linux-2.6-microblaze.git
3 years agonet: stmmac: dwmac-meson8b: fix enabling the timing-adjustment clock
Martin Blumenstingl [Wed, 6 Jan 2021 13:42:48 +0000 (14:42 +0100)]
net: stmmac: dwmac-meson8b: fix enabling the timing-adjustment clock

The timing-adjustment clock only has to be enabled when a) there is a
2ns RX delay configured using device-tree and b) the phy-mode indicates
that the RX delay should be enabled.

Only enable the RX delay if both are true, instead of (by accident) also
enabling it when there's the 2ns RX delay configured but the phy-mode
incicates that the RX delay is not used.

Fixes: 9308c47640d515 ("net: stmmac: dwmac-meson8b: add support for the RX delay configuration")
Reported-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agodt-bindings: net: dwmac-meson: use picoseconds for the RGMII RX delay
Martin Blumenstingl [Wed, 6 Jan 2021 13:42:47 +0000 (14:42 +0100)]
dt-bindings: net: dwmac-meson: use picoseconds for the RGMII RX delay

Amlogic Meson G12A, G12B and SM1 SoCs have a more advanced RGMII RX
delay register which allows picoseconds precision. Deprecate the old
"amlogic,rx-delay-ns" in favour of the generic "rx-internal-delay-ps"
property.

For older SoCs the only known supported values were 0ns and 2ns. The new
SoCs have support for RGMII RX delays between 0ps and 3000ps in 200ps
steps.

Don't carry over the description for the "rx-internal-delay-ps" property
and inherit that from ethernet-controller.yaml instead.

Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge branch 'reduce-coupling-between-dsa-and-broadcom-systemport-driver'
Jakub Kicinski [Thu, 7 Jan 2021 23:42:09 +0000 (15:42 -0800)]
Merge branch 'reduce-coupling-between-dsa-and-broadcom-systemport-driver'

Vladimir Oltean says:

====================
Reduce coupling between DSA and Broadcom SYSTEMPORT driver

Upon a quick inspection, it seems that there is some code in the generic
DSA layer that is somehow specific to the Broadcom SYSTEMPORT driver.
The challenge there is that the hardware integration is very tight between
the switch and the DSA master interface. However this does not mean that
the drivers must also be as integrated as the hardware is. We can avoid
creating a DSA notifier just for the Broadcom SYSTEMPORT, and we can
move some Broadcom-specific queue mapping helpers outside of the common
include/net/dsa.h.
====================

Link: https://lore.kernel.org/r/20210107012403.1521114-1-olteanv@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: dsa: remove the DSA specific notifiers
Vladimir Oltean [Thu, 7 Jan 2021 01:24:03 +0000 (03:24 +0200)]
net: dsa: remove the DSA specific notifiers

This effectively reverts commit 60724d4bae14 ("net: dsa: Add support for
DSA specific notifiers"). The reason is that since commit 2f1e8ea726e9
("net: dsa: link interfaces with the DSA master to get rid of lockdep
warnings"), it appears that there is a generic way to achieve the same
purpose. The only user thus far, the Broadcom SYSTEMPORT driver, was
converted to use the generic notifiers.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: systemport: use standard netdevice notifier to detect DSA presence
Vladimir Oltean [Thu, 7 Jan 2021 01:24:02 +0000 (03:24 +0200)]
net: systemport: use standard netdevice notifier to detect DSA presence

The SYSTEMPORT driver maps each port of the embedded Broadcom DSA switch
port to a certain queue of the master Ethernet controller. For that it
currently uses a dedicated notifier infrastructure which was added in
commit 60724d4bae14 ("net: dsa: Add support for DSA specific notifiers").

However, since commit 2f1e8ea726e9 ("net: dsa: link interfaces with the
DSA master to get rid of lockdep warnings"), DSA is actually an upper of
the Broadcom SYSTEMPORT as far as the netdevice adjacency lists are
concerned. So naturally, the plain NETDEV_CHANGEUPPER net device notifiers
are emitted. It looks like there is enough API exposed by DSA to the
outside world already to make the call_dsa_notifiers API redundant. So
let's convert its only user to plain netdev notifiers.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: dsa: export dsa_slave_dev_check
Vladimir Oltean [Thu, 7 Jan 2021 01:24:01 +0000 (03:24 +0200)]
net: dsa: export dsa_slave_dev_check

Using the NETDEV_CHANGEUPPER notifications, drivers can be aware when
they are enslaved to e.g. a bridge by calling netif_is_bridge_master().

Export this helper from DSA to get the equivalent functionality of
determining whether the upper interface of a CHANGEUPPER notifier is a
DSA switch interface or not.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: dsa: move the Broadcom tag information in a separate header file
Vladimir Oltean [Thu, 7 Jan 2021 01:24:00 +0000 (03:24 +0200)]
net: dsa: move the Broadcom tag information in a separate header file

It is a bit strange to see something as specific as Broadcom SYSTEMPORT
bits in the main DSA include file. Move these away into a separate
header, and have the tagger and the SYSTEMPORT driver include them.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge branch 'offload-software-learnt-bridge-addresses-to-dsa'
Jakub Kicinski [Thu, 7 Jan 2021 23:34:48 +0000 (15:34 -0800)]
Merge branch 'offload-software-learnt-bridge-addresses-to-dsa'

Vladimir Oltean says:

====================
Offload software learnt bridge addresses to DSA

This series tries to make DSA behave a bit more sanely when bridged with
"foreign" (non-DSA) interfaces and source address learning is not
supported on the hardware CPU port (which would make things work more
seamlessly without software intervention). When a station A connected to
a DSA switch port needs to talk to another station B connected to a
non-DSA port through the Linux bridge, DSA must explicitly add a route
for station B towards its CPU port.

Initial RFC was posted here:
https://patchwork.ozlabs.org/project/netdev/cover/20201108131953.2462644-1-olteanv@gmail.com/

v2 was posted here:
https://patchwork.kernel.org/project/netdevbpf/cover/20201213024018.772586-1-vladimir.oltean@nxp.com/

v3 was posted here:
https://patchwork.kernel.org/project/netdevbpf/cover/20201213140710.1198050-1-vladimir.oltean@nxp.com/

This is a resend of the previous v3 with some added Reviewed-by tags.
====================

Link: https://lore.kernel.org/r/20210106095136.224739-1-olteanv@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: dsa: ocelot: request DSA to fix up lack of address learning on CPU port
Vladimir Oltean [Wed, 6 Jan 2021 09:51:36 +0000 (11:51 +0200)]
net: dsa: ocelot: request DSA to fix up lack of address learning on CPU port

Given the following setup:

ip link add br0 type bridge
ip link set eno0 master br0
ip link set swp0 master br0
ip link set swp1 master br0
ip link set swp2 master br0
ip link set swp3 master br0

Currently, packets received on a DSA slave interface (such as swp0)
which should be routed by the software bridge towards a non-switch port
(such as eno0) are also flooded towards the other switch ports (swp1,
swp2, swp3) because the destination is unknown to the hardware switch.

This patch addresses the issue by monitoring the addresses learnt by the
software bridge on eno0, and adding/deleting them as static FDB entries
on the CPU port accordingly.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: dsa: listen for SWITCHDEV_{FDB,DEL}_ADD_TO_DEVICE on foreign bridge neighbors
Vladimir Oltean [Wed, 6 Jan 2021 09:51:35 +0000 (11:51 +0200)]
net: dsa: listen for SWITCHDEV_{FDB,DEL}_ADD_TO_DEVICE on foreign bridge neighbors

Some DSA switches (and not only) cannot learn source MAC addresses from
packets injected from the CPU. They only perform hardware address
learning from inbound traffic.

This can be problematic when we have a bridge spanning some DSA switch
ports and some non-DSA ports (which we'll call "foreign interfaces" from
DSA's perspective).

There are 2 classes of problems created by the lack of learning on
CPU-injected traffic:
- excessive flooding, due to the fact that DSA treats those addresses as
  unknown
- the risk of stale routes, which can lead to temporary packet loss

To illustrate the second class, consider the following situation, which
is common in production equipment (wireless access points, where there
is a WLAN interface and an Ethernet switch, and these form a single
bridging domain).

 AP 1:
 +------------------------------------------------------------------------+
 |                                          br0                           |
 +------------------------------------------------------------------------+
 +------------+ +------------+ +------------+ +------------+ +------------+
 |    swp0    | |    swp1    | |    swp2    | |    swp3    | |    wlan0   |
 +------------+ +------------+ +------------+ +------------+ +------------+
       |                                                       ^        ^
       |                                                       |        |
       |                                                       |        |
       |                                                    Client A  Client B
       |
       |
       |
 +------------+ +------------+ +------------+ +------------+ +------------+
 |    swp0    | |    swp1    | |    swp2    | |    swp3    | |    wlan0   |
 +------------+ +------------+ +------------+ +------------+ +------------+
 +------------------------------------------------------------------------+
 |                                          br0                           |
 +------------------------------------------------------------------------+
 AP 2

- br0 of AP 1 will know that Clients A and B are reachable via wlan0
- the hardware fdb of a DSA switch driver today is not kept in sync with
  the software entries on other bridge ports, so it will not know that
  clients A and B are reachable via the CPU port UNLESS the hardware
  switch itself performs SA learning from traffic injected from the CPU.
  Nonetheless, a substantial number of switches don't.
- the hardware fdb of the DSA switch on AP 2 may autonomously learn that
  Client A and B are reachable through swp0. Therefore, the software br0
  of AP 2 also may or may not learn this. In the example we're
  illustrating, some Ethernet traffic has been going on, and br0 from AP
  2 has indeed learnt that it can reach Client B through swp0.

One of the wireless clients, say Client B, disconnects from AP 1 and
roams to AP 2. The topology now looks like this:

 AP 1:
 +------------------------------------------------------------------------+
 |                                          br0                           |
 +------------------------------------------------------------------------+
 +------------+ +------------+ +------------+ +------------+ +------------+
 |    swp0    | |    swp1    | |    swp2    | |    swp3    | |    wlan0   |
 +------------+ +------------+ +------------+ +------------+ +------------+
       |                                                            ^
       |                                                            |
       |                                                         Client A
       |
       |
       |                                                         Client B
       |                                                            |
       |                                                            v
 +------------+ +------------+ +------------+ +------------+ +------------+
 |    swp0    | |    swp1    | |    swp2    | |    swp3    | |    wlan0   |
 +------------+ +------------+ +------------+ +------------+ +------------+
 +------------------------------------------------------------------------+
 |                                          br0                           |
 +------------------------------------------------------------------------+
 AP 2

- br0 of AP 1 still knows that Client A is reachable via wlan0 (no change)
- br0 of AP 1 will (possibly) know that Client B has left wlan0. There
  are cases where it might never find out though. Either way, DSA today
  does not process that notification in any way.
- the hardware FDB of the DSA switch on AP 1 may learn autonomously that
  Client B can be reached via swp0, if it receives any packet with
  Client 1's source MAC address over Ethernet.
- the hardware FDB of the DSA switch on AP 2 still thinks that Client B
  can be reached via swp0. It does not know that it has roamed to wlan0,
  because it doesn't perform SA learning from the CPU port.

Now Client A contacts Client B.
AP 1 routes the packet fine towards swp0 and delivers it on the Ethernet
segment.
AP 2 sees a frame on swp0 and its fdb says that the destination is swp0.
Hairpinning is disabled => drop.

This problem comes from the fact that these switches have a 'blind spot'
for addresses coming from software bridging. The generic solution is not
to assume that hardware learning can be enabled somehow, but to listen
to more bridge learning events. It turns out that the bridge driver does
learn in software from all inbound frames, in __br_handle_local_finish.
A proper SWITCHDEV_FDB_ADD_TO_DEVICE notification is emitted for the
addresses serviced by the bridge on 'foreign' interfaces. The software
bridge also does the right thing on migration, by notifying that the old
entry is deleted, so that does not need to be special-cased in DSA. When
it is deleted, we just need to delete our static FDB entry towards the
CPU too, and wait.

The problem is that DSA currently only cares about SWITCHDEV_FDB_ADD_TO_DEVICE
events received on its own interfaces, such as static FDB entries.

Luckily we can change that, and DSA can listen to all switchdev FDB
add/del events in the system and figure out if those events were emitted
by a bridge that spans at least one of DSA's own ports. In case that is
true, DSA will also offload that address towards its own CPU port, in
the eventuality that there might be bridge clients attached to the DSA
switch who want to talk to the station connected to the foreign
interface.

In terms of implementation, we need to keep the fdb_info->added_by_user
check for the case where the switchdev event was targeted directly at a
DSA switch port. But we don't need to look at that flag for snooped
events. So the check is currently too late, we need to move it earlier.
This also simplifies the code a bit, since we avoid uselessly allocating
and freeing switchdev_work.

We could probably do some improvements in the future. For example,
multi-bridge support is rudimentary at the moment. If there are two
bridges spanning a DSA switch's ports, and both of them need to service
the same MAC address, then what will happen is that the migration of one
of those stations will trigger the deletion of the FDB entry from the
CPU port while it is still used by other bridge. That could be improved
with reference counting but is left for another time.

This behavior needs to be enabled at driver level by setting
ds->assisted_learning_on_cpu_port = true. This is because we don't want
to inflict a potential performance penalty (accesses through
MDIO/I2C/SPI are expensive) to hardware that really doesn't need it
because address learning on the CPU port works there.

Reported-by: DENG Qingfang <dqfext@gmail.com>
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: dsa: exit early in dsa_slave_switchdev_event if we can't program the FDB
Vladimir Oltean [Wed, 6 Jan 2021 09:51:34 +0000 (11:51 +0200)]
net: dsa: exit early in dsa_slave_switchdev_event if we can't program the FDB

Right now, the following would happen for a switch driver that does not
implement .port_fdb_add or .port_fdb_del.

dsa_slave_switchdev_event returns NOTIFY_OK and schedules:
-> dsa_slave_switchdev_event_work
   -> dsa_port_fdb_add
      -> dsa_port_notify(DSA_NOTIFIER_FDB_ADD)
         -> dsa_switch_fdb_add
            -> if (!ds->ops->port_fdb_add) return -EOPNOTSUPP;
   -> an error is printed with dev_dbg, and
      dsa_fdb_offload_notify(switchdev_work) is not called.

We can avoid scheduling the worker for nothing and say NOTIFY_DONE.
Because we don't call dsa_fdb_offload_notify, the static FDB entry will
remain just in the software bridge.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: dsa: move switchdev event implementation under the same switch/case statement
Vladimir Oltean [Wed, 6 Jan 2021 09:51:33 +0000 (11:51 +0200)]
net: dsa: move switchdev event implementation under the same switch/case statement

We'll need to start listening to SWITCHDEV_FDB_{ADD,DEL}_TO_DEVICE
events even for interfaces where dsa_slave_dev_check returns false, so
we need that check inside the switch-case statement for SWITCHDEV_FDB_*.

This movement also avoids a useless allocation / free of switchdev_work
on the untreated "default event" case.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: dsa: don't use switchdev_notifier_fdb_info in dsa_switchdev_event_work
Vladimir Oltean [Wed, 6 Jan 2021 09:51:32 +0000 (11:51 +0200)]
net: dsa: don't use switchdev_notifier_fdb_info in dsa_switchdev_event_work

Currently DSA doesn't add FDB entries on the CPU port, because it only
does so through switchdev, which is associated with a net_device, and
there are none of those for the CPU port.

But actually FDB addresses on the CPU port have some use cases of their
own, if the switchdev operations are initiated from within the DSA
layer. There is just one problem with the existing code: it passes a
structure in dsa_switchdev_event_work which was retrieved directly from
switchdev, so it contains a net_device. We need to generalize the
contents to something that covers the CPU port as well: the "ds, port"
tuple is fine for that.

Note that the new procedure for notifying the successful FDB offload is
inspired from the rocker model.

Also, nothing was being done if added_by_user was false. Let's check for
that a lot earlier, and don't actually bother to schedule the worker
for nothing.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: dsa: be louder when a non-legacy FDB operation fails
Vladimir Oltean [Wed, 6 Jan 2021 09:51:31 +0000 (11:51 +0200)]
net: dsa: be louder when a non-legacy FDB operation fails

The dev_close() call was added in commit c9eb3e0f8701 ("net: dsa: Add
support for learning FDB through notification") "to indicate inconsistent
situation" when we could not delete an FDB entry from the port.

bridge fdb del d8:58:d7:00:ca:6d dev swp0 self master

It is a bit drastic and at the same time not helpful if the above fails
to only print with netdev_dbg log level, but on the other hand to bring
the interface down.

So increase the verbosity of the error message, and drop dev_close().

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: Jakub Kicinski <kuba@kernel.org>
3 years agonet: bridge: notify switchdev of disappearance of old FDB entry upon migration
Vladimir Oltean [Wed, 6 Jan 2021 09:51:30 +0000 (11:51 +0200)]
net: bridge: notify switchdev of disappearance of old FDB entry upon migration

Currently the bridge emits atomic switchdev notifications for
dynamically learnt FDB entries. Monitoring these notifications works
wonders for switchdev drivers that want to keep their hardware FDB in
sync with the bridge's FDB.

For example station A wants to talk to station B in the diagram below,
and we are concerned with the behavior of the bridge on the DUT device:

                   DUT
 +-------------------------------------+
 |                 br0                 |
 | +------+ +------+ +------+ +------+ |
 | |      | |      | |      | |      | |
 | | swp0 | | swp1 | | swp2 | | eth0 | |
 +-------------------------------------+
      |        |                  |
  Station A    |                  |
               |                  |
         +--+------+--+    +--+------+--+
         |  |      |  |    |  |      |  |
         |  | swp0 |  |    |  | swp0 |  |
 Another |  +------+  |    |  +------+  | Another
  switch |     br0    |    |     br0    | switch
         |  +------+  |    |  +------+  |
         |  |      |  |    |  |      |  |
         |  | swp1 |  |    |  | swp1 |  |
         +--+------+--+    +--+------+--+
                                  |
                              Station B

Interfaces swp0, swp1, swp2 are handled by a switchdev driver that has
the following property: frames injected from its control interface bypass
the internal address analyzer logic, and therefore, this hardware does
not learn from the source address of packets transmitted by the network
stack through it. So, since bridging between eth0 (where Station B is
attached) and swp0 (where Station A is attached) is done in software,
the switchdev hardware will never learn the source address of Station B.
So the traffic towards that destination will be treated as unknown, i.e.
flooded.

This is where the bridge notifications come in handy. When br0 on the
DUT sees frames with Station B's MAC address on eth0, the switchdev
driver gets these notifications and can install a rule to send frames
towards Station B's address that are incoming from swp0, swp1, swp2,
only towards the control interface. This is all switchdev driver private
business, which the notification makes possible.

All is fine until someone unplugs Station B's cable and moves it to the
other switch:

                   DUT
 +-------------------------------------+
 |                 br0                 |
 | +------+ +------+ +------+ +------+ |
 | |      | |      | |      | |      | |
 | | swp0 | | swp1 | | swp2 | | eth0 | |
 +-------------------------------------+
      |        |                  |
  Station A    |                  |
               |                  |
         +--+------+--+    +--+------+--+
         |  |      |  |    |  |      |  |
         |  | swp0 |  |    |  | swp0 |  |
 Another |  +------+  |    |  +------+  | Another
  switch |     br0    |    |     br0    | switch
         |  +------+  |    |  +------+  |
         |  |      |  |    |  |      |  |
         |  | swp1 |  |    |  | swp1 |  |
         +--+------+--+    +--+------+--+
               |
           Station B

Luckily for the use cases we care about, Station B is noisy enough that
the DUT hears it (on swp1 this time). swp1 receives the frames and
delivers them to the bridge, who enters the unlikely path in br_fdb_update
of updating an existing entry. It moves the entry in the software bridge
to swp1 and emits an addition notification towards that.

As far as the switchdev driver is concerned, all that it needs to ensure
is that traffic between Station A and Station B is not forever broken.
If it does nothing, then the stale rule to send frames for Station B
towards the control interface remains in place. But Station B is no
longer reachable via the control interface, but via a port that can
offload the bridge port learning attribute. It's just that the port is
prevented from learning this address, since the rule overrides FDB
updates. So the rule needs to go. The question is via what mechanism.

It sure would be possible for this switchdev driver to keep track of all
addresses which are sent to the control interface, and then also listen
for bridge notifier events on its own ports, searching for the ones that
have a MAC address which was previously sent to the control interface.
But this is cumbersome and inefficient. Instead, with one small change,
the bridge could notify of the address deletion from the old port, in a
symmetrical manner with how it did for the insertion. Then the switchdev
driver would not be required to monitor learn/forget events for its own
ports. It could just delete the rule towards the control interface upon
bridge entry migration. This would make hardware address learning be
possible again. Then it would take a few more packets until the hardware
and software FDB would be in sync again.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Acked-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge branch 'r8169-improve-rtl8168g-phy-suspend-quirk'
Jakub Kicinski [Thu, 7 Jan 2021 22:56:35 +0000 (14:56 -0800)]
Merge branch 'r8169-improve-rtl8168g-phy-suspend-quirk'

Heiner Kallweit says:

====================
r8169: improve RTL8168g PHY suspend quirk

According to Realtek the ERI register 0x1a8 quirk is needed to work
around a hw issue with the PHY on RTL8168g. The register needs to be
changed before powering down the PHY. Currently we don't meet this
requirement, however I'm not aware of any problems caused by this.
Therefore I see the change as an improvement.

The PHY driver has no means to access the chip ERI registers,
therefore we have to intercept MDIO writes to the BMCR register.
If the BMCR_PDOWN bit is going to be set, then let's apply the
quirk before actually powering down the PHY.
====================

Link: https://lore.kernel.org/r/9303c2cf-c521-beea-c09f-63b5dfa91b9c@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agor8169: improve RTL8168g PHY suspend quirk
Heiner Kallweit [Wed, 6 Jan 2021 10:49:50 +0000 (11:49 +0100)]
r8169: improve RTL8168g PHY suspend quirk

According to Realtek the ERI register 0x1a8 quirk is needed to work
around a hw issue with the PHY on RTL8168g. The register needs to be
changed before powering down the PHY. Currently we don't meet this
requirement, however I'm not aware of any problems caused by this.
Therefore I see the change as an improvement.

The PHY driver has no means to access the chip ERI registers,
therefore we have to intercept MDIO writes to BMCR register.
If the BMCR_PDOWN bit is going to be set, then let's apply the
quirk before actually powering down the PHY.

Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agor8169: move ERI access functions to avoid forward declaration
Heiner Kallweit [Wed, 6 Jan 2021 10:47:08 +0000 (11:47 +0100)]
r8169: move ERI access functions to avoid forward declaration

No functional change here. We just move a code block to avoid a
function forward declaration in a subsequent change.

Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: phy: replace mutex_is_locked with lockdep_assert_held in phylib
Heiner Kallweit [Wed, 6 Jan 2021 13:03:40 +0000 (14:03 +0100)]
net: phy: replace mutex_is_locked with lockdep_assert_held in phylib

Switch to lockdep_assert_held(_once), similar to what is being done
in other subsystems. One advantage is that there's zero runtime
overhead if lockdep support isn't enabled.

Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://lore.kernel.org/r/ccc40b9d-8ee0-43a1-5009-2cc95ca79c85@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: phy: bcm7xxx: Add an entry for BCM72116
Florian Fainelli [Wed, 6 Jan 2021 17:09:44 +0000 (09:09 -0800)]
net: phy: bcm7xxx: Add an entry for BCM72116

BCM72116 features a 28nm integrated EPHY, add an entry to match this PHY
OUI.

Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://lore.kernel.org/r/20210106170944.1253046-1-f.fainelli@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge branch 'udp_tunnel_nic-post-conversion-cleanup'
Jakub Kicinski [Thu, 7 Jan 2021 20:53:32 +0000 (12:53 -0800)]
Merge branch 'udp_tunnel_nic-post-conversion-cleanup'

udp_tunnel_nic: post conversion cleanup

It has been two releases since we added the common infra for UDP
tunnel port offload, and we have not heard of any major issues.
Remove the old direct driver NDOs completely, and perform minor
simplifications in the tunnel drivers.

Link: https://lore.kernel.org/r/20210106210637.1839662-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoudp_tunnel: reshuffle NETIF_F_RX_UDP_TUNNEL_PORT checks
Jakub Kicinski [Wed, 6 Jan 2021 21:06:37 +0000 (13:06 -0800)]
udp_tunnel: reshuffle NETIF_F_RX_UDP_TUNNEL_PORT checks

Move the NETIF_F_RX_UDP_TUNNEL_PORT feature check into
udp_tunnel_nic_*_port() helpers, since they're always
done right before the call.

Add similar checks before calling the notifier.
udp_tunnel_nic invokes the notifier without checking
features which could result in some wasted cycles.

Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: remove ndo_udp_tunnel_* callbacks
Jakub Kicinski [Wed, 6 Jan 2021 21:06:36 +0000 (13:06 -0800)]
net: remove ndo_udp_tunnel_* callbacks

All UDP tunnel port management is now routed via udp_tunnel_nic
infra directly. Remove the old callbacks.

Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoudp_tunnel: remove REGISTER/UNREGISTER handling from tunnel drivers
Jakub Kicinski [Wed, 6 Jan 2021 21:06:35 +0000 (13:06 -0800)]
udp_tunnel: remove REGISTER/UNREGISTER handling from tunnel drivers

udp_tunnel_nic handles REGISTER and UNREGISTER event, now that all
drivers use that infra we can drop the event handling in the tunnel
drivers.

Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoudp_tunnel: hard-wire NDOs to udp_tunnel_nic_*_port() helpers
Jakub Kicinski [Wed, 6 Jan 2021 21:06:34 +0000 (13:06 -0800)]
udp_tunnel: hard-wire NDOs to udp_tunnel_nic_*_port() helpers

All drivers use udp_tunnel_nic_*_port() helpers, prepare for
NDO removal by invoking those helpers directly.

The helpers are safe to call on all devices, they check if
device has the UDP tunnel state initialized.

Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: broadcom: Drop OF dependency from BGMAC_PLATFORM
Florian Fainelli [Wed, 6 Jan 2021 19:15:45 +0000 (11:15 -0800)]
net: broadcom: Drop OF dependency from BGMAC_PLATFORM

All of the OF code that is used has stubbed and will compile and link
just fine, keeping COMPILE_TEST is enough.

Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Link: https://lore.kernel.org/r/20210106191546.1358324-1-f.fainelli@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge branch 'bcm63xx_enet-major-makeover-of-driver'
Jakub Kicinski [Thu, 7 Jan 2021 20:39:56 +0000 (12:39 -0800)]
Merge branch 'bcm63xx_enet-major-makeover-of-driver'

Sieng Piaw Liew says:

====================
bcm63xx_enet: major makeover of driver

This patch series aim to improve the bcm63xx_enet driver by integrating the
latest networking features, i.e. batched rx processing, BQL, build_skb,
etc.

The newer enetsw SoCs are found to be able to do unaligned rx DMA by adding
NET_IP_ALIGN padding which, combined with these patches, improved packet
processing performance by ~50% on BCM6328.

Older non-enetsw SoCs still benefit mainly from rx batching. Performance
improvement of ~30% is observed on BCM6333.

The BCM63xx SoCs are designed for routers. As such, having BQL is
beneficial as well as trivial to add.

v3:
* Simplify xmit_more patch by not moving around the code needlessly.
* Fix indentation in xmit_more patch.
* Fix indentation in build_skb patch.
* Split rx ring cleanup patch from build_skb patch and precede build_skb
  patch for better understanding, as suggested by Florian Fainelli.

v2:
* Add xmit_more support and rx loop improvisation patches.
* Moved BQL netdev_reset_queue() to bcm_enet_stop()/bcm_enetsw_stop()
  functions as suggested by Florian Fainelli.
* Improved commit messages.
====================

Link: https://lore.kernel.org/r/20210106144208.1935-1-liew.s.piaw@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobcm63xx_enet: improve rx loop
Sieng Piaw Liew [Wed, 6 Jan 2021 14:42:08 +0000 (22:42 +0800)]
bcm63xx_enet: improve rx loop

Use existing rx processed count to track against budget, thereby making
budget decrement operation redundant.

rx_desc_count can be calculated outside the rx loop, making the loop a
bit smaller.

Signed-off-by: Sieng Piaw Liew <liew.s.piaw@gmail.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobcm63xx_enet: convert to build_skb
Sieng Piaw Liew [Wed, 6 Jan 2021 14:42:07 +0000 (22:42 +0800)]
bcm63xx_enet: convert to build_skb

We can increase the efficiency of rx path by using buffers to receive
packets then build SKBs around them just before passing into the network
stack. In contrast, preallocating SKBs too early reduces CPU cache
efficiency.

Check if we're in NAPI context when refilling RX. Normally we're almost
always running in NAPI context. Dispatch to napi_alloc_frag directly
instead of relying on netdev_alloc_frag which does the same but
with the overhead of local_bh_disable/enable.

Tested on BCM6328 320 MHz and iperf3 -M 512 to measure packet/sec
performance. Included netif_receive_skb_list and NET_IP_ALIGN
optimizations.

Before:
[ ID] Interval           Transfer     Bandwidth       Retr
[  4]   0.00-10.00  sec  49.9 MBytes  41.9 Mbits/sec  197         sender
[  4]   0.00-10.00  sec  49.3 MBytes  41.3 Mbits/sec            receiver

After:
[ ID] Interval           Transfer     Bandwidth       Retr
[  4]   0.00-30.00  sec   171 MBytes  47.8 Mbits/sec  272         sender
[  4]   0.00-30.00  sec   170 MBytes  47.6 Mbits/sec            receiver

Signed-off-by: Sieng Piaw Liew <liew.s.piaw@gmail.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobcm63xx_enet: consolidate rx SKB ring cleanup code
Sieng Piaw Liew [Wed, 6 Jan 2021 14:42:06 +0000 (22:42 +0800)]
bcm63xx_enet: consolidate rx SKB ring cleanup code

The rx SKB ring use the same code for cleanup at various points.
Combine them into a function to reduce lines of code.

Signed-off-by: Sieng Piaw Liew <liew.s.piaw@gmail.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobcm63xx_enet: alloc rx skb with NET_IP_ALIGN
Sieng Piaw Liew [Wed, 6 Jan 2021 14:42:05 +0000 (22:42 +0800)]
bcm63xx_enet: alloc rx skb with NET_IP_ALIGN

Use netdev_alloc_skb_ip_align on newer SoCs with integrated switch
(enetsw) when refilling RX. Increases packet processing performance
by 30% (with netif_receive_skb_list).

Non-enetsw SoCs cannot function with the extra pad so continue to use
the regular netdev_alloc_skb.

Tested on BCM6328 320 MHz and iperf3 -M 512 to measure packet/sec
performance.

Before:
[ ID] Interval Transfer Bandwidth Retr
[ 4] 0.00-30.00 sec 120 MBytes 33.7 Mbits/sec 277 sender
[ 4] 0.00-30.00 sec 120 MBytes 33.5 Mbits/sec receiver

After (+netif_receive_skb_list):
[ 4] 0.00-30.00 sec 155 MBytes 43.3 Mbits/sec 354 sender
[ 4] 0.00-30.00 sec 154 MBytes 43.1 Mbits/sec receiver

Signed-off-by: Sieng Piaw Liew <liew.s.piaw@gmail.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobcm63xx_enet: add xmit_more support
Sieng Piaw Liew [Wed, 6 Jan 2021 14:42:04 +0000 (22:42 +0800)]
bcm63xx_enet: add xmit_more support

Support bulking hardware TX queue by using netdev_xmit_more().

Signed-off-by: Sieng Piaw Liew <liew.s.piaw@gmail.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobcm63xx_enet: add BQL support
Sieng Piaw Liew [Wed, 6 Jan 2021 14:42:03 +0000 (22:42 +0800)]
bcm63xx_enet: add BQL support

Add Byte Queue Limits support to reduce/remove bufferbloat in
bcm63xx_enet.

Signed-off-by: Sieng Piaw Liew <liew.s.piaw@gmail.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobcm63xx_enet: batch process rx path
Sieng Piaw Liew [Wed, 6 Jan 2021 14:42:02 +0000 (22:42 +0800)]
bcm63xx_enet: batch process rx path

Use netif_receive_skb_list to batch process rx skb.
Tested on BCM6328 320 MHz using iperf3 -M 512, increasing performance
by 12.5%.

Before:
[ ID] Interval           Transfer     Bandwidth       Retr
[  4]   0.00-30.00  sec   120 MBytes  33.7 Mbits/sec  277         sender
[  4]   0.00-30.00  sec   120 MBytes  33.5 Mbits/sec            receiver

After:
[ ID] Interval           Transfer     Bandwidth       Retr
[  4]   0.00-30.00  sec   136 MBytes  37.9 Mbits/sec  203         sender
[  4]   0.00-30.00  sec   135 MBytes  37.7 Mbits/sec            receiver

Signed-off-by: Sieng Piaw Liew <liew.s.piaw@gmail.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoqmi_wwan: Increase headroom for QMAP SKBs
Kristian Evensen [Wed, 6 Jan 2021 12:24:03 +0000 (13:24 +0100)]
qmi_wwan: Increase headroom for QMAP SKBs

When measuring the throughput (iperf3 + TCP) while routing on a
not-so-powerful device (Mediatek MT7621, 880MHz CPU), I noticed that I
achieved significantly lower speeds with QMI-based modems than for
example a USB LAN dongle. The CPU was saturated in all of my tests.

With the dongle I got ~300 Mbit/s, while I only measured ~200 Mbit/s
with the modems. All offloads, etc.  were switched off for the dongle,
and I configured the modems to use QMAP (16k aggregation). The tests
with the dongle were performed in my local (gigabit) network, while the
LTE network the modems were connected to delivers 700-800 Mbit/s.

Profiling the kernel revealed the cause of the performance difference.
In qmimux_rx_fixup(), an SKB is allocated for each packet contained in
the URB. This SKB has too little headroom, causing the check in
skb_cow() (called from ip_forward()) to fail. pskb_expand_head() is then
called and the SKB is reallocated. In the output from perf, I see that a
significant amount of time is spent in pskb_expand_head() + support
functions.

In order to ensure that the SKB has enough headroom, this commit
increases the amount of memory allocated in qmimux_rx_fixup() by
LL_MAX_HEADER. The reason for using LL_MAX_HEADER and not a more
accurate value, is that we do not know the type of the outgoing network
interface. After making this change, I achieve the same throughput with
the modems as with the dongle.

Signed-off-by: Kristian Evensen <kristian.evensen@gmail.com>
Acked-by: Bjørn Mork <bjorn@mork.no>
Link: https://lore.kernel.org/r/20210106122403.1321180-1-kristian.evensen@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge tag 'linux-can-next-for-5.12-20210106' of git://git.kernel.org/pub/scm/linux...
Jakub Kicinski [Thu, 7 Jan 2021 18:53:31 +0000 (10:53 -0800)]
Merge tag 'linux-can-next-for-5.12-20210106' of git://git./linux/kernel/git/mkl/linux-can-next

Marc Kleine-Budde says:

====================
pull-request: can-next 2021-01-06

The first 16 patches are by me and target the tcan4x5x SPI glue driver for the
m_can CAN driver. First there are a several cleanup commits, then the SPI
regmap part is converted to 8 bits per word, to make it possible to use that
driver on SPI controllers that only support the 8 bit per word mode (such as
the SPI cores on the raspberry pi).

Oliver Hartkopp contributes a patch for the CAN_RAW protocol. The getsockopt()
for CAN_RAW_FILTER is changed to return -ERANGE if the filterset does not fit
into the provided user space buffer.

The last two patches are by Joakim Zhang and add wakeup support to the flexcan
driver for the i.MX8QM SoC. The dt-bindings docs are extended to describe the
added property.

* tag 'linux-can-next-for-5.12-20210106' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next:
  can: flexcan: add CAN wakeup function for i.MX8QM
  dt-bindings: can: fsl,flexcan: add fsl,scu-index property to indicate a resource
  can: raw: return -ERANGE when filterset does not fit into user space buffer
  can: tcan4x5x: add support for half-duplex controllers
  can: tcan4x5x: rework SPI access
  can: tcan4x5x: add {wr,rd}_table
  can: tcan4x5x: add max_raw_{read,write} of 256
  can: tcan4x5x: tcan4x5x_regmap: set reg_stride to 4
  can: tcan4x5x: fix max register value
  can: tcan4x5x: tcan4x5x_regmap_init(): use spi as context pointer
  can: tcan4x5x: tcan4x5x_regmap_write(): remove not needed casts and replace 4 by sizeof
  can: tcan4x5x: rename regmap_spi_gather_write() -> tcan4x5x_regmap_gather_write()
  can: tcan4x5x: remove regmap async support
  can: tcan4x5x: tcan4x5x_bus: remove not needed read_flag_mask
  can: tcan4x5x: mark struct regmap_bus tcan4x5x_bus as constant
  can: tcan4x5x: move regmap code into seperate file
  can: tcan4x5x: rename tcan4x5x.c -> tcan4x5x-core.c
  can: tcan4x5x: beautify indention of tcan4x5x_of_match and tcan4x5x_id_table
  can: tcan4x5x: replace DEVICE_NAME by KBUILD_MODNAME
====================

Link: https://lore.kernel.org/r/20210107094900.173046-1-mkl@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: dsa: print error on invalid port index
Rafał Miłecki [Wed, 6 Jan 2021 09:09:15 +0000 (10:09 +0100)]
net: dsa: print error on invalid port index

Looking for an -EINVAL all over the dsa code could take hours for
inexperienced DSA users.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://lore.kernel.org/r/20210106090915.21439-1-zajec5@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agocan: flexcan: add CAN wakeup function for i.MX8QM
Joakim Zhang [Fri, 6 Nov 2020 10:56:27 +0000 (18:56 +0800)]
can: flexcan: add CAN wakeup function for i.MX8QM

The System Controller Firmware (SCFW) is a low-level system function
which runs on a dedicated Cortex-M core to provide power, clock, and
resource management. It exists on some i.MX8 processors. e.g. i.MX8QM
(QM, QP), and i.MX8QX (QXP, DX). SCU driver manages the IPC interface
between host CPU and the SCU firmware running on M4.

For i.MX8QM, stop mode request is controlled by System Controller Unit(SCU)
firmware, this patch introduces FLEXCAN_QUIRK_SETUP_STOP_MODE_SCFW quirk
for this function.

Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>
Link: https://lore.kernel.org/r/20201106105627.31061-6-qiangqing.zhang@nxp.com
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agodt-bindings: can: fsl,flexcan: add fsl,scu-index property to indicate a resource
Joakim Zhang [Fri, 6 Nov 2020 10:56:24 +0000 (18:56 +0800)]
dt-bindings: can: fsl,flexcan: add fsl,scu-index property to indicate a resource

For SoCs with SCU support, need setup stop mode via SCU firmware, so this
property can help indicate a resource in SCU firmware.

Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>
Link: https://lore.kernel.org/r/20201106105627.31061-3-qiangqing.zhang@nxp.com
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: raw: return -ERANGE when filterset does not fit into user space buffer
Oliver Hartkopp [Wed, 16 Dec 2020 17:49:28 +0000 (18:49 +0100)]
can: raw: return -ERANGE when filterset does not fit into user space buffer

Multiple filters (struct can_filter) can be set with the setsockopt()
function, which was originally intended as a write-only operation.

As getsockopt() also provides a CAN_RAW_FILTER option to read back the
given filters, the caller has to provide an appropriate user space buffer.
In the case this buffer is too small the getsockopt() silently truncates
the filter information and gives no information about the needed space.
This is safe but not convenient for the programmer.

In net/core/sock.c the SO_PEERGROUPS sockopt had a similar requirement
and solved it by returning -ERANGE in the case that the provided data
does not fit into the given user space buffer and fills the required size
into optlen, so that the caller can retry with a matching buffer length.

This patch adopts this approach for CAN_RAW_FILTER getsockopt().

Reported-by: Phillip Schichtel <phillip@schich.tel>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Tested-By: Phillip Schichtel <phillip@schich.tel>
Link: https://lore.kernel.org/r/20201216174928.21663-1-socketcan@hartkopp.net
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: add support for half-duplex controllers
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:46 +0000 (00:17 +0100)]
can: tcan4x5x: add support for half-duplex controllers

This patch adds back support for half-duplex controllers, which was removed in
the last patch.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-17-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: rework SPI access
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:45 +0000 (00:17 +0100)]
can: tcan4x5x: rework SPI access

This patch reworks the SPI access and fixes several probems:
- tcan4x5x_regmap_gather_write(), tcan4x5x_regmap_read():
  Do not place variable "addr" on stack and use it as buffer for SPI
  transfer. Buffers for SPI transfers must be allocated from DMA save
  memory.
- tcan4x5x_regmap_gather_write(), tcan4x5x_regmap_read():
  Halfe number of SPI transfers by using a single buffer + memcpy().
  This improves the performance, especially on SPI controllers, which
  use interrupt based transfers.
- Use "8" bits per word, not "32". This makes it possible to use this
  driver on SoCs like the Raspberry Pi, which SPI host controller
  drivers only support 8 bits per word.

Note: this breaks half duplex only controllers. Support for them will be
re-added in the next patch.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-16-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: add {wr,rd}_table
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:44 +0000 (00:17 +0100)]
can: tcan4x5x: add {wr,rd}_table

The memory space of the chip is not fully populated, so add a regmap range
table to document this.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-15-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: add max_raw_{read,write} of 256
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:43 +0000 (00:17 +0100)]
can: tcan4x5x: add max_raw_{read,write} of 256

The tcan4x5x chip support bulk read/write, but as the length field is only 8
bits wide, the maximum is 256. A length of 0 is treated as 256.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-14-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: tcan4x5x_regmap: set reg_stride to 4
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:42 +0000 (00:17 +0100)]
can: tcan4x5x: tcan4x5x_regmap: set reg_stride to 4

This patch sets the regmap stide to 4, as the chip only supports access on 32
bit alligned access.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-13-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: fix max register value
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:41 +0000 (00:17 +0100)]
can: tcan4x5x: fix max register value

This patch fixes the max register value for the regmap.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-12-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: tcan4x5x_regmap_init(): use spi as context pointer
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:40 +0000 (00:17 +0100)]
can: tcan4x5x: tcan4x5x_regmap_init(): use spi as context pointer

This patch replaces the context pointer of the regmap callback functions by a
pointer to the spi_device. This saves one level of indirection in the
callbacks.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-11-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: tcan4x5x_regmap_write(): remove not needed casts and replace 4 by...
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:39 +0000 (00:17 +0100)]
can: tcan4x5x: tcan4x5x_regmap_write(): remove not needed casts and replace 4 by sizeof

This patch simplifies the tcan4x5x_regmap_write(0 function by removing not
needed casts and replaces hardcoded "4" by appropriate sizeof()s.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-10-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: rename regmap_spi_gather_write() -> tcan4x5x_regmap_gather_write()
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:38 +0000 (00:17 +0100)]
can: tcan4x5x: rename regmap_spi_gather_write() -> tcan4x5x_regmap_gather_write()

This patch renames the regmap_spi_gather_write() function to
tcan4x5x_regmap_gather_write(). Now it has a "tcan4x5x_" prefix as all other
functions in this driver.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-9-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: remove regmap async support
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:37 +0000 (00:17 +0100)]
can: tcan4x5x: remove regmap async support

The driver doesn't use regmap async support, so remove the stubs.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-8-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: tcan4x5x_bus: remove not needed read_flag_mask
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:36 +0000 (00:17 +0100)]
can: tcan4x5x: tcan4x5x_bus: remove not needed read_flag_mask

With C99 initializers, all non mentioned members are initialzied to 0, so
remove not needed initialization of read_flag_mask.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-7-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: mark struct regmap_bus tcan4x5x_bus as constant
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:35 +0000 (00:17 +0100)]
can: tcan4x5x: mark struct regmap_bus tcan4x5x_bus as constant

This patch marks the struct regmap_bus tcan4x5x_bus as constant.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-6-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: move regmap code into seperate file
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:34 +0000 (00:17 +0100)]
can: tcan4x5x: move regmap code into seperate file

This patch moves the regmap code into a seperate file.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-5-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: rename tcan4x5x.c -> tcan4x5x-core.c
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:33 +0000 (00:17 +0100)]
can: tcan4x5x: rename tcan4x5x.c -> tcan4x5x-core.c

This is a preparation patch to move the regmap related code into a seperate
file. This patch removes the tcan4x5x.c to tcan4x5x-core.c.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-4-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: beautify indention of tcan4x5x_of_match and tcan4x5x_id_table
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:32 +0000 (00:17 +0100)]
can: tcan4x5x: beautify indention of tcan4x5x_of_match and tcan4x5x_id_table

This patch beautifies the indention of the tcan4x5x_of_match and
tcan4x5x_id_table.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-3-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: tcan4x5x: replace DEVICE_NAME by KBUILD_MODNAME
Marc Kleine-Budde [Tue, 15 Dec 2020 23:17:31 +0000 (00:17 +0100)]
can: tcan4x5x: replace DEVICE_NAME by KBUILD_MODNAME

This patch replaces the DEVICE_NAME macro by KBUILD_MODNAME and removed the
superfluous DEVICE_NAME.

Reviewed-by: Dan Murphy <dmurphy@ti.com>
Tested-by: Sean Nyekjaer <sean@geanix.com>
Link: https://lore.kernel.org/r/20201215231746.1132907-2-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agonet-next: docs: Fix typos in snmp_counter.rst
Masanari Iida [Tue, 5 Jan 2021 14:40:29 +0000 (23:40 +0900)]
net-next: docs: Fix typos in snmp_counter.rst

This patch fixes some spelling typos in snmp_counter.rst

Signed-off-by: Masanari Iida <standby24x7@gmail.com>
Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'net-ks8851-Add-KS8851-PHY-support'
David S. Miller [Wed, 6 Jan 2021 01:05:12 +0000 (17:05 -0800)]
Merge branch 'net-ks8851-Add-KS8851-PHY-support'

Marek Vasut says:

====================
net: ks8851: Add KS8851 PHY support

The KS8851 has a reduced internal PHY, which is accessible through its
registers at offset 0xe4. The PHY is compatible with KS886x PHY present
in Micrel switches, including the PHY ID Low/High registers swap, which
is present both in the MAC and the switch.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: ks8851: Register MDIO bus and the internal PHY
Marek Vasut [Tue, 5 Jan 2021 14:11:51 +0000 (15:11 +0100)]
net: ks8851: Register MDIO bus and the internal PHY

The KS8851 has a reduced internal PHY, which is accessible through its
registers at offset 0xe4. The PHY is compatible with KS886x PHY present
in Micrel switches, except the PHY ID Low/High registers are swapped.
Register MDIO bus so this PHY can be detected and probed by phylib.

Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Heiner Kallweit <hkallweit1@gmail.com>
Cc: Lukas Wunner <lukas@wunner.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: phy: micrel: Add KS8851 PHY support
Marek Vasut [Tue, 5 Jan 2021 14:11:50 +0000 (15:11 +0100)]
net: phy: micrel: Add KS8851 PHY support

The KS8851 has a reduced internal PHY, which is accessible through its
registers at offset 0xe4. The PHY is compatible with KS886x PHY present
in Micrel switches, including the PHY ID Low/High registers swap, which
is present both in the MAC and the switch.

Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Heiner Kallweit <hkallweit1@gmail.com>
Cc: Lukas Wunner <lukas@wunner.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomacvlan: remove redundant null check on data
Yunjian Wang [Tue, 5 Jan 2021 06:31:34 +0000 (14:31 +0800)]
macvlan: remove redundant null check on data

Because macvlan_common_newlink() and macvlan_changelink() already
checked NULL data parameter, so the additional check is unnecessary,
just remove it.

Fixes: 79cf79abce71 ("macvlan: add source mode")
Signed-off-by: Yunjian Wang <wangyunjian@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoocteontx2-pf: Add RSS multi group support
Geetha sowjanya [Mon, 4 Jan 2021 07:20:39 +0000 (12:50 +0530)]
octeontx2-pf: Add RSS multi group support

Hardware supports 8 RSS groups per interface. Currently we are using
only group '0'. This patch allows user to create new RSS groups/contexts
and use the same as destination for flow steering rules.

usage:
To steer the traffic to RQ 2,3

ethtool -X eth0 weight 0 0 1 1 context new
(It will print the allocated context id number)
New RSS context is 1

ethtool -N eth0 flow-type tcp4 dst-port 80 context 1 loc 1

To delete the context
ethtool -X eth0 context 1 delete

When an RSS context is removed, the active classification
rules using this context are also removed.

Change-log:

v4
- Fixed compiletime warning.
- Address Saeed's comments on v3.

v3
- Coverted otx2_set_rxfh() to use new function.

v2
- Removed unrelated whitespace
- Coverted otx2_get_rxfh() to use new function.

Signed-off-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: nfc: nci: Change the NCI close sequence
Bongsu Jeon [Thu, 31 Dec 2020 02:59:26 +0000 (11:59 +0900)]
net: nfc: nci: Change the NCI close sequence

If there is a NCI command in work queue after closing the NCI device at
nci_unregister_device, The NCI command timer starts at flush_workqueue
function and then NCI command timeout handler would be called 5 second
after flushing the NCI command work queue and destroying the queue.
At that time, the timeout handler would try to use NCI command work queue
that is destroyed already. it will causes the problem. To avoid this
abnormal situation, change the sequence to prevent the NCI command timeout
handler from being called after destroying the NCI command work queue.

Signed-off-by: Bongsu Jeon <bongsu.jeon@samsung.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: kcm: Replace fput with sockfd_put
Zheng Yongjun [Wed, 30 Dec 2020 09:18:09 +0000 (17:18 +0800)]
net: kcm: Replace fput with sockfd_put

The function sockfd_lookup uses fget on the value that is stored in
the file field of the returned structure, so fput should ultimately be
applied to this value.  This can be done directly, but it seems better
to use the specific macro sockfd_put, which does the same thing.

Perform a source code refactoring by using the following semantic patch.

    // <smpl>
    @@
    expression s;
    @@

       s = sockfd_lookup(...)
       ...
    + sockfd_put(s);
    - fput(s->file);
    // </smpl>

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet/mlxfw: Use kzalloc for allocating only one thing
Zheng Yongjun [Wed, 30 Dec 2020 08:18:35 +0000 (16:18 +0800)]
net/mlxfw: Use kzalloc for allocating only one thing

Use kzalloc rather than kcalloc(1,...)

The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)

// <smpl>
@@
@@

- kcalloc(1,
+ kzalloc(
          ...)
// </smpl>

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge tag 'mlx5-updates-2021-01-05' of git://git.kernel.org/pub/scm/linux/kernel...
David S. Miller [Tue, 5 Jan 2021 23:44:45 +0000 (15:44 -0800)]
Merge tag 'mlx5-updates-2021-01-05' of git://git./linux/kernel/git/saeed/linux

Saeed Mahameed says:

====================
mlx5-updates-2021-01-05

SW steering, Refactor to have a device specific STE layer below dr_ste

This series introduces some improvements and refactoring by adding a new layer
below dr_ste to allow support for different devices format.

It adds a struct of device specific callbacks for STE layer below dr_ste.
Each device will implement its HW-specific function, and a common logic
from the DR code will access these functions through the new ste_ctx API.

Connect-X5-style steering format is called STE_v0.
In the next patch series we bring the Connect-X6-style format - STE_v1.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoocteontx2-af: Use kzalloc for allocating only one thing
Zheng Yongjun [Tue, 29 Dec 2020 13:53:09 +0000 (21:53 +0800)]
octeontx2-af: Use kzalloc for allocating only one thing

Use kzalloc rather than kcalloc(1,...)

The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)

// <smpl>
@@
@@

- kcalloc(1,
+ kzalloc(
          ...)
// </smpl>

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoiavf: Use kzalloc for allocating only one thing
Zheng Yongjun [Tue, 29 Dec 2020 13:53:01 +0000 (21:53 +0800)]
iavf: Use kzalloc for allocating only one thing

Use kzalloc rather than kcalloc(1,...)

The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)

// <smpl>
@@
@@

- kcalloc(1,
+ kzalloc(
          ...)
// </smpl>

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoliquidio: Use kzalloc for allocating only one thing
Zheng Yongjun [Tue, 29 Dec 2020 13:52:54 +0000 (21:52 +0800)]
liquidio: Use kzalloc for allocating only one thing

Use kzalloc rather than kcalloc(1,...)

The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)

// <smpl>
@@
@@

- kcalloc(1,
+ kzalloc(
          ...)
// </smpl>

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agobnxt_en: Use kzalloc for allocating only one thing
Zheng Yongjun [Tue, 29 Dec 2020 13:52:46 +0000 (21:52 +0800)]
bnxt_en: Use kzalloc for allocating only one thing

Use kzalloc rather than kcalloc(1,...)

The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)

// <smpl>
@@
@@

- kcalloc(1,
+ kzalloc(
          ...)
// </smpl>

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: dsa: sja1105: Use kzalloc for allocating only one thing
Zheng Yongjun [Tue, 29 Dec 2020 13:52:38 +0000 (21:52 +0800)]
net: dsa: sja1105: Use kzalloc for allocating only one thing

Use kzalloc rather than kcalloc(1,...)

The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)

// <smpl>
@@
@@

- kcalloc(1,
+ kzalloc(
          ...)
// </smpl>

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agocavium/liquidio: Use DEFINE_SPINLOCK() for spinlock
Zheng Yongjun [Tue, 29 Dec 2020 13:49:54 +0000 (21:49 +0800)]
cavium/liquidio: Use DEFINE_SPINLOCK() for spinlock

spinlock can be initialized automatically with DEFINE_SPINLOCK()
rather than explicitly calling spin_lock_init().

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: ixp4xx_eth: Use DEFINE_SPINLOCK() for spinlock
Zheng Yongjun [Tue, 29 Dec 2020 13:49:47 +0000 (21:49 +0800)]
net: ixp4xx_eth: Use DEFINE_SPINLOCK() for spinlock

spinlock can be initialized automatically with DEFINE_SPINLOCK()
rather than explicitly calling spin_lock_init().

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: usb: Use DEFINE_SPINLOCK() for spinlock
Zheng Yongjun [Tue, 29 Dec 2020 13:49:27 +0000 (21:49 +0800)]
net: usb: Use DEFINE_SPINLOCK() for spinlock

spinlock can be initialized automatically with DEFINE_SPINLOCK()
rather than explicitly calling spin_lock_init().

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: wan: Use DEFINE_SPINLOCK() for spinlock
Zheng Yongjun [Tue, 29 Dec 2020 13:49:18 +0000 (21:49 +0800)]
net: wan: Use DEFINE_SPINLOCK() for spinlock

spinlock can be initialized automatically with DEFINE_SPINLOCK()
rather than explicitly calling spin_lock_init().

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: tipc: Replace expression with offsetof()
Zheng Yongjun [Tue, 29 Dec 2020 13:49:04 +0000 (21:49 +0800)]
net: tipc: Replace expression with offsetof()

Use the existing offsetof() macro instead of duplicating code.

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: wan: Replace simple_strtol by simple_strtoul
Zheng Yongjun [Tue, 29 Dec 2020 13:48:56 +0000 (21:48 +0800)]
net: wan: Replace simple_strtol by simple_strtoul

The simple_strtol() function is deprecated, use simple_strtoul() instead.

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: mhi: Add raw IP mode support
Loic Poulain [Tue, 29 Dec 2020 09:04:54 +0000 (10:04 +0100)]
net: mhi: Add raw IP mode support

MHI net is protocol agnostic, the payload protocol depends on the modem
configuration, which can be either RMNET (IP muxing and aggregation) or
raw IP. This patch adds support for incomming IPv4/IPv6 packets, that
was previously unconditionnaly reported as RMNET packets.

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'enetc-code-cleanups'
David S. Miller [Tue, 5 Jan 2021 23:19:19 +0000 (15:19 -0800)]
Merge branch 'enetc-code-cleanups'

Michael Walle says:

====================
enetc: code cleanups

This are some code cleanups in the MDIO part of the enetc. They are
intended to make the code more readable.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoenetc: reorder macros and functions
Michael Walle [Mon, 28 Dec 2020 13:00:34 +0000 (14:00 +0100)]
enetc: reorder macros and functions

Now that there aren't any more macros with parameters, move the macros
above any functions.

Signed-off-by: Michael Walle <michael@walle.cc>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoenetc: drop MDIO_DATA() macro
Michael Walle [Mon, 28 Dec 2020 13:00:33 +0000 (14:00 +0100)]
enetc: drop MDIO_DATA() macro

value is u16, masking with 0xffff is a nop. Drop it.

Signed-off-by: Michael Walle <michael@walle.cc>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoenetc: don't use macro magic for the readx_poll_timeout() callback
Michael Walle [Mon, 28 Dec 2020 13:00:32 +0000 (14:00 +0100)]
enetc: don't use macro magic for the readx_poll_timeout() callback

The macro enetc_mdio_rd_reg() is just used in that particular case and
has a hardcoded parameter name "mdio_priv". Define a specific function
to use for readx_poll_timeout() instead. Also drop the TIMEOUT macro
since it is used just once.

Signed-off-by: Michael Walle <michael@walle.cc>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoenetc: drop unneeded indirection
Michael Walle [Mon, 28 Dec 2020 13:00:31 +0000 (14:00 +0100)]
enetc: drop unneeded indirection

Before commit 6517798dd343 ("enetc: Make MDIO accessors more generic and
export to include/linux/fsl") these macros actually had some benefits.
But after the commit it just makes the code hard to read. Drop the macro
indirections.

Signed-off-by: Michael Walle <michael@walle.cc>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet/mlx5: DR, Move STEv0 modify header logic
Yevgeny Kliteynik [Thu, 19 Nov 2020 05:15:49 +0000 (07:15 +0200)]
net/mlx5: DR, Move STEv0 modify header logic

Move HW specific modify header fields and logic to STEv0 file
and use the new STE context callbacks.
Since STEv0 and STEv1 modify actions values are different, each
version has its own implementation.

Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Alex Vesker <valex@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Add STE modify header actions per-device API
Yevgeny Kliteynik [Thu, 19 Nov 2020 05:14:27 +0000 (07:14 +0200)]
net/mlx5: DR, Add STE modify header actions per-device API

Extend the STE context struct with per-device modify header actions.

Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Move STEv0 action apply logic
Yevgeny Kliteynik [Thu, 19 Nov 2020 05:03:43 +0000 (07:03 +0200)]
net/mlx5: DR, Move STEv0 action apply logic

Use STE tx/rx actions per-device API: move HW specific
action apply logic from dr_ste to STEv0 file - STEv0 and
STEv1 actions format is different, each version should
have its own implementation.

Signed-off-by: Alex Vesker <valex@nvidia.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Add STE tx/rx actions per-device API
Yevgeny Kliteynik [Thu, 19 Nov 2020 04:48:33 +0000 (06:48 +0200)]
net/mlx5: DR, Add STE tx/rx actions per-device API

Extend the STE context struct with per-device
tx/rx actions.

Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Move STEv0 setters and getters
Yevgeny Kliteynik [Thu, 19 Nov 2020 04:44:45 +0000 (06:44 +0200)]
net/mlx5: DR, Move STEv0 setters and getters

Use the new setters and getters API for STEv0: move HW specific setter and
getters from dr_ste to STEv0 file. Since STEv0 and STEv1 format are
different each version should implemented different setters and getters.
Rename remaining static functions w/o mlx5 prefix.

Signed-off-by: Alex Vesker <valex@nvidia.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Add STE setters and getters per-device API
Yevgeny Kliteynik [Thu, 19 Nov 2020 04:41:57 +0000 (06:41 +0200)]
net/mlx5: DR, Add STE setters and getters per-device API

Extend the STE context struct with various per-device
setters and getters.

Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Move action apply logic to dr_ste
Yevgeny Kliteynik [Tue, 22 Sep 2020 01:22:31 +0000 (04:22 +0300)]
net/mlx5: DR, Move action apply logic to dr_ste

The action apply logic is device specific per STE version,
moving to the STE layer will allow implementing it for
both devices while keeping DR upper layers the same.

Signed-off-by: Alex Vesker <valex@nvidia.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Refactor ICMP STE builder
Yevgeny Kliteynik [Tue, 22 Sep 2020 01:22:24 +0000 (04:22 +0300)]
net/mlx5: DR, Refactor ICMP STE builder

Reworked ICMP tag builder to better handle ICMP v4/6 fields and avoid unneeded
code duplication and 'if' statements, removed unused macro, changed bitfield
of len 8 to u8.

Signed-off-by: Alex Vesker <valex@nvidia.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Move STEv0 look up types from mlx5_ifc_dr header
Yevgeny Kliteynik [Tue, 22 Sep 2020 01:22:12 +0000 (04:22 +0300)]
net/mlx5: DR, Move STEv0 look up types from mlx5_ifc_dr header

The lookup types are device specific and should not be
exposed to DR upper layers, matchers/tables.
Each HW STE version should keep them internal.
The lu_type size is updated to support larger lu_types as
required for STEv1.

Signed-off-by: Alex Vesker <valex@nvidia.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Merge similar DR STE SET macros
Yevgeny Kliteynik [Tue, 22 Sep 2020 01:22:02 +0000 (04:22 +0300)]
net/mlx5: DR, Merge similar DR STE SET macros

Merge DR_STE_STE macros for better code reuse, the macro
DR_STE_SET_MASK_V and DR_STE_SET_TAG are merged to avoid
tag and bit_mask function creation which are usually the
same.

Signed-off-by: Alex Vesker <valex@nvidia.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Fix STEv0 source_eswitch_owner_vhca_id support
Yevgeny Kliteynik [Wed, 28 Oct 2020 23:33:00 +0000 (01:33 +0200)]
net/mlx5: DR, Fix STEv0 source_eswitch_owner_vhca_id support

Check vport_cap only if match on source gvmi is required.

Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Alex Vesker <valex@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Remove unused macro definition from dr_ste
Yevgeny Kliteynik [Thu, 19 Nov 2020 04:20:07 +0000 (06:20 +0200)]
net/mlx5: DR, Remove unused macro definition from dr_ste

Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Move HW STEv0 match logic to a separate file
Yevgeny Kliteynik [Thu, 19 Nov 2020 04:18:01 +0000 (06:18 +0200)]
net/mlx5: DR, Move HW STEv0 match logic to a separate file

Move current STE match logic to a seprate file.
This file will be used for HW specific STEv0.

Future patches will add functionality for v1 steering.

Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Alex Vesker <valex@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Use the new HW specific STE infrastructure
Yevgeny Kliteynik [Thu, 19 Nov 2020 04:03:40 +0000 (06:03 +0200)]
net/mlx5: DR, Use the new HW specific STE infrastructure

Split the STE builders functionality into the common part and
device-specific part. All the device-specific part (with 'v0' in
the function names) is accessed through the STE context structure.

Subsequent patches will have the device-specific logic moved to a
separate file.

Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Move macros from dr_ste.c to header
Yevgeny Kliteynik [Thu, 19 Nov 2020 03:16:38 +0000 (05:16 +0200)]
net/mlx5: DR, Move macros from dr_ste.c to header

Move some macros from dr_ste.c to header - these macros
will be used by all the format-specific functions.

Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: DR, Add infrastructure for supporting several steering formats
Yevgeny Kliteynik [Thu, 19 Nov 2020 02:53:53 +0000 (04:53 +0200)]
net/mlx5: DR, Add infrastructure for supporting several steering formats

Add a struct of device specific callbacks for STE layer below dr_ste.
Each device will implement its HW-specific function, and a comon logic
from the DR code will access these functions through the new ste_ctx API.

More callbacks will follow in the subsequent patches.

Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agoMerge tag 'arc-5.11-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc
Linus Torvalds [Tue, 5 Jan 2021 20:46:27 +0000 (12:46 -0800)]
Merge tag 'arc-5.11-rc3' of git://git./linux/kernel/git/vgupta/arc

Pull ARC updates from Vineet Gupta:
 "Things are quieter on upstreaming front as we are mostly focusing on
  ARCv3/ARC64 port.

  This contains just build system updates from Masahiro Yamada"

* tag 'arc-5.11-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc:
  ARC: build: use $(READELF) instead of hard-coded readelf
  ARC: build: remove unneeded extra-y
  ARC: build: move symlink creation to arch/arc/Makefile to avoid race
  ARC: build: add boot_targets to PHONY
  ARC: build: add uImage.lzma to the top-level target
  ARC: build: remove non-existing bootpImage from KBUILD_IMAGE