Some drivers are becoming more dependent on NET_DEVLINK being selected
in configuration. With upcoming compat functions, the behavior would be
wrong in case devlink was not compiled in. So make the drivers select
NET_DEVLINK and rely on the functions being there, not just stubs.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add spinlock to protect port type and type_dev pointer consistency.
Without that, userspace may see inconsistent type and type_dev
combinations.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
v1->v2:
- rebased
Signed-off-by: David S. Miller <davem@davemloft.net>
Port needs to be registered first before the type is set. Warn and
bail-out in case it is not.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the port attributes are static and cannot change during the port
lifetime, WARN_ON if some driver calls it after registration. Also, no
need to call notifications as it is noop anyway due to check of
devlink_port->registered there.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since attrs are static during the existence of devlink port, set the
before registration of the port.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
__devlink_port_type_set() returns void, it makes no sense to pass it on,
so don't do that.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The netdevice is guaranteed to not disappear so we can rely that
devlink_port and devlink won't disappear as well. No need to take
devlink_mutex so don't take it here.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
devlink functions are in use, so include the related header file.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add missing called to mutex_destroy() for two mutexes used
in devlink code.
Signed-off-by: Jiri Pirko <jiri@mellanox.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>
We prefer static_branch_unlikely() over static_key_false() these days.
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>
With commit c5ad119fb6 ("net: sched: pfifo_fast use skb_array")
pfifo_fast no longer benefit from the TCQ_F_CAN_BYPASS optimization.
Due to retpolines the cost of the enqueue()/dequeue() pair has become
relevant and we observe measurable regression for the uncontended
scenario when the packet-rate is below line rate.
After commit 46b1c18f9d ("net: sched: put back q.qlen into a
single location") we can check for empty qdisc with a reasonably
fast operation even for nolock qdiscs.
This change extends TCQ_F_CAN_BYPASS support to nolock qdisc.
The new chunk of code mirrors closely the existing one for traditional
qdisc, leveraging a newly introduced helper to read atomically the
qdisc length.
Tested with pktgen in queue xmit mode, with pfifo_fast, a MQ
device, and MQ root qdisc:
threads vanilla patched
kpps kpps
1 2465 2889
2 4304 5188
4 7898 9589
Same as above, but with a single queue device:
threads vanilla patched
kpps kpps
1 2556 2827
2 2900 2900
4 5000 5000
8 4700 4700
No mesaurable changes in the contended scenarios, and more 10%
improvement in the uncontended ones.
v1 -> v2:
- rebased after flag name change
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Tested-by: Ivan Vecera <ivecera@redhat.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Ivan Vecera <ivecera@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The queue is marked not empty after acquiring the seqlock,
and it's up to the NOLOCK qdisc clearing such flag on dequeue.
Since the empty status lays on the same cache-line of the
seqlock, it's always hot on cache during the updates.
This makes the empty flag update a little bit loosy. Given
the lack of synchronization between enqueue and dequeue, this
is unavoidable.
v2 -> v3:
- qdisc_is_empty() has a const argument (Eric)
v1 -> v2:
- use really an 'empty' flag instead of 'not_empty', as
suggested by Eric
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Ivan Vecera <ivecera@redhat.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>
Set TCF_PROTO_OPS_DOIT_UNLOCKED for flower classifier to indicate that its
ops callbacks don't require caller to hold rtnl lock. Don't take rtnl lock
in fl_destroy_filter_work() that is executed on workqueue instead of being
called by cls API and is not affected by setting
TCF_PROTO_OPS_DOIT_UNLOCKED. Rtnl mutex is still manually taken by flower
classifier before calling hardware offloads API that has not been updated
for unlocked execution.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use 'rtnl_held' flag to track if caller holds rtnl lock. Propagate the flag
to internal functions that need to know rtnl lock state. Take rtnl lock
before calling tcf APIs that require it (hw offload, bind filter, etc.).
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
struct tcf_proto was extended with spinlock to be used by classifiers
instead of global rtnl lock. Use it to protect shared flower classifier
data structures (handle_idr, mask hashtable and list) and fields of
individual filters that can be accessed concurrently. This patch set uses
tcf_proto->lock as per instance lock that protects all filters on
tcf_proto.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Without rtnl lock protection tcf proto can be deleted concurrently. Check
tcf proto 'deleting' flag after taking tcf spinlock to verify that no
concurrent deletion is in progress. Return EAGAIN error if concurrent
deletion detected, which will cause caller to retry and possibly create new
instance of tcf proto.
Retry mechanism is a result of fine-grained locking approach used in this
and previous changes in series and is necessary to allow concurrent updates
on same chain instance. Alternative approach would be to lock the whole
chain while updating filters on any of child tp's, adding and removing
classifier instances from the chain. However, since most CPU-intensive
parts of filter update code are specifically in classifier code and its
dependencies (extensions and hw offloads), such approach would negate most
of the gains introduced by this change and previous changes in the series
when updating same chain instance.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Check if user specified a handle and another filter with the same handle
was inserted concurrently. Return EAGAIN to retry filter processing (in
case it is an overwrite request).
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Protect modifications of flower masks list with spinlock to remove
dependency on rtnl lock and allow concurrent access.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Without rtnl lock protection masks with same key can be inserted
concurrently. Insert temporary mask with reference count zero to masks
hashtable. This will cause any concurrent modifications to retry.
Wait for rcu grace period to complete after removing temporary mask from
masks hashtable to accommodate concurrent readers.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Suggested-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extend fl_flow_mask structure with reference counter to allow parallel
modification without relying on rtnl lock. Use rcu read lock to safely
lookup mask and increment reference counter in order to accommodate
concurrent deletes.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to prevent double deletion of filter by concurrent tasks when rtnl
lock is not used for synchronization, add 'deleted' filter field. Check
value of this field when modifying filters and return error if concurrent
deletion is detected.
Refactor __fl_delete() to accept pointer to 'last' boolean as argument,
and return error code as function return value instead. This is necessary
to signal concurrent filter delete to caller.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extend flower filters with reference counting in order to remove dependency
on rtnl lock in flower ops and allow to modify filters concurrently.
Reference to flower filter can be taken/released concurrently as soon as it
is marked as 'unlocked' by last patch in this series. Use atomic reference
counter type to make concurrent modifications safe.
Always take reference to flower filter while working with it:
- Modify fl_get() to take reference to filter.
- Implement tp->put() callback as fl_put() function to allow cls API to
release reference taken by fl_get().
- Modify fl_change() to assume that caller holds reference to fold and take
reference to fnew.
- Take reference to filter while using it in fl_walk().
Implement helper functions to get/put filter reference counter.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As a preparation for using classifier spinlock instead of relying on
external rtnl lock, rearrange code in fl_change. The goal is to group the
code which changes classifier state in single block in order to allow
following commits in this set to protect it from parallel modification with
tp->lock. Data structures that require tp->lock protection are mask
hashtable and filters list, and classifier handle_idr.
fl_hw_replace_filter() is a sleeping function and cannot be called while
holding a spinlock. In order to execute all sequence of changes to shared
classifier data structures atomically, call fl_hw_replace_filter() before
modifying them.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Flower classifier only changes root pointer during init and destroy. Cls
API implements reference counting for tcf_proto, so there is no danger of
concurrent access to tp when it is being destroyed, even without protection
provided by rtnl lock.
Implement new function fl_head_dereference() to dereference tp->root
without checking for rtnl lock. Use it in all flower function that obtain
head pointer instead of rtnl_dereference().
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.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>
net and null_fallback are redundant. Remove null_fallback in favor of
!net check.
Signed-off-by: David Ahern <dsahern@gmail.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>
Change addrconf_f6i_alloc to generate a fib6_config and call
ip6_route_info_create. addrconf_f6i_alloc is the last caller to
fib6_info_alloc besides ip6_route_info_create, and there is no
reason for it to do its own initialization on a fib6_info.
Host routes need to be created even if the device is down, so add a
new flag, fc_ignore_dev_down, to fib6_config and update fib6_nh_init
to not error out if device is not up.
Notes on the conversion:
- ip_fib_metrics_init is the same as fib6_config has fc_mx set to NULL
and fc_mx_len set to 0
- dst_nocount is handled by the RTF_ADDRCONF flag
- dst_host is handled by fc_dst_len = 128
nh_gw does not get set after the conversion to ip6_route_info_create
but it should not be set in addrconf_f6i_alloc since this is a host
route not a gateway route.
Everything else is a straight forward map between fib6_info and
fib6_config.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
ip6_route_info_create is a low level function for ensuring fc_metric is
set. Move the check and default setting to the 2 locations that do not
already set fc_metric before calling ip6_route_info_create. This is
required for the next patch which moves addrconf allocations to
ip6_route_info_create and want the metric for host routes to be 0.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
To free the skb in normal course of processing, consume_skb() should be
used. Only for failure paths, skb_free() is intended to be used.
https://www.kernel.org/doc/htmldocs/networking/API-consume-skb.html
Signed-off-by: Vakul Garg <vakul.garg@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
skb free-ed in:
1/ condition 1: tipc_sk_filter_rcv -> tipc_sk_proto_rcv
2/ condition 2: tipc_sk_filter_rcv -> tipc_group_filter_msg
This leads to a "use-after-free" access in the next condition.
We fix this by intializing the variable at declaration, then it is safe
to check this variable to continue processing if condition matches.
syzbot report:
==================================================================
BUG: KASAN: use-after-free in tipc_sk_filter_rcv+0x2166/0x34f0
net/tipc/socket.c:2167
Read of size 4 at addr ffff88808ea58534 by task kworker/u4:0/7
CPU: 0 PID: 7 Comm: kworker/u4:0 Not tainted 5.0.0+ #61
Hardware name: Google Google Compute Engine/Google Compute Engine,
BIOS Google 01/01/2011
Workqueue: tipc_send tipc_conn_send_work
Call Trace:
__dump_stack lib/dump_stack.c:77 [inline]
dump_stack+0x172/0x1f0 lib/dump_stack.c:113
print_address_description.cold+0x7c/0x20d mm/kasan/report.c:187
kasan_report.cold+0x1b/0x40 mm/kasan/report.c:317
__asan_report_load4_noabort+0x14/0x20 mm/kasan/generic_report.c:131
tipc_sk_filter_rcv+0x2166/0x34f0 net/tipc/socket.c:2167
tipc_sk_enqueue net/tipc/socket.c:2254 [inline]
tipc_sk_rcv+0xc45/0x25a0 net/tipc/socket.c:2305
tipc_topsrv_kern_evt+0x3b7/0x580 net/tipc/topsrv.c:610
tipc_conn_send_to_sock+0x43e/0x5f0 net/tipc/topsrv.c:283
tipc_conn_send_work+0x65/0x80 net/tipc/topsrv.c:303
process_one_work+0x98e/0x1790 kernel/workqueue.c:2269
worker_thread+0x98/0xe40 kernel/workqueue.c:2415
kthread+0x357/0x430 kernel/kthread.c:253
ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:352
Reported-by: syzbot+e863893591cc7a622e40@syzkaller.appspotmail.com
Fixes: c55c8eda ("tipc: smooth change between replicast and broadcast")
Acked-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: Hoang Le <hoang.h.le@dektech.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
In addition to icmp_echo_ignore_multicast, there is a need to also
prevent responding to pings to anycast addresses for security.
Signed-off-by: Stephen Suryaputra <ssuryaextr@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
After the previous patch, all the callers of ndo_select_queue()
provide as a 'fallback' argument netdev_pick_tx.
The only exceptions are nested calls to ndo_select_queue(),
which pass down the 'fallback' available in the current scope
- still netdev_pick_tx.
We can drop such argument and replace fallback() invocation with
netdev_pick_tx(). This avoids an indirect call per xmit packet
in some scenarios (TCP syn, UDP unconnected, XDP generic, pktgen)
with device drivers implementing such ndo. It also clean the code
a bit.
Tested with ixgbe and CONFIG_FCOE=m
With pktgen using queue xmit:
threads vanilla patched
(kpps) (kpps)
1 2334 2428
2 4166 4278
4 7895 8100
v1 -> v2:
- rebased after helper's name change
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently packet_pick_tx_queue() is the only caller of
ndo_select_queue() using a fallback argument other than
netdev_pick_tx.
Leveraging rx queue, we can obtain a similar queue selection
behavior using core helpers. After this change, ndo_select_queue()
is always invoked with netdev_pick_tx() as fallback.
We can change ndo_select_queue() signature in a followup patch,
dropping an indirect call per transmitted packet in some scenarios
(e.g. TCP syn and XDP generic xmit)
This changes slightly how af packet queue selection happens when
PACKET_QDISC_BYPASS is set. It's now more similar to plan dev_queue_xmit()
tacking in account both XPS and TC mapping.
v1 -> v2:
- rebased after helper name change
RFC -> v1:
- initialize sender_cpu to the expected value
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With the following patches, we are going to use __netdev_pick_tx() in
many modules. Rename it to netdev_pick_tx(), to make it clear is
a public API.
Also rename the existing netdev_pick_tx() to netdev_core_pick_tx(),
to avoid name clashes.
Suggested-by: Eric Dumazet <edumazet@google.com>
Suggested-by: David Miller <davem@davemloft.net>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch is to use eth_broadcast_addr() to assign broadcast address
insetad of memset().
Signed-off-by: Mao Wenan <maowenan@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Added support for AES128-CCM based record encryption. AES128-CCM is
similar to AES128-GCM. Both of them have same salt/iv/mac size. The
notable difference between the two is that while invoking AES128-CCM
operation, the salt||nonce (which is passed as IV) has to be prefixed
with a hardcoded value '2'. Further, CCM implementation in kernel
requires IV passed in crypto_aead_request() to be full '16' bytes.
Therefore, the record structure 'struct tls_rec' has been modified to
reserve '16' bytes for IV. This works for both GCM and CCM based cipher.
Signed-off-by: Vakul Garg <vakul.garg@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
IPv4 has icmp_echo_ignore_broadcast to prevent responding to broadcast pings.
IPv6 needs a similar mechanism.
v1->v2:
- Remove NET_IPV6_ICMP_ECHO_IGNORE_MULTICAST.
Signed-off-by: Stephen Suryaputra <ssuryaextr@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>
Fix sparse warning:
net/core/datagram.c:411:5: warning:
symbol '__skb_datagram_iter' was not declared. Should it be static?
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
TCP ipv6 fast path dereferences a pointer to get to the inet6
part of a tcp socket, but given the fixed memory placement,
we can do better and avoid a possible cache line miss.
This also reduces register pressure, since we let the compiler
know about this memory placement.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, a multicast stream may start out using replicast, because
there are few destinations, and then it should ideally switch to
L2/broadcast IGMP/multicast when the number of destinations grows beyond
a certain limit. The opposite should happen when the number decreases
below the limit.
To eliminate the risk of message reordering caused by method change,
a sending socket must stick to a previously selected method until it
enters an idle period of 5 seconds. Means there is a 5 seconds pause
in the traffic from the sender socket.
If the sender never makes such a pause, the method will never change,
and transmission may become very inefficient as the cluster grows.
With this commit, we allow such a switch between replicast and
broadcast without any need for a traffic pause.
Solution is to send a dummy message with only the header, also with
the SYN bit set, via broadcast or replicast. For the data message,
the SYN bit is set and sending via replicast or broadcast (inverse
method with dummy).
Then, at receiving side any messages follow first SYN bit message
(data or dummy message), they will be held in deferred queue until
another pair (dummy or data message) arrived in other link.
v2: reverse christmas tree declaration
Acked-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: Hoang Le <hoang.h.le@dektech.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
As a preparation for introducing a smooth switching between replicast
and broadcast method for multicast message, We have to introduce a new
capability flag TIPC_MCAST_RBCTL to handle this new feature.
During a cluster upgrade a node can come back with this new capabilities
which also must be reflected in the cluster capabilities field.
The new feature is only applicable if all node in the cluster supports
this new capability.
Acked-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: Hoang Le <hoang.h.le@dektech.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, a multicast stream uses either broadcast or replicast as
transmission method, based on the ratio between number of actual
destinations nodes and cluster size.
However, when an L2 interface (e.g., VXLAN) provides pseudo
broadcast support, this becomes very inefficient, as it blindly
replicates multicast packets to all cluster/subnet nodes,
irrespective of whether they host actual target sockets or not.
The TIPC multicast algorithm is able to distinguish real destination
nodes from other nodes, and hence provides a smarter and more
efficient method for transferring multicast messages than
pseudo broadcast can do.
Because of this, we now make it possible for users to force
the broadcast link to permanently switch to using replicast,
irrespective of which capabilities the bearer provides,
or pretend to provide.
Conversely, we also make it possible to force the broadcast link
to always use true broadcast. While maybe less useful in
deployed systems, this may at least be useful for testing the
broadcast algorithm in small clusters.
We retain the current AUTOSELECT ability, i.e., to let the broadcast link
automatically select which algorithm to use, and to switch back and forth
between broadcast and replicast as the ratio between destination
node number and cluster size changes. This remains the default method.
Furthermore, we make it possible to configure the threshold ratio for
such switches. The default ratio is now set to 10%, down from 25% in the
earlier implementation.
Acked-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: Hoang Le <hoang.h.le@dektech.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
A similar fix as Patch "sctp: fix ignoring asoc_id for tcp-style sockets on
SCTP_DEFAULT_SEND_PARAM sockopt" on SCTP_STREAM_SCHEDULER sockopt.
Fixes: 7efba10d6b ("sctp: add SCTP_FUTURE_ASOC and SCTP_CURRENT_ASSOC for SCTP_STREAM_SCHEDULER sockopt")
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
