Allow the gateway in a fib_nh_common to be from a different address
family than the outer fib{6}_nh. To that end, replace nhc_has_gw with
nhc_gw_family and update users of nhc_has_gw to check nhc_gw_family.
Now nhc_family is used to know if the nh_common is part of a fib_nh
or fib6_nh (used for container_of to get to route family specific data),
and nhc_gw_family represents the address family for the gateway.
Signed-off-by: David Ahern <dsahern@gmail.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
After commit a297569fe0 ("net/udp: do not touch skb->peeked unless
really needed") the 'peeked' argument of __skb_try_recv_datagram()
and friends is always equal to !!'flags & MSG_PEEK'.
Since such argument is really a boolean info, and the callers have
already 'flags & MSG_PEEK' handy, we can remove it and clean-up the
code a bit.
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch changes rhashtables to use a bit_spin_lock on BIT(1) of the
bucket pointer to lock the hash chain for that bucket.
The benefits of a bit spin_lock are:
- no need to allocate a separate array of locks.
- no need to have a configuration option to guide the
choice of the size of this array
- locking cost is often a single test-and-set in a cache line
that will have to be loaded anyway. When inserting at, or removing
from, the head of the chain, the unlock is free - writing the new
address in the bucket head implicitly clears the lock bit.
For __rhashtable_insert_fast() we ensure this always happens
when adding a new key.
- even when lockings costs 2 updates (lock and unlock), they are
in a cacheline that needs to be read anyway.
The cost of using a bit spin_lock is a little bit of code complexity,
which I think is quite manageable.
Bit spin_locks are sometimes inappropriate because they are not fair -
if multiple CPUs repeatedly contend of the same lock, one CPU can
easily be starved. This is not a credible situation with rhashtable.
Multiple CPUs may want to repeatedly add or remove objects, but they
will typically do so at different buckets, so they will attempt to
acquire different locks.
As we have more bit-locks than we previously had spinlocks (by at
least a factor of two) we can expect slightly less contention to
go with the slightly better cache behavior and reduced memory
consumption.
To enhance type checking, a new struct is introduced to represent the
pointer plus lock-bit
that is stored in the bucket-table. This is "struct rhash_lock_head"
and is empty. A pointer to this needs to be cast to either an
unsigned lock, or a "struct rhash_head *" to be useful.
Variables of this type are most often called "bkt".
Previously "pprev" would sometimes point to a bucket, and sometimes a
->next pointer in an rhash_head. As these are now different types,
pprev is NULL when it would have pointed to the bucket. In that case,
'blk' is used, together with correct locking protocol.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The non-null check on tskb is always false because it is in an else
path of a check on tskb and hence tskb is null in this code block.
This is check is therefore redundant and can be removed as well
as the label coalesc.
if (tsbk) {
...
} else {
...
if (unlikely(!skb)) {
if (tskb) /* can never be true, redundant code */
goto coalesc;
return;
}
}
Addresses-Coverity: ("Logically dead code")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: Mukesh Ojha <mojha@codeaurora.org>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Minor comment merge conflict in mlx5.
Staging driver has a fixup due to the skb->xmit_more changes
in 'net-next', but was removed in 'net'.
Signed-off-by: David S. Miller <davem@davemloft.net>
Linux currently disable ECN for incoming connections when the SYN
requests ECN and the IP header has ECT(0)/ECT(1) set, as some
networks were reportedly mangling the ToS byte, hence could later
trigger false congestion notifications.
RFC8311 §4.3 relaxes RFC3168's requirements such that ECT can be set
one TCP control packets (including SYNs). The main benefit of this
is the decreased probability of losing a SYN in a congested
ECN-capable network (i.e., it avoids the initial 1s timeout).
Additionally, this allows the development of newer TCP extensions,
such as AccECN.
This patch relaxes the previous check, by enabling ECN on incoming
connections using SYN+ECT if at least one bit of the reserved flags
of the TCP header is set. Such bit would indicate that the sender of
the SYN is using a newer TCP feature than what the host implements,
such as AccECN, and is thus implementing RFC8311. This enables
end-hosts not supporting such extensions to still negociate ECN, and
to have some of the benefits of using ECN on control packets.
Signed-off-by: Olivier Tilmans <olivier.tilmans@nokia-bell-labs.com>
Suggested-by: Bob Briscoe <research@bobbriscoe.net>
Cc: Koen De Schepper <koen.de_schepper@nokia-bell-labs.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
RFC8257 §3.5 explicitly states that "A DCTCP sender MUST react to
loss episodes in the same way as conventional TCP".
Currently, Linux DCTCP performs no cwnd reduction when losses
are encountered. Optionally, the dctcp_clamp_alpha_on_loss resets
alpha to its maximal value if a RTO happens. This behavior
is sub-optimal for at least two reasons: i) it ignores losses
triggering fast retransmissions; and ii) it causes unnecessary large
cwnd reduction in the future if the loss was isolated as it resets
the historical term of DCTCP's alpha EWMA to its maximal value (i.e.,
denoting a total congestion). The second reason has an especially
noticeable effect when using DCTCP in high BDP environments, where
alpha normally stays at low values.
This patch replace the clamping of alpha by setting ssthresh to
half of cwnd for both fast retransmissions and RTOs, at most once
per RTT. Consequently, the dctcp_clamp_alpha_on_loss module parameter
has been removed.
The table below shows experimental results where we measured the
drop probability of a PIE AQM (not applying ECN marks) at a
bottleneck in the presence of a single TCP flow with either the
alpha-clamping option enabled or the cwnd halving proposed by this
patch. Results using reno or cubic are given for comparison.
| Link | RTT | Drop
TCP CC | speed | base+AQM | probability
==================|=========|==========|============
CUBIC | 40Mbps | 7+20ms | 0.21%
RENO | | | 0.19%
DCTCP-CLAMP-ALPHA | | | 25.80%
DCTCP-HALVE-CWND | | | 0.22%
------------------|---------|----------|------------
CUBIC | 100Mbps | 7+20ms | 0.03%
RENO | | | 0.02%
DCTCP-CLAMP-ALPHA | | | 23.30%
DCTCP-HALVE-CWND | | | 0.04%
------------------|---------|----------|------------
CUBIC | 800Mbps | 1+1ms | 0.04%
RENO | | | 0.05%
DCTCP-CLAMP-ALPHA | | | 18.70%
DCTCP-HALVE-CWND | | | 0.06%
We see that, without halving its cwnd for all source of losses,
DCTCP drives the AQM to large drop probabilities in order to keep
the queue length under control (i.e., it repeatedly faces RTOs).
Instead, if DCTCP reacts to all source of losses, it can then be
controlled by the AQM using similar drop levels than cubic or reno.
Signed-off-by: Koen De Schepper <koen.de_schepper@nokia-bell-labs.com>
Signed-off-by: Olivier Tilmans <olivier.tilmans@nokia-bell-labs.com>
Cc: Bob Briscoe <research@bobbriscoe.net>
Cc: Lawrence Brakmo <brakmo@fb.com>
Cc: Florian Westphal <fw@strlen.de>
Cc: Daniel Borkmann <borkmann@iogearbox.net>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Andrew Shewmaker <agshew@gmail.com>
Cc: Glenn Judd <glenn.judd@morganstanley.com>
Acked-by: Florian Westphal <fw@strlen.de>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Just like 46cfd725c3 ("net: use kfree_skb_list() helper in more places").
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Export fib_nexthop_info and fib_add_nexthop for use by IPv6 code.
Remove rt6_nexthop_info and rt6_add_nexthop in favor of the IPv4
versions. Update fib_nexthop_info for IPv6 linkdown check and
RTA_GATEWAY for AF_INET6.
Signed-off-by: David Ahern <dsahern@gmail.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With the exception of the nexthop weight, the nexthop attributes used by
fib_nexthop_info and fib_add_nexthop come from the fib_nh_common struct.
Update both to use it and change fib_nexthop_info to check the family
as needed.
nexthop weight comes from the common struct for existing use cases, but
for nexthop groups the weight is outside of the fib_nh_common to allow
the same nexthop definition to be used in multiple groups with different
weights.
Signed-off-by: David Ahern <dsahern@gmail.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Similar to ipv6, move addition of nexthop attributes to dump
message into helpers that are called for both single path and
multipath routes. Align the new helpers to the IPv6 variant
which most notably means computing the flags argument based on
settings in nh_flags.
The RTA_FLOW argument is unique to IPv4, so it is appended after
the new fib_nexthop_info helper. The intent of a later patch is to
make both fib_nexthop_info and fib_add_nexthop usable for both IPv4
and IPv6. This patch is stepping stone in that direction.
Signed-off-by: David Ahern <dsahern@gmail.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Most of the ipv4 code only needs data from fib_nh_common. Add
fib_nh_common selection to fib_result and update users to use it.
Right now, fib_nh_common in fib_result will point to a fib_nh struct
that is embedded within a fib_info:
fib_info --> fib_nh
fib_nh
...
fib_nh
^
fib_result->nhc ----+
Later, nhc can point to a fib_nh within a nexthop struct:
fib_info --> nexthop --> fib_nh
^
fib_result->nhc ---------------+
or for a nexthop group:
fib_info --> nexthop --> nexthop --> fib_nh
nexthop --> fib_nh
...
nexthop --> fib_nh
^
fib_result->nhc ---------------------------+
In all cases nhsel within fib_result will point to which leg in the
multipath route is used.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Update fib_table_lookup tracepoint to take a fib_nh_common struct and
dump the v6 gateway address if the nexthop uses it.
Over the years saddr has not proven useful and the output of the
tracepoint produces very long lines. Since saddr is not part of
fib_nh_common, drop it. If it needs to be added later, fib_nh which
contains saddr can be obtained from a fib_nh_common via container_of.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As Eric noticed, in .sk_destruct, sk->sk_dst_cache update is prevented, and
no barrier is needed for this. So change to use rcu_dereference_protected()
instead of rcu_dereference_check() to fetch sk_dst_cache in there.
v1->v2:
- no change, repost after net-next is open.
Reported-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Configuration check to accept source route IP options should be made on
the incoming netdevice when the skb->dev is an l3mdev master. The route
lookup for the source route next hop also needs the incoming netdev.
v2->v3:
- Simplify by passing the original netdevice down the stack (per David
Ahern).
Signed-off-by: Stephen Suryaputra <ssuryaextr@gmail.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When tcp_sk_init() failed in inet_ctl_sock_create(),
'net->ipv4.tcp_congestion_control' will be left
uninitialized, but tcp_sk_exit() hasn't check for
that.
This patch add checking on 'net->ipv4.tcp_congestion_control'
in tcp_sk_exit() to prevent NULL-ptr dereference.
Fixes: 6670e15244 ("tcp: Namespace-ify sysctl_tcp_default_congestion_control")
Signed-off-by: Dust Li <dust.li@linux.alibaba.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
TCP stack relies on the fact that a freshly allocated skb
has skb->cb[] and skb_shinfo(skb)->tx_flags cleared.
When recycling tx skb, we must ensure these fields are cleared.
Fixes: 472c2e07ee ("tcp: add one skb cache for tx")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With fib_nh_common in place, move common initialization and release
code into helpers used by both ipv4 and ipv6. For the moment, the init
is just the lwt encap and the release is both the netdev reference and
the the lwt state reference. More will be added later.
Signed-off-by: David Ahern <dsahern@gmail.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add fib_nh_common struct with common nexthop attributes. Convert
fib_nh and fib6_nh to use it. Use macros to move existing
fib_nh_* references to the new nh_common.nhc_*.
Signed-off-by: David Ahern <dsahern@gmail.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Rename fib_nh entries that will be moved to a fib_nh_common struct.
Specifically, the device, oif, gateway, flags, scope, lwtstate,
nh_weight and nh_upper_bound are common with all nexthop definitions.
In the process shorten fib_nh_lwtstate to fib_nh_lws to avoid really
long lines.
Rename only; no functional change intended.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Move the fib_nh cleanup code from free_fib_info_rcu into a new helper,
fib_nh_release. Move classid accounting into fib_nh_release which is
called per fib_nh to make accounting symmetrical with fib_nh_init.
Export the helper to allow for use with nexthop objects in the
future.
Signed-off-by: David Ahern <dsahern@gmail.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Consolidate the fib_nh initialization which is duplicated between
fib_create_info for single path and fib_get_nhs for multipath.
Export the helper to allow for use with nexthop objects in the
future.
Signed-off-by: David Ahern <dsahern@gmail.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
in_dev lookup followed by IN_DEV_IGNORE_ROUTES_WITH_LINKDOWN check
is called in several places, some with the rcu lock and others with the
rtnl held.
Move the check to a helper similar to what IPv6 has. Since the helper
can be invoked from either context use rcu_dereference_rtnl to
dereference ip_ptr.
Signed-off-by: David Ahern <dsahern@gmail.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Define fib_get_nhs to return EINVAL when CONFIG_IP_ROUTE_MULTIPATH is
not enabled and remove the ifdef check for CONFIG_IP_ROUTE_MULTIPATH
in fib_create_info.
Signed-off-by: David Ahern <dsahern@gmail.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
According to Amit Klein and Benny Pinkas, IP ID generation is too weak
and might be used by attackers.
Even with recent net_hash_mix() fix (netns: provide pure entropy for net_hash_mix())
having 64bit key and Jenkins hash is risky.
It is time to switch to siphash and its 128bit keys.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Amit Klein <aksecurity@gmail.com>
Reported-by: Benny Pinkas <benny@pinkas.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
My recent patch had at least three problems :
1) TX zerocopy wants notification when skb is acknowledged,
thus we need to call skb_zcopy_clear() if the skb is
cached into sk->sk_tx_skb_cache
2) Some applications might expect precise EPOLLOUT
notifications, so we need to update sk->sk_wmem_queued
and call sk_mem_uncharge() from sk_wmem_free_skb()
in all cases. The SOCK_QUEUE_SHRUNK flag must also be set.
3) Reuse of saved skb should have used skb_cloned() instead
of simply checking if the fast clone has been freed.
Fixes: 472c2e07ee ("tcp: add one skb cache for tx")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Tested-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
An FoU socket is currently bound to the wildcard-address. While this
works fine, there are several use-cases where the use of the
wildcard-address is not desirable. For example, I use FoU on some
multi-homed servers and would like to use FoU on only one of the
interfaces.
This commit adds support for binding FoU sockets to a given source
address/interface, as well as connecting the socket to a given
destination address/port. udp_tunnel already provides the required
infrastructure, so most of the code added is for exposing and setting
the different attributes (local address, peer address, etc.).
The lookups performed when we add, delete or get an FoU-socket has also
been updated to compare all the attributes a user can set. Since the
comparison now involves several elements, I have added a separate
comparison-function instead of open-coding.
In order to test the code and ensure that the new comparison code works
correctly, I started by creating a wildcard socket bound to port 1234 on
my machine. I then tried to create a non-wildcarded socket bound to the
same port, as well as fetching and deleting the socket (including source
address, peer address or interface index in the netlink request). Both
the create, fetch and delete request failed. Deleting/fetching the
socket was only successful when my netlink request attributes matched
those used to create the socket.
I then repeated the tests, but with a socket bound to a local ip
address, a socket bound to a local address + interface, and a bound
socket that was also «connected» to a peer. Add only worked when no
socket with the matching source address/interface (or wildcard) existed,
while fetch/delete was only successful when all attributes matched.
In addition to testing that the new code work, I also checked that the
current behavior is kept. If none of the new attributes are provided,
then an FoU-socket is configured as before (i.e., wildcarded). If any
of the new attributes are provided, the FoU-socket is configured as
expected.
v1->v2:
* Fixed building with IPv6 disabled (kbuild).
* Fixed a return type warning and make the ugly comparison function more
readable (kbuild).
* Describe more in detail what has been tested (thanks David Miller).
* Make peer port required if peer address is specified.
Signed-off-by: Kristian Evensen <kristian.evensen@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Often times, recvmsg() system calls and BH handling for a particular
TCP socket are done on different cpus.
This means the incoming skb had to be allocated on a cpu,
but freed on another.
This incurs a high spinlock contention in slab layer for small rpc,
but also a high number of cache line ping pongs for larger packets.
A full size GRO packet might use 45 page fragments, meaning
that up to 45 put_page() can be involved.
More over performing the __kfree_skb() in the recvmsg() context
adds a latency for user applications, and increase probability
of trapping them in backlog processing, since the BH handler
might found the socket owned by the user.
This patch, combined with the prior one increases the rpc
performance by about 10 % on servers with large number of cores.
(tcp_rr workload with 10,000 flows and 112 threads reach 9 Mpps
instead of 8 Mpps)
This also increases single bulk flow performance on 40Gbit+ links,
since in this case there are often two cpus working in tandem :
- CPU handling the NIC rx interrupts, feeding the receive queue,
and (after this patch) freeing the skbs that were consumed.
- CPU in recvmsg() system call, essentially 100 % busy copying out
data to user space.
Having at most one skb in a per-socket cache has very little risk
of memory exhaustion, and since it is protected by socket lock,
its management is essentially free.
Note that if rps/rfs is used, we do not enable this feature, because
there is high chance that the same cpu is handling both the recvmsg()
system call and the TCP rx path, but that another cpu did the skb
allocations in the device driver right before the RPS/RFS logic.
To properly handle this case, it seems we would need to record
on which cpu skb was allocated, and use a different channel
to give skbs back to this cpu.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
On hosts with a lot of cores, RPC workloads suffer from heavy contention on slab spinlocks.
20.69% [kernel] [k] queued_spin_lock_slowpath
5.64% [kernel] [k] _raw_spin_lock
3.83% [kernel] [k] syscall_return_via_sysret
3.48% [kernel] [k] __entry_text_start
1.76% [kernel] [k] __netif_receive_skb_core
1.64% [kernel] [k] __fget
For each sendmsg(), we allocate one skb, and free it at the time ACK packet comes.
In many cases, ACK packets are handled by another cpus, and this unfortunately
incurs heavy costs for slab layer.
This patch uses an extra pointer in socket structure, so that we try to reuse
the same skb and avoid these expensive costs.
We cache at most one skb per socket so this should be safe as far as
memory pressure is concerned.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
tcp_clock_ns() (aka ktime_get_ns()) is using monotonic clock,
so the checks we had in tcp_mstamp_refresh() are no longer
relevant.
This patch removes cpu stall (when the cache line is not hot)
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since maxattr is common, the policy can't really differ sanely,
so make it common as well.
The only user that did in fact manage to make a non-common policy
is taskstats, which has to be really careful about it (since it's
still using a common maxattr!). This is no longer supported, but
we can fake it using pre_doit.
This reduces the size of e.g. nl80211.o (which has lots of commands):
text data bss dec hex filename
398745 14323 2240 415308 6564c net/wireless/nl80211.o (before)
397913 14331 2240 414484 65314 net/wireless/nl80211.o (after)
--------------------------------
-832 +8 0 -824
Which is obviously just 8 bytes for each command, and an added 8
bytes for the new policy pointer. I'm not sure why the ops list is
counted as .text though.
Most of the code transformations were done using the following spatch:
@ops@
identifier OPS;
expression POLICY;
@@
struct genl_ops OPS[] = {
...,
{
- .policy = POLICY,
},
...
};
@@
identifier ops.OPS;
expression ops.POLICY;
identifier fam;
expression M;
@@
struct genl_family fam = {
.ops = OPS,
.maxattr = M,
+ .policy = POLICY,
...
};
This also gets rid of devlink_nl_cmd_region_read_dumpit() accessing
the cb->data as ops, which we want to change in a later genl patch.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Get rid of some obsolete gc-related documentation and macros that were
missed in commit 5b7c9a8ff8 ("net: remove dst gc related code").
CC: Wei Wang <weiwan@google.com>
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Acked-by: Wei Wang <weiwan@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
fib_trie implementation calls synchronize_rcu when a certain amount of
pages are dirty from freed entries. The number of pages was determined
experimentally in 2009 (commit c3059477fc).
At the current setting, synchronize_rcu is called often -- 51 times in a
second in one test with an average of an 8 msec delay adding a fib entry.
The total impact is a lot of slow down modifying the fib. This is seen
in the output of 'time' - the difference between real time and sys+user.
For example, using 720,022 single path routes and 'ip -batch'[1]:
$ time ./ip -batch ipv4/routes-1-hops
real 0m14.214s
user 0m2.513s
sys 0m6.783s
So roughly 35% of the actual time to install the routes is from the ip
command getting scheduled out, most notably due to synchronize_rcu (this
is observed using 'perf sched timehist').
This patch makes the amount of dirty memory configurable between 64k where
the synchronize_rcu is called often (small, low end systems that are memory
sensitive) to 64M where synchronize_rcu is called rarely during a large
FIB change (for high end systems with lots of memory). The default is 512kB
which corresponds to the current setting of 128 pages with a 4kB page size.
As an example, at 16MB the worst interval shows 4 calls to synchronize_rcu
in a second blocking for up to 30 msec in a single instance, and a total
of almost 100 msec across the 4 calls in the second. The trade off is
allowing FIB entries to consume more memory in a given time window but
but with much better fib insertion rates (~30% increase in prefixes/sec).
With this patch and net.ipv4.fib_sync_mem set to 16MB, the same batch
file runs in:
$ time ./ip -batch ipv4/routes-1-hops
real 0m9.692s
user 0m2.491s
sys 0m6.769s
So the dead time is reduced to about 1/2 second or <5% of the real time.
[1] 'ip' modified to not request ACK messages which improves route
insertion times by about 20%
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the request socket is created locally, it'd make more sense to
use reqsk_free() instead of reqsk_put() in TFO and syncookies' error
path.
However, tcp_get_cookie_sock() may set ->rsk_refcnt before freeing the
socket; tcp_conn_request() may also have non-null ->rsk_refcnt because
of tcp_try_fastopen(). In both cases 'req' hasn't been exposed
to the outside world and is safe to free immediately, but that'd
trigger the WARN_ON_ONCE in reqsk_free().
Define __reqsk_free() for these situations where we know nobody's
referencing the socket, even though ->rsk_refcnt might be non-null.
Now we can consolidate the error path of tcp_get_cookie_sock() and
tcp_conn_request().
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since commit eeea10b83a ("tcp: add
tcp_v4_fill_cb()/tcp_v4_restore_cb()"), tcp_vX_fill_cb is only called
after tcp_filter(). That means, TCP_SKB_CB(skb)->end_seq still points to
the IP-part of the cb.
We thus should not mock with it, as this can trigger bugs (thanks
syzkaller):
[ 12.349396] ==================================================================
[ 12.350188] BUG: KASAN: slab-out-of-bounds in ip6_datagram_recv_specific_ctl+0x19b3/0x1a20
[ 12.351035] Read of size 1 at addr ffff88006adbc208 by task test_ip6_datagr/1799
Setting end_seq is actually no more necessary in tcp_filter as it gets
initialized later on in tcp_vX_fill_cb.
Cc: Eric Dumazet <edumazet@google.com>
Fixes: eeea10b83a ("tcp: add tcp_v4_fill_cb()/tcp_v4_restore_cb()")
Signed-off-by: Christoph Paasch <cpaasch@apple.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 7716682cc5 ("tcp/dccp: fix another race at listener
dismantle") let inet_csk_reqsk_queue_add() fail, and adjusted
{tcp,dccp}_check_req() accordingly. However, TFO and syncookies
weren't modified, thus leaking allocated resources on error.
Contrary to tcp_check_req(), in both syncookies and TFO cases,
we need to drop the request socket. Also, since the child socket is
created with inet_csk_clone_lock(), we have to unlock it and drop an
extra reference (->sk_refcount is initially set to 2 and
inet_csk_reqsk_queue_add() drops only one ref).
For TFO, we also need to revert the work done by tcp_try_fastopen()
(with reqsk_fastopen_remove()).
Fixes: 7716682cc5 ("tcp/dccp: fix another race at listener dismantle")
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The race occurs in __mkroute_output() when 2 threads lookup a dst:
CPU A CPU B
find_exception()
find_exception() [fnhe expires]
ip_del_fnhe() [fnhe is deleted]
rt_bind_exception()
In rt_bind_exception() it will bind a deleted fnhe with the new dst, and
this dst will get no chance to be freed. It causes a dev defcnt leak and
consecutive dmesg warnings:
unregister_netdevice: waiting for ethX to become free. Usage count = 1
Especially thanks Jon to identify the issue.
This patch fixes it by setting fnhe_daddr to 0 in ip_del_fnhe() to stop
binding the deleted fnhe with a new dst when checking fnhe's fnhe_daddr
and daddr in rt_bind_exception().
It works as both ip_del_fnhe() and rt_bind_exception() are protected by
fnhe_lock and the fhne is freed by kfree_rcu().
Fixes: deed49df73 ("route: check and remove route cache when we get route")
Signed-off-by: Jon Maxwell <jmaxwell37@gmail.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Returning 0 as inq to userspace indicates there is no more data to
read, and the application needs to wait for EPOLLIN. For a connection
that has received FIN from the remote peer, however, the application
must continue reading until getting EOF (return value of 0
from tcp_recvmsg) or an error, if edge-triggered epoll (EPOLLET) is
being used. Otherwise, the application will never receive a new
EPOLLIN, since there is no epoll edge after the FIN.
Return 1 when there is no data left on the queue but the
connection has received FIN, so that the applications continue
reading.
Fixes: b75eba76d3 (tcp: send in-queue bytes in cmsg upon read)
Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
sendpage was not designed for processing of the Slab pages,
in some situations it can trigger BUG_ON on receiving side.
Signed-off-by: Vasily Averin <vvs@virtuozzo.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If a non local multicast packet reaches ip_route_input_rcu() while
the ingress device IPv4 private data (in_dev) is NULL, we end up
doing a NULL pointer dereference in IN_DEV_MFORWARD().
Since the later call to ip_route_input_mc() is going to fail if
!in_dev, we can fail early in such scenario and avoid the dangerous
code path.
v1 -> v2:
- clarified the commit message, no code changes
Reported-by: Tianhao Zhao <tizhao@redhat.com>
Fixes: e58e415968 ("net: Enable support for VRF with ipv4 multicast")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When CONFIG_SYSCTL is turned off, we get a link failure for
the newly introduced tuning knob.
net/ipv6/addrconf.o: In function `addrconf_init_net':
addrconf.c:(.text+0x31dc): undefined reference to `sysctl_devconf_inherit_init_net'
Add an IS_ENABLED() check to fall back to the default behavior
(sysctl_devconf_inherit_init_net=0) here.
Fixes: 856c395cfa ("net: introduce a knob to control whether to inherit devconf config")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Christian Brauner <christian@brauner.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pablo Neira Ayuso says:
====================
Netfilter/IPVS updates for net-next
The following patchset contains Netfilter/IPVS updates for net-next:
1) Add .release_ops to properly unroll .select_ops, use it from nft_compat.
After this change, we can remove list of extensions too to simplify this
codebase.
2) Update amanda conntrack helper to support v3.4, from Florian Tham.
3) Get rid of the obsolete BUGPRINT macro in ebtables, from
Florian Westphal.
4) Merge IPv4 and IPv6 masquerading infrastructure into one single module.
From Florian Westphal.
5) Patchset to remove nf_nat_l3proto structure to get rid of
indirections, from Florian Westphal.
6) Skip unnecessary conntrack timeout updates in case the value is
still the same, also from Florian Westphal.
7) Remove unnecessary 'fall through' comments in empty switch cases,
from Li RongQing.
8) Fix lookup to fixed size hashtable sets on big endian with 32-bit keys.
9) Incorrect logic to deactivate path of fixed size hashtable sets,
element was being tested to self.
10) Remove nft_hash_key(), the bitmap set is always selected for 16-bit
keys.
11) Use boolean whenever possible in IPVS codebase, from Andrea Claudi.
12) Enter close state in conntrack if RST matches exact sequence number,
from Florian Westphal.
13) Initialize dst_cache in tunnel extension, from wenxu.
14) Pass protocol as u16 to xt_check_match and xt_check_target, from
Li RongQing.
15) SCTP header is granted to be in a linear area from IPVS NAT handler,
from Xin Long.
16) Don't steal packets coming from slave VRF device from the
ip_sabotage_in() path, from David Ahern.
17) Fix unsafe update of basechain stats, from Li RongQing.
18) Make sure CONNTRACK_LOCKS is power of 2 to let compiler optimize
modulo operation as bitwise AND, from Li RongQing.
19) Use device_attribute instead of internal definition in the IDLETIMER
target, from Sami Tolvanen.
20) Merge redir, masq and IPv4/IPv6 NAT chain types, from Florian Westphal.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
When calculating the multipath hash for input routes the flow info is
not available and therefore should not be used.
Fixes: 24ba14406c ("route: Add multipath_hash in flowi_common to make user-define hash")
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Cc: wenxu <wenxu@ucloud.cn>
Acked-by: wenxu <wenxu@ucloud.cn>
Acked-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For ip rules, we need to use 'ipproto ipv6-icmp' to match ICMPv6 headers.
But for ip -6 route, currently we only support tcp, udp and icmp.
Add ICMPv6 support so we can match ipv6-icmp rules for route lookup.
v2: As David Ahern and Sabrina Dubroca suggested, Add an argument to
rtm_getroute_parse_ip_proto() to handle ICMP/ICMPv6 with different family.
Reported-by: Jianlin Shi <jishi@redhat.com>
Fixes: eacb9384a3 ("ipv6: support sport, dport and ip_proto in RTM_GETROUTE")
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Merge the ipv4 and ipv6 nat chain type. This is the last
missing piece which allows to provide inet family support
for nat in a follow patch.
The kconfig knobs for ipv4/ipv6 nat chain are removed, the
nat chain type will be built unconditionally if NFT_NAT
expression is enabled.
Before:
text data bss dec hex filename
1576 896 0 2472 9a8 nft_chain_nat_ipv4.ko
1697 896 0 2593 a21 nft_chain_nat_ipv6.ko
After:
text data bss dec hex filename
1832 896 0 2728 aa8 nft_chain_nat.ko
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
The family specific masq modules are way too small to warrant
an extra module, just place all of them in nft_masq.
before:
text data bss dec hex filename
1001 832 0 1833 729 nft_masq.ko
766 896 0 1662 67e nft_masq_ipv4.ko
764 896 0 1660 67c nft_masq_ipv6.ko
after:
2010 960 0 2970 b9a nft_masq.ko
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
before:
text data bss dec hex filename
990 832 0 1822 71e nft_redir.ko
697 896 0 1593 639 nft_redir_ipv4.ko
713 896 0 1609 649 nft_redir_ipv6.ko
after:
text data bss dec hex filename
1910 960 0 2870 b36 nft_redir.ko
size is reduced, all helpers from nft_redir.ko can be made static.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
