The mv88e6390 uses a different bit to select between bank0 and bank1
of the statistics. So implement an ops function for this, and pass the
selector bit to the generic stats read function. Also, the histogram
selection has moved for the mv88e6390, so abstract its selection as
well.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Different families have different sets of statistics. Abstract this
using a stats_get_stats op. The mv88e6390 needs a different
implementation, which will be added later.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Different families have different sets of statistics. Abstract this
using a stats_get_sset_count and stats_get_strings op. Each stat has a
bitmap, and the ops implementer uses a bit map mask to count the
statistics which apply for the family, or return the list of strings.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
v2:
  Rename functions to avoid _ prefix.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The statistics unit on the mv88e6390 needs the histogram mode to be
configured in a different register compared to other devices. Add an
ops to do this.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
v2:
  Rename to mv88e6390_g1_stats_set_histogram
  Move into global1.c
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The MV88E6390 has a control register for what the histogram statistics
actually contain. This means the stat_snapshot method should not set
this information. So implement the 6390 stats_snapshot function without
these bits.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Knowing the family of device belongs to helps with picking the ops
implementation which is appropriate to the device. So add a comment to
each structure of ops.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Taking a stats snapshot differs between same families. Abstract this
into an ops member. At the same time, move the code into global1.[ch],
since the registers are in the global1 range.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

With the devices added to the tables, the probe will recognize the
switch. This however is not sufficient to make it work properly, other
changes are needed because of incompatibilities.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

_mv88e6xxx_stats_wait() did not check the return value from
 mv88e6xxx_g1_read(), so the compiler complained about set but unused
 err.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Reviewed-by: NVivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The switch needs to be taken out of reset before we can read its ID
register on the MDIO bus.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Reviewed-by: NVivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Declare the structure ieee802154_ops as const as it is only passed as an
argument to the function  ieee802154_alloc_hw. This argument is of type
const struct ieee802154_ops *, so ieee80254_ops structures having this
property can be declared as const.
Done using Coccinelle:

@r1 disable optional_qualifier @
identifier i;
position p;
@@
static struct ieee802154_ops i@p = {...};

@ok1@
identifier r1.i;
position p;
expression e1;
@@
ieee802154_alloc_hw(e1,&i@p)

@bad@
position p!={r1.p,ok1.p};
identifier r1.i;
@@
i@p

@depends on !bad disable optional_qualifier@
identifier r1.i;
@@
static
+const
struct ieee802154_ops  i={...};

@depends on !bad disable optional_qualifier@
identifier r1.i;
@@
+const
struct ieee802154_ops  i;

The before and after size details of the affected files are:

   text	   data	    bss	    dec	    hex	filename
   8669	   1176	     16	   9861	   2685	drivers/net/ieee802154/adf7242.o
   8805	   1048	     16	   9869	   268d	drivers/net/ieee802154/adf7242.o

   text	   data	    bss	    dec	    hex	filename
   7211	   2296	     32	   9539	   2543	drivers/net/ieee802154/atusb.o
   7339	   2160	     32	   9531	   253b	drivers/net/ieee802154/atusb.o
Signed-off-by: NBhumika Goyal <bhumirks@gmail.com>
Acked-by: NStefan Schmidt <stefan@osg.samsung.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Pravin B Shelar says:

====================
geneve: Use LWT more effectively.

Following patch series make use of geneve LWT code path for
geneve netdev type of device.
This allows us to simplify geneve module without changing any
functionality.

v2-v3:
Rebase against latest net-next.

v1-v2:
Fix warning reported by kbuild test robot.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Rather than comparing 64-bit tunnel-id, compare tunnel vni
which is 24-bit id. This also save conversion from vni
to tunnel id on each tunnel packet receive.
Signed-off-by: NPravin B Shelar <pshelar@ovn.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Geneve already has check for device socket in route
lookup function. So no need to check it in xmit
function.
Signed-off-by: NPravin B Shelar <pshelar@ovn.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There are minimal difference in building Geneve header
between ipv4 and ipv6 geneve tunnels. Following patch
refactors code to unify it.
Signed-off-by: NPravin B Shelar <pshelar@ovn.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Current geneve implementation has two separate cases to handle.
1. netdev xmit
2. LWT xmit.

In case of netdev, geneve configuration is stored in various
struct geneve_dev members. For example geneve_addr, ttl, tos,
label, flags, dst_cache, etc. For LWT ip_tunnel_info is passed
to the device in ip_tunnel_info.

Following patch uses ip_tunnel_info struct to store almost all
of configuration of a geneve netdevice. This allows us to unify
most of geneve driver code around ip_tunnel_info struct.
This dramatically simplify geneve code, since it does not
need to handle two different configuration cases. Removes
duplicate code, single code path can handle either type
of geneve devices.
Signed-off-by: NPravin B Shelar <pshelar@ovn.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Florian Westphal says:

====================
tcp: make undo_cwnd mandatory for congestion modules

highspeed, illinois, scalable, veno and yeah congestion control algorithms
don't provide a 'cwnd_undo' function.  This makes the stack default to a
'reno undo' which doubles cwnd.  However, the ssthresh implementation of
these algorithms do not halve the slowstart threshold. This causes similar
issue as the one fixed for dctcp in ce6dd233 ("dctcp: avoid bogus
doubling of cwnd after loss").

In light of this it seems better to remove the fallback and make undo_cwnd
mandatory.

First patch fixes those spots where reno undo seems incorrect by providing
.cwnd_undo functions, second patch removes the fallback.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The undo_cwnd fallback in the stack doubles cwnd based on ssthresh,
which un-does reno halving behaviour.

It seems more appropriate to let congctl algorithms pair .ssthresh
and .undo_cwnd properly. Add a 'tcp_reno_undo_cwnd' function and wire it
up for all congestion algorithms that used to rely on the fallback.

Cc: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Signed-off-by: NFlorian Westphal <fw@strlen.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

congestion control algorithms that do not halve cwnd in their .ssthresh
should provide a .cwnd_undo rather than rely on current fallback which
assumes reno halving (and thus doubles the cwnd).

All of these do 'something else' in their .ssthresh implementation, thus
store the cwnd on loss and provide .undo_cwnd to restore it again.

A followup patch will remove the fallback and all algorithms will
need to provide a .cwnd_undo function.
Signed-off-by: NFlorian Westphal <fw@strlen.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Nikolay Aleksandrov says:

====================
bridge: add support for IGMPv3 and MLDv2 querier

This patch-set adds support for IGMPv3 and MLDv2 querier in the bridge.
Two new options which can be toggled via netlink and sysfs are added that
control the version per-bridge:
 multicast_igmp_version - default 2, can be set to 3
 multicast_mld_version - default 1, can be set to 2 (this option is
                         disabled if CONFIG_IPV6=n)

Note that the names do not include "querier", I think that these options
can be re-used later as more IGMPv3 support is added to the bridge so we
can avoid adding more options to switch between v2 and v3 behaviour.

The set uses the already existing br_ip{4,6}_multicast_alloc_query
functions and adds the appropriate header based on the chosen version.

For the initial support I have removed the compatibility implementation
(RFC3376 sec 7.3.1, 7.3.2; RFC3810 sec 8.3.1, 8.3.2), because there are
some details that we need to sort out.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds basic support for MLDv2 queries, the default is MLDv1
as before. A new multicast option - multicast_mld_version, adds the
ability to change it between 1 and 2 via netlink and sysfs.
The MLD option is disabled if CONFIG_IPV6 is disabled.
Signed-off-by: NNikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds basic support for IGMPv3 queries, the default is IGMPv2
as before. A new multicast option - multicast_igmp_version, adds the
ability to change it between 2 and 3 via netlink and sysfs. The option
struct member is in a 4 byte hole in net_bridge.

There also a few minor style adjustments in br_multicast_new_group and
br_multicast_add_group.
Signed-off-by: NNikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The function macvlan_forward_source_one has already checked the flag
IFF_UP, so needn't check it outside in macvlan_forward_source too.
Signed-off-by: NGao Feng <gfree.wind@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

While stressing a 40Gbit mlx4 NIC with busy polling, I found false
sharing in mlx4 driver that can be easily avoided.

This patch brings an additional 7 % performance improvement in UDP_RR
workload.

1) If we received no frame during one mlx4_en_process_rx_cq()
   invocation, no need to call mlx4_cq_set_ci() and/or dirty ring->cons

2) Do not refill rx buffers if we have plenty of them.
   This avoids false sharing and allows some bulk/batch optimizations.
   Page allocator and its locks will thank us.

Finally, mlx4_en_poll_rx_cq() should not return 0 if it determined
cpu handling NIC IRQ should be changed. We should return budget-1
instead, to not fool net_rx_action() and its netdev_budget.

v2: keep AVG_PERF_COUNTER(... polled) even if polled is 0
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Tariq Toukan <tariqt@mellanox.com>
Reviewed-by: NTariq Toukan <tariqt@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

barrier() is a big hammer compared to READ_ONCE(),
and requires comments explaining what is protected.

READ_ONCE() is more precise and compiler should generate
better overall code.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

UDP_SKB_CB(skb)->partial_cov is located at offset 66 in skb,
requesting a cold cache line being read in cpu cache.

We can avoid this cache line miss for UDP sockets,
as partial_cov has a meaning only for UDPLite.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Daniel Borkmann says:

====================
Couple of BPF refcount fixes for mlx5

Various mlx5 bugs on eBPF refcount handling found during review.
Last patch in series adds a __must_check to BPF helpers to make
sure we won't run into it again w/o compiler complaining first.

v2 -> v3:

 - Just reworked patch 2/4 so we don't need bpf_prog_sub().
 - Rebased, rest as is.

v1 -> v2:

 - After discussion with Alexei, we agreed upon rebasing the
   patches against net-next.
 - Since net-next, I've also added the __must_check to enforce
   future users to check for errors.
 - Fixed up commit message #2.
 - Simplify assignment from patch #1 based on Saeed's feedback
   on previous set.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Helpers like bpf_prog_add(), bpf_prog_inc(), bpf_map_inc() can fail
with an error, so make sure the caller properly checks their return
value and not just ignores it, which could worst-case lead to use
after free.
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NAlexei Starovoitov <ast@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

mlx5e_xdp_set() is currently the only place where we drop reference on the
prog sitting in priv->xdp_prog when it's exchanged by a new one. We also
need to make sure that we eventually release that reference, for example,
in case the netdev is dismantled, otherwise we leak the program.

Fixes: 86994156 ("net/mlx5e: XDP fast RX drop bpf programs support")
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There are multiple issues in mlx5e_xdp_set():

1) The batched bpf_prog_add() is currently not checked for errors. When
   doing so, it should be done at an earlier point in time to makes sure
   that we cannot fail anymore at the time we want to set the program for
   each channel. The batched refs short-cut can only be performed when we
   don't need to perform a reset for changing the rq type and the device
   was in opened state. In case the device was not in opened state, then
   the next mlx5e_open_locked() will aquire the refs from the control prog
   via mlx5e_create_rq(), same when we need to perform a reset.

2) When swapping the priv->xdp_prog, then no extra reference count must be
   taken since we got that from call path via dev_change_xdp_fd() already.
   Otherwise, we'd never be able to release the program. Also, bpf_prog_add()
   without checking the return code could fail.

Fixes: 86994156 ("net/mlx5e: XDP fast RX drop bpf programs support")
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In mlx5e_create_rq(), when creating a new queue, we call bpf_prog_add() but
without checking the return value. bpf_prog_add() can fail since 92117d84
("bpf: fix refcnt overflow"), so we really must check it. Take the reference
right when we assign it to the rq from priv->xdp_prog, and just drop the
reference on error path. Destruction in mlx5e_destroy_rq() looks good, though.

Fixes: 86994156 ("net/mlx5e: XDP fast RX drop bpf programs support")
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Andrew Lunn says:

====================
Fixes for the MV88e6xxx interrupt code

The interrupt code was never tested with a board who's probing
resulted in an -EPROBE_DEFFERED. So the clean up paths never got
tested. I now do have -EPROBE_DEFFERED, and things break badly during
cleanup. These are the fixes.

This is fixing code in net-next.

v2:
Fix typo pointed out by David Miller
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Freeing interrupts requires switch register access to mask the
interrupts. Hence we must hold the register mutex.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

It is not possible to use devm_request_threaded_irq() because we have
two stacked interrupt controllers in one device. The lower interrupt
controller cannot be removed until the upper is fully removed. This
happens too late with the devm API, resulting in error messages about
removing a domain while there is still an active interrupt. Swap to
using request_threaded_irq() and manage the release of the interrupt
manually.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

On error, remask the interrupts, release all maps, and remove the
domain. This cannot be done using the mv88e6xxx_g1_irq_free() because
some of these actions are not idempotent.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix the g1 interrupt free code such that is masks any further
interrupts, and then releases the interrupt.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Trying to remove an IRQ domain that was not created results in an
Opps. Add the necessary checks that the irqs were created before
freeing them.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Simple typos, s/g2/g1/
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

1) cast to "int" is unnecessary:
   u8 will be promoted to int before decrementing,
   small positive numbers fit into "int", so their values won't be changed
   during promotion.

   Once everything is int including loop counters, signedness doesn't
   matter: 32-bit operations will stay 32-bit operations.

   But! Someone tried to make this loop smart by making everything of
   the same type apparently in an attempt to optimise it.
   Do the optimization, just differently.
   Do the cast where it matters. :^)

2) frag size is unsigned entity and sum of fragments sizes is also
   unsigned.

Make everything unsigned, leave no MOVSX instruction behind.

	add/remove: 0/0 grow/shrink: 0/3 up/down: 0/-4 (-4)
	function                                     old     new   delta
	skb_cow_data                                 835     834      -1
	ip_do_fragment                              2549    2548      -1
	ip6_fragment                                3130    3128      -2
	Total: Before=154865032, After=154865028, chg -0.00%
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Length of a netlink attribute may be u16 but lengths of basic attributes
are much smaller, so small we can save 16 bytes of .rodata and pocket
change inside .text.

16-bit is worse on x86-64 than 8-bit because of operand size override prefix.

	add/remove: 0/0 grow/shrink: 0/3 up/down: 0/-19 (-19)
	function                                     old     new   delta
	validate_nla                                 418     417      -1
	nla_policy_len                                66      64      -2
	nla_attr_minlen                               32      16     -16
	Total: Before=154865051, After=154865032, chg -0.00%
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>