Add an userdata set extension and allow the user to attach arbitrary
data to set elements. This is intended to hold TLV encoded data like
comments or DNS annotations that have no meaning to the kernel.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

Add a new "dynset" expression for dynamic set updates.

A new set op ->update() is added which, for non existant elements,
invokes an initialization callback and inserts the new element.
For both new or existing elements the extenstion pointer is returned
to the caller to optionally perform timer updates or other actions.

Element removal is not supported so far, however that seems to be a
rather exotic need and can be added later on.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

Currently a set binding is assumed to be related to a lookup and, in
case of maps, a data load.

In order to use bindings for set updates, the loop detection checks
must be restricted to map operations only. Add a flags member to the
binding struct to hold the set "action" flags such as NFT_SET_MAP,
and perform loop detection based on these.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

Use atomic operations for the element count to avoid races with async
updates.

To properly handle the transactional semantics during netlink updates,
deleted but not yet committed elements are accounted for seperately and
are treated as being already removed. This means for the duration of
a netlink transaction, the limit might be exceeded by the amount of
elements deleted. Set implementations must be prepared to handle this.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

The NFT_SET_TIMEOUT flag is ignore in nft_select_set_ops, which may
lead to selection of a set implementation that doesn't actually
support timeouts.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

nf_bridge_info->mask is used for several things, for example to
remember if skb->pkt_type was set to OTHER_HOST.

For a bridge, OTHER_HOST is expected case. For ip forward its a non-starter
though -- routing expects PACKET_HOST.

Bridge netfilter thus changes OTHER_HOST to PACKET_HOST before hook
invocation and then un-does it after hook traversal.

This information is irrelevant outside of br_netfilter.

After this change, ->mask now only contains flags that need to be
known outside of br_netfilter in fast-path.

Future patch changes mask into a 2bit state field in sk_buff, so that
we can remove skb->nf_bridge pointer for good and consider all remaining
places that access nf_bridge info content a not-so fastpath.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

->mask is a bit info field that mixes various use cases.

In particular, we have flags that are mutually exlusive, and flags that
are only used within br_netfilter while others need to be exposed to
other parts of the kernel.

Remove BRNF_8021Q/PPPoE flags.  They're mutually exclusive and only
needed within br_netfilter context.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

Don't access skb->nf_bridge directly, this pointer will be removed soon.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

Avoid skb->nf_bridge accesses where possible.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

right now we store this in the nf_bridge_info struct, accessible
via skb->nf_bridge.  This patch prepares removal of this pointer from skb:

Instead of using skb->nf_bridge->x, we use helpers to obtain the in/out
device (or ifindexes).

Followup patches to netfilter will then allow nf_bridge_info to be
obtained by a call into the br_netfilter core, rather than keeping a
pointer to it in sk_buff.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

br_netfilter maintains an extra state, nf_bridge_info, which is attached
to skb via skb->nf_bridge pointer.

Amongst other things we use skb->nf_bridge->data to store the original
mac header for every processed skb.

This is required for ip refragmentation when using conntrack
on top of bridge, because ip_fragment doesn't copy it from original skb.

However there is no need anymore to do this unconditionally.

Move this to the one place where its needed -- when br_netfilter calls
ip_fragment().

Also switch to percpu storage for this so we can handle fragmenting
without accessing nf_bridge meta data.

Only user left is neigh resolution when DNAT is detected, to hold
the original source mac address (neigh resolution builds new mac header
using bridge mac), so rename ->data and reduce its size to whats needed.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

Currently in xt_socket, we take advantage of early demuxed sockets
since commit 00028aa3 ("netfilter: xt_socket: use IP early demux")
in order to avoid a second socket lookup in the fast path, but we
only make partial use of this:

We still unnecessarily parse headers, extract proto, {s,d}addr and
{s,d}ports from the skb data, accessing possible conntrack information,
etc even though we were not even calling into the socket lookup via
xt_socket_get_sock_{v4,v6}() due to skb->sk hit, meaning those cycles
can be spared.

After this patch, we only proceed the slower, manual lookup path
when we have a skb->sk miss, thus time to match verdict for early
demuxed sockets will improve further, which might be i.e. interesting
for use cases such as mentioned in 681f130f ("netfilter: xt_socket:
add XT_SOCKET_NOWILDCARD flag").
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

We also need to save/store in forward, else br_parse_ip_options call
will zero frag_max_size as well.

Fixes: 93fdd47e ('bridge: Save frag_max_size between PRE_ROUTING and POST_ROUTING')
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

We have to stop iterating on the rule expressions if the cgroup
mismatches. Moreover, make sure a non-full socket from the input path
leads us to a crash.

Fixes: ce674173 ("netfilter: nft_meta: add cgroup support")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

While originally only being intended for outgoing traffic, commit
a00e7634 ("netfilter: x_tables: allow to use cgroup match for
LOCAL_IN nf hooks") enabled xt_cgroups for the NF_INET_LOCAL_IN hook
as well, in order to allow for nfacct accounting.

Besides being currently limited to early demuxes only, commit
a00e7634 forgot to add a check if we deal with full sockets,
i.e. in this case not with time wait sockets. TCP time wait sockets
do not have the same memory layout as full sockets, a lower memory
footprint and consequently also don't have a sk_classid member;
probing for sk_classid member there could potentially lead to a
crash.

Fixes: a00e7634 ("netfilter: x_tables: allow to use cgroup match for LOCAL_IN nf hooks")
Cc: Alexey Perevalov <a.perevalov@samsung.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

Add support for element timeouts to nft_hash. The lookup and walking
functions are changed to ignore timed out elements, a periodic garbage
collection task cleans out expired entries.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

GC is expected to happen asynchrously to the netlink interface. In the
netlink path, both insertion and removal of elements consist of two
steps, insertion followed by activation or deactivation followed by
removal, during which the element must not be freed by GC.

The synchronization helpers use an unused bit in the genmask field to
atomically mark an element as "busy", meaning it is either currently
being handled through the netlink API or by GC.

Elements being processed by GC will never survive, netlink will simply
ignore them. Elements being currently processed through netlink will be
skipped by GC and reprocessed during the next run.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

Add helpers for GC batch destruction: since element destruction needs
a RCU grace period for all set implementations, add some helper functions
for asynchronous batch destruction. Elements are collected in a batch
structure, which is asynchronously released using RCU once its full.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

Add API support for set element timeouts. Elements can have a individual
timeout value specified, overriding the sets' default.

Two new extension types are used for timeouts - the timeout value and
the expiration time. The timeout value only exists if it differs from
the default value.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

Add set timeout support to the netlink API. Sets with timeout support
enabled can have a default timeout value and garbage collection interval
specified.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

Richard Cochran says:

====================
ptp: remove open coded ns_to_timespec64 and reverse

This patch series is a follow up to the recent timespec64 work for the
PTP Hardware Clock drivers.  Arnd noticed that drivers are using open
coded implementations of ns_to_timespec64 and timespec64_to_ns.  This
series replaces the open coded logic with the helper functions.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch changes the driver to use ns_to_timespec64() and
timespec64_to_ns() instead of open coding the same logic.

Compile tested only.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch changes the driver to use ns_to_timespec64() and
timespec64_to_ns() instead of open coding the same logic.

Compile tested only.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch changes the driver to use ns_to_timespec64() instead of
open coding the same logic.

Compile tested only.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch changes the driver to use ns_to_timespec64() and
timespec64_to_ns() instead of open coding the same logic.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch changes the driver to use ns_to_timespec64() and
timespec64_to_ns() instead of open coding the same logic.

Compile tested only.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch changes the driver to use ns_to_timespec64() instead of
open coding the same logic.

Compile tested only.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch changes the driver to use ns_to_timespec64() instead of
open coding the same logic.

Compile tested only.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch changes the driver to use ns_to_timespec64() and
timespec64_to_ns() instead of open coding the same logic.

Compile tested only.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch changes the driver to use ns_to_timespec64() instead of
open coding the same logic.

Compile tested only.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Acked-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch changes the driver to use ns_to_timespec64() and
timespec64_to_ns() instead of open coding the same logic.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch changes the driver to use ns_to_timespec64() and
timespec64_to_ns() instead of open coding the same logic.

Compile tested only.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

When mii-mon discovers that the link is up, it will call
bond_3ad_handle_link_change() but we forget to add the LACP_ENABLED
flag when we discover the speed and duplex for the slave link are
normal.

Change-Id: Ie8b268ecfeea0f99bf9fdcd72706c0653f9d9e49
Signed-off-by: Mahesh Bandewar <maheshb@google.com>
Signed-off-by: Andy Gospodarek <gospo@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

AD_PORT_ACTOR_CHURN and AD_PORT_PARTNER_CHURN are already present and
essentially BOND_MONITOR_CHURNED is a combination of these two definitions.
Signed-off-by: Mahesh Bandewar <maheshb@google.com>
Signed-off-by: Andy Gospodarek <gospo@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Nicolas Ferre says:

====================
net/macb: fixes after big driver update

The recent modifications to the macb driver lead to issues with the probe
function code flow. Here are some attempt to fix them.
This time, some more issues are fixed related to the clock as reported by Boris
Brezillon.

The series is written on top of net-next.

Changes in v2:
- address Cyrille comment about exit condition of queue configuration loop
- add fixes for probe sequence related to clocks
- add ethtool register dump
- fix peripheral version test
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

As we need to check peripheral version from the hardware during probe, I
introduce a little helper to unify these tests. It would prevent to
de-synchronize the test like previously observed.
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

We currently need two checks of the peripheral version in MACB_MID register.
One of them got out of sync after modification by 8a013a9c (net: macb:
Include multi queue support for xilinx ZynqMP ethernet version).
Fix this in macb_configure_caps() so that xilinx ZynqMP will be considered
as a GEM flavor.

Fixes: 8a013a9c ("net: macb: Include multi queue support for xilinx ZynqMP
ethernet version")
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Cc: Michal Simek <michal.simek@xilinx.com>
Cc: Punnaiah Choudary Kalluri <punnaia@xilinx.com>
Cc: <stable@vger.kernel.org> #4.0 (if it doesn't make it for -final)
Signed-off-by: David S. Miller <davem@davemloft.net>

User i/o register EMAC_USRIO or GMAC_UR can be found on both macb and gem
flavors of the peripheral. By using the proper accessor, we can add it to the
register dump feature of ethtool.
Increment the version of this API so it can be noticed from user space.
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

As accessing the peripheral registers need the clocks to be set, we have to
enable them as soon as possible. Their configuration depend on the type of
device used and determined by the DT compatible string. That lead to add
another initialization function in the DT configuration structure.
As the device private structure length depend on an information read in the
registers, we have to store the clock pointers in temporary variables before
feeding the structure fields.
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

As a non-native English speaker, I would correct "capacities" of the macb
peripheral to "capabilities": correct me if I'm wrong!
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>