This extra check doesn't work for a handshake when SYN segment has
(current_key.maclen != rnext_key.maclen). It could be amended to
preserve rnext_key.maclen instead of current_key.maclen, but that
requires a lookup on listen socket.
Originally, this extra maclen check was introduced just because it was
cheap. Drop it and convert tcp_request_sock::maclen into boolean
tcp_request_sock::used_tcp_ao.
Fixes: 06b22ef295 ("net/tcp: Wire TCP-AO to request sockets")
Signed-off-by: Dmitry Safonov <dima@arista.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Use DO_ONCE_LITE_IF() and __cold attribute to put tcp_gro_dev_warn()
out of line.
This also allows the message to be printed again after a
"echo 1 > /sys/kernel/debug/clear_warn_once"
Also add a READ_ONCE() when reading device mtu, as it could
be changed concurrently.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20231130184135.4130860-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
After blamed commit, TFO client-ack-dropped-then-recovery-ms-timestamps
packetdrill test failed.
David Morley and Neal Cardwell started investigating and Neal pointed
that we had :
tcp_conn_request()
tcp_try_fastopen()
-> tcp_fastopen_create_child
-> child = inet_csk(sk)->icsk_af_ops->syn_recv_sock()
-> tcp_create_openreq_child()
-> copy req_usec_ts from req:
newtp->tcp_usec_ts = treq->req_usec_ts;
// now the new TFO server socket always does usec TS, no matter
// what the route options are...
send_synack()
-> tcp_make_synack()
// disable tcp_rsk(req)->req_usec_ts if route option is not present:
if (tcp_rsk(req)->req_usec_ts < 0)
tcp_rsk(req)->req_usec_ts = dst_tcp_usec_ts(dst);
tcp_conn_request() has the initial dst, we can initialize
tcp_rsk(req)->req_usec_ts there instead of later in send_synack();
This means tcp_rsk(req)->req_usec_ts can be a boolean.
Many thanks to David an Neal for their help.
Fixes: 614e8316aa ("tcp: add support for usec resolution in TCP TS values")
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/oe-lkp/202310302216.f79d78bc-oliver.sang@intel.com
Suggested-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: David Morley <morleyd@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Similarly to TCP-MD5, add a static key to TCP-AO that is patched out
when there are no keys on a machine and dynamically enabled with the
first setsockopt(TCP_AO) adds a key on any socket. The static key is as
well dynamically disabled later when the socket is destructed.
The lifetime of enabled static key here is the same as ao_info: it is
enabled on allocation, passed over from full socket to twsk and
destructed when ao_info is scheduled for destruction.
Signed-off-by: Dmitry Safonov <dima@arista.com>
Acked-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add Sequence Number Extension (SNE) for TCP-AO.
This is needed to protect long-living TCP-AO connections from replaying
attacks after sequence number roll-over, see RFC5925 (6.2).
Co-developed-by: Francesco Ruggeri <fruggeri@arista.com>
Signed-off-by: Francesco Ruggeri <fruggeri@arista.com>
Co-developed-by: Salam Noureddine <noureddine@arista.com>
Signed-off-by: Salam Noureddine <noureddine@arista.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Acked-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now when the new request socket is created from the listening socket,
it's recorded what MKT was used by the peer. tcp_rsk_used_ao() is
a new helper for checking if TCP-AO option was used to create the
request socket.
tcp_ao_copy_all_matching() will copy all keys that match the peer on the
request socket, as well as preparing them for the usage (creating
traffic keys).
Co-developed-by: Francesco Ruggeri <fruggeri@arista.com>
Signed-off-by: Francesco Ruggeri <fruggeri@arista.com>
Co-developed-by: Salam Noureddine <noureddine@arista.com>
Signed-off-by: Salam Noureddine <noureddine@arista.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Acked-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Introduce a helper that:
(1) shares the common code with TCP-MD5 header options parsing
(2) looks for hash signature only once for both TCP-MD5 and TCP-AO
(3) fails with -EEXIST if any TCP sign option is present twice, see
RFC5925 (2.2):
">> A single TCP segment MUST NOT have more than one TCP-AO in its
options sequence. When multiple TCP-AOs appear, TCP MUST discard
the segment."
Co-developed-by: Francesco Ruggeri <fruggeri@arista.com>
Signed-off-by: Francesco Ruggeri <fruggeri@arista.com>
Co-developed-by: Salam Noureddine <noureddine@arista.com>
Signed-off-by: Salam Noureddine <noureddine@arista.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Acked-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add traffic key calculation the way it's described in RFC5926.
Wire it up to tcp_finish_connect() and cache the new keys straight away
on already established TCP connections.
Co-developed-by: Francesco Ruggeri <fruggeri@arista.com>
Signed-off-by: Francesco Ruggeri <fruggeri@arista.com>
Co-developed-by: Salam Noureddine <noureddine@arista.com>
Signed-off-by: Salam Noureddine <noureddine@arista.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Acked-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Back in 2015, Van Jacobson suggested to use usec resolution in TCP TS values.
This has been implemented in our private kernels.
Goals were :
1) better observability of delays in networking stacks.
2) better disambiguation of events based on TSval/ecr values.
3) building block for congestion control modules needing usec resolution.
Back then we implemented a schem based on private SYN options
to negotiate the feature.
For upstream submission, we chose to use a route attribute,
because this feature is probably going to be used in private
networks [1] [2].
ip route add 10/8 ... features tcp_usec_ts
Note that RFC 7323 recommends a
"timestamp clock frequency in the range 1 ms to 1 sec per tick.",
but also mentions
"the maximum acceptable clock frequency is one tick every 59 ns."
[1] Unfortunately RFC 7323 5.5 (Outdated Timestamps) suggests
to invalidate TS.Recent values after a flow was idle for more
than 24 days. This is the part making usec_ts a problem
for peers following this recommendation for long living
idle flows.
[2] Attempts to standardize usec ts went nowhere:
https://www.ietf.org/proceedings/97/slides/slides-97-tcpm-tcp-options-for-low-latency-00.pdfhttps://datatracker.ietf.org/doc/draft-wang-tcpm-low-latency-opt/
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Before adding usec TS support, add tcp_rtt_tsopt_us() helper
to factorize code.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This helper returns a TSval from a TCP socket.
It currently calls tcp_time_stamp_ms() but will soon
be able to return a usec based TSval, depending
on an upcoming tp->tcp_usec_ts field.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This helper returns a 32bit TCP TSval from skb->tstamp.
As we are going to support usec or ms units soon, rename it
to tcp_skb_timestamp_ts() and add a boolean to select the unit.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In preparation of adding usec TCP TS values, add tcp_time_stamp_ms()
for contexts needing ms based values.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit fix wrong RTO timeout when received SACK reneging.
When an ACK arrived pointing to a SACK reneging, tcp_check_sack_reneging()
will rearm the RTO timer for min(1/2*srtt, 10ms) into to the future.
But since the commit 62d9f1a694 ("tcp: fix TLP timer not set when
CA_STATE changes from DISORDER to OPEN") merged, the tcp_set_xmit_timer()
is moved after tcp_fastretrans_alert()(which do the SACK reneging check),
so the RTO timeout will be overwrited by tcp_set_xmit_timer() with
icsk_rto instead of 1/2*srtt.
Here is a packetdrill script to check this bug:
0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
// simulate srtt to 100ms
+0 < S 0:0(0) win 32792 <mss 1000, sackOK,nop,nop,nop,wscale 7>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
+.1 < . 1:1(0) ack 1 win 1024
+0 accept(3, ..., ...) = 4
+0 write(4, ..., 10000) = 10000
+0 > P. 1:10001(10000) ack 1
// inject sack
+.1 < . 1:1(0) ack 1 win 257 <sack 1001:10001,nop,nop>
+0 > . 1:1001(1000) ack 1
// inject sack reneging
+.1 < . 1:1(0) ack 1001 win 257 <sack 9001:10001,nop,nop>
// we expect rto fired in 1/2*srtt (50ms)
+.05 > . 1001:2001(1000) ack 1
This fix remove the FLAG_SET_XMIT_TIMER from ack_flag when
tcp_check_sack_reneging() set RTO timer with 1/2*srtt to avoid
being overwrited later.
Fixes: 62d9f1a694 ("tcp: fix TLP timer not set when CA_STATE changes from DISORDER to OPEN")
Signed-off-by: Fred Chen <fred.chenchen03@gmail.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
Tested-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit changes the data receiver repath behavior to occur after
receiving a single duplicate. This can help recover ACK connectivity
quicker if a TLP was sent along a nonworking path.
For instance, consider the case where we have an initially nonworking
forward path and reverse path and subsequently switch to only working
forward paths. Before this patch we would have the following behavior.
+---------+--------+--------+----------+----------+----------+
| Event | For FL | Rev FL | FP Works | RP Works | Data Del |
+---------+--------+--------+----------+----------+----------+
| Initial | A | 1 | N | N | 0 |
+---------+--------+--------+----------+----------+----------+
| TLP | A | 1 | N | N | 0 |
+---------+--------+--------+----------+----------+----------+
| RTO 1 | B | 1 | Y | N | 1 |
+---------+--------+--------+----------+----------+----------+
| RTO 2 | C | 1 | Y | N | 2 |
+---------+--------+--------+----------+----------+----------+
| RTO 3 | D | 2 | Y | Y | 3 |
+---------+--------+--------+----------+----------+----------+
This patch gets rid of at least RTO 3, avoiding additional unnecessary
repaths of a working forward path to a (potentially) nonworking one.
In addition, this commit changes the behavior to avoid repathing upon
rx of duplicate data if the local endpoint is in CA_Loss (in which
case the RTOs will already be changing the outgoing flowlabel).
Signed-off-by: David Morley <morleyd@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Tested-by: David Morley <morleyd@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
In order to better estimate whether a data packet has been
retransmitted or is the result of a TLP, we save the last received
ipv6 flowlabel.
To make space for this field we resize the "ato" field in
inet_connection_sock as the current value of TCP_DELACK_MAX can be
fully contained in 8 bits and add a compile_time_assert ensuring this
field is the required size.
v2: addressed kernel bot feedback about dccp_delack_timer()
v3: addressed build error introduced by commit bbf80d713f ("tcp:
derive delack_max from rto_min")
Signed-off-by: David Morley <morleyd@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Tested-by: David Morley <morleyd@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
This commit fixes poor delayed ACK behavior that can cause poor TCP
latency in a particular boundary condition: when an application makes
a TCP socket write that is an exact multiple of the MSS size.
The problem is that there is painful boundary discontinuity in the
current delayed ACK behavior. With the current delayed ACK behavior,
we have:
(1) If an app reads data when > 1*MSS is unacknowledged, then
tcp_cleanup_rbuf() ACKs immediately because of:
tp->rcv_nxt - tp->rcv_wup > icsk->icsk_ack.rcv_mss ||
(2) If an app reads all received data, and the packets were < 1*MSS,
and either (a) the app is not ping-pong or (b) we received two
packets < 1*MSS, then tcp_cleanup_rbuf() ACKs immediately beecause
of:
((icsk->icsk_ack.pending & ICSK_ACK_PUSHED2) ||
((icsk->icsk_ack.pending & ICSK_ACK_PUSHED) &&
!inet_csk_in_pingpong_mode(sk))) &&
(3) *However*: if an app reads exactly 1*MSS of data,
tcp_cleanup_rbuf() does not send an immediate ACK. This is true
even if the app is not ping-pong and the 1*MSS of data had the PSH
bit set, suggesting the sending application completed an
application write.
Thus if the app is not ping-pong, we have this painful case where
>1*MSS gets an immediate ACK, and <1*MSS gets an immediate ACK, but a
write whose last skb is an exact multiple of 1*MSS can get a 40ms
delayed ACK. This means that any app that transfers data in one
direction and takes care to align write size or packet size with MSS
can suffer this problem. With receive zero copy making 4KB MSS values
more common, it is becoming more common to have application writes
naturally align with MSS, and more applications are likely to
encounter this delayed ACK problem.
The fix in this commit is to refine the delayed ACK heuristics with a
simple check: immediately ACK a received 1*MSS skb with PSH bit set if
the app reads all data. Why? If an skb has a len of exactly 1*MSS and
has the PSH bit set then it is likely the end of an application
write. So more data may not be arriving soon, and yet the data sender
may be waiting for an ACK if cwnd-bound or using TX zero copy. Thus we
set ICSK_ACK_PUSHED in this case so that tcp_cleanup_rbuf() will send
an ACK immediately if the app reads all of the data and is not
ping-pong. Note that this logic is also executed for the case where
len > MSS, but in that case this logic does not matter (and does not
hurt) because tcp_cleanup_rbuf() will always ACK immediately if the
app reads data and there is more than an MSS of unACKed data.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Cc: Xin Guo <guoxin0309@gmail.com>
Link: https://lore.kernel.org/r/20231001151239.1866845-2-ncardwell.sw@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
SO_MAX_PACING_RATE setsockopt() does not need to hold
the socket lock, because sk->sk_pacing_rate readers
can run fine if the value is changed by other threads,
after adding READ_ONCE() accessors.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The 2023 SIGCOMM paper "Improving Network Availability with Protective
ReRoute" has indicated Linux TCP's RTO-triggered txhash rehashing can
effectively reduce application disruption during outages. To better
measure the efficacy of this feature, this patch adds three more
detailed stats during RTO recovery and exports via TCP_INFO.
Applications and monitoring systems can leverage this data to measure
the network path diversity and end-to-end repair latency during network
outages to improve their network infrastructure.
The following counters are added to tcp_sock in order to track RTO
events over the lifetime of a TCP socket.
1. u16 total_rto - Counts the total number of RTO timeouts.
2. u16 total_rto_recoveries - Counts the total number of RTO recoveries.
3. u32 total_rto_time - Counts the total time spent (ms) in RTO
recoveries. (time spent in CA_Loss and
CA_Recovery states)
To compute total_rto_time, we add a new u32 rto_stamp field to
tcp_sock. rto_stamp records the start timestamp (ms) of the last RTO
recovery (CA_Loss).
Corresponding fields are also added to the tcp_info struct.
Signed-off-by: Aananth V <aananthv@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For passive TCP Fast Open sockets that had SYN/ACK timeout and did not
send more data in SYN_RECV, upon receiving the final ACK in 3WHS, the
congestion state may awkwardly stay in CA_Loss mode unless the CA state
was undone due to TCP timestamp checks. However, if
tcp_rcv_synrecv_state_fastopen() decides not to undo, then we should
enter CA_Open, because at that point we have received an ACK covering
the retransmitted SYNACKs. Currently, the icsk_ca_state is only set to
CA_Open after we receive an ACK for a data-packet. This is because
tcp_ack does not call tcp_fastretrans_alert (and tcp_process_loss) if
!prior_packets
Note that tcp_process_loss() calls tcp_try_undo_recovery(), so having
tcp_rcv_synrecv_state_fastopen() decide that if we're in CA_Loss we
should call tcp_try_undo_recovery() is consistent with that, and
low risk.
Fixes: dad8cea7ad ("tcp: fix TFO SYNACK undo to avoid double-timestamp-undo")
Signed-off-by: Aananth V <aananthv@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This idea came after a particular workload requested
the quickack attribute set on routes, and a performance
drop was noticed for large bulk transfers.
For high throughput flows, it is best to use one cpu
running the user thread issuing socket system calls,
and a separate cpu to process incoming packets from BH context.
(With TSO/GRO, bottleneck is usually the 'user' cpu)
Problem is the user thread can spend a lot of time while holding
the socket lock, forcing BH handler to queue most of incoming
packets in the socket backlog.
Whenever the user thread releases the socket lock, it must first
process all accumulated packets in the backlog, potentially
adding latency spikes. Due to flood mitigation, having too many
packets in the backlog increases chance of unexpected drops.
Backlog processing unfortunately shifts a fair amount of cpu cycles
from the BH cpu to the 'user' cpu, thus reducing max throughput.
This patch takes advantage of the backlog processing,
and the fact that ACK are mostly cumulative.
The idea is to detect we are in the backlog processing
and defer all eligible ACK into a single one,
sent from tcp_release_cb().
This saves cpu cycles on both sides, and network resources.
Performance of a single TCP flow on a 200Gbit NIC:
- Throughput is increased by 20% (100Gbit -> 120Gbit).
- Number of generated ACK per second shrinks from 240,000 to 40,000.
- Number of backlog drops per second shrinks from 230 to 0.
Benchmark context:
- Regular netperf TCP_STREAM (no zerocopy)
- Intel(R) Xeon(R) Platinum 8481C (Saphire Rapids)
- MAX_SKB_FRAGS = 17 (~60KB per GRO packet)
This feature is guarded by a new sysctl, and enabled by default:
/proc/sys/net/ipv4/tcp_backlog_ack_defer
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Dave Taht <dave.taht@gmail.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
IP_TRANSPARENT socket option can now be set/read
without locking the socket.
v2: removed unused issk variable in mptcp_setsockopt_sol_ip_set_transparent()
v4: rebased after commit 3f326a821b ("mptcp: change the mpc check helper to return a sk")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Paolo Abeni <pabeni@redhat.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fow now, an ACK can update the window in following case, according to
the tcp_may_update_window():
1. the ACK acknowledged new data
2. the ACK has new data
3. the ACK expand the window and the seq of it is valid
Now, we allow the ACK update the window if the window is 0, and the
seq/ack of it is valid. This is for the case that the receiver replies
an zero-window ACK when it is under memory stress and can't queue the new
data.
Signed-off-by: Menglong Dong <imagedong@tencent.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For now, skb will be dropped when no memory, which makes client keep
retrans util timeout and it's not friendly to the users.
In this patch, we reply an ACK with zero-window in this case to update
the snd_wnd of the sender to 0. Therefore, the sender won't timeout the
connection and will probe the zero-window with the retransmits.
Signed-off-by: Menglong Dong <imagedong@tencent.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
rskq_defer_accept field can be read/written without
the need of holding the socket lock.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
tp->linger2 can be set locklessly as long as readers
use READ_ONCE().
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since commit 9ea88a1530 ("tcp: md5: check md5 signature without socket
lock"), the MD5 option is checked in tcp_v[46]_rcv().
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20230803224552.69398-3-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
tcp_sequence() uses two conditions to decide to drop a packet,
and we currently report generic TCP_INVALID_SEQUENCE drop reason.
Duplicates are common, we need to distinguish them from
the other case.
I chose to not reuse TCP_OLD_DATA, and instead added
TCP_OLD_SEQUENCE drop reason.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20230719064754.2794106-1-edumazet@google.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
After commit d2ccd7bc8a ("tcp: avoid resetting ACK timer in DCTCP"),
tcp_enter_quickack_mode() is only used from net/ipv4/tcp_input.c.
Fixes: d2ccd7bc8a ("tcp: avoid resetting ACK timer in DCTCP")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Link: https://lore.kernel.org/r/20230718162049.1444938-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
With modern NIC drivers shifting to full page allocations per
received frame, we face the following issue:
TCP has one per-netns sysctl used to tweak how to translate
a memory use into an expected payload (RWIN), in RX path.
tcp_win_from_space() implementation is limited to few cases.
For hosts dealing with various MSS, we either under estimate
or over estimate the RWIN we send to the remote peers.
For instance with the default sysctl_tcp_adv_win_scale value,
we expect to store 50% of payload per allocated chunk of memory.
For the typical use of MTU=1500 traffic, and order-0 pages allocations
by NIC drivers, we are sending too big RWIN, leading to potential
tcp collapse operations, which are extremely expensive and source
of latency spikes.
This patch makes sysctl_tcp_adv_win_scale obsolete, and instead
uses a per socket scaling factor, so that we can precisely
adjust the RWIN based on effective skb->len/skb->truesize ratio.
This patch alone can double TCP receive performance when receivers
are too slow to drain their receive queue, or by allowing
a bigger RWIN when MSS is close to PAGE_SIZE.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Link: https://lore.kernel.org/r/20230717152917.751987-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
request sockets are lockless, __tcp_oow_rate_limited() could be called
on the same object from different cpus. This is harmless.
Add READ_ONCE()/WRITE_ONCE() annotations to avoid a KCSAN report.
Fixes: 4ce7e93cb3 ("tcp: rate limit ACK sent by SYN_RECV request sockets")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In this patch, we mainly try to handle sending a compressed ack
correctly if it's deferred.
Here are more details in the old logic:
When sack compression is triggered in the tcp_compressed_ack_kick(),
if the sock is owned by user, it will set TCP_DELACK_TIMER_DEFERRED
and then defer to the release cb phrase. Later once user releases
the sock, tcp_delack_timer_handler() should send a ack as expected,
which, however, cannot happen due to lack of ICSK_ACK_TIMER flag.
Therefore, the receiver would not sent an ack until the sender's
retransmission timeout. It definitely increases unnecessary latency.
Fixes: 5d9f4262b7 ("tcp: add SACK compression")
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: fuyuanli <fuyuanli@didiglobal.com>
Signed-off-by: Jason Xing <kerneljasonxing@gmail.com>
Link: https://lore.kernel.org/netdev/20230529113804.GA20300@didi-ThinkCentre-M920t-N000/
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20230531080150.GA20424@didi-ThinkCentre-M920t-N000
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Our Network Load Balancer (NLB) [0] has multiple nodes with different
IP addresses, and each node forwards TCP flows from clients to backend
targets. NLB has an option to preserve the client's source IP address
and port when routing packets to backend targets. [1]
When a client connects to two different NLB nodes, they may select the
same backend target. Then, if the client has used the same source IP
and port, the two flows at the backend side will have the same 4-tuple.
While testing around such cases, I saw these sequences on the backend
target.
IP 10.0.0.215.60000 > 10.0.3.249.10000: Flags [S], seq 2819965599, win 62727, options [mss 8365,sackOK,TS val 1029816180 ecr 0,nop,wscale 7], length 0
IP 10.0.3.249.10000 > 10.0.0.215.60000: Flags [S.], seq 3040695044, ack 2819965600, win 62643, options [mss 8961,sackOK,TS val 1224784076 ecr 1029816180,nop,wscale 7], length 0
IP 10.0.0.215.60000 > 10.0.3.249.10000: Flags [.], ack 1, win 491, options [nop,nop,TS val 1029816181 ecr 1224784076], length 0
IP 10.0.0.215.60000 > 10.0.3.249.10000: Flags [S], seq 2681819307, win 62727, options [mss 8365,sackOK,TS val 572088282 ecr 0,nop,wscale 7], length 0
IP 10.0.3.249.10000 > 10.0.0.215.60000: Flags [.], ack 1, win 490, options [nop,nop,TS val 1224794914 ecr 1029816181,nop,nop,sack 1 {4156821004:4156821005}], length 0
It seems to be working correctly, but the last ACK was generated by
tcp_send_dupack() and PAWSEstab was increased. This is because the
second connection has a smaller timestamp than the first one.
In this case, we should send a dup ACK in tcp_send_challenge_ack()
to increase the correct counter and rate-limit it properly.
Let's check the SYN flag after the PAWS tests to avoid adding unnecessary
overhead for most packets.
Link: https://docs.aws.amazon.com/elasticloadbalancing/latest/network/introduction.html [0]
Link: https://docs.aws.amazon.com/elasticloadbalancing/latest/network/load-balancer-target-groups.html#client-ip-preservation [1]
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Jason Xing <kerneljasonxing@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We can change tcp_sk() to propagate its argument const qualifier,
thanks to container_of_const().
We have two places where a const sock pointer has to be upgraded
to a write one. We have been using const qualifier for lockless
listeners to clearly identify points where writes could happen.
Add tcp_sk_rw() helper to better document these.
tcp_inbound_md5_hash(), __tcp_grow_window(), tcp_reset_check()
and tcp_rack_reo_wnd() get an additional const qualififer
for their @tp local variables.
smc_check_reset_syn_req() also needs a similar change.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
tcp_poll() reads sk->sk_err without socket lock held/owned.
We should used READ_ONCE() here, and update writers
to use WRITE_ONCE().
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This field can be read/written without lock synchronization.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Core
----
- Allow live renaming when an interface is up
- Add retpoline wrappers for tc, improving considerably the
performances of complex queue discipline configurations.
- Add inet drop monitor support.
- A few GRO performance improvements.
- Add infrastructure for atomic dev stats, addressing long standing
data races.
- De-duplicate common code between OVS and conntrack offloading
infrastructure.
- A bunch of UBSAN_BOUNDS/FORTIFY_SOURCE improvements.
- Netfilter: introduce packet parser for tunneled packets
- Replace IPVS timer-based estimators with kthreads to scale up
the workload with the number of available CPUs.
- Add the helper support for connection-tracking OVS offload.
BPF
---
- Support for user defined BPF objects: the use case is to allocate
own objects, build own object hierarchies and use the building
blocks to build own data structures flexibly, for example, linked
lists in BPF.
- Make cgroup local storage available to non-cgroup attached BPF
programs.
- Avoid unnecessary deadlock detection and failures wrt BPF task
storage helpers.
- A relevant bunch of BPF verifier fixes and improvements.
- Veristat tool improvements to support custom filtering, sorting,
and replay of results.
- Add LLVM disassembler as default library for dumping JITed code.
- Lots of new BPF documentation for various BPF maps.
- Add bpf_rcu_read_{,un}lock() support for sleepable programs.
- Add RCU grace period chaining to BPF to wait for the completion
of access from both sleepable and non-sleepable BPF programs.
- Add support storing struct task_struct objects as kptrs in maps.
- Improve helper UAPI by explicitly defining BPF_FUNC_xxx integer
values.
- Add libbpf *_opts API-variants for bpf_*_get_fd_by_id() functions.
Protocols
---------
- TCP: implement Protective Load Balancing across switch links.
- TCP: allow dynamically disabling TCP-MD5 static key, reverting
back to fast[er]-path.
- UDP: Introduce optional per-netns hash lookup table.
- IPv6: simplify and cleanup sockets disposal.
- Netlink: support different type policies for each generic
netlink operation.
- MPTCP: add MSG_FASTOPEN and FastOpen listener side support.
- MPTCP: add netlink notification support for listener sockets
events.
- SCTP: add VRF support, allowing sctp sockets binding to VRF
devices.
- Add bridging MAC Authentication Bypass (MAB) support.
- Extensions for Ethernet VPN bridging implementation to better
support multicast scenarios.
- More work for Wi-Fi 7 support, comprising conversion of all
the existing drivers to internal TX queue usage.
- IPSec: introduce a new offload type (packet offload) allowing
complete header processing and crypto offloading.
- IPSec: extended ack support for more descriptive XFRM error
reporting.
- RXRPC: increase SACK table size and move processing into a
per-local endpoint kernel thread, reducing considerably the
required locking.
- IEEE 802154: synchronous send frame and extended filtering
support, initial support for scanning available 15.4 networks.
- Tun: bump the link speed from 10Mbps to 10Gbps.
- Tun/VirtioNet: implement UDP segmentation offload support.
Driver API
----------
- PHY/SFP: improve power level switching between standard
level 1 and the higher power levels.
- New API for netdev <-> devlink_port linkage.
- PTP: convert existing drivers to new frequency adjustment
implementation.
- DSA: add support for rx offloading.
- Autoload DSA tagging driver when dynamically changing protocol.
- Add new PCP and APPTRUST attributes to Data Center Bridging.
- Add configuration support for 800Gbps link speed.
- Add devlink port function attribute to enable/disable RoCE and
migratable.
- Extend devlink-rate to support strict prioriry and weighted fair
queuing.
- Add devlink support to directly reading from region memory.
- New device tree helper to fetch MAC address from nvmem.
- New big TCP helper to simplify temporary header stripping.
New hardware / drivers
----------------------
- Ethernet:
- Marvel Octeon CNF95N and CN10KB Ethernet Switches.
- Marvel Prestera AC5X Ethernet Switch.
- WangXun 10 Gigabit NIC.
- Motorcomm yt8521 Gigabit Ethernet.
- Microchip ksz9563 Gigabit Ethernet Switch.
- Microsoft Azure Network Adapter.
- Linux Automation 10Base-T1L adapter.
- PHY:
- Aquantia AQR112 and AQR412.
- Motorcomm YT8531S.
- PTP:
- Orolia ART-CARD.
- WiFi:
- MediaTek Wi-Fi 7 (802.11be) devices.
- RealTek rtw8821cu, rtw8822bu, rtw8822cu and rtw8723du USB
devices.
- Bluetooth:
- Broadcom BCM4377/4378/4387 Bluetooth chipsets.
- Realtek RTL8852BE and RTL8723DS.
- Cypress.CYW4373A0 WiFi + Bluetooth combo device.
Drivers
-------
- CAN:
- gs_usb: bus error reporting support.
- kvaser_usb: listen only and bus error reporting support.
- Ethernet NICs:
- Intel (100G):
- extend action skbedit to RX queue mapping.
- implement devlink-rate support.
- support direct read from memory.
- nVidia/Mellanox (mlx5):
- SW steering improvements, increasing rules update rate.
- Support for enhanced events compression.
- extend H/W offload packet manipulation capabilities.
- implement IPSec packet offload mode.
- nVidia/Mellanox (mlx4):
- better big TCP support.
- Netronome Ethernet NICs (nfp):
- IPsec offload support.
- add support for multicast filter.
- Broadcom:
- RSS and PTP support improvements.
- AMD/SolarFlare:
- netlink extened ack improvements.
- add basic flower matches to offload, and related stats.
- Virtual NICs:
- ibmvnic: introduce affinity hint support.
- small / embedded:
- FreeScale fec: add initial XDP support.
- Marvel mv643xx_eth: support MII/GMII/RGMII modes for Kirkwood.
- TI am65-cpsw: add suspend/resume support.
- Mediatek MT7986: add RX wireless wthernet dispatch support.
- Realtek 8169: enable GRO software interrupt coalescing per
default.
- Ethernet high-speed switches:
- Microchip (sparx5):
- add support for Sparx5 TC/flower H/W offload via VCAP.
- Mellanox mlxsw:
- add 802.1X and MAC Authentication Bypass offload support.
- add ip6gre support.
- Embedded Ethernet switches:
- Mediatek (mtk_eth_soc):
- improve PCS implementation, add DSA untag support.
- enable flow offload support.
- Renesas:
- add rswitch R-Car Gen4 gPTP support.
- Microchip (lan966x):
- add full XDP support.
- add TC H/W offload via VCAP.
- enable PTP on bridge interfaces.
- Microchip (ksz8):
- add MTU support for KSZ8 series.
- Qualcomm 802.11ax WiFi (ath11k):
- support configuring channel dwell time during scan.
- MediaTek WiFi (mt76):
- enable Wireless Ethernet Dispatch (WED) offload support.
- add ack signal support.
- enable coredump support.
- remain_on_channel support.
- Intel WiFi (iwlwifi):
- enable Wi-Fi 7 Extremely High Throughput (EHT) PHY capabilities.
- 320 MHz channels support.
- RealTek WiFi (rtw89):
- new dynamic header firmware format support.
- wake-over-WLAN support.
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
-----BEGIN PGP SIGNATURE-----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=CbJC
-----END PGP SIGNATURE-----
Merge tag 'net-next-6.2' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next
Pull networking updates from Paolo Abeni:
"Core:
- Allow live renaming when an interface is up
- Add retpoline wrappers for tc, improving considerably the
performances of complex queue discipline configurations
- Add inet drop monitor support
- A few GRO performance improvements
- Add infrastructure for atomic dev stats, addressing long standing
data races
- De-duplicate common code between OVS and conntrack offloading
infrastructure
- A bunch of UBSAN_BOUNDS/FORTIFY_SOURCE improvements
- Netfilter: introduce packet parser for tunneled packets
- Replace IPVS timer-based estimators with kthreads to scale up the
workload with the number of available CPUs
- Add the helper support for connection-tracking OVS offload
BPF:
- Support for user defined BPF objects: the use case is to allocate
own objects, build own object hierarchies and use the building
blocks to build own data structures flexibly, for example, linked
lists in BPF
- Make cgroup local storage available to non-cgroup attached BPF
programs
- Avoid unnecessary deadlock detection and failures wrt BPF task
storage helpers
- A relevant bunch of BPF verifier fixes and improvements
- Veristat tool improvements to support custom filtering, sorting,
and replay of results
- Add LLVM disassembler as default library for dumping JITed code
- Lots of new BPF documentation for various BPF maps
- Add bpf_rcu_read_{,un}lock() support for sleepable programs
- Add RCU grace period chaining to BPF to wait for the completion of
access from both sleepable and non-sleepable BPF programs
- Add support storing struct task_struct objects as kptrs in maps
- Improve helper UAPI by explicitly defining BPF_FUNC_xxx integer
values
- Add libbpf *_opts API-variants for bpf_*_get_fd_by_id() functions
Protocols:
- TCP: implement Protective Load Balancing across switch links
- TCP: allow dynamically disabling TCP-MD5 static key, reverting back
to fast[er]-path
- UDP: Introduce optional per-netns hash lookup table
- IPv6: simplify and cleanup sockets disposal
- Netlink: support different type policies for each generic netlink
operation
- MPTCP: add MSG_FASTOPEN and FastOpen listener side support
- MPTCP: add netlink notification support for listener sockets events
- SCTP: add VRF support, allowing sctp sockets binding to VRF devices
- Add bridging MAC Authentication Bypass (MAB) support
- Extensions for Ethernet VPN bridging implementation to better
support multicast scenarios
- More work for Wi-Fi 7 support, comprising conversion of all the
existing drivers to internal TX queue usage
- IPSec: introduce a new offload type (packet offload) allowing
complete header processing and crypto offloading
- IPSec: extended ack support for more descriptive XFRM error
reporting
- RXRPC: increase SACK table size and move processing into a
per-local endpoint kernel thread, reducing considerably the
required locking
- IEEE 802154: synchronous send frame and extended filtering support,
initial support for scanning available 15.4 networks
- Tun: bump the link speed from 10Mbps to 10Gbps
- Tun/VirtioNet: implement UDP segmentation offload support
Driver API:
- PHY/SFP: improve power level switching between standard level 1 and
the higher power levels
- New API for netdev <-> devlink_port linkage
- PTP: convert existing drivers to new frequency adjustment
implementation
- DSA: add support for rx offloading
- Autoload DSA tagging driver when dynamically changing protocol
- Add new PCP and APPTRUST attributes to Data Center Bridging
- Add configuration support for 800Gbps link speed
- Add devlink port function attribute to enable/disable RoCE and
migratable
- Extend devlink-rate to support strict prioriry and weighted fair
queuing
- Add devlink support to directly reading from region memory
- New device tree helper to fetch MAC address from nvmem
- New big TCP helper to simplify temporary header stripping
New hardware / drivers:
- Ethernet:
- Marvel Octeon CNF95N and CN10KB Ethernet Switches
- Marvel Prestera AC5X Ethernet Switch
- WangXun 10 Gigabit NIC
- Motorcomm yt8521 Gigabit Ethernet
- Microchip ksz9563 Gigabit Ethernet Switch
- Microsoft Azure Network Adapter
- Linux Automation 10Base-T1L adapter
- PHY:
- Aquantia AQR112 and AQR412
- Motorcomm YT8531S
- PTP:
- Orolia ART-CARD
- WiFi:
- MediaTek Wi-Fi 7 (802.11be) devices
- RealTek rtw8821cu, rtw8822bu, rtw8822cu and rtw8723du USB
devices
- Bluetooth:
- Broadcom BCM4377/4378/4387 Bluetooth chipsets
- Realtek RTL8852BE and RTL8723DS
- Cypress.CYW4373A0 WiFi + Bluetooth combo device
Drivers:
- CAN:
- gs_usb: bus error reporting support
- kvaser_usb: listen only and bus error reporting support
- Ethernet NICs:
- Intel (100G):
- extend action skbedit to RX queue mapping
- implement devlink-rate support
- support direct read from memory
- nVidia/Mellanox (mlx5):
- SW steering improvements, increasing rules update rate
- Support for enhanced events compression
- extend H/W offload packet manipulation capabilities
- implement IPSec packet offload mode
- nVidia/Mellanox (mlx4):
- better big TCP support
- Netronome Ethernet NICs (nfp):
- IPsec offload support
- add support for multicast filter
- Broadcom:
- RSS and PTP support improvements
- AMD/SolarFlare:
- netlink extened ack improvements
- add basic flower matches to offload, and related stats
- Virtual NICs:
- ibmvnic: introduce affinity hint support
- small / embedded:
- FreeScale fec: add initial XDP support
- Marvel mv643xx_eth: support MII/GMII/RGMII modes for Kirkwood
- TI am65-cpsw: add suspend/resume support
- Mediatek MT7986: add RX wireless wthernet dispatch support
- Realtek 8169: enable GRO software interrupt coalescing per
default
- Ethernet high-speed switches:
- Microchip (sparx5):
- add support for Sparx5 TC/flower H/W offload via VCAP
- Mellanox mlxsw:
- add 802.1X and MAC Authentication Bypass offload support
- add ip6gre support
- Embedded Ethernet switches:
- Mediatek (mtk_eth_soc):
- improve PCS implementation, add DSA untag support
- enable flow offload support
- Renesas:
- add rswitch R-Car Gen4 gPTP support
- Microchip (lan966x):
- add full XDP support
- add TC H/W offload via VCAP
- enable PTP on bridge interfaces
- Microchip (ksz8):
- add MTU support for KSZ8 series
- Qualcomm 802.11ax WiFi (ath11k):
- support configuring channel dwell time during scan
- MediaTek WiFi (mt76):
- enable Wireless Ethernet Dispatch (WED) offload support
- add ack signal support
- enable coredump support
- remain_on_channel support
- Intel WiFi (iwlwifi):
- enable Wi-Fi 7 Extremely High Throughput (EHT) PHY capabilities
- 320 MHz channels support
- RealTek WiFi (rtw89):
- new dynamic header firmware format support
- wake-over-WLAN support"
* tag 'net-next-6.2' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (2002 commits)
ipvs: fix type warning in do_div() on 32 bit
net: lan966x: Remove a useless test in lan966x_ptp_add_trap()
net: ipa: add IPA v4.7 support
dt-bindings: net: qcom,ipa: Add SM6350 compatible
bnxt: Use generic HBH removal helper in tx path
IPv6/GRO: generic helper to remove temporary HBH/jumbo header in driver
selftests: forwarding: Add bridge MDB test
selftests: forwarding: Rename bridge_mdb test
bridge: mcast: Support replacement of MDB port group entries
bridge: mcast: Allow user space to specify MDB entry routing protocol
bridge: mcast: Allow user space to add (*, G) with a source list and filter mode
bridge: mcast: Add support for (*, G) with a source list and filter mode
bridge: mcast: Avoid arming group timer when (S, G) corresponds to a source
bridge: mcast: Add a flag for user installed source entries
bridge: mcast: Expose __br_multicast_del_group_src()
bridge: mcast: Expose br_multicast_new_group_src()
bridge: mcast: Add a centralized error path
bridge: mcast: Place netlink policy before validation functions
bridge: mcast: Split (*, G) and (S, G) addition into different functions
bridge: mcast: Do not derive entry type from its filter mode
...
The cited commit caused the following build break when CONFIG_IPV6 was
disabled
net/ipv4/tcp_input.c: In function ‘tcp_syn_flood_action’:
include/net/sock.h:387:37: error: ‘const struct sock_common’ has no member named ‘skc_v6_rcv_saddr’; did you mean ‘skc_rcv_saddr’?
Fix by using inet6_rcv_saddr() macro which handles this situation
nicely.
Fixes: d9282e48c6 ("tcp: Add listening address to SYN flood message")
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Reported-by: Geert Uytterhoeven <geert+renesas@glider.be>
CC: Matthieu Baerts <matthieu.baerts@tessares.net>
CC: Jamie Bainbridge <jamie.bainbridge@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20221122184158.170798-1-saeed@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
These cases were done with this Coccinelle:
@@
expression H;
expression L;
@@
- (get_random_u32_below(H) + L)
+ get_random_u32_inclusive(L, H + L - 1)
@@
expression H;
expression L;
expression E;
@@
get_random_u32_inclusive(L,
H
- + E
- - E
)
@@
expression H;
expression L;
expression E;
@@
get_random_u32_inclusive(L,
H
- - E
- + E
)
@@
expression H;
expression L;
expression E;
expression F;
@@
get_random_u32_inclusive(L,
H
- - E
+ F
- + E
)
@@
expression H;
expression L;
expression E;
expression F;
@@
get_random_u32_inclusive(L,
H
- + E
+ F
- - E
)
And then subsequently cleaned up by hand, with several automatic cases
rejected if it didn't make sense contextually.
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com> # for infiniband
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
This is a simple mechanical transformation done by:
@@
expression E;
@@
- prandom_u32_max
+ get_random_u32_below
(E)
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: Darrick J. Wong <djwong@kernel.org> # for xfs
Reviewed-by: SeongJae Park <sj@kernel.org> # for damon
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com> # for infiniband
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk> # for arm
Acked-by: Ulf Hansson <ulf.hansson@linaro.org> # for mmc
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Annotate the lockless read of queue->synflood_warned.
Following xchg() has the needed data-race resolution.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The SYN flood message prints the listening port number, but with many
processes bound to the same port on different IPs, it's impossible to
tell which socket is the problem.
Add the listen IP address to the SYN flood message.
For IPv6 use "[IP]:port" as per RFC-5952 and to provide ease of
copy-paste to "ss" filters. For IPv4 use "IP:port" to match.
Each protcol's "any" address and a host address now look like:
Possible SYN flooding on port 0.0.0.0:9001.
Possible SYN flooding on port 127.0.0.1:9001.
Possible SYN flooding on port [::]:9001.
Possible SYN flooding on port [fc00::1]:9001.
Signed-off-by: Jamie Bainbridge <jamie.bainbridge@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
Link: https://lore.kernel.org/r/4fedab7ce54a389aeadbdc639f6b4f4988e9d2d7.1668386107.git.jamie.bainbridge@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This commit fixes a bug that can cause a TCP data sender to repeatedly
defer RTOs when encountering SACK reneging.
The bug is that when we're in fast recovery in a scenario with SACK
reneging, every time we get an ACK we call tcp_check_sack_reneging()
and it can note the apparent SACK reneging and rearm the RTO timer for
srtt/2 into the future. In some SACK reneging scenarios that can
happen repeatedly until the receive window fills up, at which point
the sender can't send any more, the ACKs stop arriving, and the RTO
fires at srtt/2 after the last ACK. But that can take far too long
(O(10 secs)), since the connection is stuck in fast recovery with a
low cwnd that cannot grow beyond ssthresh, even if more bandwidth is
available.
This fix changes the logic in tcp_check_sack_reneging() to only rearm
the RTO timer if data is cumulatively ACKed, indicating forward
progress. This avoids this kind of nearly infinite loop of RTO timer
re-arming. In addition, this meets the goals of
tcp_check_sack_reneging() in handling Windows TCP behavior that looks
temporarily like SACK reneging but is not really.
Many thanks to Jakub Kicinski and Neil Spring, who reported this issue
and provided critical packet traces that enabled root-causing this
issue. Also, many thanks to Jakub Kicinski for testing this fix.
Fixes: 5ae344c949 ("tcp: reduce spurious retransmits due to transient SACK reneging")
Reported-by: Jakub Kicinski <kuba@kernel.org>
Reported-by: Neil Spring <ntspring@fb.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Tested-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20221021170821.1093930-1-ncardwell.kernel@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Fix a bug reported and analyzed by Nagaraj Arankal, where the handling
of a spurious non-SACK RTO could cause a connection to fail to clear
retrans_stamp, causing a later RTO to very prematurely time out the
connection with ETIMEDOUT.
Here is the buggy scenario, expanding upon Nagaraj Arankal's excellent
report:
(*1) Send one data packet on a non-SACK connection
(*2) Because no ACK packet is received, the packet is retransmitted
and we enter CA_Loss; but this retransmission is spurious.
(*3) The ACK for the original data is received. The transmitted packet
is acknowledged. The TCP timestamp is before the retrans_stamp,
so tcp_may_undo() returns true, and tcp_try_undo_loss() returns
true without changing state to Open (because tcp_is_sack() is
false), and tcp_process_loss() returns without calling
tcp_try_undo_recovery(). Normally after undoing a CA_Loss
episode, tcp_fastretrans_alert() would see that the connection
has returned to CA_Open and fall through and call
tcp_try_to_open(), which would set retrans_stamp to 0. However,
for non-SACK connections we hold the connection in CA_Loss, so do
not fall through to call tcp_try_to_open() and do not set
retrans_stamp to 0. So retrans_stamp is (erroneously) still
non-zero.
At this point the first "retransmission event" has passed and
been recovered from. Any future retransmission is a completely
new "event". However, retrans_stamp is erroneously still
set. (And we are still in CA_Loss, which is correct.)
(*4) After 16 minutes (to correspond with tcp_retries2=15), a new data
packet is sent. Note: No data is transmitted between (*3) and
(*4) and we disabled keep alives.
The socket's timeout SHOULD be calculated from this point in
time, but instead it's calculated from the prior "event" 16
minutes ago (step (*2)).
(*5) Because no ACK packet is received, the packet is retransmitted.
(*6) At the time of the 2nd retransmission, the socket returns
ETIMEDOUT, prematurely, because retrans_stamp is (erroneously)
too far in the past (set at the time of (*2)).
This commit fixes this bug by ensuring that we reuse in
tcp_try_undo_loss() the same careful logic for non-SACK connections
that we have in tcp_try_undo_recovery(). To avoid duplicating logic,
we factor out that logic into a new
tcp_is_non_sack_preventing_reopen() helper and call that helper from
both undo functions.
Fixes: da34ac7626 ("tcp: only undo on partial ACKs in CA_Loss")
Reported-by: Nagaraj Arankal <nagaraj.p.arankal@hpe.com>
Link: https://lore.kernel.org/all/SJ0PR84MB1847BE6C24D274C46A1B9B0EB27A9@SJ0PR84MB1847.NAMPRD84.PROD.OUTLOOK.COM/
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20220903121023.866900-1-ncardwell.kernel@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Because per host rate limiting has been proven problematic (side channel
attacks can be based on it), per host rate limiting of challenge acks ideally
should be per netns and turned off by default.
This is a long due followup of following commits:
083ae30828 ("tcp: enable per-socket rate limiting of all 'challenge acks'")
f2b2c582e8 ("tcp: mitigate ACK loops for connections as tcp_sock")
75ff39ccc1 ("tcp: make challenge acks less predictable")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Jason Baron <jbaron@akamai.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
challenge_timestamp can be read an written by concurrent threads.
This was expected, but we need to annotate the race to avoid potential issues.
Following patch moves challenge_timestamp and challenge_count
to per-netns storage to provide better isolation.
Fixes: 354e4aa391 ("tcp: RFC 5961 5.2 Blind Data Injection Attack Mitigation")
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
While reading sysctl_tcp_comp_sack_nr, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its reader.
Fixes: 9c21d2fc41 ("tcp: add tcp_comp_sack_nr sysctl")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_comp_sack_slack_ns, it can be changed
concurrently. Thus, we need to add READ_ONCE() to its reader.
Fixes: a70437cc09 ("tcp: add hrtimer slack to sack compression")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_comp_sack_delay_ns, it can be changed
concurrently. Thus, we need to add READ_ONCE() to its reader.
Fixes: 6d82aa2420 ("tcp: add tcp_comp_sack_delay_ns sysctl")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading these sysctl variables, they can be changed concurrently.
Thus, we need to add READ_ONCE() to their readers.
- .sysctl_rmem
- .sysctl_rwmem
- .sysctl_rmem_offset
- .sysctl_wmem_offset
- sysctl_tcp_rmem[1, 2]
- sysctl_tcp_wmem[1, 2]
- sysctl_decnet_rmem[1]
- sysctl_decnet_wmem[1]
- sysctl_tipc_rmem[1]
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_pacing_(ss|ca)_ratio, they can be changed
concurrently. Thus, we need to add READ_ONCE() to their readers.
Fixes: 43e122b014 ("tcp: refine pacing rate determination")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_invalid_ratelimit, it can be changed
concurrently. Thus, we need to add READ_ONCE() to its reader.
Fixes: 032ee42369 ("tcp: helpers to mitigate ACK loops by rate-limiting out-of-window dupacks")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_min_rtt_wlen, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its reader.
Fixes: f672258391 ("tcp: track min RTT using windowed min-filter")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_challenge_ack_limit, it can be changed
concurrently. Thus, we need to add READ_ONCE() to its reader.
Fixes: 282f23c6ee ("tcp: implement RFC 5961 3.2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_moderate_rcvbuf, it can be changed
concurrently. Thus, we need to add READ_ONCE() to its readers.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_frto, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its reader.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_app_win, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its reader.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_dsack, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_max_reordering, it can be changed
concurrently. Thus, we need to add READ_ONCE() to its readers.
Fixes: dca145ffaa ("tcp: allow for bigger reordering level")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_stdurg, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its reader.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_recovery, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
Fixes: 4f41b1c58a ("tcp: use RACK to detect losses")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading these knobs, they can be changed concurrently.
Thus, we need to add READ_ONCE() to their readers.
- tcp_sack
- tcp_window_scaling
- tcp_timestamps
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_max_syn_backlog, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_reordering, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_syncookies, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_ecn, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann says:
====================
pull-request: bpf-next 2022-06-17
We've added 72 non-merge commits during the last 15 day(s) which contain
a total of 92 files changed, 4582 insertions(+), 834 deletions(-).
The main changes are:
1) Add 64 bit enum value support to BTF, from Yonghong Song.
2) Implement support for sleepable BPF uprobe programs, from Delyan Kratunov.
3) Add new BPF helpers to issue and check TCP SYN cookies without binding to a
socket especially useful in synproxy scenarios, from Maxim Mikityanskiy.
4) Fix libbpf's internal USDT address translation logic for shared libraries as
well as uprobe's symbol file offset calculation, from Andrii Nakryiko.
5) Extend libbpf to provide an API for textual representation of the various
map/prog/attach/link types and use it in bpftool, from Daniel Müller.
6) Provide BTF line info for RV64 and RV32 JITs, and fix a put_user bug in the
core seen in 32 bit when storing BPF function addresses, from Pu Lehui.
7) Fix libbpf's BTF pointer size guessing by adding a list of various aliases
for 'long' types, from Douglas Raillard.
8) Fix bpftool to readd setting rlimit since probing for memcg-based accounting
has been unreliable and caused a regression on COS, from Quentin Monnet.
9) Fix UAF in BPF cgroup's effective program computation triggered upon BPF link
detachment, from Tadeusz Struk.
10) Fix bpftool build bootstrapping during cross compilation which was pointing
to the wrong AR process, from Shahab Vahedi.
11) Fix logic bug in libbpf's is_pow_of_2 implementation, from Yuze Chi.
12) BPF hash map optimization to avoid grabbing spinlocks of all CPUs when there
is no free element. Also add a benchmark as reproducer, from Feng Zhou.
13) Fix bpftool's codegen to bail out when there's no BTF, from Michael Mullin.
14) Various minor cleanup and improvements all over the place.
* https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next: (72 commits)
bpf: Fix bpf_skc_lookup comment wrt. return type
bpf: Fix non-static bpf_func_proto struct definitions
selftests/bpf: Don't force lld on non-x86 architectures
selftests/bpf: Add selftests for raw syncookie helpers in TC mode
bpf: Allow the new syncookie helpers to work with SKBs
selftests/bpf: Add selftests for raw syncookie helpers
bpf: Add helpers to issue and check SYN cookies in XDP
bpf: Allow helpers to accept pointers with a fixed size
bpf: Fix documentation of th_len in bpf_tcp_{gen,check}_syncookie
selftests/bpf: add tests for sleepable (uk)probes
libbpf: add support for sleepable uprobe programs
bpf: allow sleepable uprobe programs to attach
bpf: implement sleepable uprobes by chaining gps
bpf: move bpf_prog to bpf.h
libbpf: Fix internal USDT address translation logic for shared libraries
samples/bpf: Check detach prog exist or not in xdp_fwd
selftests/bpf: Avoid skipping certain subtests
selftests/bpf: Fix test_varlen verification failure with latest llvm
bpftool: Do not check return value from libbpf_set_strict_mode()
Revert "bpftool: Use libbpf 1.0 API mode instead of RLIMIT_MEMLOCK"
...
====================
Link: https://lore.kernel.org/r/20220617220836.7373-1-daniel@iogearbox.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The new helpers bpf_tcp_raw_{gen,check}_syncookie_ipv{4,6} allow an XDP
program to generate SYN cookies in response to TCP SYN packets and to
check those cookies upon receiving the first ACK packet (the final
packet of the TCP handshake).
Unlike bpf_tcp_{gen,check}_syncookie these new helpers don't need a
listening socket on the local machine, which allows to use them together
with synproxy to accelerate SYN cookie generation.
Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://lore.kernel.org/r/20220615134847.3753567-4-maximmi@nvidia.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Currently, tcp_memory_allocated can hit tcp_mem[] limits quite fast.
Each TCP socket can forward allocate up to 2 MB of memory, even after
flow became less active.
10,000 sockets can have reserved 20 GB of memory,
and we have no shrinker in place to reclaim that.
Instead of trying to reclaim the extra allocations in some places,
just keep sk->sk_forward_alloc values as small as possible.
This should not impact performance too much now we have per-cpu
reserves: Changes to tcp_memory_allocated should not be too frequent.
For sockets not using SO_RESERVE_MEM:
- idle sockets (no packets in tx/rx queues) have zero forward alloc.
- non idle sockets have a forward alloc smaller than one page.
Note:
- Removal of SK_RECLAIM_CHUNK and SK_RECLAIM_THRESHOLD
is left to MPTCP maintainers as a follow up.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Due to memcg interface, SK_MEM_QUANTUM is effectively PAGE_SIZE.
This might change in the future, but it seems better to avoid the
confusion.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This patch improves TCP PRR loss recovery behavior for a corner
case. Previously during PRR conservation-bound mode, it strictly
sends the amount equals to the amount newly acked or s/acked.
The patch changes s.t. PRR may send additional amount that was banked
previously (e.g. application-limited) in the conservation-bound
mode, similar to the slow-start mode. This unifies and simplifies the
algorithm further and may improve the recovery latency. This change
still follow the general packet conservation design principle and
always keep inflight/cwnd below the slow start threshold set
by the congestion control module.
PRR is described in RFC 6937. We'll include this change in the
latest revision rfc6937-bis as well.
Reported-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20220519003410.2531936-1-ycheng@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In commit f84af32cbc ("net: ip_queue_rcv_skb() helper")
I dropped the skb dst in tcp_data_queue().
This only dealt with so-called TCP input slow path.
When fast path is taken, tcp_rcv_established() calls
tcp_queue_rcv() while skb still has a dst.
This was mostly fine, because most dsts at this point
are not refcounted (thanks to early demux)
However, TCP packets sent over loopback have refcounted dst.
Then commit 68822bdf76 ("net: generalize skb freeing
deferral to per-cpu lists") came and had the effect
of delaying skb freeing for an arbitrary time.
If during this time the involved netns is dismantled, cleanup_net()
frees the struct net with embedded net->ipv6.ip6_dst_ops.
Then when eventually dst_destroy_rcu() is called,
if (dst->ops->destroy) ... triggers an use-after-free.
It is not clear if ip6_route_net_exit() lacks a rcu_barrier()
as syzbot reported similar issues before the blamed commit.
( https://groups.google.com/g/syzkaller-bugs/c/CofzW4eeA9A/m/009WjumTAAAJ )
Fixes: 68822bdf76 ("net: generalize skb freeing deferral to per-cpu lists")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
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
0cbe6a8f08 ("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: c9bee3b7fd ("tcp: TCP_NOTSENT_LOWAT socket option")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Doug Porter <dsp@fb.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Kernel test robot reported:
smatch warnings:
net/ipv4/tcp_input.c:5966 tcp_rcv_established() warn: unsigned 'reason' is never less than zero.
I actually had one packetdrill failing because of this bug,
and was about to send the fix :)
v2: Andreas Schwab also pointed out that @reason needs to be negated
before we reach tcp_drop_reason()
Fixes: 4b506af9c5 ("tcp: add two drop reasons for tcp_ack()")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Andreas Schwab <schwab@linux-m68k.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
packets in OFO queue might be redundant, and dropped.
tcp_drop() is no longer needed.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
1) A valid RST packet should be consumed, to not confuse drop monitor.
2) Same remark for packet validating cross syn setup,
even if we might ignore part of it.
3) When third packet of 3WHS is delayed, do not pretend
the SYNACK was dropped.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add one reason for packets dropped from OFO queue because
of memory pressure.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add TCP_TOO_OLD_ACK and TCP_ACK_UNSENT_DATA drop
reasons so that tcp_rcv_established() can report
them.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add basic support for drop reasons in tcp_rcv_state_process()
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
tcp_rcv_state_process() incorrectly drops packets
instead of consuming it, making drop monitor very noisy,
if not unusable.
Calling tcp_time_wait() or tcp_done() is part
of standard behavior, packets triggering these actions
were not dropped.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Creates four new drop reasons for the following cases:
1) packet being rejected by RFC 7323 PAWS check
2) packet being rejected by SEQUENCE check
3) Invalid RST packet
4) Invalid SYN packet
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Small cleanup in tcp_validate_incoming(), no need for rst_seq_match
setting and testing.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Whenever tcp_validate_incoming() handles a valid RST packet,
we should not pretend the packet was dropped.
Create a special section at the end of tcp_validate_incoming()
to handle this case.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We had various bugs over the years with code
breaking the assumption that tp->snd_cwnd is greater
than zero.
Lately, syzbot reported the WARN_ON_ONCE(!tp->prior_cwnd) added
in commit 8b8a321ff7 ("tcp: fix zero cwnd in tcp_cwnd_reduction")
can trigger, and without a repro we would have to spend
considerable time finding the bug.
Instead of complaining too late, we want to catch where
and when tp->snd_cwnd is set to an illegal value.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Suggested-by: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Link: https://lore.kernel.org/r/20220405233538.947344-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Replace tcp_drop() used in tcp_data_queue_ofo with tcp_drop_reason().
Following drop reasons are introduced:
SKB_DROP_REASON_TCP_OFOMERGE
Reviewed-by: Mengen Sun <mengensun@tencent.com>
Reviewed-by: Hao Peng <flyingpeng@tencent.com>
Signed-off-by: Menglong Dong <imagedong@tencent.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Replace tcp_drop() used in tcp_data_queue() with tcp_drop_reason().
Following drop reasons are introduced:
SKB_DROP_REASON_TCP_ZEROWINDOW
SKB_DROP_REASON_TCP_OLD_DATA
SKB_DROP_REASON_TCP_OVERWINDOW
SKB_DROP_REASON_TCP_OLD_DATA is used for the case that end_seq of skb
less than the left edges of receive window. (Maybe there is a better
name?)
Reviewed-by: Mengen Sun <mengensun@tencent.com>
Reviewed-by: Hao Peng <flyingpeng@tencent.com>
Signed-off-by: Menglong Dong <imagedong@tencent.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Replace tcp_drop() used in tcp_rcv_established() with tcp_drop_reason().
Following drop reasons are added:
SKB_DROP_REASON_TCP_FLAGS
Reviewed-by: Mengen Sun <mengensun@tencent.com>
Reviewed-by: Hao Peng <flyingpeng@tencent.com>
Signed-off-by: Menglong Dong <imagedong@tencent.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
For TCP protocol, tcp_drop() is used to free the skb when it needs
to be dropped. To make use of kfree_skb_reason() and pass the drop
reason to it, introduce the function tcp_drop_reason(). Meanwhile,
make tcp_drop() an inline call to tcp_drop_reason().
Reviewed-by: Mengen Sun <mengensun@tencent.com>
Reviewed-by: Hao Peng <flyingpeng@tencent.com>
Signed-off-by: Menglong Dong <imagedong@tencent.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch intends to provide a mechanism to put constraint on SMC
connections visit according to the pressure of SMC handshake process.
At present, frequent visits will cause the incoming connections to be
backlogged in SMC handshake queue, raise the connections established
time. Which is quite unacceptable for those applications who base on
short lived connections.
There are two ways to implement this mechanism:
1. Put limitation after TCP established.
2. Put limitation before TCP established.
In the first way, we need to wait and receive CLC messages that the
client will potentially send, and then actively reply with a decline
message, in a sense, which is also a sort of SMC handshake, affect the
connections established time on its way.
In the second way, the only problem is that we need to inject SMC logic
into TCP when it is about to reply the incoming SYN, since we already do
that, it's seems not a problem anymore. And advantage is obvious, few
additional processes are required to complete the constraint.
This patch use the second way. After this patch, connections who beyond
constraint will not informed any SMC indication, and SMC will not be
involved in any of its subsequent processes.
Link: https://lore.kernel.org/all/1641301961-59331-1-git-send-email-alibuda@linux.alibaba.com/
Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
tcp_shift_skb_data() might collapse three packets into a larger one.
P_A, P_B, P_C -> P_ABC
Historically, it used a single tcp_skb_can_collapse_to(P_A) call,
because it was enough.
In commit 8571248411 ("tcp: coalesce/collapse must respect MPTCP extensions"),
this call was replaced by a call to tcp_skb_can_collapse(P_A, P_B)
But the now needed test over P_C has been missed.
This probably broke MPTCP.
Then later, commit 9b65b17db7 ("net: avoid double accounting for pure zerocopy skbs")
added an extra condition to tcp_skb_can_collapse(), but the missing call
from tcp_shift_skb_data() is also breaking TCP zerocopy, because P_A and P_C
might have different skb_zcopy_pure() status.
Fixes: 8571248411 ("tcp: coalesce/collapse must respect MPTCP extensions")
Fixes: 9b65b17db7 ("net: avoid double accounting for pure zerocopy skbs")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Mat Martineau <mathew.j.martineau@linux.intel.com>
Cc: Talal Ahmad <talalahmad@google.com>
Cc: Arjun Roy <arjunroy@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Link: https://lore.kernel.org/r/20220201184640.756716-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When setting RTO through BPF program, some SYN ACK packets were unaffected
and continued to use TCP_TIMEOUT_INIT constant. This patch adds timeout
option to struct request_sock. Option is initialized with TCP_TIMEOUT_INIT
and is reassigned through BPF using tcp_timeout_init call. SYN ACK
retransmits now use newly added timeout option.
Signed-off-by: Akhmat Karakotov <hmukos@yandex-team.ru>
Acked-by: Martin KaFai Lau <kafai@fb.com>
v2:
- Add timeout option to struct request_sock. Do not call
tcp_timeout_init on every syn ack retransmit.
v3:
- Use unsigned long for min. Bound tcp_timeout_init to TCP_RTO_MAX.
v4:
- Refactor duplicate code by adding reqsk_timeout function.
Signed-off-by: David S. Miller <davem@davemloft.net>
After this parameter is passed in, there is no usage, and deleting it will
not bring any impact.
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Benjamin Yim <yan2228598786@gmail.com>
Link: https://lore.kernel.org/r/20220109130824.2776-1-yan2228598786@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
tcp_poll() and tcp_ioctl() are reading tp->urg_data without socket lock
owned.
Also, it is faster to first check tp->urg_data in tcp_poll(),
then tp->urg_seq == tp->copied_seq, because tp->urg_seq is
located in a different/cold cache line.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When user sets SO_RESERVE_MEM socket option, in order to utilize the
reserved memory when in memory pressure state, we adjust rcv_ssthresh
according to the available reserved memory for the socket, instead of
using 4 * advmss always.
Signed-off-by: Wei Wang <weiwan@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If user sets SO_RESERVE_MEM socket option, in order to fully utilize the
reserved memory in memory pressure state on the tx path, we modify the
logic in sk_stream_moderate_sndbuf() to set sk_sndbuf according to
available reserved memory, instead of MIN_SOCK_SNDBUF, and adjust it
when new data is acked.
Signed-off-by: Wei Wang <weiwan@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to track CE marks per rate sample (one round trip), TCP needs a
per-skb header field to record the tp->delivered_ce count when the skb
was sent. To make space, we replace the "last_in_flight" field which is
used exclusively for NV congestion control. The stat needed by NV can be
alternatively approximated by existing stats tcp_sock delivered and
mss_cache.
This patch counts the number of packets delivered which have CE marks in
the rate sample, using similar approach of delivery accounting.
Cc: Lawrence Brakmo <brakmo@fb.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Luke Hsiao <lukehsiao@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
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>
Previously, a DSACK could expand the RACK reordering window when no
reordering has been seen, and/or when the DSACK was due to an
unnecessary TLP retransmit (rather than a spurious fast recovery due
to reordering). This could result in unnecessarily growing the RACK
reordering window and thus unnecessarily delaying RACK-based fast
recovery episodes.
To avoid these issues, this commit tightens the conditions under which
a DSACK triggers the RACK reordering window to grow, so that a
connection only expands its RACK reordering window if:
(a) reordering has been seen in the connection
(b) a DSACKed range does not match the most recent TLP retransmit
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Priyaranjan Jha <priyarjha@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Previously TLP is considered spurious if the sender receives any
DSACK during a TLP episode. This patch further checks the DSACK
sequences match the TLP's to improve accuracy.
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Priyaranjan Jha <priyarjha@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
tcp_grow_window() is using skb->len/skb->truesize to increase tp->rcv_ssthresh
which has a direct impact on advertized window sizes.
We added TCP coalescing in linux-3.4 & linux-3.5:
Instead of storing skbs with one or two MSS in receive queue (or OFO queue),
we try to append segments together to reduce memory overhead.
High performance network drivers tend to cook skb with 3 parts :
1) sk_buff structure (256 bytes)
2) skb->head contains room to copy headers as needed, and skb_shared_info
3) page fragment(s) containing the ~1514 bytes frame (or more depending on MTU)
Once coalesced into a previous skb, 1) and 2) are freed.
We can therefore tweak the way we compute len/truesize ratio knowing
that skb->truesize is inflated by 1) and 2) soon to be freed.
This is done only for in-order skb, or skb coalesced into OFO queue.
The result is that low rate flows no longer pay the memory price of having
low GRO aggregation factor. Same result for drivers not using GRO.
This is critical to allow a big enough receiver window,
typically tcp_rmem[2] / 2.
We have been using this at Google for about 5 years, it is due time
to make it upstream.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For tcp sockets, sk->sk_write_space is most probably sk_stream_write_space().
Other sk->sk_write_space() calls in TCP are slow path and do not deserve
any change.
Signed-off-by: Eric Dumazet <edumazet@google.com>
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>
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>
This patch introduces a function wrapper to call the sk_error_report
callback. That will prepare to add additional handling whenever
sk_error_report is called, for example to trace socket errors.
Signed-off-by: Alexander Aring <aahringo@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch aims to improve the situation when reordering and loss are
ocurring in the same flight of packets.
Previously the reordering would first induce a spurious recovery, then
the subsequent ACK may undo the cwnd (based on the timestamps e.g.).
However the current loss recovery does not proceed to invoke
RACK to install a reordering timer. If some packets are also lost, this
may lead to a long RTO-based recovery. An example is
https://groups.google.com/g/bbr-dev/c/OFHADvJbTEI
The solution is to after reverting the recovery, always invoke RACK
to either mount the RACK timer to fast retransmit after the reordering
window, or restarts the recovery if new loss is identified. Hence
it is possible the sender may go from Recovery to Disorder/Open to
Recovery again in one ACK.
Reported-by: mingkun bian <bianmingkun@gmail.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a tracepoint for capturing TCP segments with
a bad checksum. This makes it easy to identify
sources of bad frames in the fleet (e.g. machines
with faulty NICs).
It should also help tools like IOvisor's tcpdrop.py
which are used today to get detailed information
about such packets.
We don't have a socket in many cases so we must
open code the address extraction based just on
the skb.
v2: add missing export for ipv6=m
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jakub reported Data included in a Fastopen SYN that had to be
retransmit would have to wait for an RTO if TX completions are slow,
even with prior fix.
This is because tcp_rcv_fastopen_synack() does not use standard
rtx logic, meaning TSQ handler exits early in tcp_tsq_write()
because tp->lost_out == tp->retrans_out
Lets make tcp_rcv_fastopen_synack() use standard rtx logic,
by using tcp_mark_skb_lost() on the skb thats needs to be
sent again.
Not this raised a warning in tcp_fastretrans_alert() during my tests
since we consider the data not being aknowledged
by the receiver does not mean packet was lost on the network.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Jakub Kicinski <kuba@kernel.org>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
My prior cleanup missed that tcp_data_ready() has to look at SOCK_DONE.
Otherwise, an application using SO_RCVLOWAT will not get EPOLLIN event
if a FIN is received in the middle of expected payload.
The reason SOCK_DONE is not examined in tcp_epollin_ready()
is that tcp_poll() catches the FIN because tcp_fin()
is also setting RCV_SHUTDOWN into sk->sk_shutdown
Fixes: 05dc72aba3 ("tcp: factorize logic into tcp_epollin_ready()")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Wei Wang <weiwan@google.com>
Cc: Arjun Roy <arjunroy@google.com>
Reviewed-by: Wei Wang <weiwan@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Both tcp_data_ready() and tcp_stream_is_readable() share the same logic.
Add tcp_epollin_ready() helper to avoid duplication.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Arjun Roy <arjunroy@google.com>
Cc: Wei Wang <weiwan@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
drivers/net/can/dev.c
b552766c87 ("can: dev: prevent potential information leak in can_fill_info()")
3e77f70e73 ("can: dev: move driver related infrastructure into separate subdir")
0a042c6ec9 ("can: dev: move netlink related code into seperate file")
Code move.
drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
57ac4a31c4 ("net/mlx5e: Correctly handle changing the number of queues when the interface is down")
214baf2287 ("net/mlx5e: Support HTB offload")
Adjacent code changes
net/switchdev/switchdev.c
20776b465c ("net: switchdev: don't set port_obj_info->handled true when -EOPNOTSUPP")
ffb68fc58e ("net: switchdev: remove the transaction structure from port object notifiers")
bae33f2b5a ("net: switchdev: remove the transaction structure from port attributes")
Transaction parameter gets dropped otherwise keep the fix.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Upon receiving a cumulative ACK that changes the congestion state from
Disorder to Open, the TLP timer is not set. If the sender is app-limited,
it can only wait for the RTO timer to expire and retransmit.
The reason for this is that the TLP timer is set before the congestion
state changes in tcp_ack(), so we delay the time point of calling
tcp_set_xmit_timer() until after tcp_fastretrans_alert() returns and
remove the FLAG_SET_XMIT_TIMER from ack_flag when the RACK reorder timer
is set.
This commit has two additional benefits:
1) Make sure to reset RTO according to RFC6298 when receiving ACK, to
avoid spurious RTO caused by RTO timer early expires.
2) Reduce the xmit timer reschedule once per ACK when the RACK reorder
timer is set.
Fixes: df92c8394e ("tcp: fix xmit timer to only be reset if data ACKed/SACKed")
Link: https://lore.kernel.org/netdev/1611311242-6675-1-git-send-email-yangpc@wangsu.com
Signed-off-by: Pengcheng Yang <yangpc@wangsu.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Cc: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/1611464834-23030-1-git-send-email-yangpc@wangsu.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The TCP_USER_TIMEOUT is checked by the 0-window probe timer. As the
timer has backoff with a max interval of about two minutes, the
actual timeout for TCP_USER_TIMEOUT can be off by up to two minutes.
In this patch the TCP_USER_TIMEOUT is made more accurate by taking it
into account when computing the timer value for the 0-window probes.
This patch is similar to and builds on top of the one that made
TCP_USER_TIMEOUT accurate for RTOs in commit b701a99e43 ("tcp: Add
tcp_clamp_rto_to_user_timeout() helper to improve accuracy").
Fixes: 9721e709fa ("tcp: simplify window probe aborting on USER_TIMEOUT")
Signed-off-by: Enke Chen <enchen@paloaltonetworks.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20210122191306.GA99540@localhost.localdomain
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This patch adds TCP_NLA_TTL to SCM_TIMESTAMPING_OPT_STATS that exports
the time-to-live or hop limit of the latest incoming packet with
SCM_TSTAMP_ACK. The value exported may not be from the packet that acks
the sequence when incoming packets are aggregated. Exporting the
time-to-live or hop limit value of incoming packets helps to estimate
the hop count of the path of the flow that may change over time.
Signed-off-by: Yousuk Seung <ysseung@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Link: https://lore.kernel.org/r/20210120204155.552275-1-ysseung@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The previous commit 32efcc06d2 ("tcp: export count for rehash attempts")
would mis-account rehashing SNMP and socket stats:
a. During handshake of an active open, only counts the first
SYN timeout
b. After handshake of passive and active open, stop updating
after (roughly) TCP_RETRIES1 recurring RTOs
c. After the socket aborts, over count timeout_rehash by 1
This patch fixes this by checking the rehash result from sk_rethink_txhash.
Fixes: 32efcc06d2 ("tcp: export count for rehash attempts")
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Link: https://lore.kernel.org/r/20210119192619.1848270-1-ycheng@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The TCP session does not terminate with TCP_USER_TIMEOUT when data
remain untransmitted due to zero window.
The number of unanswered zero-window probes (tcp_probes_out) is
reset to zero with incoming acks irrespective of the window size,
as described in tcp_probe_timer():
RFC 1122 4.2.2.17 requires the sender to stay open indefinitely
as long as the receiver continues to respond probes. We support
this by default and reset icsk_probes_out with incoming ACKs.
This counter, however, is the wrong one to be used in calculating the
duration that the window remains closed and data remain untransmitted.
Thanks to Jonathan Maxwell <jmaxwell37@gmail.com> for diagnosing the
actual issue.
In this patch a new timestamp is introduced for the socket in order to
track the elapsed time for the zero-window probes that have not been
answered with any non-zero window ack.
Fixes: 9721e709fa ("tcp: simplify window probe aborting on USER_TIMEOUT")
Reported-by: William McCall <william.mccall@gmail.com>
Co-developed-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Enke Chen <enchen@paloaltonetworks.com>
Reviewed-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20210115223058.GA39267@localhost.localdomain
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
There are cases where a fastopen SYN may trigger either a ICMP_TOOBIG
message in the case of IPv6 or a fragmentation request in the case of
IPv4. This results in the socket stalling for a second or more as it does
not respond to the message by retransmitting the SYN frame.
Normally a SYN frame should not be able to trigger a ICMP_TOOBIG or
ICMP_FRAG_NEEDED however in the case of fastopen we can have a frame that
makes use of the entire MSS. In the case of fastopen it does, and an
additional complication is that the retransmit queue doesn't contain the
original frames. As a result when tcp_simple_retransmit is called and
walks the list of frames in the queue it may not mark the frames as lost
because both the SYN and the data packet each individually are smaller than
the MSS size after the adjustment. This results in the socket being stalled
until the retransmit timer kicks in and forces the SYN frame out again
without the data attached.
In order to resolve this we can reduce the MSS the packets are compared
to in tcp_simple_retransmit to -1 for cases where we are still in the
TCP_SYN_SENT state for a fastopen socket. Doing this we will mark all of
the packets related to the fastopen SYN as lost.
Signed-off-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Link: https://lore.kernel.org/r/160780498125.3272.15437756269539236825.stgit@localhost.localdomain
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Because TCP-level resets only affect the subflow, there is a MPTCP
option to indicate that the MPTCP-level connection should be closed
immediately without a mptcp-level fin exchange.
This is the 'MPTCP fast close option'. It can be carried on ack
segments or TCP resets. In the latter case, its needed to parse mptcp
options also for reset packets so that MPTCP can act accordingly.
Next patch will add receive side fastclose support in MPTCP.
Acked-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
xdp_return_frame_bulk() needs to pass a xdp_buff
to __xdp_return().
strlcpy got converted to strscpy but here it makes no
functional difference, so just keep the right code.
Conflicts:
net/netfilter/nf_tables_api.c
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Before commit a337531b94 ("tcp: up initial rmem to 128KB and SYN rwin to around 64KB")
small tcp_rmem[1] values were overridden by tcp_fixup_rcvbuf() to accommodate various MSS.
This is no longer the case, and Hazem Mohamed Abuelfotoh reported
that DRS would not work for MTU 9000 endpoints receiving regular (1500 bytes) frames.
Root cause is that tcp_init_buffer_space() uses tp->rcv_wnd for upper limit
of rcvq_space.space computation, while it can select later a smaller
value for tp->rcv_ssthresh and tp->window_clamp.
ss -temoi on receiver would show :
skmem:(r0,rb131072,t0,tb46080,f0,w0,o0,bl0,d0) rcv_space:62496 rcv_ssthresh:56596
This means that TCP can not increase its window in tcp_grow_window(),
and that DRS can never kick.
Fix this by making sure that rcvq_space.space is not bigger than number of bytes
that can be held in TCP receive queue.
People unable/unwilling to change their kernel can work around this issue by
selecting a bigger tcp_rmem[1] value as in :
echo "4096 196608 6291456" >/proc/sys/net/ipv4/tcp_rmem
Based on an initial report and patch from Hazem Mohamed Abuelfotoh
https://lore.kernel.org/netdev/20201204180622.14285-1-abuehaze@amazon.com/
Fixes: a337531b94 ("tcp: up initial rmem to 128KB and SYN rwin to around 64KB")
Fixes: 041a14d267 ("tcp: start receiver buffer autotuning sooner")
Reported-by: Hazem Mohamed Abuelfotoh <abuehaze@amazon.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The Multipath-TCP standard (RFC 8684) says that an MPTCP host should send
a TCP reset if the token in a MP_JOIN request is unknown.
At this time we don't do this, the 3whs completes and the 'new subflow'
is reset afterwards. There are two ways to allow MPTCP to send the
reset.
1. override 'send_synack' callback and emit the rst from there.
The drawback is that the request socket gets inserted into the
listeners queue just to get removed again right away.
2. Send the reset from the 'route_req' function instead.
This avoids the 'add&remove request socket', but route_req lacks the
skb that is required to send the TCP reset.
Instead of just adding the skb to that function for MPTCP sake alone,
Paolo suggested to merge init_req and route_req functions.
This saves one indirection from syn processing path and provides the skb
to the merged function at the same time.
'send reset on unknown mptcp join token' is added in next patch.
Suggested-by: Paolo Abeni <pabeni@redhat.com>
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
During TCP fast recovery, the congestion control in charge is by
default the Proportional Rate Reduction (PRR) unless the congestion
control module specified otherwise (e.g. BBR).
Previously when tcp_packets_in_flight() is below snd_ssthresh PRR
would slow start upon receiving an ACK that
1) cumulatively acknowledges retransmitted data
and
2) does not detect further lost retransmission
Such conditions indicate the repair is in good steady progress
after the first round trip of recovery. Otherwise PRR adopts the
packet conservation principle to send only the amount that was
newly delivered (indicated by this ACK).
This patch generalizes the previous design principle to include
also the newly sent data beside retransmission: as long as
the delivery is making good progress, both retransmission and
new data should be accounted to make PRR more cautious in slow
starting.
Suggested-by: Matt Mathis <mattmathis@google.com>
Suggested-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20201031013412.1973112-1-ycheng@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
With SO_RCVLOWAT, under memory pressure,
it is possible to enter a state where:
1. We have not received enough bytes to satisfy SO_RCVLOWAT.
2. We have not entered buffer pressure (see tcp_rmem_pressure()).
3. But, we do not have enough buffer space to accept more packets.
In this case, we advertise 0 rwnd (due to #3) but the application does
not drain the receive queue (no wakeup because of #1 and #2) so the
flow stalls.
Modify the heuristic for SO_RCVLOWAT so that, if we are advertising
rwnd<=rcv_mss, force a wakeup to prevent a stall.
Without this patch, setting tcp_rmem to 6143 and disabling TCP
autotune causes a stalled flow. With this patch, no stall occurs. This
is with RPC-style traffic with large messages.
Fixes: 03f45c883c ("tcp: avoid extra wakeups for SO_RCVLOWAT users")
Signed-off-by: Arjun Roy <arjunroy@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20201023184709.217614-1-arjunroy.kdev@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In the header prediction fast path for a bulk data receiver, if no
data is newly acknowledged then we do not call tcp_ack() and do not
call tcp_ack_update_window(). This means that a bulk receiver that
receives large amounts of data can have the incoming sequence numbers
wrap, so that the check in tcp_may_update_window fails:
after(ack_seq, tp->snd_wl1)
If the incoming receive windows are zero in this state, and then the
connection that was a bulk data receiver later wants to send data,
that connection can find itself persistently rejecting the window
updates in incoming ACKs. This means the connection can persistently
fail to discover that the receive window has opened, which in turn
means that the connection is unable to send anything, and the
connection's sending process can get permanently "stuck".
The fix is to update snd_wl1 in the header prediction fast path for a
bulk data receiver, so that it keeps up and does not see wrapping
problems.
This fix is based on a very nice and thorough analysis and diagnosis
by Apollon Oikonomopoulos (see link below).
This is a stable candidate but there is no Fixes tag here since the
bug predates current git history. Just for fun: looks like the bug
dates back to when header prediction was added in Linux v2.1.8 in Nov
1996. In that version tcp_rcv_established() was added, and the code
only updates snd_wl1 in tcp_ack(), and in the new "Bulk data transfer:
receiver" code path it does not call tcp_ack(). This fix seems to
apply cleanly at least as far back as v3.2.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Reported-by: Apollon Oikonomopoulos <apoikos@dmesg.gr>
Tested-by: Apollon Oikonomopoulos <apoikos@dmesg.gr>
Link: https://www.spinics.net/lists/netdev/msg692430.html
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20201022143331.1887495-1-ncardwell.kernel@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Replace commas with semicolons. Commas introduce unnecessary
variability in the code structure and are hard to see. What is done
is essentially described by the following Coccinelle semantic patch
(http://coccinelle.lip6.fr/):
// <smpl>
@@ expression e1,e2; @@
e1
-,
+;
e2
... when any
// </smpl>
Signed-off-by: Julia Lawall <Julia.Lawall@inria.fr>
Link: https://lore.kernel.org/r/1602412498-32025-4-git-send-email-Julia.Lawall@inria.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Rejecting non-native endian BTF overlapped with the addition
of support for it.
The rest were more simple overlapping changes, except the
renesas ravb binding update, which had to follow a file
move as well as a YAML conversion.
Signed-off-by: David S. Miller <davem@davemloft.net>
The retransmission refactoring patch
686989700c ("tcp: simplify tcp_mark_skb_lost")
does not properly update the total lost packet counter which may
break the policer mode in BBR. This patch fixes it.
Fixes: 686989700c ("tcp: simplify tcp_mark_skb_lost")
Reported-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
tcp_skb_mark_lost is used by RFC6675-SACK and can easily be replaced
with the new tcp_mark_skb_lost handler.
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch consolidates and simplifes the loss marking logic used
by a few loss detections (RACK, RFC6675, NewReno). Previously
each detection uses a subset of several intertwined subroutines.
This unncessary complexity has led to bugs (and fixes of bug fixes).
tcp_mark_skb_lost now is the single one routine to mark a packet loss
when a loss detection caller deems an skb ist lost:
1. rewind tp->retransmit_hint_skb if skb has lower sequence or
all lost ones have been retransmitted.
2. book-keeping: adjust flags and counts depending on if skb was
retransmitted or not.
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A pure refactor to move tcp_mark_skb_lost to tcp_input.c to prepare
for the later loss marking consolidation.
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
tcp_simple_retransmit() used for path MTU discovery may not adjust
the retransmit hint properly by deducting retrans_out before checking
it to adjust the hint. This patch fixes this by a correct routine
tcp_mark_skb_lost() already used by the RACK loss detection.
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since commit cfde141ea3 ("mptcp: move option parsing into
mptcp_incoming_options()"), the 3rd function argument is no longer used.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, we use length of DSACKed range to compute number of
delivered packets. And if sequence range in DSACK is corrupted,
we can get bogus dsacked/acked count, and bogus cwnd.
This patch put bounds on DSACKed range to skip update of data
delivery and spurious retransmission information, if the DSACK
is unlikely caused by sender's action:
- DSACKed range shouldn't be greater than maximum advertised rwnd.
- Total no. of DSACKed segments shouldn't be greater than total
no. of retransmitted segs. Unlike spurious retransmits, network
duplicates or corrupted DSACKs shouldn't be counted as delivery.
Signed-off-by: Priyaranjan Jha <priyarjha@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Alexei Starovoitov says:
====================
pull-request: bpf-next 2020-09-23
The following pull-request contains BPF updates for your *net-next* tree.
We've added 95 non-merge commits during the last 22 day(s) which contain
a total of 124 files changed, 4211 insertions(+), 2040 deletions(-).
The main changes are:
1) Full multi function support in libbpf, from Andrii.
2) Refactoring of function argument checks, from Lorenz.
3) Make bpf_tail_call compatible with functions (subprograms), from Maciej.
4) Program metadata support, from YiFei.
5) bpf iterator optimizations, from Yonghong.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
SOCK_QUEUE_SHRUNK is currently used by TCP as a temporary state
that remembers if some room has been made in the rtx queue
by an incoming ACK packet.
This is later used from tcp_check_space() before
considering to send EPOLLOUT.
Problem is: If we receive SACK packets, and no packet
is removed from RTX queue, we can send fresh packets, thus
moving them from write queue to rtx queue and eventually
empty the write queue.
This stall can happen if TCP_NOTSENT_LOWAT is used.
With this fix, we no longer risk stalling sends while holes
are repaired, and we can fully use socket sndbuf.
This also removes a cache line dirtying for typical RPC
workloads.
Fixes: c9bee3b7fd ("tcp: TCP_NOTSENT_LOWAT socket option")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
