When query/report packets are received, mld module processes them.
But they are processed under BH context so it couldn't use sleepable
functions. So, in order to switch context, the two workqueues are
added which processes query and report event.
In the struct inet6_dev, mc_{query | report}_queue are added so it
is per-interface queue.
And mc_{query | report}_work are workqueue structure.
When the query or report event is received, skb is queued to proper
queue and worker function is scheduled immediately.
Workqueues and queues are protected by spinlock, which is
mc_{query | report}_lock, and worker functions are protected by RTNL.
Suggested-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The ifmcaddr6 has been protected by inet6_dev->lock(rwlock) so that
the critical section is atomic context. In order to switch this context,
changing locking is needed. The ifmcaddr6 actually already protected by
RTNL So if it's converted to use RCU, its control path context can be
switched to sleepable.
Suggested-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The ip6_sf_list has been protected by mca_lock(spin_lock) so that the
critical section is atomic context. In order to switch this context,
changing locking is needed. The ip6_sf_list actually already protected
by RTNL So if it's converted to use RCU, its control path context can
be switched to sleepable.
But It doesn't remove mca_lock yet because ifmcaddr6 isn't converted
to RCU yet. So, It's not fully converted to the sleepable context.
Suggested-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The sflist has been protected by rwlock so that the critical section
is atomic context.
In order to switch this context, changing locking is needed.
The sflist actually already protected by RTNL So if it's converted
to use RCU, its control path context can be switched to sleepable.
Suggested-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The purpose of mc_lock is to protect inet6_dev->mc_tomb.
But mc_tomb is already protected by RTNL and all functions,
which manipulate mc_tomb are called under RTNL.
So, mc_lock is not needed.
Furthermore, it is spinlock so the critical section is atomic.
In order to reduce atomic context, it should be removed.
Suggested-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
mcast.c has several timers for delaying works.
Timer's expire handler is working under atomic context so it can't use
sleepable things such as GFP_KERNEL, mutex, etc.
In order to use sleepable APIs, it converts from timers to delayed work.
But there are some critical sections, which is used by both process
and BH context. So that it still uses spin_lock_bh() and rwlock.
Suggested-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Many tcp sysctls are either bools or small ints that can fit into u8.
Reducing space taken by sysctls can save few cache line misses
when sending/receiving data while cpu caches are empty,
for example after cpu idle period.
This is hard to measure with typical network performance tests,
but after this patch, struct netns_ipv4 has shrunk
by three cache lines.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For these sysctls, their dedicated helpers have
to use proc_dou8vec_minmax().
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This sysctl uses ip_fwd_update_priority() helper,
so the conversion needs to change it.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
These sysctls that can fit in one byte instead of one int
are converted to save space and thus reduce cache line misses.
- icmp_echo_ignore_all, icmp_echo_ignore_broadcasts,
- icmp_ignore_bogus_error_responses, icmp_errors_use_inbound_ifaddr
- tcp_ecn, tcp_ecn_fallback
- ip_default_ttl, ip_no_pmtu_disc, ip_fwd_use_pmtu
- ip_nonlocal_bind, ip_autobind_reuse
- ip_dynaddr, ip_early_demux, raw_l3mdev_accept
- nexthop_compat_mode, fwmark_reflect
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
After commit 098a697b49 ("tcp_metrics: Use a single hash table
for all network namespaces."), tcpm_hash_bucket is local to
net/ipv4/tcp_metrics.c
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add an action to represent the PPPoE hardware offload support that
includes the session ID.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The switch might have already added the VLAN tag through PVID hardware
offload. Keep this extra VLAN in the flowtable but skip it on egress.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
If there is a forward path to reach an ethernet device and hardware
offload is enabled, then use the direct xmit path.
Moreover, store the real device in the direct xmit path info since
software datapath uses dev_hard_header() to push the layer encapsulation
headers while hardware offload refers to the real device.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add the vlan id and protocol to the flow tuple to uniquely identify
flows from the receive path. For the transmit path, dev_hard_header() on
the vlan device push the headers. This patch includes support for two
vlan headers (QinQ) from the ingress path.
Add a generic encap field to the flowtable entry which stores the
protocol and the tag id. This allows to reuse these fields in the PPPoE
support coming in a later patch.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The egress device in the tuple is obtained from route. Use
dev_fill_forward_path() instead to provide the real egress device for
this flow whenever this is available.
The new FLOW_OFFLOAD_XMIT_DIRECT type uses dev_queue_xmit() to transmit
ethernet frames. Cache the source and destination hardware address to
use dev_queue_xmit() to transfer packets.
The FLOW_OFFLOAD_XMIT_DIRECT replaces FLOW_OFFLOAD_XMIT_NEIGH if
dev_fill_forward_path() finds a direct transmit path.
In case of topology updates, if peer is moved to different bridge port,
the connection will time out, reconnect will result in a new entry with
the correct path. Snooping fdb updates would allow for cleaning up stale
flowtable entries.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Obtain the ingress device in the tuple from the route in the reply
direction. Use dev_fill_forward_path() instead to get the real ingress
device for this flow.
Fall back to use the ingress device that the IP forwarding route
provides if:
- dev_fill_forward_path() finds no real ingress device.
- the ingress device that is obtained is not part of the flowtable
devices.
- this route has a xfrm policy.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add the xmit_type field that defines the two supported xmit paths in the
flowtable data plane, which are the neighbour and the xfrm xmit paths.
This patch prepares for new flowtable xmit path types to come.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
I have a system with DSA ports, and udhcpcd is configured to bring
interfaces up as soon as they are created.
I create a bridge as follows:
ip link add br0 type bridge
As soon as I create the bridge and udhcpcd brings it up, I also have
avahi which automatically starts sending IPv6 packets to advertise some
local services, and because of that, the br0 bridge joins the following
IPv6 groups due to the code path detailed below:
33:33:ff:6d:c1:9c vid 0
33:33:00:00:00:6a vid 0
33:33:00:00:00:fb vid 0
br_dev_xmit
-> br_multicast_rcv
-> br_ip6_multicast_add_group
-> __br_multicast_add_group
-> br_multicast_host_join
-> br_mdb_notify
This is all fine, but inside br_mdb_notify we have br_mdb_switchdev_host
hooked up, and switchdev will attempt to offload the host joined groups
to an empty list of ports. Of course nobody offloads them.
Then when we add a port to br0:
ip link set swp0 master br0
the bridge doesn't replay the host-joined MDB entries from br_add_if,
and eventually the host joined addresses expire, and a switchdev
notification for deleting it is emitted, but surprise, the original
addition was already completely missed.
The strategy to address this problem is to replay the MDB entries (both
the port ones and the host joined ones) when the new port joins the
bridge, similar to what vxlan_fdb_replay does (in that case, its FDB can
be populated and only then attached to a bridge that you offload).
However there are 2 possibilities: the addresses can be 'pushed' by the
bridge into the port, or the port can 'pull' them from the bridge.
Considering that in the general case, the new port can be really late to
the party, and there may have been many other switchdev ports that
already received the initial notification, we would like to avoid
delivering duplicate events to them, since they might misbehave. And
currently, the bridge calls the entire switchdev notifier chain, whereas
for replaying it should just call the notifier block of the new guy.
But the bridge doesn't know what is the new guy's notifier block, it
just knows where the switchdev notifier chain is. So for simplification,
we make this a driver-initiated pull for now, and the notifier block is
passed as an argument.
To emulate the calling context for mdb objects (deferred and put on the
blocking notifier chain), we must iterate under RCU protection through
the bridge's mdb entries, queue them, and only call them once we're out
of the RCU read-side critical section.
There was some opportunity for reuse between br_mdb_switchdev_host_port,
br_mdb_notify and the newly added br_mdb_queue_one in how the switchdev
mdb object is created, so a helper was created.
Suggested-by: Ido Schimmel <idosch@idosch.org>
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Acked-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Problem:
The "lapb_t1timer_running" function in "lapb_timer.c" is used in only
one place: in the "lapb_kick" function in "lapb_out.c". "lapb_kick" calls
"lapb_t1timer_running" to check if the timer is already pending, and if
it is not, schedule it to run.
However, if the timer has already fired and is running, and is waiting to
get the "lapb->lock" lock, "lapb_t1timer_running" will not detect this,
and "lapb_kick" will then schedule a new timer. The old timer will then
abort when it sees a new timer pending.
I think this is not right. The purpose of "lapb_kick" should be ensuring
that the actual work of the timer function is scheduled to be done.
If the timer function is already running but waiting for the lock,
"lapb_kick" should not abort and reschedule it.
Changes made:
I added a new field "t1timer_running" in "struct lapb_cb" for
"lapb_t1timer_running" to use. "t1timer_running" will accurately reflect
whether the actual work of the timer is pending. If the timer has fired
but is still waiting for the lock, "t1timer_running" will still correctly
reflect whether the actual work is waiting to be done.
The old "t1timer_stop" field, whose only responsibility is to ask a timer
(that is already running but waiting for the lock) to abort, is no longer
needed, because the new "t1timer_running" field can fully take over its
responsibility. Therefore "t1timer_stop" is deleted.
"t1timer_running" is not simply a negation of the old "t1timer_stop".
At the end of the timer function, if it does not reschedule itself,
"t1timer_running" is set to false to indicate that the timer is stopped.
For consistency of the code, I also added "t2timer_running" and deleted
"t2timer_stop".
Signed-off-by: Xie He <xie.he.0141@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pablo Neira Ayuso says:
====================
Netfilter updates for net-next
The following batch contains Netfilter updates for net-next:
1) Split flowtable workqueues per events, from Oz Shlomo.
2) fall-through warnings for clang, from Gustavo A. R. Silva
3) Remove unused declaration in conntrack, from YueHaibing.
4) Consolidate skb_try_make_writable() in flowtable datapath,
simplify some of the existing codebase.
5) Call dst_check() to fall back to static classic forwarding path.
6) Update table flags from commit phase.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
call_gro_receive() is used to limit GRO recursion, but it works only
with callback pointers.
There's a combined version of call_gro_receive() + INDIRECT_CALL_2()
in <net/inet_common.h>, but it doesn't check for IPv6 modularity.
Add a similar new helper to cover both of these. It can and will be
used to avoid retpoline overhead when IP header lies behind another
offloaded proto.
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
If some source file includes <net/gro.h>, but doesn't include
<linux/indirect_call_wrapper.h>:
In file included from net/8021q/vlan_core.c:7:
./include/net/gro.h:6:1: warning: data definition has no type or storage class
6 | INDIRECT_CALLABLE_DECLARE(struct sk_buff *ipv6_gro_receive(struct list_head *,
| ^~~~~~~~~~~~~~~~~~~~~~~~~
./include/net/gro.h:6:1: error: type defaults to ‘int’ in declaration of ‘INDIRECT_CALLABLE_DECLARE’ [-Werror=implicit-int]
[...]
Include <linux/indirect_call_wrapper.h> directly. It's small and
won't pull lots of dependencies.
Also add some incomplete struct declarations to be fully stacked.
Fixes: 04f00ab227 ("net/core: move gro function declarations to separate header ")
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pablo Neira Ayuso says:
====================
Netfilter fixes for net
1) Several patches to testore use of memory barriers instead of RCU to
ensure consistent access to ruleset, from Mark Tomlinson.
2) Fix dump of expectation via ctnetlink, from Florian Westphal.
3) GRE helper works for IPv6, from Ludovic Senecaux.
4) Set error on unsupported flowtable flags.
5) Use delayed instead of deferrable workqueue in the flowtable,
from Yinjun Zhang.
6) Fix spurious EEXIST in case of add-after-delete flowtable in
the same batch.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
In order for a driver to be able to query a bridge for information
about itself, e.g. reading out port flags, it has to use a netdev that
is known to the bridge. In the simple case, that is just the netdev
representing the port, e.g. swp0 or swp1 in this example:
br0
/ \
swp0 swp1
But in the case of an offloaded lag, this will be the bond or team
interface, e.g. bond0 in this example:
br0
/
bond0
/ \
swp0 swp1
Add a helper that hides some of this complexity from the
drivers. Then, redefine dsa_port_offloads_bridge_port using the helper
to avoid double accounting of the set of possible offloaded uppers.
Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Do not update table flags from the preparation phase. Store the flags
update into the transaction, then update the flags from the commit
phase.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Simplify existing fast NAT routines by returning void. After the
skb_try_make_writable() call consolidation, these routines cannot ever
fail.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This allows to remove the default case which should not ever happen and
that was added to avoid gcc warnings on unhandled FLOW_OFFLOAD_DIR_MAX
enumeration case.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
commit e97c3e278e ("tproxy: split off ipv6 defragmentation to a separate
module") left behind this.
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Honor flowtable flags from the control update path. Disallow disabling
to toggle hardware offload support though.
Fixes: 8bb69f3b29 ("netfilter: nf_tables: add flowtable offload control plane")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Add support for legacy Broadcom tags, which are similar to DSA_TAG_PROTO_BRCM.
These tags are used on BCM5325, BCM5365 and BCM63xx switches.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a non-initial netns is destroyed, the usual policy is to delete
all virtual network interfaces contained, but move physical interfaces
back to the initial netns. This keeps the physical interface visible
on the system.
CAN devices are somewhat special, as they define rtnl_link_ops even
if they are physical devices. If a CAN interface is moved into a
non-initial netns, destroying that netns lets the interface vanish
instead of moving it back to the initial netns. default_device_exit()
skips CAN interfaces due to having rtnl_link_ops set. Reproducer:
ip netns add foo
ip link set can0 netns foo
ip netns delete foo
WARNING: CPU: 1 PID: 84 at net/core/dev.c:11030 ops_exit_list+0x38/0x60
CPU: 1 PID: 84 Comm: kworker/u4:2 Not tainted 5.10.19 #1
Workqueue: netns cleanup_net
[<c010e700>] (unwind_backtrace) from [<c010a1d8>] (show_stack+0x10/0x14)
[<c010a1d8>] (show_stack) from [<c086dc10>] (dump_stack+0x94/0xa8)
[<c086dc10>] (dump_stack) from [<c086b938>] (__warn+0xb8/0x114)
[<c086b938>] (__warn) from [<c086ba10>] (warn_slowpath_fmt+0x7c/0xac)
[<c086ba10>] (warn_slowpath_fmt) from [<c0629f20>] (ops_exit_list+0x38/0x60)
[<c0629f20>] (ops_exit_list) from [<c062a5c4>] (cleanup_net+0x230/0x380)
[<c062a5c4>] (cleanup_net) from [<c0142c20>] (process_one_work+0x1d8/0x438)
[<c0142c20>] (process_one_work) from [<c0142ee4>] (worker_thread+0x64/0x5a8)
[<c0142ee4>] (worker_thread) from [<c0148a98>] (kthread+0x148/0x14c)
[<c0148a98>] (kthread) from [<c0100148>] (ret_from_fork+0x14/0x2c)
To properly restore physical CAN devices to the initial netns on owning
netns exit, introduce a flag on rtnl_link_ops that can be set by drivers.
For CAN devices setting this flag, default_device_exit() considers them
non-virtual, applying the usual namespace move.
The issue was introduced in the commit mentioned below, as at that time
CAN devices did not have a dellink() operation.
Fixes: e008b5fc8d ("net: Simplfy default_device_exit and improve batching.")
Link: https://lore.kernel.org/r/20210302122423.872326-1-martin@strongswan.org
Signed-off-by: Martin Willi <martin@strongswan.org>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Currently tcp_check_req can be called with obsolete req socket for which big
socket have been already created (because of CPU race or early demux
assigning req socket to multiple packets in gro batch).
Commit e0f9759f53 ("tcp: try to keep packet if SYN_RCV race
is lost") added retry in case when tcp_check_req is called for PSH|ACK packet.
But if client sends RST+ACK immediatly after connection being
established (it is performing healthcheck, for example) retry does not
occur. In that case tcp_check_req tries to close req socket,
leaving big socket active.
Fixes: e0f9759f53 ("tcp: try to keep packet if SYN_RCV race is lost")
Signed-off-by: Alexander Ovechkin <ovov@yandex-team.ru>
Reported-by: Oleg Senin <olegsenin@yandex-team.ru>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extend psample to report the following attributes when available:
* Output traffic class as a 16-bit value
* Output traffic class occupancy in bytes as a 64-bit value
* End-to-end latency of the packet in nanoseconds resolution
* Software timestamp in nanoseconds resolution (always available)
* Packet's protocol. Needed for packet dissection in user space (always
available)
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, callers of psample_sample_packet() pass three metadata
attributes: Ingress port, egress port and truncated size. Subsequent
patches are going to add more attributes (e.g., egress queue occupancy),
which also need an indication whether they are valid or not.
Encapsulate packet metadata in a struct in order to keep the number of
arguments reasonable.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Flow Dissector code never modifies the input buffer, neither skb nor
raw data.
Make 'data' argument const for all of the Flow dissector's functions.
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
BPF Flow dissection programs are read-only and don't touch input
buffers.
Mark 'data' and 'data_end' in struct bpf_flow_dissector as const
in preparation for global input constifying.
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
Allow a policer action to enforce a rate-limit based on packets-per-second,
configurable using a packet-per-second rate and burst parameters.
e.g.
tc filter add dev tap1 parent ffff: u32 match \
u32 0 0 police pkts_rate 3000 pkts_burst 1000
Testing was unable to uncover a performance impact of this change on
existing features.
Signed-off-by: Baowen Zheng <baowen.zheng@corigine.com>
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: Louis Peens <louis.peens@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Allow flow_offload API to configure packet-per-second policing using rate
and burst parameters.
Dummy implementations of tcf_police_rate_pkt_ps() and
tcf_police_burst_pkt() are supplied which return 0, the unconfigured state.
This is to facilitate splitting the offload, driver, and TC code portion of
this feature into separate patches with the aim of providing a logical flow
for review. And the implementation of these helpers will be filled out by a
follow-up patch.
Signed-off-by: Xingfeng Hu <xingfeng.hu@corigine.com>
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: Louis Peens <louis.peens@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch defined a new struct mptcp_rm_list, the ids field was an
array of the removing address ids, the nr field was the valid number of
removing address ids in the array. The array size was definced as a new
macro MPTCP_RM_IDS_MAX. Changed the member rm_id of struct
mptcp_out_options to rm_list.
In mptcp_established_options_rm_addr, invoked mptcp_pm_rm_addr_signal to
get the rm_list. According the number of addresses in it, calculated
the padded RM_ADDR suboption length. And saved the ids array in struct
mptcp_out_options's rm_list member.
In mptcp_write_options, iterated each address id from struct
mptcp_out_options's rm_list member, set the invalid ones as TCPOPT_NOP,
then filled them into the RM_ADDR suboption.
Changed TCPOLEN_MPTCP_RM_ADDR_BASE from 4 to 3.
Signed-off-by: Geliang Tang <geliangtang@gmail.com>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The "backlog" argument in listen() specifies
the maximom length of pending connections,
so the accept queue should be considered full
if there are exactly "backlog" elements.
Signed-off-by: liuyacan <yacanliu@163.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The kernel periodically checks the idle time of nexthop buckets to
determine if they are idle and can be re-populated with a new nexthop.
When the resilient nexthop group is offloaded to hardware, the kernel
will not see activity on nexthop buckets unless it is reported from
hardware.
Add a function that can be periodically called by device drivers to
report activity on nexthop buckets after querying it from the underlying
device.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a function that can be called by device drivers to set "offload" or
"trap" indication on nexthop buckets following nexthop notifications and
other changes such as a neighbour becoming invalid.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add data structures that will be used for in-kernel notifications about
addition / deletion of a resilient nexthop group and about changes to a
hash bucket within a resilient group.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
At this moment, there is only one type of next-hop group: an mpath group,
which implements the hash-threshold algorithm.
To select a next hop, hash-threshold algorithm first assigns a range of
hashes to each next hop in the group, and then selects the next hop by
comparing the SKB hash with the individual ranges. When a next hop is
removed from the group, the ranges are recomputed, which leads to
reassignment of parts of hash space from one next hop to another. While
there will usually be some overlap between the previous and the new
distribution, some traffic flows change the next hop that they resolve to.
That causes problems e.g. as established TCP connections are reset, because
the traffic is forwarded to a server that is not familiar with the
connection.
Resilient hashing is a technique to address the above problem. Resilient
next-hop group has another layer of indirection between the group itself
and its constituent next hops: a hash table. The selection algorithm uses a
straightforward modulo operation to choose a hash bucket, and then reads
the next hop that this bucket contains, and forwards traffic there.
This indirection brings an important feature. In the hash-threshold
algorithm, the range of hashes associated with a next hop must be
continuous. With a hash table, mapping between the hash table buckets and
the individual next hops is arbitrary. Therefore when a next hop is deleted
the buckets that held it are simply reassigned to other next hops. When
weights of next hops in a group are altered, it may be possible to choose a
subset of buckets that are currently not used for forwarding traffic, and
use those to satisfy the new next-hop distribution demands, keeping the
"busy" buckets intact. This way, established flows are ideally kept being
forwarded to the same endpoints through the same paths as before the
next-hop group change.
In a nutshell, the algorithm works as follows. Each next hop has a number
of buckets that it wants to have, according to its weight and the number of
buckets in the hash table. In case of an event that might cause bucket
allocation change, the numbers for individual next hops are updated,
similarly to how ranges are updated for mpath group next hops. Following
that, a new "upkeep" algorithm runs, and for idle buckets that belong to a
next hop that is currently occupying more buckets than it wants (it is
"overweight"), it migrates the buckets to one of the next hops that has
fewer buckets than it wants (it is "underweight"). If, after this, there
are still underweight next hops, another upkeep run is scheduled to a
future time.
Chances are there are not enough "idle" buckets to satisfy the new demands.
The algorithm has knobs to select both what it means for a bucket to be
idle, and for whether and when to forcefully migrate buckets if there keeps
being an insufficient number of idle buckets.
There are three users of the resilient data structures.
- The forwarding code accesses them under RCU, and does not modify them
except for updating the time a selected bucket was last used.
- Netlink code, running under RTNL, which may modify the data.
- The delayed upkeep code, which may modify the data. This runs unlocked,
and mutual exclusion between the RTNL code and the delayed upkeep is
maintained by canceling the delayed work synchronously before the RTNL
code touches anything. Later it restarts the delayed work if necessary.
The RTNL code has to implement next-hop group replacement, next hop
removal, etc. For removal, the mpath code uses a neat trick of having a
backup next hop group structure, doing the necessary changes offline, and
then RCU-swapping them in. However, the hash tables for resilient hashing
are about an order of magnitude larger than the groups themselves (the size
might be e.g. 4K entries), and it was felt that keeping two of them is an
overkill. Both the primary next-hop group and the spare therefore use the
same resilient table, and writers are careful to keep all references valid
for the forwarding code. The hash table references next-hop group entries
from the next-hop group that is currently in the primary role (i.e. not
spare). During the transition from primary to spare, the table references a
mix of both the primary group and the spare. When a next hop is deleted,
the corresponding buckets are not set to NULL, but instead marked as empty,
so that the pointer is valid and can be used by the forwarding code. The
buckets are then migrated to a new next-hop group entry during upkeep. The
only times that the hash table is invalid is the very beginning and very
end of its lifetime. Between those points, it is always kept valid.
This patch introduces the core support code itself. It does not handle
notifications towards drivers, which are kept as if the group were an mpath
one. It does not handle netlink either. The only bit currently exposed to
user space is the new next-hop group type, and that is currently bounced.
There is therefore no way to actually access this code.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
With the introduction of resilient nexthop groups, there will be two types
of multipath groups: the current hash-threshold "mpath" ones, and resilient
groups. Both are multipath, but to determine the fact, the system needs to
consider two flags. This might prove costly in the datapath. Therefore,
introduce a new flag, that should be set for next-hop groups that have more
than one nexthop, and should be considered multipath.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
As specified in IETF RFC 8754, section 4.3.1.2, if the upper layer
header is IPv4 or IPv6, perform IPv6 decapsulation and resubmit the
decapsulated packet to the IPv4 or IPv6 module.
Only IPv6 decapsulation was implemented. This patch adds support for IPv4
decapsulation.
Link: https://tools.ietf.org/html/rfc8754#section-4.3.1.2
Signed-off-by: Julien Massonneau <julien.massonneau@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
