The phonet_protocol structs don't need to be written by anyone and
so can be marked as const.
Signed-off-by: Lin Zhang <xiaolou4617@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, most of the rt6_stats are not hooked up correctly. As the
last part of this patch series, hook up all existing rt6_stats and add
one new stat fib_rt_uncache to indicate the number of routes in the
uncached list.
For details of the stats, please refer to the comments added in
include/net/ip6_fib.h.
Note: fib_rt_alloc and fib_rt_uncache are not guaranteed to be modified
under a lock. So atomic_t is used for them.
Signed-off-by: Wei Wang <weiwan@google.com>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With all the preparation work before, we are now ready to replace rwlock
with rcu and spinlock in fib6_table.
That means now all fib6_node in fib6_table are protected by rcu. And
when freeing fib6_node, call_rcu() is used to wait for the rcu grace
period before releasing the memory.
When accessing fib6_node, corresponding rcu APIs need to be used.
And all previous sessions protected by the write lock will now be
protected by the spin lock per table.
All previous sessions protected by read lock will now be protected by
rcu_read_lock().
A couple of things to note here:
1. As part of the work of replacing rwlock with rcu, the linked list of
fn->leaf now has to be rcu protected as well. So both fn->leaf and
rt->dst.rt6_next are now __rcu tagged and corresponding rcu APIs are
used when manipulating them.
2. For fn->rr_ptr, first of all, it also needs to be rcu protected now
and is tagged with __rcu and rcu APIs are used in corresponding places.
Secondly, fn->rr_ptr is changed in rt6_select() which is a reader
thread. This makes the issue a bit complicated. We think a valid
solution for it is to let rt6_select() grab the tb6_lock if it decides
to change it. As it is not in the normal operation and only happens when
there is no valid neighbor cache for the route, we think the performance
impact should be low.
3. fib6_walk_continue() has to be called with tb6_lock held even in the
route dumping related functions, e.g. inet6_dump_fib(),
fib6_tables_dump() and ipv6_route_seq_ops. It is because
fib6_walk_continue() makes modifications to the walker structure, and so
are fib6_repair_tree() and fib6_del_route(). In order to do proper
syncing between them, we need to let fib6_walk_continue() hold the lock.
We may be able to do further improvement on the way we do the tree walk
to get rid of the need for holding the spin lock. But not for now.
4. When fib6_del_route() removes a route from the tree, we no longer
mark rt->dst.rt6_next to NULL to make simultaneous reader be able to
further traverse the list with rcu. However, rt->dst.rt6_next is only
valid within this same rcu period. No one should access it later.
5. All the operation of atomic_inc(rt->rt6i_ref) is changed to be
performed before we publish this route (either by linking it to fn->leaf
or insert it in the list pointed by fn->leaf) just to be safe because as
soon as we publish the route, some read thread will be able to access it.
Signed-off-by: Wei Wang <weiwan@google.com>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
fib6_add() logic currently calls fib6_add_1() to figure out what node
should be used for the newly added route and then call
fib6_add_rt2node() to insert the route to the node.
And during the call of fib6_add_1(), fn_sernum is updated for all nodes
that share the same prefix as the new route.
This does not have issue in the current code because reader thread will
not be able to access the tree while writer thread is inserting new
route to it. However, it is not the case once we transition to use RCU.
Reader thread could potentially see the new fn_sernum before the new
route is inserted. As a result, reader thread's route lookup will return
a stale route with the new fn_sernum.
In order to solve this issue, we remove all the update of fn_sernum in
fib6_add_1(), and instead, introduce a new function that updates fn_sernum
for all related nodes and call this functions once the route is
successfully inserted to the tree.
Also, smp_wmb() is used after a route is successfully inserted into the
fib tree and right before the updated of fn->sernum. And smp_rmb() is
used right after fn->sernum is accessed in rt6_get_cookie_safe(). This
is to guarantee that when the reader thread sees the new fn->sernum, the
new route is already inserted in the tree in memory.
Signed-off-by: Wei Wang <weiwan@google.com>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit makes use of the exception hash table implementation to
store dst caches created by pmtu discovery and ip redirect into the hash
table under the rt_info and no longer inserts these routes into fib6
tree.
This makes the fib6 tree only contain static configured routes and could
now be protected by rcu instead of a rw lock.
With this change, in the route lookup related functions, after finding
the rt6_info with the longest prefix, we also need to search for the
exception table before doing backtracking.
In the route delete function, if the route being deleted is not a dst
cache, deletion of this route also need to flush the whole hash table
under it. If it is a dst cache, then only delete the cached dst in the
hash table.
Note: for fib6_walk_continue() function, w->root now is always pointing
to a root node considering that fib6_prune_clones() is removed from the
code. So we add a WARN_ON() msg to make sure w->root always points to a
root node and also removed the update of w->root in fib6_repair_tree().
This is a prerequisite for later patch because we don't need to make
w->root as rcu protected when replacing rwlock with RCU.
Also, we remove all prune related variables as it is no longer used.
Signed-off-by: Wei Wang <weiwan@google.com>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
fib6_locate() is used to find the fib6_node according to the passed in
prefix address key. It currently tries to find the fib6_node with the
exact match of the passed in key. However, when we move cached routes
into the exception table, fib6_locate() will fail to find the fib6_node
for it as the cached routes will be stored in the exception table under
the fib6_node with the longest prefix match of the cache's dst addr key.
This commit adds a new parameter to let the caller specify if it needs
exact match or longest prefix match.
Right now, all callers still does exact match when calling
fib6_locate(). It will be changed in later commit where exception table
is hooked up to store cached routes.
Signed-off-by: Wei Wang <weiwan@google.com>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If all dst cache entries are stored in the exception table under the
main route, we have to go through them during fib6_age() when doing
garbage collecting.
Introduce a new function rt6_age_exception() which goes through all dst
entries in the exception table and remove those entries that are expired.
This function is called in fib6_age() so that all dst caches are also
garbage collected.
Signed-off-by: Wei Wang <weiwan@google.com>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a hash table into struct rt6_info in order to store dst caches
created by pmtu discovery and ip redirect in ipv6 routing code.
APIs to add dst cache, delete dst cache, find dst cache and update
dst cache in the hash table are implemented and will be used in later
commits.
This is a preparation work to move all cache routes into the exception
table instead of getting inserted into the fib6 tree.
Signed-off-by: Wei Wang <weiwan@google.com>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This function takes a route as input and tries to update the sernum in
the fib6_node this route is associated with. It will be used in later
commit when adding a cached route into the exception table under that
route.
Signed-off-by: Wei Wang <weiwan@google.com>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Using a linear list to store all skbs in write queue has been okay
for quite a while : O(N) is not too bad when N < 500.
Things get messy when N is the order of 100,000 : Modern TCP stacks
want 10Gbit+ of throughput even with 200 ms RTT flows.
40 ns per cache line miss means a full scan can use 4 ms,
blowing away CPU caches.
SACK processing often can use various hints to avoid parsing
whole retransmit queue. But with high packet losses and/or high
reordering, hints no longer work.
Sender has to process thousands of unfriendly SACK, accumulating
a huge socket backlog, burning a cpu and massively dropping packets.
Using an rb-tree for retransmit queue has been avoided for years
because it added complexity and overhead, but now is the time
to be more resistant and say no to quadratic behavior.
1) RTX queue is no longer part of the write queue : already sent skbs
are stored in one rb-tree.
2) Since reaching the head of write queue no longer needs
sk->sk_send_head, we added an union of sk_send_head and tcp_rtx_queue
Tested:
On receiver :
netem on ingress : delay 150ms 200us loss 1
GRO disabled to force stress and SACK storms.
for f in `seq 1 10`
do
./netperf -H lpaa6 -l30 -- -K bbr -o THROUGHPUT|tail -1
done | awk '{print $0} {sum += $0} END {printf "%7u\n",sum}'
Before patch :
323.87
351.48
339.59
338.62
306.72
204.07
304.93
291.88
202.47
176.88
2840
After patch:
1700.83
2207.98
2070.17
1544.26
2114.76
2124.89
1693.14
1080.91
2216.82
1299.94
18053
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the upcoming rtx rbtree will add some extra code,
it is time to not inline this fat function anymore.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
commit cc71b7b071 ("net/ipv6: remove unused err variable on
icmpv6_push_pending_frames") exposed icmpv6_push_pending_frames
return value not being used.
Remove now unnecessary int err declarations and uses.
Miscellanea:
o Remove unnecessary goto and out: labels
o Realign arguments
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds a new queue (list) that tracks the sent but not yet
acked or SACKed skbs for a TCP connection. The list is chronologically
ordered by skb->skb_mstamp (the head is the oldest sent skb).
This list will be used to optimize TCP Rack recovery, which checks
an skb's timestamp to judge if it has been lost and needs to be
retransmitted. Since TCP write queue is ordered by sequence instead
of sent time, RACK has to scan over the write queue to catch all
eligible packets to detect lost retransmission, and iterates through
SACKed skbs repeatedly.
Special cares for rare events:
1. TCP repair fakes skb transmission so the send queue needs adjusted
2. SACK reneging would require re-inserting SACKed skbs into the
send queue. For now I believe it's not worth the complexity to
make RACK work perfectly on SACK reneging, so we do nothing here.
3. Fast Open: currently for non-TFO, send-queue correctly queues
the pure SYN packet. For TFO which queues a pure SYN and
then a data packet, send-queue only queues the data packet but
not the pure SYN due to the structure of TFO code. This is okay
because the SYN receiver would never respond with a SACK on a
missing SYN (i.e. SYN is never fast-retransmitted by SACK/RACK).
In order to not grow sk_buff, we use an union for the new list and
_skb_refdst/destructor fields. This is a bit complicated because
we need to make sure _skb_refdst and destructor are properly zeroed
before skb is cloned/copied at transmit, and before being freed.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently in the TCP code, the initialization sequence for cached
metrics, congestion control, BPF, etc, after successful connection
is very inconsistent. This introduces inconsistent bevhavior and is
prone to bugs. The current call sequence is as follows:
(1) for active case (tcp_finish_connect() case):
tcp_mtup_init(sk);
icsk->icsk_af_ops->rebuild_header(sk);
tcp_init_metrics(sk);
tcp_call_bpf(sk, BPF_SOCK_OPS_ACTIVE_ESTABLISHED_CB);
tcp_init_congestion_control(sk);
tcp_init_buffer_space(sk);
(2) for passive case (tcp_rcv_state_process() TCP_SYN_RECV case):
icsk->icsk_af_ops->rebuild_header(sk);
tcp_call_bpf(sk, BPF_SOCK_OPS_PASSIVE_ESTABLISHED_CB);
tcp_init_congestion_control(sk);
tcp_mtup_init(sk);
tcp_init_buffer_space(sk);
tcp_init_metrics(sk);
(3) for TFO passive case (tcp_fastopen_create_child()):
inet_csk(child)->icsk_af_ops->rebuild_header(child);
tcp_init_congestion_control(child);
tcp_mtup_init(child);
tcp_init_metrics(child);
tcp_call_bpf(child, BPF_SOCK_OPS_PASSIVE_ESTABLISHED_CB);
tcp_init_buffer_space(child);
This commit uniforms the above functions to have the following sequence:
tcp_mtup_init(sk);
icsk->icsk_af_ops->rebuild_header(sk);
tcp_init_metrics(sk);
tcp_call_bpf(sk, BPF_SOCK_OPS_ACTIVE/PASSIVE_ESTABLISHED_CB);
tcp_init_congestion_control(sk);
tcp_init_buffer_space(sk);
This sequence is the same as the (1) active case. We pick this sequence
because this order correctly allows BPF to override the settings
including congestion control module and initial cwnd, etc from
the route, and then allows the CC module to see those settings.
Suggested-by: Neal Cardwell <ncardwell@google.com>
Tested-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Wei Wang <weiwan@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There are two state fields: socket->state and sock->sk_state. The
socket->state field uses SS_UNCONNECTED, SS_CONNECTED, etc while the
sock->sk_state typically uses values that match TCP state constants
(TCP_CLOSE, TCP_ESTABLISHED). AF_VSOCK does not follow this convention
and instead uses SS_* constants for both fields.
The sk_state field will be exposed to userspace through the vsock_diag
interface for ss(8), netstat(8), and other programs.
This patch switches sk_state to TCP state constants so that the meaning
of this field is consistent with other address families. Not just
AF_INET and AF_INET6 use the TCP constants, AF_UNIX and others do too.
The following mapping was used to convert the code:
SS_FREE -> TCP_CLOSE
SS_UNCONNECTED -> TCP_CLOSE
SS_CONNECTING -> TCP_SYN_SENT
SS_CONNECTED -> TCP_ESTABLISHED
SS_DISCONNECTING -> TCP_CLOSING
VSOCK_SS_LISTEN -> TCP_LISTEN
In __vsock_create() the sk_state initialization was dropped because
sock_init_data() already initializes sk_state to TCP_CLOSE.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The vsock_diag.ko module will need to check socket table membership.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The socket table symbols need to be exported from vsock.ko so that the
vsock_diag.ko module will be able to traverse sockets.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pass extack to do_set_master and down to ndo_add_slave
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
no users in the tree.
Signed-off-by: Florian Westphal <fw@strlen.de>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch introduces RFC Draft ndata section 3.2 Priority Based
Scheduler (SCTP_SS_RR).
Works by maintaining a list of enqueued streams and tracking the last
one used to send data. When the datamsg is done, it switches to the next
stream.
See-also: https://tools.ietf.org/html/draft-ietf-tsvwg-sctp-ndata-13
Tested-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch introduces RFC Draft ndata section 3.4 Priority Based
Scheduler (SCTP_SS_PRIO).
It works by having a struct sctp_stream_priority for each priority
configured. This struct is then enlisted on a queue ordered per priority
if, and only if, there is a stream with data queued, so that dequeueing
is very straightforward: either finish current datamsg or simply dequeue
from the highest priority queued, which is the next stream pointed, and
that's it.
If there are multiple streams assigned with the same priority and with
data queued, it will do round robin amongst them while respecting
datamsgs boundaries (when not using idata chunks), to be reasonably
fair.
We intentionally don't maintain a list of priorities nor a list of all
streams with the same priority to save memory. The first would mean at
least 2 other pointers per priority (which, for 1000 priorities, that
can mean 16kB) and the second would also mean 2 other pointers but per
stream. As SCTP supports up to 65535 streams on a given asoc, that's
1MB. This impacts when giving a priority to some stream, as we have to
find out if the new priority is already being used and if we can free
the old one, and also when tearing down.
The new fields in struct sctp_stream_out_ext and sctp_stream are added
under a union because that memory is to be shared with other schedulers.
See-also: https://tools.ietf.org/html/draft-ietf-tsvwg-sctp-ndata-13
Tested-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch introduces the hooks necessary to do stream scheduling, as
per RFC Draft ndata. It also introduces the first scheduler, which is
what we do today but now factored out: first come first served (FCFS).
With stream scheduling now we have to track which chunk was enqueued on
which stream and be able to select another other than the in front of
the main outqueue. So we introduce a list on sctp_stream_out_ext
structure for this purpose.
We reuse sctp_chunk->transmitted_list space for the list above, as the
chunk cannot belong to the two lists at the same time. By using the
union in there, we can have distinct names for these moments.
sctp_sched_ops are the operations expected to be implemented by each
scheduler. The dequeueing is a bit particular to this implementation but
it is to match how we dequeue packets today. We first dequeue and then
check if it fits the packet and if not, we requeue it at head. Thus why
we don't have a peek operation but have dequeue_done instead, which is
called once the chunk can be safely considered as transmitted.
The check removed from sctp_outq_flush is now performed by
sctp_stream_outq_migrate, which is only called during assoc setup.
(sctp_sendmsg() also checks for it)
The only operation that is foreseen but not yet added here is a way to
signalize that a new packet is starting or that the packet is done, for
round robin scheduler per packet, but is intentionally left to the
patch that actually implements it.
Support for I-DATA chunks, also described in this RFC, with user message
interleaving is straightforward as it just requires the schedulers to
probe for the feature and ignore datamsg boundaries when dequeueing.
See-also: https://tools.ietf.org/html/draft-ietf-tsvwg-sctp-ndata-13
Tested-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to fetch the stream number from a given chunk.
Tested-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With the stream schedulers, sctp_stream_out will become too big to be
allocated by kmalloc and as we need to allocate with BH disabled, we
cannot use __vmalloc in sctp_stream_init().
This patch moves out the stats from sctp_stream_out to
sctp_stream_out_ext, which will be allocated only when the application
tries to sendmsg something on it.
Just the introduction of sctp_stream_out_ext would already fix the issue
described above by splitting the allocation in two. Moving the stats
to it also reduces the pressure on the allocator as we will ask for less
memory atomically when creating the socket and we will use GFP_KERNEL
later.
Then, for stream schedulers, we will just use sctp_stream_out_ext.
Tested-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Make the skb parameter of skb_metadata_dst() and skb_tunnel_info()
const as they are not modified. This is in preparation for using
them in call-sites where skb is const.
Signed-off-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This function is unused, and furthermore it is buggy since it suffers
from the same issue that requires IP6_ECN_set_ce() to take a pointer
to the skb so that it may (in case of CHECKSUM_COMPLETE) update skb->csum
Instead of fixing it, let's just outright remove it.
Tested: builds, and 'git grep IP6_ECN_clear' comes up empty
Signed-off-by: Maciej Żenczykowski <maze@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Different namespace application might require different time period in
second to disable Fastopen on active TCP sockets.
Tested:
Simulate following similar situation that the server's data gets dropped
after 3WHS.
C ---- syn-data ---> S
C <--- syn/ack ----- S
C ---- ack --------> S
S (accept & write)
C? X <- data ------ S
[retry and timeout]
And then print netstat of TCPFastOpenBlackhole, the counter increased as
expected when the firewall blackhole issue is detected and active TFO is
disabled.
# cat /proc/net/netstat | awk '{print $91}'
TCPFastOpenBlackhole
1
Signed-off-by: Haishuang Yan <yanhaishuang@cmss.chinamobile.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Different namespace application might require different tcp_fastopen_key
independently of the host.
David Miller pointed out there is a leak without releasing the context
of tcp_fastopen_key during netns teardown. So add the release action in
exit_batch path.
Tested:
1. Container namespace:
# cat /proc/sys/net/ipv4/tcp_fastopen_key:
2817fff2-f803cf97-eadfd1f3-78c0992b
cookie key in tcp syn packets:
Fast Open Cookie
Kind: TCP Fast Open Cookie (34)
Length: 10
Fast Open Cookie: 1e5dd82a8c492ca9
2. Host:
# cat /proc/sys/net/ipv4/tcp_fastopen_key:
107d7c5f-68eb2ac7-02fb06e6-ed341702
cookie key in tcp syn packets:
Fast Open Cookie
Kind: TCP Fast Open Cookie (34)
Length: 10
Fast Open Cookie: e213c02bf0afbc8a
Signed-off-by: Haishuang Yan <yanhaishuang@cmss.chinamobile.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The 'publish' logic is not necessary after commit dfea2aa654 ("tcp:
Do not call tcp_fastopen_reset_cipher from interrupt context"), because
in tcp_fastopen_cookie_gen，it wouldn't call tcp_fastopen_init_key_once.
Signed-off-by: Haishuang Yan <yanhaishuang@cmss.chinamobile.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Different namespace application might require enable TCP Fast Open
feature independently of the host.
This patch series continues making more of the TCP Fast Open related
sysctl knobs be per net-namespace.
Reported-by: Luca BRUNO <lucab@debian.org>
Signed-off-by: Haishuang Yan <yanhaishuang@cmss.chinamobile.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now that the dsa_ptr is a dsa_port instance, there is no need to keep
the tag operations in the dsa_switch_tree structure. Remove it.
Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In preparation to make DSA master devices point to their corresponding
CPU port instead of the whole tree, add copies of dst and rcv in the
dsa_port structure so that we keep fast access in the receive hot path.
Also keep the copies at the beginning of the dsa_port structure in order
to ensure they are available in cacheline 1.
Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The DSA tagging protocol operations are specific to each CPU port,
thus the dsa_device_ops pointer belongs to the dsa_port structure.
>From now on assign a slave's xmit copy from its CPU port tagging
operations. This will ease the future support for multiple CPU ports.
Also keep the tag_ops at the beginning of the dsa_port structure so that
we ensure copies for hot path are in cacheline 1.
Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The UDP early demux can leverate the rx dst cache even for
multicast unconnected sockets.
In such scenario the ipv4 source address is validated only on
the first packet in the given flow. After that, when we fetch
the dst entry from the socket rx cache, we stop enforcing
the rp_filter and we even start accepting any kind of martian
addresses.
Disabling the dst cache for unconnected multicast socket will
cause large performace regression, nearly reducing by half the
max ingress tput.
Instead we factor out a route helper to completely validate an
skb source address for multicast packets and we call it from
the UDP early demux for mcast packets landing on unconnected
sockets, after successful fetching the related cached dst entry.
This still gives a measurable, but limited performance
regression:
rp_filter = 0 rp_filter = 1
edmux disabled: 1182 Kpps 1127 Kpps
edmux before: 2238 Kpps 2238 Kpps
edmux after: 2037 Kpps 2019 Kpps
The above figures are on top of current net tree.
Applying the net-next commit 6e617de84e ("net: avoid a full
fib lookup when rp_filter is disabled.") the delta with
rp_filter == 0 will decrease even more.
Fixes: 421b3885bf ("udp: ipv4: Add udp early demux")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently no error is emitted, but this infrastructure will
used by the next patch to allow source address validation
for mcast sockets.
Since early demux can do a route lookup and an ipv4 route
lookup can return an error code this is consistent with the
current ipv4 route infrastructure.
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
fib_weight in fib_info is set but not used. Remove it and the
helpers for setting it.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Make the ipmr module register as a FIB notifier. To do that, implement both
the ipmr_seq_read and ipmr_dump ops.
The ipmr_seq_read op returns a sequence counter that is incremented on
every notification related operation done by the ipmr. To implement that,
add a sequence counter in the netns_ipv4 struct and increment it whenever a
new MFC route or VIF are added or deleted. The sequence operations are
protected by the RTNL lock.
The ipmr_dump iterates the list of MFC routes and the list of VIF entries
and sends notifications about them. The entries dump is done under RCU
where the VIF dump uses the mrt_lock too, as the vif->dev field can change
under RCU.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order for an interface to forward packets according to the kernel
multicast routing table, it must be configured with a VIF index according
to the mroute user API. The VIF index is then used to refer to that
interface in the mroute user API, for example, to set the iif and oifs of
an MFC entry.
In order to allow drivers to be aware and offload multicast routes, they
have to be aware of the VIF add and delete notifications.
Due to the fact that a specific VIF can be deleted and re-added pointing to
another netdevice, and the MFC routes that point to it will forward the
matching packets to the new netdevice, a driver willing to offload MFC
cache entries must be aware of the VIF add and delete events in addition to
MFC routes notifications.
Signed-off-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Introduce a helper called is_tcf_gact_pass which could be used to
tell if the action is gact pass or not.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Key length can't be negative.
Leave comparisons against nla_len() signed just in case truncated attribute
can sneak in there.
Space savings:
add/remove: 0/0 grow/shrink: 0/7 up/down: 0/-7 (-7)
function old new delta
pneigh_delete 273 272 -1
mlx5e_rep_netevent_event 1415 1414 -1
mlx5e_create_encap_header_ipv6 1194 1193 -1
mlx5e_create_encap_header_ipv4 1071 1070 -1
cxgb4_l2t_get 1104 1103 -1
__pneigh_lookup 69 68 -1
__neigh_create 2452 2451 -1
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When CONFIG_KASAN is enabled, the "--param asan-stack=1" causes rather large
stack frames in some functions. This goes unnoticed normally because
CONFIG_FRAME_WARN is disabled with CONFIG_KASAN by default as of commit
3f181b4d86 ("lib/Kconfig.debug: disable -Wframe-larger-than warnings with
KASAN=y").
The kernelci.org build bot however has the warning enabled and that led
me to investigate it a little further, as every build produces these warnings:
net/wireless/nl80211.c:4389:1: warning: the frame size of 2240 bytes is larger than 2048 bytes [-Wframe-larger-than=]
net/wireless/nl80211.c:1895:1: warning: the frame size of 3776 bytes is larger than 2048 bytes [-Wframe-larger-than=]
net/wireless/nl80211.c:1410:1: warning: the frame size of 2208 bytes is larger than 2048 bytes [-Wframe-larger-than=]
net/bridge/br_netlink.c:1282:1: warning: the frame size of 2544 bytes is larger than 2048 bytes [-Wframe-larger-than=]
Most of this problem is now solved in gcc-8, which can consolidate
the stack slots for the inline function arguments. On older compilers
we can add a workaround by declaring a local variable in each function
to pass the inline function argument.
Cc: stable@vger.kernel.org
Link: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81715
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
In linux-4.13, Wei worked hard to convert dst to a traditional
refcounted model, removing GC.
We now want to make sure a dst refcount can not transition from 0 back
to 1.
The problem here is that input path attached a not refcounted dst to an
skb. Then later, because packet is forwarded and hits skb_dst_force()
before exiting RCU section, we might try to take a refcount on one dst
that is about to be freed, if another cpu saw 1 -> 0 transition in
dst_release() and queued the dst for freeing after one RCU grace period.
Lets unify skb_dst_force() and skb_dst_force_safe(), since we should
always perform the complete check against dst refcount, and not assume
it is not zero.
Bugzilla : https://bugzilla.kernel.org/show_bug.cgi?id=197005
[ 989.919496] skb_dst_force+0x32/0x34
[ 989.919498] __dev_queue_xmit+0x1ad/0x482
[ 989.919501] ? eth_header+0x28/0xc6
[ 989.919502] dev_queue_xmit+0xb/0xd
[ 989.919504] neigh_connected_output+0x9b/0xb4
[ 989.919507] ip_finish_output2+0x234/0x294
[ 989.919509] ? ipt_do_table+0x369/0x388
[ 989.919510] ip_finish_output+0x12c/0x13f
[ 989.919512] ip_output+0x53/0x87
[ 989.919513] ip_forward_finish+0x53/0x5a
[ 989.919515] ip_forward+0x2cb/0x3e6
[ 989.919516] ? pskb_trim_rcsum.part.9+0x4b/0x4b
[ 989.919518] ip_rcv_finish+0x2e2/0x321
[ 989.919519] ip_rcv+0x26f/0x2eb
[ 989.919522] ? vlan_do_receive+0x4f/0x289
[ 989.919523] __netif_receive_skb_core+0x467/0x50b
[ 989.919526] ? tcp_gro_receive+0x239/0x239
[ 989.919529] ? inet_gro_receive+0x226/0x238
[ 989.919530] __netif_receive_skb+0x4d/0x5f
[ 989.919532] netif_receive_skb_internal+0x5c/0xaf
[ 989.919533] napi_gro_receive+0x45/0x81
[ 989.919536] ixgbe_poll+0xc8a/0xf09
[ 989.919539] ? kmem_cache_free_bulk+0x1b6/0x1f7
[ 989.919540] net_rx_action+0xf4/0x266
[ 989.919543] __do_softirq+0xa8/0x19d
[ 989.919545] irq_exit+0x5d/0x6b
[ 989.919546] do_IRQ+0x9c/0xb5
[ 989.919548] common_interrupt+0x93/0x93
[ 989.919548] </IRQ>
Similarly dst_clone() can use dst_hold() helper to have additional
debugging, as a follow up to commit 44ebe79149 ("net: add debug
atomic_inc_not_zero() in dst_hold()")
In net-next we will convert dst atomic_t to refcount_t for peace of
mind.
Fixes: a4c2fd7f78 ("net: remove DST_NOCACHE flag")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Wei Wang <weiwan@google.com>
Reported-by: Paweł Staszewski <pstaszewski@itcare.pl>
Bisected-by: Paweł Staszewski <pstaszewski@itcare.pl>
Acked-by: Wei Wang <weiwan@google.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
> net/ipv4/fib_frontend.c: In function 'fib_validate_source':
> net/ipv4/fib_frontend.c:411:16: error: 'struct netns_ipv4' has no member named 'fib_has_custom_local_routes'
> if (net->ipv4.fib_has_custom_local_routes)
> ^
> net/ipv4/fib_frontend.c: In function 'inet_rtm_newroute':
> net/ipv4/fib_frontend.c:773:12: error: 'struct netns_ipv4' has no member named 'fib_has_custom_local_routes'
> net->ipv4.fib_has_custom_local_routes = true;
> ^
Fixes: 6e617de84e ("net: avoid a full fib lookup when rp_filter is disabled.")
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since commit 1dced6a854 ("ipv4: Restore accept_local behaviour
in fib_validate_source()") a full fib lookup is needed even if
the rp_filter is disabled, if accept_local is false - which is
the default.
What we really need in the above scenario is just checking
that the source IP address is not local, and in most case we
can do that is a cheaper way looking up the ifaddr hash table.
This commit adds a helper for such lookup, and uses it to
validate the src address when rp_filter is disabled and no
'local' routes are created by the user space in the relevant
namespace.
A new ipv4 netns flag is added to account for such routes.
We need that to preserve the same behavior we had before this
patch.
It also drops the checks to bail early from __fib_validate_source,
added by the commit 1dced6a854 ("ipv4: Restore accept_local
behaviour in fib_validate_source()") they do not give any
measurable performance improvement: if we do the lookup with are
on a slower path.
This improves UDP performances for unconnected sockets
when rp_filter is disabled by 5% and also gives small but
measurable performance improvement for TCP flood scenarios.
v1 -> v2:
- use the ifaddr lookup helper in __ip_dev_find(), as suggested
by Eric
- fall-back to full lookup if custom local routes are present
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Having a global list of labels do not scale to thousands of
netns in the cloud era. This causes quadratic behavior on
netns creation and deletion.
This is time having a per netns list of ~10 labels.
Tested:
$ time perf record (for f in `seq 1 3000` ; do ip netns add tast$f; done)
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 3.637 MB perf.data (~158898 samples) ]
real 0m20.837s # instead of 0m24.227s
user 0m0.328s
sys 0m20.338s # instead of 0m23.753s
16.17% ip [kernel.kallsyms] [k] netlink_broadcast_filtered
12.30% ip [kernel.kallsyms] [k] netlink_has_listeners
6.76% ip [kernel.kallsyms] [k] _raw_spin_lock_irqsave
5.78% ip [kernel.kallsyms] [k] memset_erms
5.77% ip [kernel.kallsyms] [k] kobject_uevent_env
5.18% ip [kernel.kallsyms] [k] refcount_sub_and_test
4.96% ip [kernel.kallsyms] [k] _raw_read_lock
3.82% ip [kernel.kallsyms] [k] refcount_inc_not_zero
3.33% ip [kernel.kallsyms] [k] _raw_spin_unlock_irqrestore
2.11% ip [kernel.kallsyms] [k] unmap_page_range
1.77% ip [kernel.kallsyms] [k] __wake_up
1.69% ip [kernel.kallsyms] [k] strlen
1.17% ip [kernel.kallsyms] [k] __wake_up_common
1.09% ip [kernel.kallsyms] [k] insert_header
1.04% ip [kernel.kallsyms] [k] page_remove_rmap
1.01% ip [kernel.kallsyms] [k] consume_skb
0.98% ip [kernel.kallsyms] [k] netlink_trim
0.51% ip [kernel.kallsyms] [k] kernfs_link_sibling
0.51% ip [kernel.kallsyms] [k] filemap_map_pages
0.46% ip [kernel.kallsyms] [k] memcpy_erms
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
