We used to queue tx packets in sk_receive_queue, this is less
efficient since it requires spinlocks to synchronize between producer
and consumer.

This patch tries to address this by:

- switch from sk_receive_queue to a skb_array, and resize it when
  tx_queue_len was changed.
- introduce a new proto_ops peek_len which was used for peeking the
  skb length.
- implement a tun version of peek_len for vhost_net to use and convert
  vhost_net to use peek_len if possible.

Pktgen test shows about 15.3% improvement on guest receiving pps for small
buffers:

Before: ~1300000pps
After : ~1500000pps
Signed-off-by: NJason Wang <jasowang@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The commit 34166093 ("tuntap: use common code for virtio_net_hdr
and skb GSO conversion") replaced the tun code for header manipulation
with the generic helpers. While doing so, it implictly moved the
skb_partial_csum_set() invocation after eth_type_trans(), which
invalidate the current gso start/offset values.
Fix it by moving the helper invocation before the mac pulling.

Fixes: 34166093 ("tuntap: use common code for virtio_net_hdr and skb GSO conversion")
Signed-off-by: NPaolo Abeni <pabeni@redhat.com>
Acked-by: NMike Rapoport <rppt@linux.vnet.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Replace open coded conversion between virtio_net_hdr to skb GSO info with
virtio_net_hdr_{from,to}_skb
Signed-off-by: NMike Rapoport <rppt@linux.vnet.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We used to check dev->reg_state against NETREG_REGISTERED after each
time we are woke up. But after commit 9e641bdc ("net-tun:
restructure tun_do_read for better sleep/wakeup efficiency"), it uses
skb_recv_datagram() which does not check dev->reg_state. This will
result if we delete a tun/tap device after a process is blocked in the
reading. The device will wait for the reference count which was held
by that process for ever.

Fixes this by using RCV_SHUTDOWN which will be checked during
sk_recv_datagram() before trying to wake up the process during uninit.

Fixes: 9e641bdc ("net-tun: restructure tun_do_read for better
sleep/wakeup efficiency")
Cc: Eric Dumazet <edumazet@google.com>
Cc: Xi Wang <xii@google.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: NJason Wang <jasowang@redhat.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Acked-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There's no need to calculate rps hash if it was not enabled. So this
patch export rps_needed and check it before trying to get rps
hash. Tests (using pktgen to inject packets to guest) shows this can
improve pps about 13% (when rps is disabled).

Before:
~1150000 pps
After:
~1300000 pps

Cc: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: NJason Wang <jasowang@redhat.com>
----
Changes from V1:
- Fix build when CONFIG_RPS is not set
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The current tun_net_xmit() implementation don't need any external
lock since it relies on rcu protection for the tun data structure
and on socket queue lock for skb queuing.

This patch set the NETIF_F_LLTX feature bit in the tun device, so
that on xmit, in absence of qdisc, no serialization lock is acquired
by the caller.

The user space can remove the default tun qdisc with:

tc qdisc replace dev <tun device name> root noqueue
Signed-off-by: NPaolo Abeni <pabeni@redhat.com>
Acked-by: NHannes Frederic Sowa <hannes@stressinduktion.org>
Acked-by: NEric Dumazet <edumazet@google.com>
Acked-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently the tun device accounting uses dev->stats without applying any
kind of protection, regardless that accounting happens in preemptible
process context.
This patch move the tun stats to a per cpu data structure, and protect
the updates with  u64_stats_update_begin()/u64_stats_update_end() or
this_cpu_inc according to the stat type. The per cpu stats are
aggregated by the newly added ndo_get_stats64 ops.
Signed-off-by: NPaolo Abeni <pabeni@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

After commit f84bb1ea ("net: fix IFF_NO_QUEUE for drivers using
alloc_netdev"), default qdisc was changed to noqueue because
tuntap does not set tx_queue_len during .setup(). This patch restores
default qdisc by setting tx_queue_len in tun_setup().

Fixes: f84bb1ea ("net: fix IFF_NO_QUEUE for drivers using alloc_netdev")
Cc: Phil Sutter <phil@nwl.cc>
Signed-off-by: NJason Wang <jasowang@redhat.com>
Acked-by: NMichael S. Tsirkin <mst@redhat.com>
Acked-by: NPhil Sutter <phil@nwl.cc>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This reverts commit 5a5abb1f ("tun, bpf: fix suspicious RCU usage
in tun_{attach, detach}_filter") and replaces it to use lock_sock around
sk_{attach,detach}_filter. The checks inside filter.c are updated with
lockdep_sock_is_held to check for proper socket locks.

It keeps the code cleaner by ensuring that only one lock governs the
socket filter instead of two independent locks.

Cc: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: NHannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, SOL_TIMESTAMPING can only be enabled using setsockopt.
This is very costly when users want to sample writes to gather
tx timestamps.

Add support for enabling SO_TIMESTAMPING via control messages by
using tsflags added in `struct sockcm_cookie` (added in the previous
patches in this series) to set the tx_flags of the last skb created in
a sendmsg. With this patch, the timestamp recording bits in tx_flags
of the skbuff is overridden if SO_TIMESTAMPING is passed in a cmsg.

Please note that this is only effective for overriding the recording
timestamps flags. Users should enable timestamp reporting (e.g.,
SOF_TIMESTAMPING_SOFTWARE | SOF_TIMESTAMPING_OPT_ID) using
socket options and then should ask for SOF_TIMESTAMPING_TX_*
using control messages per sendmsg to sample timestamps for each
write.
Signed-off-by: NSoheil Hassas Yeganeh <soheil@google.com>
Acked-by: NWillem de Bruijn <willemb@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Sasha Levin reported a suspicious rcu_dereference_protected() warning
found while fuzzing with trinity that is similar to this one:

  [   52.765684] net/core/filter.c:2262 suspicious rcu_dereference_protected() usage!
  [   52.765688] other info that might help us debug this:
  [   52.765695] rcu_scheduler_active = 1, debug_locks = 1
  [   52.765701] 1 lock held by a.out/1525:
  [   52.765704]  #0:  (rtnl_mutex){+.+.+.}, at: [<ffffffff816a64b7>] rtnl_lock+0x17/0x20
  [   52.765721] stack backtrace:
  [   52.765728] CPU: 1 PID: 1525 Comm: a.out Not tainted 4.5.0+ #264
  [...]
  [   52.765768] Call Trace:
  [   52.765775]  [<ffffffff813e488d>] dump_stack+0x85/0xc8
  [   52.765784]  [<ffffffff810f2fa5>] lockdep_rcu_suspicious+0xd5/0x110
  [   52.765792]  [<ffffffff816afdc2>] sk_detach_filter+0x82/0x90
  [   52.765801]  [<ffffffffa0883425>] tun_detach_filter+0x35/0x90 [tun]
  [   52.765810]  [<ffffffffa0884ed4>] __tun_chr_ioctl+0x354/0x1130 [tun]
  [   52.765818]  [<ffffffff8136fed0>] ? selinux_file_ioctl+0x130/0x210
  [   52.765827]  [<ffffffffa0885ce3>] tun_chr_ioctl+0x13/0x20 [tun]
  [   52.765834]  [<ffffffff81260ea6>] do_vfs_ioctl+0x96/0x690
  [   52.765843]  [<ffffffff81364af3>] ? security_file_ioctl+0x43/0x60
  [   52.765850]  [<ffffffff81261519>] SyS_ioctl+0x79/0x90
  [   52.765858]  [<ffffffff81003ba2>] do_syscall_64+0x62/0x140
  [   52.765866]  [<ffffffff817d563f>] entry_SYSCALL64_slow_path+0x25/0x25

Same can be triggered with PROVE_RCU (+ PROVE_RCU_REPEATEDLY) enabled
from tun_attach_filter() when user space calls ioctl(tun_fd, TUN{ATTACH,
DETACH}FILTER, ...) for adding/removing a BPF filter on tap devices.

Since the fix in f91ff5b9 ("net: sk_{detach|attach}_filter() rcu
fixes") sk_attach_filter()/sk_detach_filter() now dereferences the
filter with rcu_dereference_protected(), checking whether socket lock
is held in control path.

Since its introduction in 99405162 ("tun: socket filter support"),
tap filters are managed under RTNL lock from __tun_chr_ioctl(). Thus the
sock_owned_by_user(sk) doesn't apply in this specific case and therefore
triggers the false positive.

Extend the BPF API with __sk_attach_filter()/__sk_detach_filter() pair
that is used by tap filters and pass in lockdep_rtnl_is_held() for the
rcu_dereference_protected() checks instead.
Reported-by: NSasha Levin <sasha.levin@oracle.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

ndo_set_rx_headroom controls the align value used by tun devices to
allocate skbs on frame reception.
When the xmit device adds a large encapsulation, this avoids an skb
head reallocation on forwarding.

The measured improvement when forwarding towards a vxlan dev with
frame size below the egress device MTU is as follow:

vxlan over ipv6, bridged: +6%
vxlan over ipv6, ovs: +7%

In case of ipv4 tunnels there is no improvement, since the tun
device default alignment provides enough headroom to avoid the skb
head reallocation.
Signed-off-by: NPaolo Abeni <pabeni@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If a tun interface is turned down, we should not allow packet injection
into the kernel.

Kernel does not send packets to the tun already.

TUNATTACHFILTER can not be used as only tun_net_xmit() is taking care
of it.
Reported-by: NCurt Wohlgemuth <curtw@google.com>
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch is a cleanup to make following patch easier to
review.

Goal is to move SOCK_ASYNC_NOSPACE and SOCK_ASYNC_WAITDATA
from (struct socket)->flags to a (struct socket_wq)->flags
to benefit from RCU protection in sock_wake_async()

To ease backports, we rename both constants.

Two new helpers, sk_set_bit(int nr, struct sock *sk)
and sk_clear_bit(int net, struct sock *sk) are added so that
following patch can change their implementation.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

timewait or request sockets are small and do not contain sk->sk_tsflags

Without this fix, we might read garbage, and crash later in

__skb_complete_tx_timestamp()
 -> sock_queue_err_skb()

(These pseudo sockets do not have an error queue either)

Fixes: ca6fb065 ("tcp: attach SYNACK messages to request sockets instead of listener")
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Acked-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Tap devices don't need to segment multiple tagged packets.
Signed-off-by: NToshiaki Makita <makita.toshiaki@lab.ntt.co.jp>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The VNET_LE flag was introduced to fix accesses to virtio 1.0 headers
that are always little-endian. It can also be used to handle the special
case of a legacy little-endian device implemented by a big-endian host.

Let's add a flag and ioctls for big-endian devices as well. If both flags
are set, little-endian wins.

Since this is isn't a common usecase, the feature is controlled by a kernel
config option (not set by default).

Both macvtap and tun are covered by this patch since they share the same
API with userland.
Signed-off-by: NGreg Kurz <gkurz@linux.vnet.ibm.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>

The current memory accessors logic is:
- little endian if little_endian
- native endian (i.e. no byteswap) if !little_endian

If we want to fully support cross-endian vhost, we also need to be
able to convert to big endian.

Instead of changing the little_endian argument to some 3-value enum, this
patch changes the logic to:
- little endian if little_endian
- big endian if !little_endian

The native endian case is handled by all users with a trivial helper. This
patch doesn't change any functionality, nor it does add overhead.
Signed-off-by: NGreg Kurz <gkurz@linux.vnet.ibm.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>

Signed-off-by: NGreg Kurz <gkurz@linux.vnet.ibm.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Acked-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>

In preparation for changing how struct net is refcounted
on kernel sockets pass the knowledge that we are creating
a kernel socket from sock_create_kern through to sk_alloc.
Signed-off-by: N"Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There is no need for tun to do the weird network namespace refcounting.
The existing network namespace refcounting in tfile has almost exactly
the same lifetime.  So rewrite the code to use the struct sock network
namespace refcounting and remove the unnecessary hand rolled network
namespace refcounting and the unncesary tfile->net.

This change allows the tun code to directly call sock_put bypassing
sock_release and making SOCK_EXTERNALLY_ALLOCATED unnecessary.

Remove the now unncessary tun_release so that if anything tries to use
the sock_release code path the kernel will oops, and let us know about
the bug.

The macvtap code already uses it's internal socket this way.
Signed-off-by: N"Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

All places outside of core VFS that checked ->read and ->write for being NULL or
called the methods directly are gone now, so NULL {read,write} with non-NULL
{read,write}_iter will do the right thing in all cases.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

After TIPC doesn't depend on iocb argument in its internal
implementations of sendmsg() and recvmsg() hooks defined in proto
structure, no any user is using iocb argument in them at all now.
Then we can drop the redundant iocb argument completely from kinds of
implementations of both sendmsg() and recvmsg() in the entire
networking stack.

Cc: Christoph Hellwig <hch@lst.de>
Suggested-by: NAl Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: NYing Xue <ying.xue@windriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Receive Flow Steering is a nice solution but suffers from
hash collisions when a mix of connected and unconnected traffic
is received on the host, when flow hash table is populated.

Also, clearing flow in inet_release() makes RFS not very good
for short lived flows, as many packets can follow close().
(FIN , ACK packets, ...)

This patch extends the information stored into global hash table
to not only include cpu number, but upper part of the hash value.

I use a 32bit value, and dynamically split it in two parts.

For host with less than 64 possible cpus, this gives 6 bits for the
cpu number, and 26 (32-6) bits for the upper part of the hash.

Since hash bucket selection use low order bits of the hash, we have
a full hash match, if /proc/sys/net/core/rps_sock_flow_entries is big
enough.

If the hash found in flow table does not match, we fallback to RPS (if
it is enabled for the rxqueue).

This means that a packet for an non connected flow can avoid the
IPI through a unrelated/victim CPU.

This also means we no longer have to clear the table at socket
close time, and this helps short lived flows performance.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Acked-by: NTom Herbert <therbert@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Instead of manual calls of device_create_file() and
device_remove_files(), assign the static attribute groups to netdev
groups array.  This simplifies the code and avoids the possible
races.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This reverts commit 3d0ad094.

Now that GSO functionality can correctly track if the fragment
id has been selected and select a fragment id if necessary,
we can re-enable UFO on tap/macvap and virtio devices.
Signed-off-by: NVladislav Yasevich <vyasevic@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This reverts commit 5188cd44.

Now that GSO layer can track if fragment id has been selected
and can allocate one if necessary, we don't need to do this in
tap and macvtap.  This reverts most of the code and only keeps
the new ipv6 fragment id generation function that is still needed.

Fixes: 3d0ad094 (drivers/net: Disable UFO through virtio)
Signed-off-by: NVladislav Yasevich <vyasevic@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The same macros are used for rx as well. So rename it.
Signed-off-by: NJiri Pirko <jiri@resnulli.us>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Networking under kvm works best if we allocate a per-vCPU RX and TX
queue in a virtual NIC. This requires a per-vCPU queue on the host side.

It is now safe to increase the maximum number of queues.
Preceding patch: 'net: allow large number of rx queues'
made sure this won't cause failures due to high order memory
allocations. Increase it to 256: this is the max number of vCPUs
KVM supports.

Size of tun_struct changes from 8512 to 10496 after this patch. This keeps
pages allocated for tun_struct before and after the patch to 3.
Signed-off-by: NPankaj Gupta <pagupta@redhat.com>
Reviewed-by: NDavid Gibson <dgibson@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Instead of -1 with EAGAIN, read on a O_NONBLOCK tun fd will return 0.  This
fixes this by properly returning the error code from __skb_recv_datagram.
Signed-off-by: NAlex Gartrell <agartrell@fb.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Validated that this was actually using the unsigned comparison with gdb.
Signed-off-by: NAlex Gartrell <agartrell@fb.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use TUNSETVNETLE/TUNGETVNETLE instead.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Note that the code _using_ ->msg_iter at that point will be very
unhappy with anything other than unshifted iovec-backed iov_iter.
We still need to convert users to proper primitives.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Pretty straight-forward: convert all fields to/from
virtio endian-ness.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Reviewed-by: NJason Wang <jasowang@redhat.com>

It's just as easy to use IFF_ flags directly,
there's no point in adding our own defines.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

TUN_ flags are internal and never exposed
to userspace. Any application using it is almost
certainly buggy.

Move them out to tun.c.

Note: we remove these completely in follow-up patches,
this code movement is split out for ease of review.
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>

To be more friendly with drop monitor, we should only call kfree_skb() when
the packets were dropped and use consume_skb() in other cases.

Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NJason Wang <jasowang@redhat.com>
Acked-by: NMichael S. Tsirkin <mst@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When we write the GSO meta-data in tun_get_user we end up advancing
the IO vector twice, thus exhausting the user buffer before we can
finish writing the packet.

Fixes: f5ff53b4 ("{macvtap,tun}_get_user(): switch to iov_iter")
Reported-by: NMarcelo Ricardo Leitner <mleitner@redhat.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Acked-by: NJason Wang <jasowang@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

allows to switch macvtap and tun from ->aio_write() to ->write_iter()
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>