Hide away the memmove used by DSA EtherType header taggers to shift the
MAC SA and DA to the left to make room for the header, after they've
called skb_push(). The call to skb_push() is still left explicit in
drivers, to be symmetric with dsa_strip_etype_header, and because not
all callers can be refactored to do it (for example, brcm_tag_xmit_ll
has common code for a pre-Ethernet DSA tag and an EtherType DSA tag).
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

All header taggers open-code a memmove that is fairly not all that
obvious, and we can hide the details behind a helper function, since the
only thing specific to the driver is the length of the header tag.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Parav Pandit says:

====================
devlink: Control auxiliary devices

Currently, for mlx5 multi-function device, a user is not able to control
which functionality to enable/disable. For example, each PCI
PF, VF, SF function by default has netdevice, RDMA and vdpa-net
devices always enabled.

Hence, enable user to control which device functionality to enable/disable.

This is achieved by using existing devlink params [1] to
enable/disable eth, rdma and vdpa net functionality control knob.

For example user interested in only vdpa device function: performs,

$ devlink dev param set pci/0000:06:00.0 name enable_rdma value false \
                   cmode driverinit
$ devlink dev param set pci/0000:06:00.0 name enable_eth value false \
                   cmode driverinit
$ devlink dev param set pci/0000:06:00.0 name enable_vnet value true \
                   cmode driverinit

$ devlink dev reload pci/0000:06:00.0

Reload command honors parameters set, initializes the device that user
has composed using devlink dev params and resources.
Devices before reload:

            mlx5_core.sf.4
         (subfunction device)
                  /\
                 /| \
                / |  \
               /  |   \
 mlx5_core.eth.4  |  mlx5_core.rdma.4
(SF eth aux dev)  |  (SF rdma aux dev)
    |             |        |
    |             |        |
 enp6s0f0s88      |      mlx5_0
 (SF netdev)      |  (SF rdma device)
                  |
         mlx5_core.vnet.4
         (SF vnet aux dev)
                 |
                 |
        auxiliary/mlx5_core.sf.4
        (vdpa net mgmt device)

Above example reconfigures the device with only VDPA functionality.
Devices after reload:

            mlx5_core.sf.4
         (subfunction device)
                  /\
                 /  \
                /    \
               /      \
 mlx5_core.vnet.4     no eth, no rdma aux devices
 (SF vnet aux dev)

Above parameters enable user to compose the device as needed based
on the use case.

Since devlink params are done on the devlink instance, these
knobs are uniformly usable for PCI PF, VF and SF devices.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Enable user to disable VDPA net auxiliary device so that when it is not
required, user can disable it.

For example,

$ devlink dev param set pci/0000:06:00.0 \
              name enable_vnet value false cmode driverinit
$ devlink dev reload pci/0000:06:00.0

At this point devlink instance do not create auxiliary device
mlx5_core.vnet.2 for the VDPA net functionality.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Enable user to disable RDMA auxiliary device so that when it is not
required, user can disable it.

For example,

$ devlink dev param set pci/0000:06:00.0 \
              name enable_rdma value false cmode driverinit
$ devlink dev reload pci/0000:06:00.0

At this point devlink instance do not create auxiliary device
mlx5_core.rdma.2 for the RDMA functionality.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Enable user to disable Ethernet auxiliary device so that when it is not
required, user can disable it.

For example,

$ devlink dev param set pci/0000:06:00.0 \
              name enable_eth value false cmode driverinit
$ devlink dev reload pci/0000:06:00.0

At this point devlink instance do not create mlx5_core.eth.2 auxiliary
device for the Ethernet functionality.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Cleanup routine missed to unpublish the parameters. Add it.

Fixes: e890acd5 ("net/mlx5: Add devlink flow_steering_mode parameter")
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Enable drivers to publish/unpublish individual parameter.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Currently device configuration parameters can be registered as an array.
Due to this a constant array must be registered. A single driver
supporting multiple devices each with different device capabilities end
up registering all parameters even if it doesn't support it.

One possible workaround a driver can do is, it registers multiple single
entry arrays to overcome such limitation.

Better is to provide a API that enables driver to register/unregister a
single parameter. This also further helps in two ways.
(1) to reduce the memory of devlink_param_entry by avoiding in registering
parameters which are not supported by the device.
(2) avoid generating multiple parameter add, delete, publish, unpublish,
init value notifications for such unsupported parameters
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Create and use a helper function for one parameter registration.
Subsequent patch also will reuse this for driver facing routine to
register a single parameter.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add new device generic parameter to enable/disable creation of
VDPA net auxiliary device and associated device functionality
in the devlink instance.

User who prefers to disable such functionality can disable it using below
example.

$ devlink dev param set pci/0000:06:00.0 \
              name enable_vnet value false cmode driverinit
$ devlink dev reload pci/0000:06:00.0

At this point devlink instance do not create auxiliary device for the
VDPA net functionality.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add new device generic parameter to enable/disable creation of
RDMA auxiliary device and associated device functionality
in the devlink instance.

User who prefers to disable such functionality can disable it using below
example.

$ devlink dev param set pci/0000:06:00.0 \
              name enable_rdma value false cmode driverinit
$ devlink dev reload pci/0000:06:00.0

At this point devlink instance do not create auxiliary device for the
RDMA functionality.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add new device generic parameter to enable/disable creation of
Ethernet auxiliary device and associated device functionality
in the devlink instance.

User who prefers to disable such functionality can disable it using below
example.

$ devlink dev param set pci/0000:06:00.0 \
              name enable_eth value false cmode driverinit
$ devlink dev reload pci/0000:06:00.0

At this point devlink instance do not create auxiliary device for the
Ethernet functionality.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Nikolay Aleksandrov says:

====================
net: bridge: vlan: add global mcast options

This is the first follow-up set after the support for per-vlan multicast
contexts which extends global vlan options to support bridge's multicast
config per-vlan, it enables user-space to change and dump the already
existing bridge vlan multicast context options. The global option patches
(01 - 09 and 12-13) follow a similar pattern of changing current mcast
functions to take multicast context instead of a port/bridge directly.
Option equality checks have been added for dumping vlan range compression.
The last 2 patches extend the mcast router dump support so it can be
re-used when dumping vlan config.

patches 01 - 09: add support for various mcast options
patches 10 - 11: prepare for per-vlan querier control
patches 12 - 13: add support for querier control and router control
patches 14 - 15: add support for dumping per-vlan router ports

Next patch-sets:
 - per-port/vlan router option config
 - iproute2 support for all new vlan options
 - selftests
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Embed the standard multicast router port export by br_rports_fill_info()
into a new global vlan attribute BRIDGE_VLANDB_GOPTS_MCAST_ROUTER_PORTS.
In order to have the same format for the global bridge mcast context and
the per-vlan mcast context we need a double-nesting:
 - BRIDGE_VLANDB_GOPTS_MCAST_ROUTER_PORTS
   - MDBA_ROUTER

Currently we don't compare router lists, if any router port exists in
the bridge mcast contexts we consider their option sets as different and
export them separately.

In addition we export the router port vlan id when dumping similar to
the router port notification format.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

When we are dumping the router ports of a vlan mcast context we need to
use the bridge/vlan and port/vlan's multicast contexts to check if
IPv4/IPv6 router port is present and later to dump the vlan id.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support to change and retrieve global vlan multicast router state
which is used for the bridge itself. We just need to pass multicast context
to br_multicast_set_router instead of bridge device and the rest of the
logic remains the same.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support to change and retrieve global vlan multicast querier state.
We just need to pass multicast context to br_multicast_set_querier
instead of bridge device and the rest of the logic remains the same.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

It is a minor optimization and better behaviour to make sure querier and
query sending routines affect only the matching multicast context
depending if vlan snooping is enabled (vlan ctx vs bridge ctx).
It also avoids sending unnecessary extra query packets.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

We need to have the querier state per multicast context in order to have
per-vlan control, so remove the internal option bit and move it to the
multicast context. Also annotate the lockless reads of the new variable.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support to change and retrieve global vlan multicast startup query
interval option.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support to change and retrieve global vlan multicast query response
interval option.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support to change and retrieve global vlan multicast query interval
option.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support to change and retrieve global vlan multicast querier interval
option.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support to change and retrieve global vlan multicast membership
interval option.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support to change and retrieve global vlan multicast last member
interval option.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support to change and retrieve global vlan multicast startup query
count option.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support to change and retrieve global vlan multicast last member
count option.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support to change and retrieve global vlan IGMP/MLD versions.
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Alex Elder says:

====================
net: ipa: use runtime PM reference counting

This series does further rework of the IPA clock code so that we
rely on some of the core runtime power management code (including
its referencing counting) instead.

The first patch makes ipa_clock_get() act like pm_runtime_get_sync().

The second patch makes system suspend occur regardless of the
current reference count value, which is again more like how the
runtime PM core code behaves.

The third patch creates functions to encapsulate all hardware
suspend and resume activity.  The fourth uses those functions as
the ->runtime_suspend and ->runtime_resume power callbacks.  With
that in place, ipa_clock_get() and ipa_clock_put() are changed to
use runtime PM get and put functions when needed.

The fifth patch eliminates an extra clock reference previously used
to control system suspend.  The sixth eliminates the "IPA clock"
reference count and mutex.

The final patch replaces the one call to ipa_clock_get_additional()
with a call to pm_runtime_get_if_active(), making the former
unnecessary.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Now that ipa_clock_get_additional() is a trivial wrapper around
pm_runtime_get_if_active(), just open-code it in its only caller
and delete the function.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

The runtime power management core code maintains a usage count.  This
count mirrors the IPA clock reference count, and there's no need to
maintain both.  So get rid of the IPA clock reference count and just
rely on the runtime PM usage count to determine when the hardware
should be suspended or resumed.

Use pm_runtime_get_if_active() in ipa_clock_get_additional().  We
care whether power is active, regardless of whether it's in use, so
pass true for its ign_usage_count argument.

The IPA clock mutex is just used to make enabling/disabling the
clock and updating the reference count occur atomically.  Without
the reference count, there's no need for the mutex, so get rid of
that too.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Suspending the IPA hardware is now managed by the runtime PM core
code.  The ->runtime_idle callback returns a non-zero value, so it
will never suspend except when forced.  As a result, there's no need
to take an extra "do not suspend" clock reference.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use the runtime power management core to cause hardware suspend and
resume to occur.  Enable it in ipa_clock_init() (without autosuspend),
and disable it in ipa_clock_exit().

Use ipa_runtime_suspend() as the ->runtime_suspend power operation,
and arrange for it to be called by having ipa_clock_get() call
pm_runtime_get_sync() when the first clock reference is taken.
Similarly, use ipa_runtime_resume() as the ->runtime_resume power
operation, and pm_runtime_put() when the last IPA clock reference
is dropped.

Introduce ipa_runtime_idle() as the ->runtime_idle power operation,
and have it return a non-zero value; this way suspend will never
occur except when forced.

Use pm_runtime_force_suspend() and pm_runtime_force_resume() as the
system suspend and resume callbacks, and remove ipa_suspend() and
ipa_resume().

Store a pointer to the device structure passed to ipa_clock_init(),
so it can be used by ipa_clock_exit() to disable runtime power
management.

For now we preserve IPA clock reference counting.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Introduce ipa_runtime_suspend() and ipa_runtime_resume(), which
encapsulate the activities necessary for suspending and resuming
the IPA hardware.  Call these functions from ipa_clock_get() and
ipa_clock_put() when the first reference is taken or last one is
dropped.

When the very first clock reference is taken (for ipa_config()),
setup isn't complete yet, so (as before) only the core clock gets
enabled.

When the last clock reference is dropped (after ipa_deconfig()),
ipa_teardown() will have made the setup_complete flag false, so
there too, the core clock will be stopped without affecting GSI
or the endpoints.

Otherwise these new functions will perform the desired suspend and
resume actions once setup is complete.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Disable the IPA clock rather than dropping a reference to it in the
system suspend callback.  This forces the suspend to occur without
affecting existing references.

Similarly, enable the clock rather than taking a reference in
ipa_resume(), forcing a resume without changing the reference count.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

We currently assume no errors occur when enabling or disabling the
IPA core clock and interconnects.  And although this commit exposes
errors that could occur, we generally assume this won't happen in
practice.

This commit changes ipa_clock_get() and ipa_clock_put() so each
returns a value.  The values returned are meant to mimic what the
runtime power management functions return, so we can set up error
handling here before we make the switch.  Have ipa_clock_get()
increment the reference count even if it returns an error, to match
the behavior of pm_runtime_get().

More details follow.

When taking a reference in ipa_clock_get(), return 0 for the first
reference, 1 for subsequent references, or a negative error code if
an error occurs.  Note that if ipa_clock_get() returns an error, we
must not touch hardware; in some cases such errors now cause entire
blocks of code to be skipped.

When dropping a reference in ipa_clock_put(), we return 0 or an
error code.  The error would come from ipa_clock_disable(), which
now returns what ipa_interconnect_disable() returns (either 0 or a
negative error code).  For now, callers ignore the return value;
if an error occurs, a message will have already been logged, and
little more can actually be done to improve the situation.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pablo Neira Ayuso says:

====================
Netfilter updates for net-next

The following patchset contains Netfilter updates for net-next:

1) Use nfnetlink_unicast() instead of netlink_unicast() in nft_compat.

2) Remove call to nf_ct_l4proto_find() in flowtable offload timeout
   fixup.

3) CLUSTERIP registers ARP hook on demand, from Florian.

4) Use clusterip_net to store pernet warning, also from Florian.

5) Remove struct netns_xt, from Florian Westphal.

6) Enable ebtables hooks in initns on demand, from Florian.

7) Allow to filter conntrack netlink dump per status bits,
   from Florian Westphal.

8) Register x_tables hooks in initns on demand, from Florian.

9) Remove queue_handler from per-netns structure, again from Florian.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Currently there's support for filtering neighbours/links for interfaces
which have a specific master device (using the IFLA_MASTER/NDA_MASTER
attributes).

This patch adds support for filtering interfaces/neighbours dump for
interfaces that *don't* have a master.
Signed-off-by: Lahav Schlesinger <lschlesinger@drivenets.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/20210810090658.2778960-1-lschlesinger@drivenets.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

tc_new_tfilter() can replay a request if it got EAGAIN. The cited commit
didn't account for this when it converted TC action ->init() API
to use flags instead of parameters. This can lead to passing stale flags
down the call chain which results in trying to lock rtnl when it's
already locked, deadlocking the entire system.

Fix by making sure to reset flags on each replay.

============================================
WARNING: possible recursive locking detected
5.14.0-rc3-custom-49011-g3d2bbb4f104d #447 Not tainted
--------------------------------------------
tc/37605 is trying to acquire lock:
ffffffff841df2f0 (rtnl_mutex){+.+.}-{3:3}, at: tc_setup_cb_add+0x14b/0x4d0

but task is already holding lock:
ffffffff841df2f0 (rtnl_mutex){+.+.}-{3:3}, at: tc_new_tfilter+0xb12/0x22e0

other info that might help us debug this:
 Possible unsafe locking scenario:
       CPU0
       ----
  lock(rtnl_mutex);
  lock(rtnl_mutex);

 *** DEADLOCK ***
 May be due to missing lock nesting notation
1 lock held by tc/37605:
 #0: ffffffff841df2f0 (rtnl_mutex){+.+.}-{3:3}, at: tc_new_tfilter+0xb12/0x22e0

stack backtrace:
CPU: 0 PID: 37605 Comm: tc Not tainted 5.14.0-rc3-custom-49011-g3d2bbb4f104d #447
Hardware name: Mellanox Technologies Ltd. MSN2010/SA002610, BIOS 5.6.5 08/24/2017
Call Trace:
 dump_stack_lvl+0x8b/0xb3
 __lock_acquire.cold+0x175/0x3cb
 lock_acquire+0x1a4/0x4f0
 __mutex_lock+0x136/0x10d0
 fl_hw_replace_filter+0x458/0x630 [cls_flower]
 fl_change+0x25f2/0x4a64 [cls_flower]
 tc_new_tfilter+0xa65/0x22e0
 rtnetlink_rcv_msg+0x86c/0xc60
 netlink_rcv_skb+0x14d/0x430
 netlink_unicast+0x539/0x7e0
 netlink_sendmsg+0x84d/0xd80
 ____sys_sendmsg+0x7ff/0x970
 ___sys_sendmsg+0xf8/0x170
 __sys_sendmsg+0xea/0x1b0
 do_syscall_64+0x35/0x80
 entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x7f7b93b6c0a7
Code: 0c 00 f7 d8 64 89 02 48 c7 c0 ff ff ff ff eb b7 0f 1f 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 b8 2e 00 00 00 0f 05 <48>
RSP: 002b:00007ffe365b3818 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f7b93b6c0a7
RDX: 0000000000000000 RSI: 00007ffe365b3880 RDI: 0000000000000003
RBP: 00000000610a75f6 R08: 0000000000000001 R09: 0000000000000000
R10: fffffffffffff3a9 R11: 0000000000000246 R12: 0000000000000001
R13: 0000000000000000 R14: 00007ffe365b7b58 R15: 00000000004822c0

Fixes: 695176bf ("net_sched: refactor TC action init API")
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Vlad Buslov <vladbu@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Tested-by: Ido Schimmel <idosch@nvidia.com>
Link: https://lore.kernel.org/r/20210810034305.63997-1-mbloch@nvidia.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>