| Age | Commit message (Collapse) | Author |
|---|
|
When sockmap sock with TLS enabled is removed we cleanup bpf/psock state
and call tcp_update_ulp() to push updates to TLS ULP on top. However, we
don't push the write_space callback up and instead simply overwrite the
op with the psock stored previous op. This may or may not be correct so
to ensure we don't overwrite the TLS write space hook pass this field to
the ULP and have it fixup the ctx.
This completes a previous fix that pushed the ops through to the ULP
but at the time missed doing this for write_space, presumably because
write_space TLS hook was added around the same time.
Fixes: 95fa145479fbc ("bpf: sockmap/tls, close can race with map free")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Acked-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/bpf/20200111061206.8028-4-john.fastabend@gmail.com
|
|
When the packet pointed to by retransmit_skb_hint is unlinked by ACK,
retransmit_skb_hint will be set to NULL in tcp_clean_rtx_queue().
If packet loss is detected at this time, retransmit_skb_hint will be set
to point to the current packet loss in tcp_verify_retransmit_hint(),
then the packets that were previously marked lost but not retransmitted
due to the restriction of cwnd will be skipped and cannot be
retransmitted.
To fix this, when retransmit_skb_hint is NULL, retransmit_skb_hint can
be reset only after all marked lost packets are retransmitted
(retrans_out >= lost_out), otherwise we need to traverse from
tcp_rtx_queue_head in tcp_xmit_retransmit_queue().
Packetdrill to demonstrate:
// Disable RACK and set max_reordering to keep things simple
0 `sysctl -q net.ipv4.tcp_recovery=0`
+0 `sysctl -q net.ipv4.tcp_max_reordering=3`
// Establish a connection
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+.1 < S 0:0(0) win 32792 <mss 1000,sackOK,nop,nop,nop,wscale 7>
+0 > S. 0:0(0) ack 1 <...>
+.01 < . 1:1(0) ack 1 win 257
+0 accept(3, ..., ...) = 4
// Send 8 data segments
+0 write(4, ..., 8000) = 8000
+0 > P. 1:8001(8000) ack 1
// Enter recovery and 1:3001 is marked lost
+.01 < . 1:1(0) ack 1 win 257 <sack 3001:4001,nop,nop>
+0 < . 1:1(0) ack 1 win 257 <sack 5001:6001 3001:4001,nop,nop>
+0 < . 1:1(0) ack 1 win 257 <sack 5001:7001 3001:4001,nop,nop>
// Retransmit 1:1001, now retransmit_skb_hint points to 1001:2001
+0 > . 1:1001(1000) ack 1
// 1001:2001 was ACKed causing retransmit_skb_hint to be set to NULL
+.01 < . 1:1(0) ack 2001 win 257 <sack 5001:8001 3001:4001,nop,nop>
// Now retransmit_skb_hint points to 4001:5001 which is now marked lost
// BUG: 2001:3001 was not retransmitted
+0 > . 2001:3001(1000) ack 1
Signed-off-by: Pengcheng Yang <yangpc@wangsu.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Tested-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Commit 9b42c1f179a6 ("xfrm: Extend the output_mark") added output_mark
support but missed ESP offload support.
xfrm_smark_get() is not called within xfrm_input() for packets coming
from esp4_gro_receive() or esp6_gro_receive(). Therefore call
xfrm_smark_get() directly within these functions.
Fixes: 9b42c1f179a6 ("xfrm: Extend the output_mark to support input direction and masking.")
Signed-off-by: Ulrich Weber <ulrich.weber@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
|
|
When performing L3 offload, routes and nexthops are usually programmed
into two different tables in the underlying device. Therefore, the fact
that a nexthop resides in hardware does not necessarily mean that all
the associated routes also reside in hardware and vice-versa.
While the kernel can signal to user space the presence of a nexthop in
hardware (via 'RTNH_F_OFFLOAD'), it does not have a corresponding flag
for routes. In addition, the fact that a route resides in hardware does
not necessarily mean that the traffic is offloaded. For example,
unreachable routes (i.e., 'RTN_UNREACHABLE') are programmed to trap
packets to the CPU so that the kernel will be able to generate the
appropriate ICMP error packet.
This patch adds an "offload" and "trap" indications to IPv4 routes, so
that users will have better visibility into the offload process.
'struct fib_alias' is extended with two new fields that indicate if the
route resides in hardware or not and if it is offloading traffic from
the kernel or trapping packets to it. Note that the new fields are added
in the 6 bytes hole and therefore the struct still fits in a single
cache line [1].
Capable drivers are expected to invoke fib_alias_hw_flags_set() with the
route's key in order to set the flags.
The indications are dumped to user space via a new flags (i.e.,
'RTM_F_OFFLOAD' and 'RTM_F_TRAP') in the 'rtm_flags' field in the
ancillary header.
v2:
* Make use of 'struct fib_rt_info' in fib_alias_hw_flags_set()
[1]
struct fib_alias {
struct hlist_node fa_list; /* 0 16 */
struct fib_info * fa_info; /* 16 8 */
u8 fa_tos; /* 24 1 */
u8 fa_type; /* 25 1 */
u8 fa_state; /* 26 1 */
u8 fa_slen; /* 27 1 */
u32 tb_id; /* 28 4 */
s16 fa_default; /* 32 2 */
u8 offload:1; /* 34: 0 1 */
u8 trap:1; /* 34: 1 1 */
u8 unused:6; /* 34: 2 1 */
/* XXX 5 bytes hole, try to pack */
struct callback_head rcu __attribute__((__aligned__(8))); /* 40 16 */
/* size: 56, cachelines: 1, members: 12 */
/* sum members: 50, holes: 1, sum holes: 5 */
/* sum bitfield members: 8 bits (1 bytes) */
/* forced alignments: 1, forced holes: 1, sum forced holes: 5 */
/* last cacheline: 56 bytes */
} __attribute__((__aligned__(8)));
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
fib_dump_info() is used to prepare RTM_{NEW,DEL}ROUTE netlink messages
using the passed arguments. Currently, the function takes 11 arguments,
6 of which are attributes of the route being dumped (e.g., prefix, TOS).
The next patch will need the function to also dump to user space an
indication if the route is present in hardware or not. Instead of
passing yet another argument, change the function to take a struct
containing the different route attributes.
v2:
* Name last argument of fib_dump_info()
* Move 'struct fib_rt_info' to include/net/ip_fib.h so that it could
later be passed to fib_alias_hw_flags_set()
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Subsequent patches will add an offload / trap indication to routes which
will signal if the route is present in hardware or not.
After programming the route to the hardware, drivers will have to ask
the IPv4 code to set the flags by passing the route's key.
In the case of route replace, the new route is notified before it is
actually inserted into the FIB alias list. This can prevent simple
drivers (e.g., netdevsim) that program the route to the hardware in the
same context it is notified in from being able to set the flag.
Solve this by first inserting the new route to the list and rollback the
operation in case the route was vetoed.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This is a straight-forward conversion case for the new function, keeping
the flow of the existing code as intact as possible.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This is a straight-forward conversion case for the new function,
iterating over the return value from udp_rcv_segment, which actually is
a wrapper around skb_gso_segment.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
With an ebpf program that redirects packets through a vti[6] interface,
the packets are dropped because no dst is attached.
This could also be reproduced with an AF_PACKET socket, with the following
python script (vti1 is an ip_vti interface):
import socket
send_s = socket.socket(socket.AF_PACKET, socket.SOCK_RAW, 0)
# scapy
# p = IP(src='10.100.0.2', dst='10.200.0.1')/ICMP(type='echo-request')
# raw(p)
req = b'E\x00\x00\x1c\x00\x01\x00\x00@\x01e\xb2\nd\x00\x02\n\xc8\x00\x01\x08\x00\xf7\xff\x00\x00\x00\x00'
send_s.sendto(req, ('vti1', 0x800, 0, 0))
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
|
|
An earlier commit (1b789577f655060d98d20e,
"netfilter: arp_tables: init netns pointer in xt_tgchk_param struct")
fixed missing net initialization for arptables, but turns out it was
incomplete. We can get a very similar struct net NULL deref during
error unwinding:
general protection fault: 0000 [#1] PREEMPT SMP KASAN
RIP: 0010:xt_rateest_put+0xa1/0x440 net/netfilter/xt_RATEEST.c:77
xt_rateest_tg_destroy+0x72/0xa0 net/netfilter/xt_RATEEST.c:175
cleanup_entry net/ipv4/netfilter/arp_tables.c:509 [inline]
translate_table+0x11f4/0x1d80 net/ipv4/netfilter/arp_tables.c:587
do_replace net/ipv4/netfilter/arp_tables.c:981 [inline]
do_arpt_set_ctl+0x317/0x650 net/ipv4/netfilter/arp_tables.c:1461
Also init the netns pointer in xt_tgdtor_param struct.
Fixes: add67461240c1d ("netfilter: add struct net * to target parameters")
Reported-by: syzbot+91bdd8eece0f6629ec8b@syzkaller.appspotmail.com
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
Sven-Haegar reported looping on fib dumps when 255.255.255.255 route has
been added to a table. The looping is caused by the key rolling over from
FFFFFFFF to 0. When dumping a specific table only, we need a means to detect
when the table dump is done. The key and count saved to cb args are both 0
only at the start of the table dump. If key is 0 and count > 0, then we are
in the rollover case. Detect and return to avoid looping.
This only affects dumps of a specific table; for dumps of all tables
(the case prior to the change in the Fixes tag) inet_dump_fib moved
the entry counter to the next table and reset the cb args used by
fib_table_dump and fn_trie_dump_leaf, so the rollover ffffffff back
to 0 did not cause looping with the dumps.
Fixes: effe67926624 ("net: Enable kernel side filtering of route dumps")
Reported-by: Sven-Haegar Koch <haegar@sdinet.de>
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Update the SACK check to work with zero option space available, a case
that's possible with MPTCP but not MD5+TS. Maintained only one
conditional branch for insufficient SACK space.
v1 -> v2:
- Moves the check inside the SACK branch by taking recent SACK fix:
9424e2e7ad93 (tcp: md5: fix potential overestimation of TCP option space)
in to account, but modifies it to work in MPTCP scenarios beyond the
MD5+TS corner case.
Co-developed-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
MPTCP will make use of tcp_send_mss() and tcp_push() when sending
data to specific TCP subflows.
tcp_request_sock_ipvX_ops and ipvX_specific will be referenced
during TCP subflow creation.
Co-developed-by: Peter Krystad <peter.krystad@linux.intel.com>
Signed-off-by: Peter Krystad <peter.krystad@linux.intel.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Coalesce and collapse of packets carrying MPTCP extensions is allowed
when the newer packet has no extension or the extensions carried by both
packets are equal.
This allows merging of TSO packet trains and even cross-TSO packets, and
does not require any additional action when moving data into existing
SKBs.
v3 -> v4:
- allow collapsing, under mptcp_skb_can_collapse() constraint
v5 -> v6:
- clarify MPTCP skb extensions must always be cleared at allocation
time
Co-developed-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
If ULP is used on a listening socket, icsk_ulp_ops and icsk_ulp_data are
copied when the listener is cloned. Sometimes the clone is immediately
deleted, which will invoke the release op on the clone and likely
corrupt the listening socket's icsk_ulp_data.
The clone operation is invoked immediately after the clone is copied and
gives the ULP type an opportunity to set up the clone socket and its
icsk_ulp_data.
The MPTCP ULP clone will silently fallback to plain TCP on allocation
failure, so 'clone()' does not need to return an error code.
v6 -> v7:
- move and rename ulp clone helper to make it inline-friendly
v5 -> v6:
- clarified MPTCP clone usage in commit message
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Right now in tcp_bpf_recvmsg, sock read data first from sk_receive_queue
if not empty than psock->ingress_msg otherwise. If a FIN packet arrives
and there's also some data in psock->ingress_msg, the data in
psock->ingress_msg will be purged. It is always happen when request to a
HTTP1.0 server like python SimpleHTTPServer since the server send FIN
packet after data is sent out.
Fixes: 604326b41a6fb ("bpf, sockmap: convert to generic sk_msg interface")
Reported-by: Arika Chen <eaglesora@gmail.com>
Suggested-by: Arika Chen <eaglesora@gmail.com>
Signed-off-by: Lingpeng Chen <forrest0579@gmail.com>
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Song Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20200109014833.18951-1-forrest0579@gmail.com
|
|
The ungrafting from PRIO bug fixes in net, when merged into net-next,
merge cleanly but create a build failure. The resolution used here is
from Petr Machata.
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add a helper to send out a tcp-ack. It will be used in the later
bpf_dctcp implementation that requires to send out an ack
when the CE state changed.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20200109004551.3900448-1-kafai@fb.com
|
|
This patch makes "struct tcp_congestion_ops" to be the first user
of BPF STRUCT_OPS. It allows implementing a tcp_congestion_ops
in bpf.
The BPF implemented tcp_congestion_ops can be used like
regular kernel tcp-cc through sysctl and setsockopt. e.g.
[root@arch-fb-vm1 bpf]# sysctl -a | egrep congestion
net.ipv4.tcp_allowed_congestion_control = reno cubic bpf_cubic
net.ipv4.tcp_available_congestion_control = reno bic cubic bpf_cubic
net.ipv4.tcp_congestion_control = bpf_cubic
There has been attempt to move the TCP CC to the user space
(e.g. CCP in TCP). The common arguments are faster turn around,
get away from long-tail kernel versions in production...etc,
which are legit points.
BPF has been the continuous effort to join both kernel and
userspace upsides together (e.g. XDP to gain the performance
advantage without bypassing the kernel). The recent BPF
advancements (in particular BTF-aware verifier, BPF trampoline,
BPF CO-RE...) made implementing kernel struct ops (e.g. tcp cc)
possible in BPF. It allows a faster turnaround for testing algorithm
in the production while leveraging the existing (and continue growing)
BPF feature/framework instead of building one specifically for
userspace TCP CC.
This patch allows write access to a few fields in tcp-sock
(in bpf_tcp_ca_btf_struct_access()).
The optional "get_info" is unsupported now. It can be added
later. One possible way is to output the info with a btf-id
to describe the content.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andriin@fb.com>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20200109003508.3856115-1-kafai@fb.com
|
|
Pablo Neira Ayuso says:
====================
Netfilter fixes for net
The following patchset contains Netfilter fixes for net:
1) Missing netns context in arp_tables, from Florian Westphal.
2) Underflow in flowtable reference counter, from wenxu.
3) Fix incorrect ethernet destination address in flowtable offload,
from wenxu.
4) Check for status of neighbour entry, from wenxu.
5) Fix NAT port mangling, from wenxu.
6) Unbind callbacks from destroy path to cleanup hardware properly
on flowtable removal.
7) Fix missing casting statistics timestamp, add nf_flowtable_time_stamp
and use it.
8) NULL pointer exception when timeout argument is null in conntrack
dccp and sctp protocol helpers, from Florian Westphal.
9) Possible nul-dereference in ipset with IPSET_ATTR_LINENO, also from
Florian.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The argument is always ignored, so remove it.
Signed-off-by: Li RongQing <lirongqing@baidu.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
REXMIT_NEW is a macro for "FRTO-style
transmit of unsent/new packets", this patch
makes it more readable.
Signed-off-by: Mao Wenan <maowenan@huawei.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add support for userspace to specify a device index to limit the scope
of an entry via the TCP_MD5SIG_EXT setsockopt. The existing __tcpm_pad
is renamed to tcpm_ifindex and the new field is only checked if the new
TCP_MD5SIG_FLAG_IFINDEX is set in tcpm_flags. For now, the device index
must point to an L3 master device (e.g., VRF). The API and error
handling are setup to allow the constraint to be relaxed in the future
to any device index.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add l3index to tcp_md5sig_key to represent the L3 domain of a key, and
add l3index to tcp_md5_do_add and tcp_md5_do_del to fill in the key.
With the key now based on an l3index, add the new parameter to the
lookup functions and consider the l3index when looking for a match.
The l3index comes from the skb when processing ingress packets leveraging
the helpers created for socket lookups, tcp_v4_sdif and inet_iif (and the
v6 variants). When the sdif index is set it means the packet ingressed a
device that is part of an L3 domain and inet_iif points to the VRF device.
For egress, the L3 domain is determined from the socket binding and
sk_bound_dev_if.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The original ingress device index is saved to the cb space of the skb
and the cb is moved during tcp processing. Since tcp_v4_inbound_md5_hash
can be called before and after the cb move, pass dif and sdif to it so
the caller can save both prior to the cb move. Both are used by a later
patch.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Extract the typecast to (union tcp_md5_addr *) to a local variable
rather than the current long, inline declaration with function calls.
No functional change intended.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When we receive a D-SACK, where the sequence number satisfies:
undo_marker <= start_seq < end_seq <= prior_snd_una
we consider this is a valid D-SACK and tcp_is_sackblock_valid()
returns true, then this D-SACK is discarded as "old stuff",
but the variable first_sack_index is not marked as negative
in tcp_sacktag_write_queue().
If this D-SACK also carries a SACK that needs to be processed
(for example, the previous SACK segment was lost), this SACK
will be treated as a D-SACK in the following processing of
tcp_sacktag_write_queue(), which will eventually lead to
incorrect updates of undo_retrans and reordering.
Fixes: fd6dad616d4f ("[TCP]: Earlier SACK block verification & simplify access to them")
Signed-off-by: Pengcheng Yang <yangpc@wangsu.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Simple overlapping changes in bpf land wrt. bpf_helper_defs.h
handling.
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
>From commit 50895b9de1d3 ("tcp: highest_sack fix"), the logic about
setting tp->highest_sack to the head of the send queue was removed.
Of course the logic is error prone, but it is logical. Before we
remove the pointer to the highest sack skb and use the seq instead,
we need to set tp->highest_sack to NULL when there is no skb after
the last sack, and then replace NULL with the real skb when new skb
inserted into the rtx queue, because the NULL means the highest sack
seq is tp->snd_nxt. If tp->highest_sack is NULL and new data sent,
the next ACK with sack option will increase tp->reordering unexpectedly.
This patch sets tp->highest_sack to the tail of the rtx queue if
it's NULL and new data is sent. The patch keeps the rule that the
highest_sack can only be maintained by sack processing, except for
this only case.
Fixes: 50895b9de1d3 ("tcp: highest_sack fix")
Signed-off-by: Cambda Zhu <cambda@linux.alibaba.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Neal Cardwell suggested to not change ca->delay_min
and apply the ack delay cushion only when Hystart ACK train
is still under consideration. This should avoid a 64bit
divide unless needed.
Tested:
40Gbit(mlx4) testbed (with sch_fq as packet scheduler)
$ echo -n 'file tcp_cubic.c +p' >/sys/kernel/debug/dynamic_debug/control
$ nstat -n;for f in {1..10}; do ./super_netperf 1 -H lpaa24 -l -4000000; done;nstat|egrep "Hystart"
14815
16280
15293
15563
11574
15145
14789
18548
16972
12520
TcpExtTCPHystartTrainDetect 10 0.0
TcpExtTCPHystartTrainCwnd 1396 0.0
$ dmesg | tail -10
[ 4873.951350] hystart_ack_train (116 > 93) delay_min 24 (+ ack_delay 69) cwnd 80
[ 4875.155379] hystart_ack_train (55 > 50) delay_min 21 (+ ack_delay 29) cwnd 160
[ 4876.333921] hystart_ack_train (69 > 62) delay_min 23 (+ ack_delay 39) cwnd 130
[ 4877.519037] hystart_ack_train (69 > 60) delay_min 22 (+ ack_delay 38) cwnd 130
[ 4878.701559] hystart_ack_train (87 > 63) delay_min 24 (+ ack_delay 39) cwnd 160
[ 4879.844597] hystart_ack_train (93 > 50) delay_min 21 (+ ack_delay 29) cwnd 216
[ 4880.956650] hystart_ack_train (74 > 67) delay_min 20 (+ ack_delay 47) cwnd 108
[ 4882.098500] hystart_ack_train (61 > 57) delay_min 23 (+ ack_delay 34) cwnd 130
[ 4883.262056] hystart_ack_train (72 > 67) delay_min 21 (+ ack_delay 46) cwnd 130
[ 4884.418760] hystart_ack_train (74 > 67) delay_min 29 (+ ack_delay 38) cwnd 152
10Gbit(bnx2x) testbed (with sch_fq as packet scheduler)
$ echo -n 'file tcp_cubic.c +p' >/sys/kernel/debug/dynamic_debug/control
$ nstat -n;for f in {1..10}; do ./super_netperf 1 -H lpk52 -l -4000000; done;nstat|egrep "Hystart"
7050
7065
7100
6900
7202
7263
7189
6869
7463
7034
TcpExtTCPHystartTrainDetect 10 0.0
TcpExtTCPHystartTrainCwnd 3199 0.0
$ dmesg | tail -10
[ 176.920012] hystart_ack_train (161 > 141) delay_min 83 (+ ack_delay 58) cwnd 264
[ 179.144645] hystart_ack_train (164 > 159) delay_min 120 (+ ack_delay 39) cwnd 444
[ 181.354527] hystart_ack_train (214 > 168) delay_min 125 (+ ack_delay 43) cwnd 436
[ 183.539565] hystart_ack_train (170 > 147) delay_min 96 (+ ack_delay 51) cwnd 326
[ 185.727309] hystart_ack_train (177 > 160) delay_min 61 (+ ack_delay 99) cwnd 128
[ 187.947142] hystart_ack_train (184 > 167) delay_min 123 (+ ack_delay 44) cwnd 367
[ 190.166680] hystart_ack_train (230 > 153) delay_min 116 (+ ack_delay 37) cwnd 444
[ 192.327285] hystart_ack_train (210 > 206) delay_min 86 (+ ack_delay 120) cwnd 152
[ 194.511392] hystart_ack_train (173 > 151) delay_min 94 (+ ack_delay 57) cwnd 239
[ 196.736023] hystart_ack_train (149 > 146) delay_min 105 (+ ack_delay 41) cwnd 399
Fixes: 42f3a8aaae66 ("tcp_cubic: tweak Hystart detection for short RTT flows")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Neal Cardwell <ncardwell@google.com>
Link: https://www.spinics.net/lists/netdev/msg621886.html
Link: https://www.spinics.net/lists/netdev/msg621797.html
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
We get crash when the targets checkentry function tries to make
use of the network namespace pointer for arptables.
When the net pointer got added back in 2010, only ip/ip6/ebtables were
changed to initialize it, so arptables has this set to NULL.
This isn't a problem for normal arptables because no existing
arptables target has a checkentry function that makes use of par->net.
However, direct users of the setsockopt interface can provide any
target they want as long as its registered for ARP or UNPSEC protocols.
syzkaller managed to send a semi-valid arptables rule for RATEEST target
which is enough to trigger NULL deref:
kasan: GPF could be caused by NULL-ptr deref or user memory access
general protection fault: 0000 [#1] PREEMPT SMP KASAN
RIP: xt_rateest_tg_checkentry+0x11d/0xb40 net/netfilter/xt_RATEEST.c:109
[..]
xt_check_target+0x283/0x690 net/netfilter/x_tables.c:1019
check_target net/ipv4/netfilter/arp_tables.c:399 [inline]
find_check_entry net/ipv4/netfilter/arp_tables.c:422 [inline]
translate_table+0x1005/0x1d70 net/ipv4/netfilter/arp_tables.c:572
do_replace net/ipv4/netfilter/arp_tables.c:977 [inline]
do_arpt_set_ctl+0x310/0x640 net/ipv4/netfilter/arp_tables.c:1456
Fixes: add67461240c1d ("netfilter: add struct net * to target parameters")
Reported-by: syzbot+d7358a458d8a81aee898@syzkaller.appspotmail.com
Signed-off-by: Florian Westphal <fw@strlen.de>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
For years we disabled Hystart ACK train detection at Google
because it was fooled by TCP pacing.
ACK train detection uses a simple heuristic, detecting if
we receive ACK past half the RTT, to exit slow start before
hitting the bottleneck and experience massive drops.
But pacing by design might delay packets up to RTT/2,
so we need to tweak the Hystart logic to be aware of this
extra delay.
Tested:
Added a 100 usec delay at receiver.
Before:
nstat -n;for f in {1..10}; do ./super_netperf 1 -H lpaa24 -l -4000000; done;nstat|egrep "Hystart"
9117
7057
9553
8300
7030
6849
9533
10126
6876
8473
TcpExtTCPHystartTrainDetect 10 0.0
TcpExtTCPHystartTrainCwnd 1230 0.0
After :
nstat -n;for f in {1..10}; do ./super_netperf 1 -H lpaa24 -l -4000000; done;nstat|egrep "Hystart"
9845
10103
10866
11096
11936
11487
11773
12188
11066
11894
TcpExtTCPHystartTrainDetect 10 0.0
TcpExtTCPHystartTrainCwnd 6462 0.0
Disabling Hystart ACK Train detection gives similar numbers
echo 2 >/sys/module/tcp_cubic/parameters/hystart_detect
nstat -n;for f in {1..10}; do ./super_netperf 1 -H lpaa24 -l -4000000; done;nstat|egrep "Hystart"
11173
10954
12455
10627
11578
11583
11222
10880
10665
11366
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
After switching ca->delay_min to usec resolution, we exit
slow start prematurely for very low RTT flows, setting
snd_ssthresh to 20.
The reason is that delay_min is fed with RTT of small packet
trains. Then as cwnd is increased, TCP sends bigger TSO packets.
LRO/GRO aggregation and/or interrupt mitigation strategies
on receiver tend to inflate RTT samples.
Fix this by adding to delay_min the expected delay of
two TSO packets, given current pacing rate.
Tested:
Sender uses pfifo_fast qdisc
Before :
$ nstat -n;for f in {1..10}; do ./super_netperf 1 -H lpaa24 -l -4000000; done;nstat|egrep "Hystart"
11348
11707
11562
11428
11773
11534
9878
11693
10597
10968
TcpExtTCPHystartTrainDetect 10 0.0
TcpExtTCPHystartTrainCwnd 200 0.0
After :
$ nstat -n;for f in {1..10}; do ./super_netperf 1 -H lpaa24 -l -4000000; done;nstat|egrep "Hystart"
14877
14517
15797
18466
17376
14833
17558
17933
16039
18059
TcpExtTCPHystartTrainDetect 10 0.0
TcpExtTCPHystartTrainCwnd 1670 0.0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Current 1ms clock feeds ca->round_start, ca->delay_min,
ca->last_ack.
This is quite problematic for data-center flows, where delay_min
is way below 1 ms.
This means Hystart Train detection triggers every time jiffies value
is updated, since "((s32)(now - ca->round_start) > ca->delay_min >> 4)"
expression becomes true.
This kind of random behavior can be solved by reusing the existing
usec timestamp that TCP keeps in tp->tcp_mstamp
Note that a followup patch will tweak things a bit, because
during slow start, GRO aggregation on receivers naturally
increases the RTT as TSO packets gradually come to ~64KB size.
To recap, right after this patch CUBIC Hystart train detection
is more aggressive, since short RTT flows might exit slow start at
cwnd = 20, instead of being possibly unbounded.
Following patch will address this problem.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
If we initialize ca->curr_rtt to ~0U, we do not need to test
for zero value in hystart_update()
We only read ca->curr_rtt if at least HYSTART_MIN_SAMPLES have
been processed, and thus ca->curr_rtt will have a sane value.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
We do not care which bit in ca->found is set.
We avoid accessing hystart and hystart_detect unless really needed,
possibly avoiding one cache line miss.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When do IPv6 tunnel PMTU update and calls __ip6_rt_update_pmtu() in the end,
we should not call dst_confirm_neigh() as there is no two-way communication.
Although vti and vti6 are immune to this problem because they are IFF_NOARP
interfaces, as Guillaume pointed. There is still no sense to confirm neighbour
here.
v5: Update commit description.
v4: No change.
v3: Do not remove dst_confirm_neigh, but add a new bool parameter in
dst_ops.update_pmtu to control whether we should do neighbor confirm.
Also split the big patch to small ones for each area.
v2: Remove dst_confirm_neigh in __ip6_rt_update_pmtu.
Reviewed-by: Guillaume Nault <gnault@redhat.com>
Acked-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When do tunnel PMTU update and calls __ip6_rt_update_pmtu() in the end,
we should not call dst_confirm_neigh() as there is no two-way communication.
v5: No Change.
v4: Update commit description
v3: Do not remove dst_confirm_neigh, but add a new bool parameter in
dst_ops.update_pmtu to control whether we should do neighbor confirm.
Also split the big patch to small ones for each area.
v2: Remove dst_confirm_neigh in __ip6_rt_update_pmtu.
Fixes: 0dec879f636f ("net: use dst_confirm_neigh for UDP, RAW, ICMP, L2TP")
Reviewed-by: Guillaume Nault <gnault@redhat.com>
Tested-by: Guillaume Nault <gnault@redhat.com>
Acked-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The MTU update code is supposed to be invoked in response to real
networking events that update the PMTU. In IPv6 PMTU update function
__ip6_rt_update_pmtu() we called dst_confirm_neigh() to update neighbor
confirmed time.
But for tunnel code, it will call pmtu before xmit, like:
- tnl_update_pmtu()
- skb_dst_update_pmtu()
- ip6_rt_update_pmtu()
- __ip6_rt_update_pmtu()
- dst_confirm_neigh()
If the tunnel remote dst mac address changed and we still do the neigh
confirm, we will not be able to update neigh cache and ping6 remote
will failed.
So for this ip_tunnel_xmit() case, _EVEN_ if the MTU is changed, we
should not be invoking dst_confirm_neigh() as we have no evidence
of successful two-way communication at this point.
On the other hand it is also important to keep the neigh reachability fresh
for TCP flows, so we cannot remove this dst_confirm_neigh() call.
To fix the issue, we have to add a new bool parameter for dst_ops.update_pmtu
to choose whether we should do neigh update or not. I will add the parameter
in this patch and set all the callers to true to comply with the previous
way, and fix the tunnel code one by one on later patches.
v5: No change.
v4: No change.
v3: Do not remove dst_confirm_neigh, but add a new bool parameter in
dst_ops.update_pmtu to control whether we should do neighbor confirm.
Also split the big patch to small ones for each area.
v2: Remove dst_confirm_neigh in __ip6_rt_update_pmtu.
Suggested-by: David Miller <davem@davemloft.net>
Reviewed-by: Guillaume Nault <gnault@redhat.com>
Acked-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When the size of the receive buffer for a socket is close to 2^31 when
computing if we have enough space in the buffer to copy a packet from
the queue to the buffer we might hit an integer overflow.
When an user set net.core.rmem_default to a value close to 2^31 UDP
packets are dropped because of this overflow. This can be visible, for
instance, with failure to resolve hostnames.
This can be fixed by casting sk_rcvbuf (which is an int) to unsigned
int, similarly to how it is done in TCP.
Signed-off-by: Antonio Messina <amessina@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Mere overlapping changes in the conflicts here.
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Pull networking fixes from David Miller:
1) Several nf_flow_table_offload fixes from Pablo Neira Ayuso,
including adding a missing ipv6 match description.
2) Several heap overflow fixes in mwifiex from qize wang and Ganapathi
Bhat.
3) Fix uninit value in bond_neigh_init(), from Eric Dumazet.
4) Fix non-ACPI probing of nxp-nci, from Stephan Gerhold.
5) Fix use after free in tipc_disc_rcv(), from Tuong Lien.
6) Enforce limit of 33 tail calls in mips and riscv JIT, from Paul
Chaignon.
7) Multicast MAC limit test is off by one in qede, from Manish Chopra.
8) Fix established socket lookup race when socket goes from
TCP_ESTABLISHED to TCP_LISTEN, because there lacks an intervening
RCU grace period. From Eric Dumazet.
9) Don't send empty SKBs from tcp_write_xmit(), also from Eric Dumazet.
10) Fix active backup transition after link failure in bonding, from
Mahesh Bandewar.
11) Avoid zero sized hash table in gtp driver, from Taehee Yoo.
12) Fix wrong interface passed to ->mac_link_up(), from Russell King.
13) Fix DSA egress flooding settings in b53, from Florian Fainelli.
14) Memory leak in gmac_setup_txqs(), from Navid Emamdoost.
15) Fix double free in dpaa2-ptp code, from Ioana Ciornei.
16) Reject invalid MTU values in stmmac, from Jose Abreu.
17) Fix refcount leak in error path of u32 classifier, from Davide
Caratti.
18) Fix regression causing iwlwifi firmware crashes on boot, from Anders
Kaseorg.
19) Fix inverted return value logic in llc2 code, from Chan Shu Tak.
20) Disable hardware GRO when XDP is attached to qede, frm Manish
Chopra.
21) Since we encode state in the low pointer bits, dst metrics must be
at least 4 byte aligned, which is not necessarily true on m68k. Add
annotations to fix this, from Geert Uytterhoeven.
* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (160 commits)
sfc: Include XDP packet headroom in buffer step size.
sfc: fix channel allocation with brute force
net: dst: Force 4-byte alignment of dst_metrics
selftests: pmtu: fix init mtu value in description
hv_netvsc: Fix unwanted rx_table reset
net: phy: ensure that phy IDs are correctly typed
mod_devicetable: fix PHY module format
qede: Disable hardware gro when xdp prog is installed
net: ena: fix issues in setting interrupt moderation params in ethtool
net: ena: fix default tx interrupt moderation interval
net/smc: unregister ib devices in reboot_event
net: stmmac: platform: Fix MDIO init for platforms without PHY
llc2: Fix return statement of llc_stat_ev_rx_null_dsap_xid_c (and _test_c)
net: hisilicon: Fix a BUG trigered by wrong bytes_compl
net: dsa: ksz: use common define for tag len
s390/qeth: don't return -ENOTSUPP to userspace
s390/qeth: fix promiscuous mode after reset
s390/qeth: handle error due to unsupported transport mode
cxgb4: fix refcount init for TC-MQPRIO offload
tc-testing: initial tdc selftests for cls_u32
...
|
|
sk->sk_pacing_shift can be read and written without lock
synchronization. This patch adds annotations to
document this fact and avoid future syzbot complains.
This might also avoid unexpected false sharing
in sk_pacing_shift_update(), as the compiler
could remove the conditional check and always
write over sk->sk_pacing_shift :
if (sk->sk_pacing_shift != val)
sk->sk_pacing_shift = val;
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Unlike mlxsw, the other listeners to the FIB notification chain do not
require any special modifications as they never considered multiple
identical routes.
This patch removes the old route notifications and converts all the
listeners to use the new replace / delete notifications.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When a new listener is registered to the FIB notification chain it
receives a dump of all the available routes in the system. Instead, make
sure to only replay the IPv4 routes that are actually used in the data
path and are of any interest to the new listener.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
In a similar fashion to previous patch, when a route is deleted as part
of table flushing, promote the next route in the list, if exists.
Otherwise, simply emit a delete notification.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When a route is deleted we potentially need to promote the next route in
the FIB alias list (e.g., with an higher metric). In case we find such a
route, a replace notification is emitted. Otherwise, a delete
notification for the deleted route.
v2:
* Convert to use fib_find_alias() instead of fib_find_first_alias()
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When a route is added, it should only be notified in case it is the
first route in the FIB alias list with the given {prefix, prefix length,
table ID}. Otherwise, it is not used in the data path and should not be
considered by switch drivers.
v2:
* Convert to use fib_find_alias() instead of fib_find_first_alias()
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When replacing a route, its replacement should only be notified in case
the replaced route is of any interest to listeners. In other words, if
the replaced route is currently used in the data path, which means it is
the first route in the FIB alias list with the given {prefix, prefix
length, table ID}.
v2:
* Convert to use fib_find_alias() instead of fib_find_first_alias()
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Extend the function with another argument, 'find_first'. When set, the
function returns the first FIB alias with the matching {prefix, prefix
length, table ID}. The TOS and priority parameters are ignored. Current
callers are converted to pass 'false' in order to maintain existing
behavior.
This will be used by subsequent patches in the series.
v2:
* New patch
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Suggested-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
