and mm (gup, pagealloc, slab,
slub, kmemleak, dax, debug, pagecache, gup, swap, memcg, pagemap,
mprotect, bootmem, dma, tracing, vmalloc, kasan, initialization,
pagealloc, and memory-failure)"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>: (191 commits)
mm,hwpoison: make get_hwpoison_page() call get_any_page()
mm,hwpoison: send SIGBUS with error virutal address
mm/page_alloc: split pcp->high across all online CPUs for cpuless nodes
mm/page_alloc: allow high-order pages to be stored on the per-cpu lists
mm: replace CONFIG_FLAT_NODE_MEM_MAP with CONFIG_FLATMEM
mm: replace CONFIG_NEED_MULTIPLE_NODES with CONFIG_NUMA
docs: remove description of DISCONTIGMEM
arch, mm: remove stale mentions of DISCONIGMEM
mm: remove CONFIG_DISCONTIGMEM
m68k: remove support for DISCONTIGMEM
arc: remove support for DISCONTIGMEM
arc: update comment about HIGHMEM implementation
alpha: remove DISCONTIGMEM and NUMA
mm/page_alloc: move free_the_page
mm/page_alloc: fix counting of managed_pages
mm/page_alloc: improve memmap_pages dbg msg
mm: drop SECTION_SHIFT in code comments
mm/page_alloc: introduce vm.percpu_pagelist_high_fraction
mm/page_alloc: limit the number of pages on PCP lists when reclaim is active
mm/page_alloc: scale the number of pages that are batch freed
...
Trivial conflict in net/netfilter/nf_tables_api.c.
Duplicate fix in tools/testing/selftests/net/devlink_port_split.py
- take the net-next version.
skmsg, and L4 bpf - keep the bpf code but remove the flags
and err params.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
