linux.git/drivers/net/virtio_net.c, branch v5.8

xdp: Rename convert_to_xdp_frame in xdp_convert_buff_to_frame

2020-06-01T22:02:53+00:00

In order to use standard 'xdp' prefix, rename convert_to_xdp_frame
utility routine in xdp_convert_buff_to_frame and replace all the
occurrences

Signed-off-by: Lorenzo Bianconi 
Signed-off-by: Alexei Starovoitov 
Acked-by: Jesper Dangaard Brouer 
Link: https://lore.kernel.org/bpf/6344f739be0d1a08ab2b9607584c4d5478c8c083.1590698295.git.lorenzo@kernel.org

Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net

2020-05-15T20:48:59+00:00

Move the bpf verifier trace check into the new switch statement in
HEAD.

Resolve the overlapping changes in hinic, where bug fixes overlap
the addition of VF support.

Signed-off-by: David S. Miller

virtio_net: Add XDP frame size in two code paths

2020-05-15T04:21:55+00:00

The virtio_net driver is running inside the guest-OS. There are two
XDP receive code-paths in virtio_net, namely receive_small() and
receive_mergeable(). The receive_big() function does not support XDP.

In receive_small() the frame size is available in buflen. The buffer
backing these frames are allocated in add_recvbuf_small() with same
size, except for the headroom, but tailroom have reserved room for
skb_shared_info. The headroom is encoded in ctx pointer as a value.

In receive_mergeable() the frame size is more dynamic. There are two
basic cases: (1) buffer size is based on a exponentially weighted
moving average (see DECLARE_EWMA) of packet length. Or (2) in case
virtnet_get_headroom() have any headroom then buffer size is
PAGE_SIZE. The ctx pointer is this time used for encoding two values;
the buffer len "truesize" and headroom. In case (1) if the rx buffer
size is underestimated, the packet will have been split over more
buffers (num_buf info in virtio_net_hdr_mrg_rxbuf placed in top of
buffer area). If that happens the XDP path does a xdp_linearize_page
operation.

V3: Adjust frame_sz in receive_mergeable() case, spotted by Jason Wang.

The code is really hard to follow, so some hints to reviewers.
The receive_mergeable() case gets frames that were allocated in
add_recvbuf_mergeable() which uses headroom=virtnet_get_headroom(),
and 'buf' ptr is advanced this headroom.  The headroom can only
be 0 or VIRTIO_XDP_HEADROOM, as virtnet_get_headroom is really
simple:

  static unsigned int virtnet_get_headroom(struct virtnet_info *vi)
  {
	return vi->xdp_queue_pairs ? VIRTIO_XDP_HEADROOM : 0;
  }

As frame_sz is an offset size from xdp.data_hard_start, reviewers
should notice how this is calculated in receive_mergeable():

  int offset = buf - page_address(page);
  [...]
  data = page_address(xdp_page) + offset;
  xdp.data_hard_start = data - VIRTIO_XDP_HEADROOM + vi->hdr_len;

The calculated offset will always be VIRTIO_XDP_HEADROOM when
reaching this code.  Thus, xdp.data_hard_start will be page-start
address plus vi->hdr_len.  Given this xdp.frame_sz need to be
reduced with vi->hdr_len size.

IMHO a followup patch should cleanup this code to make it easier
to maintain and understand, but it is outside the scope of this
patchset.

Signed-off-by: Jesper Dangaard Brouer 
Signed-off-by: Alexei Starovoitov 
Acked-by: Michael S. Tsirkin 
Acked-by: Jason Wang 
Link: https://lore.kernel.org/bpf/158945344436.97035.9445115070189151680.stgit@firesoul

virtio_net: fix lockdep warning on 32 bit

2020-05-07T14:57:21+00:00

When we fill up a receive VQ, try_fill_recv currently tries to count
kicks using a 64 bit stats counter. Turns out, on a 32 bit kernel that
uses a seqcount. sequence counts are "lock" constructs where you need to
make sure that writers are serialized.

In turn, this means that we mustn't run two try_fill_recv concurrently.
Which of course we don't. We do run try_fill_recv sometimes from a
softirq napi context, and sometimes from a fully preemptible context,
but the later always runs with napi disabled.

However, when it comes to the seqcount, lockdep is trying to enforce the
rule that the same lock isn't accessed from preemptible and softirq
context - it doesn't know about napi being enabled/disabled. This causes
a false-positive warning:

WARNING: inconsistent lock state
...
inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage.

As a work around, shut down the warning by switching
to u64_stats_update_begin_irqsave - that works by disabling
interrupts on 32 bit only, is a NOP on 64 bit.

Reported-by: Thomas Gleixner 
Suggested-by: Eric Dumazet 
Signed-off-by: Michael S. Tsirkin 
Signed-off-by: David S. Miller

virtio_net: reject unsupported coalescing params

2020-03-05T20:12:35+00:00

Set ethtool_ops->supported_coalesce_params to let
the core reject unsupported coalescing parameters.

This driver correctly rejects all unsupported parameters.
As a side effect of these changes the error code for
unsupported params changes from EINVAL to EOPNOTSUPP.

v2: correctly handle rx-frames (and adjust the commit msg)
v3: adjust commit message for new error code and member name

Signed-off-by: Jakub Kicinski 
Signed-off-by: David S. Miller

net/ethtool: Introduce link_ksettings API for virtual network devices

2020-03-01T05:48:55+00:00

With the ethtool_virtdev_set_link_ksettings function in core/ethtool.c,
ibmveth, netvsc, and virtio now use the core's helper function.

Funtionality changes that pertain to ibmveth driver include:

  1. Changed the initial hardcoded link speed to 1GB.

  2. Added support for allowing a user to change the reported link
  speed via ethtool.

Functionality changes to the netvsc driver include:

  1. When netvsc_get_link_ksettings is called, it will defer to the VF
  device if it exists to pull accelerated networking values, otherwise
  pull default or user-defined values.

  2. Similarly, if netvsc_set_link_ksettings called and a VF device
  exists, the real values of speed and duplex are changed.

Signed-off-by: Cris Forno 
Signed-off-by: David S. Miller

virtio_net: Add XDP meta data support

2020-02-25T21:50:55+00:00

Implement support for transferring XDP meta data into skb for
virtio_net driver; before calling into the program, xdp.data_meta points
to xdp.data, where on program return with pass verdict, we call
into skb_metadata_set().

Tested with the script at
https://github.com/higebu/virtio_net-xdp-metadata-test.

Signed-off-by: Yuya Kusakabe 
Signed-off-by: Daniel Borkmann 
Acked-by: Jason Wang 
Acked-by: Michael S. Tsirkin 
Link: https://lore.kernel.org/bpf/20200225033212.437563-2-yuya.kusakabe@gmail.com

virtio_net: Keep vnet header zeroed if XDP is loaded for small buffer

2020-02-25T21:50:55+00:00

We do not want to care about the vnet header in receive_small() if XDP
is loaded, since we can not know whether or not the packet is modified
by XDP.

Fixes: f6b10209b90d ("virtio-net: switch to use build_skb() for small buffer")
Signed-off-by: Yuya Kusakabe 
Signed-off-by: Daniel Borkmann 
Acked-by: Jason Wang 
Acked-by: Michael S. Tsirkin 
Link: https://lore.kernel.org/bpf/20200225033212.437563-1-yuya.kusakabe@gmail.com

bpf, xdp: virtio_net use access ptr macro for xdp enable check

2020-01-27T10:16:25+00:00

virtio_net currently relies on rcu critical section to access the xdp
program in its xdp_xmit handler. However, the pointer to the xdp program
is only used to do a NULL pointer comparison to determine if xdp is
enabled or not.

Use rcu_access_pointer() instead of rcu_dereference() to reflect this.
Then later when we drop rcu_read critical section virtio_net will not
need in special handling.

Signed-off-by: John Fastabend 
Signed-off-by: Daniel Borkmann 
Acked-by: Jesper Dangaard Brouer 
Link: https://lore.kernel.org/bpf/1580084042-11598-3-git-send-email-john.fastabend@gmail.com

xdp: Use bulking for non-map XDP_REDIRECT and consolidate code paths

2020-01-17T04:03:34+00:00

Since the bulk queue used by XDP_REDIRECT now lives in struct net_device,
we can re-use the bulking for the non-map version of the bpf_redirect()
helper. This is a simple matter of having xdp_do_redirect_slow() queue the
frame on the bulk queue instead of sending it out with __bpf_tx_xdp().

Unfortunately we can't make the bpf_redirect() helper return an error if
the ifindex doesn't exit (as bpf_redirect_map() does), because we don't
have a reference to the network namespace of the ingress device at the time
the helper is called. So we have to leave it as-is and keep the device
lookup in xdp_do_redirect_slow().

Since this leaves less reason to have the non-map redirect code in a
separate function, so we get rid of the xdp_do_redirect_slow() function
entirely. This does lose us the tracepoint disambiguation, but fortunately
the xdp_redirect and xdp_redirect_map tracepoints use the same tracepoint
entry structures. This means both can contain a map index, so we can just
amend the tracepoint definitions so we always emit the xdp_redirect(_err)
tracepoints, but with the map ID only populated if a map is present. This
means we retire the xdp_redirect_map(_err) tracepoints entirely, but keep
the definitions around in case someone is still listening for them.

With this change, the performance of the xdp_redirect sample program goes
from 5Mpps to 8.4Mpps (a 68% increase).

Since the flush functions are no longer map-specific, rename the flush()
functions to drop _map from their names. One of the renamed functions is
the xdp_do_flush_map() callback used in all the xdp-enabled drivers. To
keep from having to update all drivers, use a #define to keep the old name
working, and only update the virtual drivers in this patch.

Signed-off-by: Toke Høiland-Jørgensen 
Signed-off-by: Alexei Starovoitov 
Acked-by: John Fastabend 
Link: https://lore.kernel.org/bpf/157918768505.1458396.17518057312953572912.stgit@toke.dk