Add necessary plumbing at the slave network device level to have switch
drivers implement ndo_setup_tc() and most particularly the cls_matchall
classifier. We add support for two switch operations:

port_add_mirror and port_del_mirror() which configure, on a per-port
basis the mirror parameters requested from the cls_matchall classifier.

Code is largely borrowed from the Mellanox Spectrum switch driver.
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Nothing about lwt state requires a device reference, so remove the
input argument.
Signed-off-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Packets arriving in a VRF currently are delivered to UDP sockets that
aren't bound to any interface. TCP defaults to not delivering packets
arriving in a VRF to unbound sockets. IP route lookup and socket
transmit both assume that unbound means using the default table and
UDP applications that haven't been changed to be aware of VRFs may not
function correctly in this case since they may not be able to handle
overlapping IP address ranges, or be able to send packets back to the
original sender if required.

So add a sysctl, udp_l3mdev_accept, to control this behaviour with it
being analgous to the existing tcp_l3mdev_accept, namely to allow a
process to have a VRF-global listen socket. Have this default to off
as this is the behaviour that users will expect, given that there is
no explicit mechanism to set unmodified VRF-unaware application into a
default VRF.
Signed-off-by: NRobert Shearman <rshearma@brocade.com>
Acked-by: NDavid Ahern <dsa@cumulusnetworks.com>
Tested-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In preparation for adding support for CFP/TCAMP in the bcm_sf2 driver add the
plumbing to call into driver specific {get,set}_rxnfc operations.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds devm_alloc_etherdev_mqs function and devm_alloc_etherdev
macro. These can be used for simpler netdev allocation without having to
care about calling free_netdev.

Thanks to this change drivers, their error paths and removal paths may
get simpler by a bit.
Signed-off-by: NRafał Miłecki <rafal@milecki.pl>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently the retransmission stats are not incremented if the
retransmit fails locally. But we always increment the other packet
counters that track total packet/bytes sent.  Awkwardly while we
don't count these failed retransmits in RETRANSSEGS, we do count
them in FAILEDRETRANS.

If the qdisc is dropping many packets this could under-estimate
TCP retransmission rate substantially from both SNMP or per-socket
TCP_INFO stats. This patch changes this by always incrementing
retransmission stats on retransmission attempts and failures.

Another motivation is to properly track retransmists in
SCM_TIMESTAMPING_OPT_STATS. Since SCM_TSTAMP_SCHED collection is
triggered in tcp_transmit_skb(), If tp->total_retrans is incremented
after the function, we'll always mis-count by the amount of the
latest retransmission.
Signed-off-by: NYuchung Cheng <ycheng@google.com>
Signed-off-by: NSoheil Hassas Yeganeh <soheil@google.com>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add two stats in SCM_TIMESTAMPING_OPT_STATS:

TCP_NLA_DATA_SEGS_OUT: total data packets sent including retransmission
TCP_NLA_TOTAL_RETRANS: total data packets retransmitted

The names are picked to be consistent with corresponding fields in
TCP_INFO. This allows applications that are using the timestamping
API to measure latency stats to also retrive retransmission rate
of application write.
Signed-off-by: NYuchung Cheng <ycheng@google.com>
Signed-off-by: NSoheil Hassas Yeganeh <soheil@google.com>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

do_execute_actions() implements a worthwhile optimization: in case
an output action is the last action in an action list, skb_clone()
can be avoided by outputing the current skb. However, the
implementation is more complicated than necessary.  This patch
simplify this logic.
Signed-off-by: NAndy Zhou <azhou@ovn.org>
Acked-by: NPravin B Shelar <pshelar@ovn.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Upon reception of the NETDEV_CHANGEUPPER, a leaving port is already
unbridged, so reflect this by assigning the port's bridge_dev pointer to
NULL before calling the port_bridge_leave DSA driver operation.

Now that the bridge_dev pointer is exposed to the drivers, reflecting
the current state of the DSA switch fabric is necessary for the drivers
to adjust their port based VLANs correctly.

Pass the bridge device pointer to the port_bridge_leave operation so
that drivers have all information to re-program their chips properly,
and do not need to cache it anymore.
Signed-off-by: NVivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Move the bridge_dev pointer from dsa_slave_priv to dsa_port so that DSA
drivers can access this information and remove the need to cache it.
Signed-off-by: NVivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Store a pointer to the dsa_port structure in the dsa_slave_priv
structure, instead of the switch/port index.

This will allow to store more information such as the bridge device,
needed in DSA drivers for multi-chip configuration.
Signed-off-by: NVivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add the physical switch instance and port index a DSA port belongs to to
the dsa_port structure.

That can be used later to retrieve information about a physical port
when configuring a switch fabric, or lighten up struct dsa_slave_priv.
Signed-off-by: NVivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The dsa_switch structure contains the number of ports. Use it where the
structure is valid instead of the DSA_MAX_PORTS value.
Signed-off-by: NVivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Change the ports[DSA_MAX_PORTS] array of the dsa_switch structure for a
zero-length array, allocated at the same time as the dsa_switch
structure itself. A dsa_switch_alloc() helper is provided for that.

This commit brings no functional change yet since we pass DSA_MAX_PORTS
as the number of ports for the moment. Future patches can update the DSA
drivers separately to support dynamic number of ports.
Signed-off-by: NVivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

IS_ERR/ERR_PTR are not defined in linux/device.h but in linux/err.h. The
files using these macros therefore have to include the correct one.
Reported-by: NLinus Luessing <linus.luessing@web.de>
Signed-off-by: NSven Eckelmann <sven@narfation.org>
Signed-off-by: NSimon Wunderlich <sw@simonwunderlich.de>

The net_xmit_eval has side effects because it is not making sure that e
isn't evaluated twice.

    #define net_xmit_eval(e)        ((e) == NET_XMIT_CN ? 0 : (e))

The code requested by David Miller [1]

    return net_xmit_eval(dev_queue_xmit(skb));

will get transformed into

    return ((dev_queue_xmit(skb)) == NET_XMIT_CN ? 0 : (dev_queue_xmit(skb)))

dev_queue_xmit will therefore be tried again (with an already consumed skb)
whenever the return code is not NET_XMIT_CN.

[1] https://lkml.kernel.org/r/20170125.225624.965229145391320056.davem@davemloft.net

Fixes: c3370518 ("batman-adv: Treat NET_XMIT_CN as transmit successfully")
Signed-off-by: NSven Eckelmann <sven@narfation.org>
Signed-off-by: NSimon Wunderlich <sw@simonwunderlich.de>

Slava Shwartsman reported a warning in skb_try_coalesce(), when we
detect skb->truesize is completely wrong.

In his case, issue came from IPv6 reassembly coping with malicious
datagrams, that forced various pskb_may_pull() to reallocate a bigger
skb->head than the one allocated by NIC driver before entering GRO
layer.

Current code does not change skb->truesize, leaving this burden to
callers if they care enough.

Blindly changing skb->truesize in pskb_expand_head() is not
easy, as some producers might track skb->truesize, for example
in xmit path for back pressure feedback (sk->sk_wmem_alloc)

We can detect the cases where it should be safe to change
skb->truesize :

1) skb is not attached to a socket.
2) If it is attached to a socket, destructor is sock_edemux()

My audit gave only two callers doing their own skb->truesize
manipulation.

I had to remove skb parameter in sock_edemux macro when
CONFIG_INET is not set to avoid a compile error.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Reported-by: NSlava Shwartsman <slavash@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Unlike ipv4, this control socket is shared by all cpus so we cannot use
it as scratchpad area to annotate the mark that we pass to ip6_xmit().

Add a new parameter to ip6_xmit() to indicate the mark. The SCTP socket
family caches the flowi6 structure in the sctp_transport structure, so
we cannot use to carry the mark unless we later on reset it back, which
I discarded since it looks ugly to me.

Fixes: bf99b4de ("tcp: fix mark propagation with fwmark_reflect enabled")
Suggested-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NPablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NFelix Jia <felix.jia@alliedtelesis.co.nz>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The address generation mode for IPv6 link-local can only be configured
by netlink messages. This patch adds the ability to change the address
generation mode via sysctl.

v1 -> v2
Removed the rtnl lock and switch to use RCU lock to iterate through
the netdev list.

v2 -> v3
Removed the addrgenmode variable from the idev structure and use the
systcl storage for the flag.

Simplifed the logic for sysctl handling by removing the supported
for all operation.

Added support for more types of tunnel interfaces for link-local
address generation.

Based the patches from net-next.

v3 -> v4
Removed unnecessary whitespace changes.
Signed-off-by: NFelix Jia <felix.jia@alliedtelesis.co.nz>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

lkp-robot reported a BUG:
[   10.151226] BUG: unable to handle kernel NULL pointer dereference at 00000198
[   10.152525] IP: rt6_fill_node+0x164/0x4b8
[   10.153307] *pdpt = 0000000012ee5001 *pde = 0000000000000000
[   10.153309]
[   10.154492] Oops: 0000 [#1]
[   10.154987] CPU: 0 PID: 909 Comm: netifd Not tainted 4.10.0-rc4-00722-g41e8c70e-dirty #10
[   10.156482] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.7.5-20140531_083030-gandalf 04/01/2014
[   10.158254] task: d0deb000 task.stack: d0e0c000
[   10.159059] EIP: rt6_fill_node+0x164/0x4b8
[   10.159780] EFLAGS: 00010296 CPU: 0
[   10.160404] EAX: 00000000 EBX: d10c2358 ECX: c1f7c6cc EDX: c1f6ff44
[   10.161469] ESI: 00000000 EDI: c2059900 EBP: d0e0dc4c ESP: d0e0dbe4
[   10.162534]  DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 0068
[   10.163482] CR0: 80050033 CR2: 00000198 CR3: 10d94660 CR4: 000006b0
[   10.164535] Call Trace:
[   10.164993]  ? paravirt_sched_clock+0x9/0xd
[   10.165727]  ? sched_clock+0x9/0xc
[   10.166329]  ? sched_clock_cpu+0x19/0xe9
[   10.166991]  ? lock_release+0x13e/0x36c
[   10.167652]  rt6_dump_route+0x4c/0x56
[   10.168276]  fib6_dump_node+0x1d/0x3d
[   10.168913]  fib6_walk_continue+0xab/0x167
[   10.169611]  fib6_walk+0x2a/0x40
[   10.170182]  inet6_dump_fib+0xfb/0x1e0
[   10.170855]  netlink_dump+0xcd/0x21f

This happens when the loopback device is set down and a ipv6 fib route
dump is requested.

ip6_null_entry is the root of all ipv6 fib tables making it integrated
into the table and hence passed to the ipv6 route dump code. The
null_entry route uses the loopback device for dst.dev but may not have
rt6i_idev set because of the order in which initializations are done --
ip6_route_net_init is run before addrconf_init has initialized the
loopback device. Fixing the initialization order is a much bigger problem
with no obvious solution thus far.

The BUG is triggered when the loopback is set down and the netif_running
check added by a1a22c12 fails. The fill_node descends to checking
rt->rt6i_idev for ignore_routes_with_linkdown and since rt6i_idev is
NULL it faults.

The null_entry route should not be processed in a dump request. Catch
and ignore. This check is done in rt6_dump_route as it is the highest
place in the callchain with knowledge of both the route and the network
namespace.

Fixes: a1a22c12("net: ipv6: Keep nexthop of multipath route on admin down")
Signed-off-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Remove skb_reserve and skb_reset_mac_header from inet6_rtm_getroute. The
allocated skb is not passed through the routing engine (like it is for
IPv4) and has not since the beginning of git time.
Signed-off-by: NDavid Ahern <dsa@cumulusnetworks.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Move the assignment of ports in _dsa_register_switch() closer to where
it is checked, no functional change. Re-order declarations to be
preserve the inverted christmas tree style.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Make it clear that these functions take a device_node structure pointer
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In preparation for allowing platform data, and therefore no valid
device_node pointer, make most DSA functions takes a pointer to a
dsa_port structure whenever possible. While at it, introduce a
dsa_port_is_valid() helper function which checks whether port->dn is
NULL or not at the moment.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In preparation for allowing dsa_register_switch() to be supplied with
device/platform data, pass down a struct device pointer instead of a
struct device_node.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The tc could return NET_XMIT_CN as one congestion notification, but
it does not mean the packet is lost. Other modules like ipvlan,
macvlan, and others treat NET_XMIT_CN as success too.

So batman-adv should handle NET_XMIT_CN also as NET_XMIT_SUCCESS.
Signed-off-by: NGao Feng <gfree.wind@gmail.com>
[sven@narfation.org: Moved NET_XMIT_CN handling to batadv_send_skb_packet]
Signed-off-by: NSven Eckelmann <sven@narfation.org>
Signed-off-by: NSimon Wunderlich <sw@simonwunderlich.de>

It could decrease one condition check to collect some statements in the
first condition block.
Signed-off-by: NGao Feng <gfree.wind@gmail.com>
Signed-off-by: NSven Eckelmann <sven@narfation.org>
Signed-off-by: NSimon Wunderlich <sw@simonwunderlich.de>

Signed-off-by: NSven Eckelmann <sven@narfation.org>
Signed-off-by: NSimon Wunderlich <sw@simonwunderlich.de>

Signed-off-by: NSven Eckelmann <sven@narfation.org>
Signed-off-by: NSimon Wunderlich <sw@simonwunderlich.de>

The bridge loop avoidance (BLA) feature of batman-adv sends packets to
probe for Mesh/LAN packet loops. Those packets are not sent by real
clients and should therefore not be added to the translation table (TT).
Signed-off-by: NSimon Wunderlich <simon.wunderlich@open-mesh.com>

The only caller of this new function is inside of an #ifdef checking
for CONFIG_BRIDGE_IGMP_SNOOPING, so we should move the implementation
there too, in order to avoid this harmless warning:

net/bridge/br_forward.c:177:13: error: 'maybe_deliver_addr' defined but not used [-Werror=unused-function]

Fixes: 6db6f0ea ("bridge: multicast to unicast")
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Acked-by: NStephen Hemminger <stephen@networkplumber.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 448b4482 ("net: dsa: Add lockdep class to tx queues to avoid
lockdep splat") removed the netif_device_detach() call done in
dsa_slave_suspend() which is necessary, and paired with a corresponding
netif_device_attach(), bring it back.

Fixes: 448b4482 ("net: dsa: Add lockdep class to tx queues to avoid lockdep splat")
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Reviewed-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Without TFO, any subsequent connect() call after a successful one returns
-1 EISCONN. The last API update ensured that __inet_stream_connect() can
return -1 EINPROGRESS in response to sendmsg() when TFO is in use to
indicate that the connection is now in progress. Unfortunately since this
function is used both for connect() and sendmsg(), it has the undesired
side effect of making connect() now return -1 EINPROGRESS as well after
a successful call, while at the same time poll() returns POLLOUT. This
can confuse some applications which happen to call connect() and to
check for -1 EISCONN to ensure the connection is usable, and for which
EINPROGRESS indicates a need to poll, causing a loop.

This problem was encountered in haproxy where a call to connect() is
precisely used in certain cases to confirm a connection's readiness.
While arguably haproxy's behaviour should be improved here, it seems
important to aim at a more robust behaviour when the goal of the new
API is to make it easier to implement TFO in existing applications.

This patch simply ensures that we preserve the same semantics as in
the non-TFO case on the connect() syscall when using TFO, while still
returning -1 EINPROGRESS on sendmsg(). For this we simply tell
__inet_stream_connect() whether we're doing a regular connect() or in
fact connecting for a sendmsg() call.

Cc: Wei Wang <weiwan@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: NWilly Tarreau <w@1wt.eu>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds a new socket option, TCP_FASTOPEN_CONNECT, as an
alternative way to perform Fast Open on the active side (client). Prior
to this patch, a client needs to replace the connect() call with
sendto(MSG_FASTOPEN). This can be cumbersome for applications who want
to use Fast Open: these socket operations are often done in lower layer
libraries used by many other applications. Changing these libraries
and/or the socket call sequences are not trivial. A more convenient
approach is to perform Fast Open by simply enabling a socket option when
the socket is created w/o changing other socket calls sequence:
  s = socket()
    create a new socket
  setsockopt(s, IPPROTO_TCP, TCP_FASTOPEN_CONNECT …);
    newly introduced sockopt
    If set, new functionality described below will be used.
    Return ENOTSUPP if TFO is not supported or not enabled in the
    kernel.

  connect()
    With cookie present, return 0 immediately.
    With no cookie, initiate 3WHS with TFO cookie-request option and
    return -1 with errno = EINPROGRESS.

  write()/sendmsg()
    With cookie present, send out SYN with data and return the number of
    bytes buffered.
    With no cookie, and 3WHS not yet completed, return -1 with errno =
    EINPROGRESS.
    No MSG_FASTOPEN flag is needed.

  read()
    Return -1 with errno = EWOULDBLOCK/EAGAIN if connect() is called but
    write() is not called yet.
    Return -1 with errno = EWOULDBLOCK/EAGAIN if connection is
    established but no msg is received yet.
    Return number of bytes read if socket is established and there is
    msg received.

The new API simplifies life for applications that always perform a write()
immediately after a successful connect(). Such applications can now take
advantage of Fast Open by merely making one new setsockopt() call at the time
of creating the socket. Nothing else about the application's socket call
sequence needs to change.
Signed-off-by: NWei Wang <weiwan@google.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Acked-by: NYuchung Cheng <ycheng@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Remove __sk_dst_reset() in the failure handling because __sk_dst_reset()
will eventually get called when sk is released. No need to handle it in
the protocol specific connect call.
This is also to make the code path consistent with ipv4.
Signed-off-by: NWei Wang <weiwan@google.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Refactor the cookie check logic in tcp_send_syn_data() into a function.
This function will be called else where in later changes.
Signed-off-by: NWei Wang <weiwan@google.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Acked-by: NYuchung Cheng <ycheng@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

sock_reset_flag() maps to __clear_bit() not the atomic version clear_bit().
Thus, we need smp_mb(), smp_mb__after_atomic() is not sufficient.

Fixes: 3c715127 ("tcp: add memory barriers to write space paths")
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: NJason Baron <jbaron@akamai.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Reported-by: NOleg Nesterov <oleg@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

tcp_add_backlog() can use skb_condense() helper to get better
gains and less SKB_TRUESIZE() magic. This only happens when socket
backlog has to be used.

Some attacks involve specially crafted out of order tiny TCP packets,
clogging the ofo queue of (many) sockets.
Then later, expensive collapse happens, trying to copy all these skbs
into single ones.
This unfortunately does not work if each skb has no neighbor in TCP
sequence order.

By using skb_condense() if the skb could not be coalesced to a prior
one, we defeat these kind of threats, potentially saving 4K per skb
(or more, since this is one page fragment).

A typical NAPI driver allocates gro packets with GRO_MAX_HEAD bytes
in skb->head, meaning the copy done by skb_condense() is limited to
about 200 bytes.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We shuffled some code around and added some new case statements here and
now "res" isn't initialized on all paths.

Fixes: 01fd12bb ("tipc: make replicast a user selectable option")
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>