Bloom filter is shared within multiple regions. For updates, it needs to
be guarded by a separate mutex. Do that in order to not rely on RTNL
mutex.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In order to remove dependency on RTNL, introduce a mutex
to guard vregion structure, list of chunks and list of entries in
chunks.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Refactor existing _vchunk_assoc/_vchunk_deassoc() functions into
_vregion_get()/_vregion_put() to make the code simpler and prepared for
vregion locking.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In order to remove RTNL lock dependency, it is needed to protect
the regions list in a group. Introduce a mutex to do the job.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Make the existing group structure to contain fields needed for HW region
list manipulations. Move the rest of the fields into new vgroup struct.
This makes layering cleaner as the vgroup struct is on higher level than
low-level group struct. Also, this makes it possible to introduce
fine-grained locking.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Never used, remove it.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Tristram Ha says:

====================
net: dsa: microchip: add MIB counters support

This series of patches is to modify the KSZ9477 DSA driver to read MIB
counters periodically to avoid overflow.

The MIB counters should be read only when there is link.  Otherwise it is
a waste of time as hardware never increases the counters.

Functions are added to check the port link status so that MIB counters
read call is used efficiently.

v4
- Use readx_poll_timeout
- Fix using mutex in a timer callback function problem
- use dp->slave directly instead of checking whether it is valid
- Add port_cleanup function in a separate patch
- Add a mutex so that changing device variables is safe

v3
- Use netif_carrier_ok instead of checking the phy device pointer

v2
- Create macro similar to readx_poll_timeout to use with switch
- Create ksz_port_cleanup function so that variables like on_ports and
  live_ports can be updated inside it

v1
- Use readx_poll_timeout
- Do not clear MIB counters when port is enabled
- Do not advertise 1000 half-duplex mode when port is enabled
- Do not use freeze function as MIB counters may miss counts
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Add port_cleanup function to reset some device variables when the port is
disabled.  Add a mutex to make sure changing those variables is
thread-safe.
Signed-off-by: Tristram Ha <Tristram.Ha@microchip.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Remove unnecessary header include.
Signed-off-by: Tristram Ha <Tristram.Ha@microchip.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>

Get port link status to know whether to read MIB counters when the link
is going down.
Signed-off-by: Tristram Ha <Tristram.Ha@microchip.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>

Add background MIB counter reading support.

Port MIB counters should only be read when there is link.  Otherwise it is
a waste of time as hardware never increases those counters.  There are
exceptions as some switches keep track of dropped counts no matter what.
Signed-off-by: Tristram Ha <Tristram.Ha@microchip.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>

Prepare PHY for proper advertisement as sometimes the PHY in the switch
has its own problems even though it may share the PHY id from regular PHY
but the fixes in the PHY driver do not apply.
Signed-off-by: Tristram Ha <Tristram.Ha@microchip.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>

Maxime Chevallier says:

====================
net: phy: marvell10g: Add 2.5GBaseT support

This series adds the missing bits necessary to fully support 2.5GBaseT
in the Marvell Alaska PHYs.

The main points for that support are :

 - Making use of the .get_features call, recently introduced by Heiner
   and Andrew, that allows having a fully populated list of supported
   modes, including 2500BaseT.

 - Configuring the MII to 2500BaseX when establishing a link at 2.5G

 - Adding a small quirk to take into account the fact that some PHYs in
   the family won't report the correct supported abilities

The rest of the series consists of small cosmetic improvements such as
using the correct helper to set a linkmode bit and adding macros for the
PHY ids.

We also add support for the 88E2110 PHY, which doesn't require the
quirk, and support for 2500BaseT in the PPv2 driver, in order to have a
fully working setup on the MacchiatoBin board.

Changes since V1 : Fixed formatting issue in patch 01, rebased.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch adds support for the 88x2110 PHY, which is similar to the
already supported 88x3310 PHY without the SFP interface.

It supports 10/100/1000BASET along with 2.5GBASET, 5GBASET and 10GBASET,
with the same interface modes that are used by the 3310.

This PHY don't have the same issue as the 88x3310 regarding 2.5/5G
abilities, and correctly follows the 802.3bz standard to list the
supported abilities.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Suggested-by: Antoine Tenart <antoine.tenart@bootlin.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

The PPv2 controller is able to support 2.5G speeds, allowing to use
2.5GBASET in conjunction with PHYs that use 2500BASEX as their MII
interface when using this mode.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

As per 802.3bz, if bit 14 of (1.11) "PMA Extended Abilities" indicates
whether or not we should read register (1.21) "2.52/5G PMA Extended
Abilities", which contains information on the support of 2.5GBASET and
5GBASET.

After testing on several variants of PHYS of this family, it appears
that bit 14 in (1.11) isn't always set when it should be.

PHYs 88X3310 (on MacchiatoBin) and 88E2010 do support 2.5G and 5GBASET,
but don't have 1.11.14 set. Their register 1.21 is filled with the
correct values, indicating 2.5G and 5G support.

PHYs 88E2110 do have their 1.11.14 bit set, as it should.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

The PHY ID corresponding to the 88X3310 is also used for other PHYs in
the same family, such as the 88E2010. Use a #define for the PHY id, that
ignores the last nibble.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

The Marvell Alaska family of PHYs supports 2.5GBaseT and 5GBaseT modes,
as defined in the 802.3bz specification.

Upon establishing a 2.5GBASET link, the PHY will reconfigure it's MII
interface to 2500BASEX.

At 5G, the PHY will reconfigure it's interface to 5GBASE-R, but this
mode isn't supported by any MAC for now.

This was tested with :
 - The 88X3310, which is on the MacchiatoBin
 - The 88E2010, an Alaska PHY that has no fiber interfaces, and is
   limited to 5G maximum speed.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Cosmetic patch making use of helpers dedicated to linkmodes handling.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

The Alaska family of 10G PHYs has more abilities than the ones listed in
PHY_10GBIT_FULL_FEATURES, the exact list depending on the model.

Make use of the newly introduced .get_features call to build this list,
using genphy_c45_pma_read_abilities to build the list of supported
linkmodes, and adding autoneg ability based on what's reported by the AN
MMD.

.config_init is still used to validate the interface_mode.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

The current code reports a link as up if all devices (except a few
blacklisted ones) report the link as up. This breaks Aquantia AQCS109
for lower speeds because on this PHY the PCS link status reflects a
10G link only. For Marvell there's a similar issue, therefore PHYXS
device isn't checked.

There may be more PHYs where depending on the mode the link status
of only selected devices is relevant.

For now it seems to be sufficient to check the link status of the
PMAPMD device only. Leave the loop in the code to be prepared in
case we have to add functionality to check more than one device,
depending on the mode.

Successfully tested on a board with an AQCS109.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Florian Fainelli says:

====================
net: switchdev.h inclusion removal

This targets a few drivers that no longer to have net/switchdev.h
included.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

This is no longer necessary after eca59f69 ("net: Remove support for bridge bypass ndos from stacked devices")
Suggested-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Andy Gospodarek <andy@greyhouse.net>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This is no longer necessary after a5084bb7 ("nfp: Implement
ndo_get_port_parent_id()")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

eth_change_mtu() is not needed any more, the networking subsystem will
call it automatically when this callback is not implemented.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

The way to define __ETHTOOL_LINK_MODE_MASK_NBITS seems to be overly
complicated, go with a standard approach instead.
Whilst we're at it, move the comment to the right place.

v2:
- rebased
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Andy Roulin says:

====================
net: protodown support for macvlan and vxlan

This patch series adds dev_change_proto_down_generic, a generic
implementation of ndo_change_proto_down, which sets the netdev carrier
state according to the new proto_down value.

This handler adds the ability to set protodown on macvlan and vxlan
interfaces in a generic way for use by control protocols like VRRPD.

Patch (1) introduces the handler in net/code/dev.c. Patch (2) and (3) add
support for change_proto_down in macvlan and vxlan drivers, respectively,
using the new function.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Add ndo_change_proto_down support through dev_change_proto_down_generic
for use by control protocols like VRRPD.
Signed-off-by: Andy Roulin <aroulin@cumulusnetworks.com>
Acked-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add ndo_change_proto_down support through dev_change_proto_down_generic
for use by control protocols like VRRPD.
Signed-off-by: Andy Roulin <aroulin@cumulusnetworks.com>
Acked-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Introduce dev_change_proto_down_generic, a generic ndo_change_proto_down
implementation, which sets the netdev carrier state according to proto_down.

This adds the ability to set protodown on vxlan and macvlan devices in a
generic way for use by control protocols like VRRPD.
Signed-off-by: Andy Roulin <aroulin@cumulusnetworks.com>
Acked-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Vlad Buslov says:

====================
Add tests for unlocked flower classifier implementation

Implement tests for tdc testsuite to verify concurrent rules update with
rtnl-unlocked flower classifier implementation. The goal of these tests
is to verify general flower classifier correctness by updating filters
on same classifier instance in parallel and to verify its atomicity by
concurrently updating filters in same handle range. All three filter
update operations (add, replace, delete) are tested.

Existing script tdc_batch.py is re-used for batch file generation. It is
extended with several optional CLI arguments that are needed for
concurrency tests. Thin wrapper tdc_multibatch.py is implemented on top
of tdc_batch.py to simplify its usage when generating multiple batch
files for several test configurations.

Parallelism in tests is implemented by running multiple instances of tc
in batch mode with xargs tool. Xargs is chosen for its ease of use and
because it is available by default on most modern Linux distributions.
====================
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Implement test that runs 5 instances of tc replace filter in parallel with
5 instances of tc del filter from same tp instance. Each instance uses its
own filter handle and key range.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Implement test that runs 5 instances of tc add filter in parallel with 5
instances of tc del filter from same tp instance. Each instance uses its
own filter handle and key range.

Extend tdc_multibatch.py with additional options required to implement the
test: common prefix for all generated batch files, first value of filter
handle range, MAC address prefix modifier. These are necessary to allow
creating batch files with unique keys and handle ranges with multiple
invocation of tdc_multibatch.py helper script.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Implement test that verifies concurrent deletion of rules by executing 10
tc instances that delete flower filters in same handle range. In this case
only one tc instance succeeds in deleting a filter with particular handle.
To mitigate expected failures of all other instances, run tc with 'force'
option to continue processing batch file in case of errors and expect xargs
to return code '123' that indicates that invocation of command(s) exited
with error in range 1-125.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Implement test that verifies concurrent replacement of rules by executing
10 tc instances that replace flower filters in same handle range.

Extend tdc_multibatch.py script with new optional CLI argument that is used
to generate all batch files with same filter handle range.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Implement test that verifies parallel rules replacement by adding 1 million
flower filters and then replacing them with 10 concurrent tc instances.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Implement test that verifies parallel rules deletion by adding 1 million
flower filters and then deleting them with 10 concurrent tc instances.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Implement test that verifies parallel rules insertion by adding 1 million
flower filters with 10 concurrent tc instances. Put it to standalone
'concurrency' category.

Implement tdc_multibatch.py helper script that is used to generate multiple
batch files for concurrent tc execution. Extend config with new 'BATCH_DIR'
variable to specify temporary output directory that is used to store batch
files generated by tdc_multibatch.py.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Extend tdc_batch.py with several optional CLI arguments that are used for
implementation of concurrency tests in following patches in this set:

- Add optional argument to specify range of filter handles used in batch
  file [fitler_handle, filter_handle+number). This is needed for testing
  filter deletion where it is necessary to know exact handles of configured
  filters.

- Add optional argument to specify filter operation type (possible values
  are ['add', 'del', 'replace']) instead of hardcoded "add" value. This
  allows generation of batches for filter addition, deletion and
  replacement.

- Add optional argument to allow user to change mac address prefix that is
  used for all filters in batch. This is necessary to allow generating
  multiple batches with unique flower classifier keys.
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

qdisc_pkt_len_init expects transport_header to be set for GSO packets.
Patch [1] skips transport_header validation for GSO packets that don't
have network_header set at the moment of calling virtio_net_hdr_to_skb,
and allows them to pass into the stack. After patch [2] no placeholder
value is assigned to transport_header if dissection fails, so this patch
adds a check to the place where the value of transport_header is used.

[1] https://patchwork.ozlabs.org/patch/1044429/
[2] https://patchwork.ozlabs.org/patch/1046122/Signed-off-by: Maxim Mikityanskiy <maximmi@mellanox.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>