The introduction of IPPROTO_SMC enables eBPF programs to determine
whether to use SMC based on the context of socket creation, such as
network namespaces, PID and comm name, etc.
As a subsequent enhancement, to introduce a new generic hook that
allows decisions on whether to use SMC or not at runtime, including
but not limited to local/remote IP address or ports.
User can write their own implememtion via bpf_struct_ops now to choose
whether to use SMC or not before TCP 3rd handshake to be comleted.
Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Reviewed-by: Dust Li <dust.li@linux.alibaba.com>
Link: https://patch.msgid.link/20251107035632.115950-3-alibuda@linux.alibaba.com
TCP SACK compression has been added in 2018 in commit
5d9f4262b7ea ("tcp: add SACK compression").
It is working great for WAN flows (with large RTT).
Wifi in particular gets a significant boost _when_ ACK are suppressed.
Add a new sysctl so that we can tune the very conservative 5 % value
that has been used so far in this formula, so that small RTT flows
can benefit from this feature.
delay = min ( 5 % of RTT, 1 ms)
This patch adds new tcp_comp_sack_rtt_percent sysctl
to ease experiments and tuning.
Given that we cap the delay to 1ms (tcp_comp_sack_delay_ns sysctl),
set the default value to 33 %.
Quoting Neal Cardwell ( https://lore.kernel.org/netdev/CADVnQymZ1tFnEA1Q=vtECs0=Db7zHQ8=+WCQtnhHFVbEOzjVnQ@mail.gmail.com/ )
The rationale for 33% is basically to try to facilitate pipelining,
where there are always at least 3 ACKs and 3 GSO/TSO skbs per SRTT, so
that the path can maintain a budget for 3 full-sized GSO/TSO skbs "in
flight" at all times:
+ 1 skb in the qdisc waiting to be sent by the NIC next
+ 1 skb being sent by the NIC (being serialized by the NIC out onto the wire)
+ 1 skb being received and aggregated by the receiver machine's
aggregation mechanism (some combination of LRO, GRO, and sack
compression)
Note that this is basically the same magic number (3) and the same
rationales as:
(a) tcp_tso_should_defer() ensuring that we defer sending data for no
longer than cwnd/tcp_tso_win_divisor (where tcp_tso_win_divisor = 3),
and
(b) bbr_quantization_budget() ensuring that cwnd is at least 3 GSO/TSO
skbs to maintain pipelining and full throughput at low RTTs
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
Link: https://patch.msgid.link/20251106115236.3450026-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
MPLS (re)uses RTNL to protect net->mpls.platform_label,
but the lock does not need to be RTNL at all.
Let's protect net->mpls.platform_label with a dedicated
per-netns mutex.
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Guillaume Nault <gnault@redhat.com>
Link: https://patch.msgid.link/20251029173344.2934622-13-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Think SMC_WR_BUF_CNT_SEND := SMC_WR_BUF_CNT used in send context and
SMC_WR_BUF_CNT_RECV := 3 * SMC_WR_BUF_CNT used in recv context. Those
get replaced with lgr->max_send_wr and lgr->max_recv_wr respective.
Please note that although with the default sysctl values
qp_attr.cap.max_send_wr == qp_attr.cap.max_recv_wr is maintained but
can not be assumed to be generally true any more. I see no downside to
that, but my confidence level is rather modest.
Signed-off-by: Halil Pasic <pasic@linux.ibm.com>
Reviewed-by: Sidraya Jayagond <sidraya@linux.ibm.com>
Reviewed-by: Dust Li <dust.li@linux.alibaba.com>
Tested-by: Mahanta Jambigi <mjambigi@linux.ibm.com>
Link: https://patch.msgid.link/20251027224856.2970019-2-pasic@linux.ibm.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Add the ability to append the incoming IP interface information to
ICMPv6 error messages in accordance with RFC 5837 and RFC 4884. This is
required for more meaningful traceroute results in unnumbered networks.
The feature is disabled by default and controlled via a new sysctl
("net.ipv6.icmp.errors_extension_mask") which accepts a bitmask of ICMP
extensions to append to ICMP error messages. Currently, only a single
value is supported, but the interface and the implementation should be
able to support more extensions, if needed.
Clone the skb and copy the relevant data portions before modifying the
skb as the caller of icmp6_send() still owns the skb after the function
returns. This should be fine since by default ICMP error messages are
rate limited to 1000 per second and no more than 1 per second per
specific host.
Trim or pad the packet to 128 bytes before appending the ICMP extension
structure in order to be compatible with legacy applications that assume
that the ICMP extension structure always starts at this offset (the
minimum length specified by RFC 4884).
Since commit 20e1954fe238 ("ipv6: RFC 4884 partial support for SIT/GRE
tunnels") it is possible for icmp6_send() to be called with an skb that
already contains ICMP extensions. This can happen when we receive an
ICMPv4 message with extensions from a tunnel and translate it to an
ICMPv6 message towards an IPv6 host in the overlay network. I could not
find an RFC that supports this behavior, but it makes sense to not
overwrite the original extensions that were appended to the packet.
Therefore, avoid appending extensions if the length field in the
provided ICMPv6 header is already filled.
Export netdev_copy_name() using EXPORT_IPV6_MOD_GPL() to make it
available to IPv6 when it is built as a module.
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20251027082232.232571-3-idosch@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add the ability to append the incoming IP interface information to
ICMPv4 error messages in accordance with RFC 5837 and RFC 4884. This is
required for more meaningful traceroute results in unnumbered networks.
The feature is disabled by default and controlled via a new sysctl
("net.ipv4.icmp_errors_extension_mask") which accepts a bitmask of ICMP
extensions to append to ICMP error messages. Currently, only a single
value is supported, but the interface and the implementation should be
able to support more extensions, if needed.
Clone the skb and copy the relevant data portions before modifying the
skb as the caller of __icmp_send() still owns the skb after the function
returns. This should be fine since by default ICMP error messages are
rate limited to 1000 per second and no more than 1 per second per
specific host.
Trim or pad the packet to 128 bytes before appending the ICMP extension
structure in order to be compatible with legacy applications that assume
that the ICMP extension structure always starts at this offset (the
minimum length specified by RFC 4884).
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20251027082232.232571-2-idosch@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
If a socket has sk->sk_bypass_prot_mem flagged, the socket opts out
of the global protocol memory accounting.
Let's control the flag by a new sysctl knob.
The flag is written once during socket(2) and is inherited to child
sockets.
Tested with a script that creates local socket pairs and send()s a
bunch of data without recv()ing.
Setup:
# mkdir /sys/fs/cgroup/test
# echo $$ >> /sys/fs/cgroup/test/cgroup.procs
# sysctl -q net.ipv4.tcp_mem="1000 1000 1000"
# ulimit -n 524288
Without net.core.bypass_prot_mem, charged to tcp_mem & memcg
# python3 pressure.py &
# cat /sys/fs/cgroup/test/memory.stat | grep sock
sock 22642688 <-------------------------------------- charged to memcg
# cat /proc/net/sockstat| grep TCP
TCP: inuse 2006 orphan 0 tw 0 alloc 2008 mem 5376 <-- charged to tcp_mem
# ss -tn | head -n 5
State Recv-Q Send-Q Local Address:Port Peer Address:Port
ESTAB 2000 0 127.0.0.1:34479 127.0.0.1:53188
ESTAB 2000 0 127.0.0.1:34479 127.0.0.1:49972
ESTAB 2000 0 127.0.0.1:34479 127.0.0.1:53868
ESTAB 2000 0 127.0.0.1:34479 127.0.0.1:53554
# nstat | grep Pressure || echo no pressure
TcpExtTCPMemoryPressures 1 0.0
With net.core.bypass_prot_mem=1, charged to memcg only:
# sysctl -q net.core.bypass_prot_mem=1
# python3 pressure.py &
# cat /sys/fs/cgroup/test/memory.stat | grep sock
sock 2757468160 <------------------------------------ charged to memcg
# cat /proc/net/sockstat | grep TCP
TCP: inuse 2006 orphan 0 tw 0 alloc 2008 mem 0 <- NOT charged to tcp_mem
# ss -tn | head -n 5
State Recv-Q Send-Q Local Address:Port Peer Address:Port
ESTAB 111000 0 127.0.0.1:36019 127.0.0.1:49026
ESTAB 110000 0 127.0.0.1:36019 127.0.0.1:45630
ESTAB 110000 0 127.0.0.1:36019 127.0.0.1:44870
ESTAB 111000 0 127.0.0.1:36019 127.0.0.1:45274
# nstat | grep Pressure || echo no pressure
no pressure
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Reviewed-by: Shakeel Butt <shakeel.butt@linux.dev>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Acked-by: Roman Gushchin <roman.gushchin@linux.dev>
Link: https://patch.msgid.link/20251014235604.3057003-4-kuniyu@google.com
Add a new sysctl to control how often a queue reselection
can happen even if a flow has a persistent queue of skbs
in a Qdisc or NIC queue.
A value of zero means the feature is disabled.
Default is 1000 (1 second).
This sysctl is used in the following patch.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20251013152234.842065-4-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Instead of sending the option in every ACK, limit sending to
those ACKs where the option is necessary:
- Handshake
- "Change-triggered ACK" + the ACK following it. The
2nd ACK is necessary to unambiguously indicate which
of the ECN byte counters in increasing. The first
ACK has two counters increasing due to the ecnfield
edge.
- ACKs with CE to allow CEP delta validations to take
advantage of the option.
- Force option to be sent every at least once per 2^22
bytes. The check is done using the bit edges of the
byte counters (avoids need for extra variables).
- AccECN option beacon to send a few times per RTT even if
nothing in the ECN state requires that. The default is 3
times per RTT, and its period can be set via
sysctl_tcp_ecn_option_beacon.
Below are the pahole outcomes before and after this patch,
in which the group size of tcp_sock_write_tx is increased
from 89 to 97 due to the new u64 accecn_opt_tstamp member:
[BEFORE THIS PATCH]
struct tcp_sock {
[...]
u64 tcp_wstamp_ns; /* 2488 8 */
struct list_head tsorted_sent_queue; /* 2496 16 */
[...]
__cacheline_group_end__tcp_sock_write_tx[0]; /* 2521 0 */
__cacheline_group_begin__tcp_sock_write_txrx[0]; /* 2521 0 */
u8 nonagle:4; /* 2521: 0 1 */
u8 rate_app_limited:1; /* 2521: 4 1 */
/* XXX 3 bits hole, try to pack */
/* Force alignment to the next boundary: */
u8 :0;
u8 received_ce_pending:4;/* 2522: 0 1 */
u8 unused2:4; /* 2522: 4 1 */
u8 accecn_minlen:2; /* 2523: 0 1 */
u8 est_ecnfield:2; /* 2523: 2 1 */
u8 unused3:4; /* 2523: 4 1 */
[...]
__cacheline_group_end__tcp_sock_write_txrx[0]; /* 2628 0 */
[...]
/* size: 3200, cachelines: 50, members: 171 */
}
[AFTER THIS PATCH]
struct tcp_sock {
[...]
u64 tcp_wstamp_ns; /* 2488 8 */
u64 accecn_opt_tstamp; /* 2596 8 */
struct list_head tsorted_sent_queue; /* 2504 16 */
[...]
__cacheline_group_end__tcp_sock_write_tx[0]; /* 2529 0 */
__cacheline_group_begin__tcp_sock_write_txrx[0]; /* 2529 0 */
u8 nonagle:4; /* 2529: 0 1 */
u8 rate_app_limited:1; /* 2529: 4 1 */
/* XXX 3 bits hole, try to pack */
/* Force alignment to the next boundary: */
u8 :0;
u8 received_ce_pending:4;/* 2530: 0 1 */
u8 unused2:4; /* 2530: 4 1 */
u8 accecn_minlen:2; /* 2531: 0 1 */
u8 est_ecnfield:2; /* 2531: 2 1 */
u8 accecn_opt_demand:2; /* 2531: 4 1 */
u8 prev_ecnfield:2; /* 2531: 6 1 */
[...]
__cacheline_group_end__tcp_sock_write_txrx[0]; /* 2636 0 */
[...]
/* size: 3200, cachelines: 50, members: 173 */
}
Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Co-developed-by: Ilpo Järvinen <ij@kernel.org>
Signed-off-by: Ilpo Järvinen <ij@kernel.org>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20250916082434.100722-8-chia-yu.chang@nokia-bell-labs.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The Accurate ECN allows echoing back the sum of bytes for
each IP ECN field value in the received packets using
AccECN option. This change implements AccECN option tx & rx
side processing without option send control related features
that are added by a later change.
Based on specification:
https://tools.ietf.org/id/draft-ietf-tcpm-accurate-ecn-28.txt
(Some features of the spec will be added in the later changes
rather than in this one).
A full-length AccECN option is always attempted but if it does
not fit, the minimum length is selected based on the counters
that have changed since the last update. The AccECN option
(with 24-bit fields) often ends in odd sizes so the option
write code tries to take advantage of some nop used to pad
the other TCP options.
The delivered_ecn_bytes pairs with received_ecn_bytes similar
to how delivered_ce pairs with received_ce. In contrast to
ACE field, however, the option is not always available to update
delivered_ecn_bytes. For ACK w/o AccECN option, the delivered
bytes calculated based on the cumulative ACK+SACK information
are assigned to one of the counters using an estimation
heuristic to select the most likely ECN byte counter. Any
estimation error is corrected when the next AccECN option
arrives. It may occur that the heuristic gets too confused
when there are enough different byte counter deltas between
ACKs with the AccECN option in which case the heuristic just
gives up on updating the counters for a while.
tcp_ecn_option sysctl can be used to select option sending
mode for AccECN: TCP_ECN_OPTION_DISABLED, TCP_ECN_OPTION_MINIMUM,
and TCP_ECN_OPTION_FULL.
This patch increases the size of tcp_info struct, as there is
no existing holes for new u32 variables. Below are the pahole
outcomes before and after this patch:
[BEFORE THIS PATCH]
struct tcp_info {
[...]
__u32 tcpi_total_rto_time; /* 244 4 */
/* size: 248, cachelines: 4, members: 61 */
}
[AFTER THIS PATCH]
struct tcp_info {
[...]
__u32 tcpi_total_rto_time; /* 244 4 */
__u32 tcpi_received_ce; /* 248 4 */
__u32 tcpi_delivered_e1_bytes; /* 252 4 */
__u32 tcpi_delivered_e0_bytes; /* 256 4 */
__u32 tcpi_delivered_ce_bytes; /* 260 4 */
__u32 tcpi_received_e1_bytes; /* 264 4 */
__u32 tcpi_received_e0_bytes; /* 268 4 */
__u32 tcpi_received_ce_bytes; /* 272 4 */
/* size: 280, cachelines: 5, members: 68 */
}
This patch uses the existing 1-byte holes in the tcp_sock_write_txrx
group for new u8 members, but adds a 4-byte hole in tcp_sock_write_rx
group after the new u32 delivered_ecn_bytes[3] member. Therefore, the
group size of tcp_sock_write_rx is increased from 96 to 112. Below
are the pahole outcomes before and after this patch:
[BEFORE THIS PATCH]
struct tcp_sock {
[...]
u8 received_ce_pending:4; /* 2522: 0 1 */
u8 unused2:4; /* 2522: 4 1 */
/* XXX 1 byte hole, try to pack */
[...]
u32 rcv_rtt_last_tsecr; /* 2668 4 */
[...]
__cacheline_group_end__tcp_sock_write_rx[0]; /* 2728 0 */
[...]
/* size: 3200, cachelines: 50, members: 167 */
}
[AFTER THIS PATCH]
struct tcp_sock {
[...]
u8 received_ce_pending:4;/* 2522: 0 1 */
u8 unused2:4; /* 2522: 4 1 */
u8 accecn_minlen:2; /* 2523: 0 1 */
u8 est_ecnfield:2; /* 2523: 2 1 */
u8 unused3:4; /* 2523: 4 1 */
[...]
u32 rcv_rtt_last_tsecr; /* 2668 4 */
u32 delivered_ecn_bytes[3];/* 2672 12 */
/* XXX 4 bytes hole, try to pack */
[...]
__cacheline_group_end__tcp_sock_write_rx[0]; /* 2744 0 */
[...]
/* size: 3200, cachelines: 50, members: 171 */
}
Signed-off-by: Ilpo Järvinen <ij@kernel.org>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Co-developed-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Signed-off-by: Chia-Yu Chang <chia-yu.chang@nokia-bell-labs.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20250916082434.100722-7-chia-yu.chang@nokia-bell-labs.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Cross-merge networking fixes after downstream PR (net-6.17-rc6).
Conflicts:
net/netfilter/nft_set_pipapo.c
net/netfilter/nft_set_pipapo_avx2.c
c4eaca2e1052 ("netfilter: nft_set_pipapo: don't check genbit from packetpath lookups")
84c1da7b38d9 ("netfilter: nft_set_pipapo: use avx2 algorithm for insertions too")
Only trivial adjacent changes (in a doc and a Makefile).
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This will soon be read from packet path around same time as the gencursor.
Both gencursor and base_seq get incremented almost at the same time, so
it makes sense to place them in the same structure.
This doesn't increase struct net size on 64bit due to padding.
Signed-off-by: Florian Westphal <fw@strlen.de>
Provide isolation between netns for ping idents.
Randomize initial ping_port_rover value at netns creation.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://patch.msgid.link/20250829153054.474201-4-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Convert SCTP cookies to use HMAC-SHA256, instead of the previous choice
of the legacy algorithms HMAC-MD5 and HMAC-SHA1. Simplify and optimize
the code by using the HMAC-SHA256 library instead of crypto_shash, and
by preparing the HMAC key when it is generated instead of per-operation.
This doesn't break compatibility, since the cookie format is an
implementation detail, not part of the SCTP protocol itself.
Note that the cookie size doesn't change either. The HMAC field was
already 32 bytes, even though previously at most 20 bytes were actually
compared. 32 bytes exactly fits an untruncated HMAC-SHA256 value. So,
although we could safely truncate the MAC to something slightly shorter,
for now just keep the cookie size the same.
I also considered SipHash, but that would generate only 8-byte MACs. An
8-byte MAC *might* suffice here. However, there's quite a lot of
information in the SCTP cookies: more than in TCP SYN cookies. So
absent an analysis that occasional forgeries of all that information is
okay in SCTP, I errored on the side of caution.
Remove HMAC-MD5 and HMAC-SHA1 as options, since the new HMAC-SHA256
option is just better. It's faster as well as more secure. For
example, benchmarking on x86_64, cookie authentication is now nearly 3x
as fast as the previous default choice and implementation of HMAC-MD5.
Also just make the kernel always support cookie authentication if SCTP
is supported at all, rather than making it optional in the build. (It
was sort of optional before, but it didn't really work properly. E.g.,
a kernel with CONFIG_SCTP_COOKIE_HMAC_MD5=n still supported HMAC-MD5
cookie authentication if CONFIG_CRYPTO_HMAC and CONFIG_CRYPTO_MD5
happened to be enabled in the kconfig for other reasons.)
Acked-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Link: https://patch.msgid.link/20250818205426.30222-5-ebiggers@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Ensure that a specific EID (remote or local) bind will match in
preference to a MCTP_ADDR_ANY bind.
This adds infrastructure for binding a socket to receive messages from a
specific remote peer address, a future commit will expose an API for
this.
Signed-off-by: Matt Johnston <matt@codeconstruct.com.au>
Link: https://patch.msgid.link/20250710-mctp-bind-v4-5-8ec2f6460c56@codeconstruct.com.au
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The DCCP socket family has now been removed from this tree, see:
8bb3212be4b4 ("Merge branch 'net-retire-dccp-socket'")
Remove connection tracking and NAT support for this protocol, this
should not pose a problem because no DCCP traffic is expected to be seen
on the wire.
As for the code for matching on dccp header for iptables and nftables,
mark it as deprecated and keep it in place. Ruleset restoration is an
atomic operation. Without dccp matching support, an astray match on dccp
could break this operation leaving your computer with no policy in
place, so let's follow a more conservative approach for matches.
Add CONFIG_NFT_EXTHDR_DCCP which is set to 'n' by default to deprecate
dccp extension support. Similarly, label CONFIG_NETFILTER_XT_MATCH_DCCP
as deprecated too and also set it to 'n' by default.
Code to match on DCCP protocol from ebtables also remains in place, this
is just a few checks on IPPROTO_DCCP from _check() path which is
exercised when ruleset is loaded. There is another use of IPPROTO_DCCP
from the _check() path in the iptables multiport match. Another check
for IPPROTO_DCCP from the packet in the reject target is also removed.
So let's schedule removal of the dccp matching for a second stage, this
should not interfer with the dccp retirement since this is only matching
on the dccp header.
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Paolo Abeni <pabeni@redhat.com>
Cc: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
We will get rid of RTNL from RTM_NEWROUTE and SIOCADDRT.
If the request specifies a new table ID, fib6_new_table() is
called to create a new routing table.
Two concurrent requests could specify the same table ID, so we
need a lock to protect net->ipv6.fib_table_hash[h].
Let's add a spinlock to protect the hash bucket linkage.
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Link: https://patch.msgid.link/20250418000443.43734-13-kuniyu@amazon.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Most UDP tunnels bind a socket to a local port, with ANY address, no
peer and no interface index specified.
Additionally it's quite common to have a single tunnel device per
namespace.
Track in each namespace the UDP tunnel socket respecting the above.
When only a single one is present, store a reference in the netns.
When such reference is not NULL, UDP tunnel GRO lookup just need to
match the incoming packet destination port vs the socket local port.
The tunnel socket never sets the reuse[port] flag[s]. When bound to no
address and interface, no other socket can exist in the same netns
matching the specified local port.
Matching packets with non-local destination addresses will be
aggregated, and eventually segmented as needed - no behavior changes
intended.
Restrict the optimization to kernel sockets only: it covers all the
relevant use-cases, and user-space owned sockets could be disconnected
and rebound after setup_udp_tunnel_sock(), breaking the uniqueness
assumption
Note that the UDP tunnel socket reference is stored into struct
netns_ipv4 for both IPv4 and IPv6 tunnels. That is intentional to keep
all the fastpath-related netns fields in the same struct and allow
cacheline-based optimization. Currently both the IPv4 and IPv6 socket
pointer share the same cacheline as the `udp_table` field.
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/41d16bc8d1257d567f9344c445b4ae0b4a91ede4.1744040675.git.pabeni@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Revert "udp_tunnel: use static call for GRO hooks when possible"
This reverts commit 311b36574ceaccfa3f91b74054a09cd4bb877702.
Revert "udp_tunnel: create a fastpath GRO lookup."
This reverts commit 8d4880db378350f8ed8969feea13bdc164564fc1.
There are multiple small issues with the series. In the interest
of unblocking the merge window let's opt for a revert.
Link: https://lore.kernel.org/cover.1742557254.git.pabeni@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Most UDP tunnels bind a socket to a local port, with ANY address, no
peer and no interface index specified.
Additionally it's quite common to have a single tunnel device per
namespace.
Track in each namespace the UDP tunnel socket respecting the above.
When only a single one is present, store a reference in the netns.
When such reference is not NULL, UDP tunnel GRO lookup just need to
match the incoming packet destination port vs the socket local port.
The tunnel socket never sets the reuse[port] flag[s]. When bound to no
address and interface, no other socket can exist in the same netns
matching the specified local port.
Matching packets with non-local destination addresses will be
aggregated, and eventually segmented as needed - no behavior changes
intended.
Note that the UDP tunnel socket reference is stored into struct
netns_ipv4 for both IPv4 and IPv6 tunnels. That is intentional to keep
all the fastpath-related netns fields in the same struct and allow
cacheline-based optimization. Currently both the IPv4 and IPv6 socket
pointer share the same cacheline as the `udp_table` field.
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/4d5c319c4471161829f50cb8436841de81a5edae.1741718157.git.pabeni@redhat.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
We will convert RTM_NEWROUTE and RTM_DELROUTE to per-netns RTNL.
Then, we need to have per-netns hash tables for struct fib_info.
Let's allocate the hash tables per netns.
fib_info_hash, fib_info_hash_bits, and fib_info_cnt are now moved
to struct netns_ipv4 and accessed with net->ipv4.fib_XXX.
Also, the netns checks are removed from fib_find_info_nh() and
fib_find_info().
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://patch.msgid.link/20250228042328.96624-9-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Previous patch added a TCP_RTO_MAX_MS socket option
to tune a TCP socket max RTO value.
Many setups prefer to change a per netns sysctl.
This patch adds /proc/sys/net/ipv4/tcp_rto_max_ms
Its initial value is 120000 (120 seconds).
Keep in mind that a decrease of tcp_rto_max_ms
means shorter overall timeouts, unless tcp_retries2
sysctl is increased.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jason Xing <kerneljasonxing@gmail.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Today we have a hardcoded delay of 1 sec before a TIME-WAIT socket can be
reused by reopening a connection. This is a safe choice based on an
assumption that the other TCP timestamp clock frequency, which is unknown
to us, may be as low as 1 Hz (RFC 7323, section 5.4).
However, this means that in the presence of short lived connections with an
RTT of couple of milliseconds, the time during which a 4-tuple is blocked
from reuse can be orders of magnitude longer that the connection lifetime.
Combined with a reduced pool of ephemeral ports, when using
IP_LOCAL_PORT_RANGE to share an egress IP address between hosts [1], the
long TIME-WAIT reuse delay can lead to port exhaustion, where all available
4-tuples are tied up in TIME-WAIT state.
Turn the reuse delay into a per-netns setting so that sysadmins can make
more aggressive assumptions about remote TCP timestamp clock frequency and
shorten the delay in order to allow connections to reincarnate faster.
Note that applications can completely bypass the TIME-WAIT delay protection
already today by locking the local port with bind() before connecting. Such
immediate connection reuse may result in PAWS failing to detect old
duplicate segments, leaving us with just the sequence number check as a
safety net.
This new configurable offers a trade off where the sysadmin can balance
between the risk of PAWS detection failing to act versus exhausting ports
by having sockets tied up in TIME-WAIT state for too long.
[1] https://lpc.events/event/16/contributions/1349/
Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jason Xing <kerneljasonxing@gmail.com>
Link: https://patch.msgid.link/20241209-jakub-krn-909-poc-msec-tw-tstamp-v2-2-66aca0eed03e@cloudflare.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Now that we can have percpu xfrm states, the number of active
states might increase. To get a better lookup performance,
we add a percpu cache to cache the used inbound xfrm states.
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Tested-by: Antony Antony <antony.antony@secunet.com>
Tested-by: Tobias Brunner <tobias@strongswan.org>
Using RTNL to protect ops->fib_rules_seq reads seems a big hammer.
Writes are protected by RTNL.
We can use READ_ONCE() when reading it.
Constify 'struct net' argument of fib4_rules_seq_read()
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://patch.msgid.link/20241009184405.3752829-3-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
sysctl_tcp_l3mdev_accept is read from TCP receive fast path from
tcp_v6_early_demux(),
__inet6_lookup_established,
inet_request_bound_dev_if().
Move it to netns_ipv4_read_rx.
Remove the '#ifdef CONFIG_NET_L3_MASTER_DEV' that was guarding
its definition.
Note this adds a hole of three bytes that could be filled later.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Cc: Wei Wang <weiwan@google.com>
Cc: Coco Li <lixiaoyan@google.com>
Link: https://patch.msgid.link/20241010034100.320832-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Each IPv4 address could have a lifetime, which is useful for DHCP,
and GC is periodically executed as check_lifetime_work.
check_lifetime() does the actual GC under RTNL.
1. Acquire RTNL
2. Iterate inet_addr_lst
3. Remove IPv4 address if expired
4. Release RTNL
Namespacifying the GC is required for per-netns RTNL, but using the
per-netns hash table will shorten the time on the hash bucket iteration
under RTNL.
Let's add per-netns GC work and use the per-netns hash table.
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://patch.msgid.link/20241008172906.1326-4-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
As a prep for per-netns RTNL conversion, we want to namespacify
the IPv4 address hash table and the GC work.
Let's allocate the per-netns IPv4 address hash table to
net->ipv4.inet_addr_lst and link IPv4 addresses into it.
The actual users will be converted later.
Note that the IPv6 address hash table is already namespacified.
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://patch.msgid.link/20241008172906.1326-2-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Let it be tuned in per netns by admins.
Signed-off-by: Jason Xing <kernelxing@tencent.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20241005222609.94980-1-kerneljasonxing@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Previous patch made ICMP rate limits per netns, it makes sense
to allow each netns to change the associated sysctl.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://patch.msgid.link/20240829144641.3880376-4-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Host wide ICMP ratelimiter should be per netns, to provide better isolation.
Following patch in this series makes the sysctl per netns.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://patch.msgid.link/20240829144641.3880376-3-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
-----BEGIN PGP SIGNATURE-----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=KyYL
-----END PGP SIGNATURE-----
Merge tag 'ipsec-next-2024-07-13' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec-next
Steffen Klassert says:
====================
pull request (net-next): ipsec-next 2024-07-13
1) Support sending NAT keepalives in ESP in UDP states.
Userspace IKE daemon had to do this before, but the
kernel can better keep track of it.
From Eyal Birger.
2) Support IPsec crypto offload for IPv6 ESP and IPv4 UDP-encapsulated
ESP data paths. Currently, IPsec crypto offload is enabled for GRO
code path only. This patchset support UDP encapsulation for the non
GRO path. From Mike Yu.
* tag 'ipsec-next-2024-07-13' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec-next:
xfrm: Support crypto offload for outbound IPv4 UDP-encapsulated ESP packet
xfrm: Support crypto offload for inbound IPv4 UDP-encapsulated ESP packet
xfrm: Allow UDP encapsulation in crypto offload control path
xfrm: Support crypto offload for inbound IPv6 ESP packets not in GRO path
xfrm: support sending NAT keepalives in ESP in UDP states
====================
Link: https://patch.msgid.link/20240713102416.3272997-1-steffen.klassert@secunet.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add the ability to send out RFC-3948 NAT keepalives from the xfrm stack.
To use, Userspace sets an XFRM_NAT_KEEPALIVE_INTERVAL integer property when
creating XFRM outbound states which denotes the number of seconds between
keepalive messages.
Keepalive messages are sent from a per net delayed work which iterates over
the xfrm states. The logic is guarded by the xfrm state spinlock due to the
xfrm state walk iterator.
Possible future enhancements:
- Adding counters to keep track of sent keepalives.
- deduplicate NAT keepalives between states sharing the same nat keepalive
parameters.
- provisioning hardware offloads for devices capable of implementing this.
- revise xfrm state list to use an rcu list in order to avoid running this
under spinlock.
Suggested-by: Paul Wouters <paul.wouters@aiven.io>
Tested-by: Paul Wouters <paul.wouters@aiven.io>
Tested-by: Antony Antony <antony.antony@secunet.com>
Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Cross-merge networking fixes after downstream PR.
Conflicts:
drivers/net/ethernet/broadcom/bnxt/bnxt.c
1e7962114c10 ("bnxt_en: Restore PTP tx_avail count in case of skb_pad() error")
165f87691a89 ("bnxt_en: add timestamping statistics support")
No adjacent changes.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Currently, the sysctl net.netfilter.nf_hooks_lwtunnel depends on the
nf_conntrack module, but the nf_conntrack module is not always loaded.
Therefore, accessing net.netfilter.nf_hooks_lwtunnel may have an error.
Move sysctl nf_hooks_lwtunnel into the netfilter core.
Fixes: 7a3f5b0de364 ("netfilter: add netfilter hooks to SRv6 data plane")
Suggested-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Jianguo Wu <wujianguo@chinatelecom.cn>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
When calculating hashes for the purpose of multipath forwarding, both IPv4
and IPv6 code currently fall back on flow_hash_from_keys(). That uses a
randomly-generated seed. That's a fine choice by default, but unfortunately
some deployments may need a tighter control over the seed used.
In this patch, make the seed configurable by adding a new sysctl key,
net.ipv4.fib_multipath_hash_seed to control the seed. This seed is used
specifically for multipath forwarding and not for the other concerns that
flow_hash_from_keys() is used for, such as queue selection. Expose the knob
as sysctl because other such settings, such as headers to hash, are also
handled that way. Like those, the multipath hash seed is a per-netns
variable.
Despite being placed in the net.ipv4 namespace, the multipath seed sysctl
is used for both IPv4 and IPv6, similarly to e.g. a number of TCP
variables.
The seed used by flow_hash_from_keys() is a 128-bit quantity. However it
seems that usually the seed is a much more modest value. 32 bits seem
typical (Cisco, Cumulus), some systems go even lower. For that reason, and
to decouple the user interface from implementation details, go with a
32-bit quantity, which is then quadruplicated to form the siphash key.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/20240607151357.421181-3-petrm@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Adding a sysctl knob to allow user to specify a default
rto_min at socket init time, other than using the hard
coded 200ms default rto_min.
Note that the rto_min route option has the highest precedence
for configuring this setting, followed by the TCP_BPF_RTO_MIN
socket option, followed by the tcp_rto_min_us sysctl.
Signed-off-by: Kevin Yang <yyd@google.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Tony Lu <tonylu@linux.alibaba.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
optmem_max being used in tx zerocopy,
we want to be able to control it on a netns basis.
Following patch changes two tests.
Tested:
oqq130:~# cat /proc/sys/net/core/optmem_max
131072
oqq130:~# echo 1000000 >/proc/sys/net/core/optmem_max
oqq130:~# cat /proc/sys/net/core/optmem_max
1000000
oqq130:~# unshare -n
oqq130:~# cat /proc/sys/net/core/optmem_max
131072
oqq130:~# exit
logout
oqq130:~# cat /proc/sys/net/core/optmem_max
1000000
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 227b60f5102cd added a seqlock to ensure that the low and high
port numbers were always updated together.
This is overkill because the two 16bit port numbers can be held in
a u32 and read/written in a single instruction.
More recently 91d0b78c5177f added support for finer per-socket limits.
The user-supplied value is 'high << 16 | low' but they are held
separately and the socket options protected by the socket lock.
Use a u32 containing 'high << 16 | low' for both the 'net' and 'sk'
fields and use READ_ONCE()/WRITE_ONCE() to ensure both values are
always updated together.
Change (the now trival) inet_get_local_port_range() to a static inline
to optimise the calling code.
(In particular avoiding returning integers by reference.)
Signed-off-by: David Laight <david.laight@aculab.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Acked-by: Mat Martineau <martineau@kernel.org>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://lore.kernel.org/r/4e505d4198e946a8be03fb1b4c3072b0@AcuMS.aculab.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reorganize fast path variables on tx-txrx-rx order.
Fastpath cacheline ends after sysctl_tcp_rmem.
There are only read-only variables here. (write is on the control path
and not considered in this case)
Below data generated with pahole on x86 architecture.
Fast path variables span cache lines before change: 4
Fast path variables span cache lines after change: 2
Suggested-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Wei Wang <weiwan@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: Coco Li <lixiaoyan@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a new sysctl: net.smc.smcr_max_conns_per_lgr, which is
used to control the preferred max connections per lgr for
SMC-R v2.1. The default value of this sysctl is 255, and
the acceptable value ranges from 16 to 255.
Signed-off-by: Guangguan Wang <guangguan.wang@linux.alibaba.com>
Reviewed-by: Dust Li <dust.li@linux.alibaba.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a new sysctl: net.smc.smcr_max_links_per_lgr, which is
used to control the preferred max links per lgr for SMC-R
v2.1. The default value of this sysctl is 2, and the acceptable
value ranges from 1 to 2.
Signed-off-by: Guangguan Wang <guangguan.wang@linux.alibaba.com>
Reviewed-by: Dust Li <dust.li@linux.alibaba.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The spinlock is back from the day when connabels did not have
a fixed size and reallocation had to be supported.
Remove it. This change also allows to call the helpers from
softirq or timers without deadlocks.
Also add WARN()s to catch refcounting imbalances.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
TCP pingpong threshold is 1 by default. But some applications, like SQL DB
may prefer a higher pingpong threshold to activate delayed acks in quick
ack mode for better performance.
The pingpong threshold and related code were changed to 3 in the year
2019 in:
commit 4a41f453bedf ("tcp: change pingpong threshold to 3")
And reverted to 1 in the year 2022 in:
commit 4d8f24eeedc5 ("Revert "tcp: change pingpong threshold to 3"")
There is no single value that fits all applications.
Add net.ipv4.tcp_pingpong_thresh sysctl tunable, so it can be tuned for
optimal performance based on the application needs.
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://lore.kernel.org/r/1697056244-21888-1-git-send-email-haiyangz@microsoft.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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>
