platform/kernel/linux-starfive.git
2 years agoaf_vsock: SOCK_SEQPACKET receive timeout test
Krasnov Arseniy Vladimirovich [Thu, 17 Mar 2022 08:31:49 +0000 (08:31 +0000)]
af_vsock: SOCK_SEQPACKET receive timeout test

Test for receive timeout check: connection is established,
receiver sets timeout, but sender does nothing. Receiver's
'read()' call must return EAGAIN.

Signed-off-by: Krasnov Arseniy Vladimirovich <AVKrasnov@sberdevices.ru>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoMerge tag 'wireless-next-2022-03-18' of git://git.kernel.org/pub/scm/linux/kernel...
David S. Miller [Fri, 18 Mar 2022 15:11:31 +0000 (15:11 +0000)]
Merge tag 'wireless-next-2022-03-18' of git://git./linux/kernel/git/wireless/wireless-next

Kalle Valo says:

====================
wireless-next patches for v5.18

Third set of patches for v5.18. Smaller set this time, support for
mt7921u and some work on MBSSID support. Also a workaround for rfkill
userspace event.

Major changes:

mac80211

* MBSSID beacon handling in AP mode

rfkill

* make new event layout opt-in to workaround buggy user space

rtlwifi

* support On Networks N150 device id

mt76

* mt7915: MBSSID and 6 GHz band support

* new driver mt7921u
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoMerge branch 'lan743x-PCI11010-#PCI11414'
David S. Miller [Fri, 18 Mar 2022 12:53:23 +0000 (12:53 +0000)]
Merge branch 'lan743x-PCI11010-#PCI11414'

Raju Lakkaraju says:

====================
net: lan743x: PCI11010 / PCI11414 devices

This patch series continues with the addition of supported features
for the Ethernet function of the PCI11010 / PCI11414 devices to
the LAN743x driver.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: lan743x: Add support for PTP-IO Event Output (Periodic Output)
Raju Lakkaraju [Thu, 17 Mar 2022 10:43:10 +0000 (16:13 +0530)]
net: lan743x: Add support for PTP-IO Event Output (Periodic Output)

Add support for PTP-IO Event Output (Periodic Output - perout) for
PCI11010/PCI11414 chips

Signed-off-by: Raju Lakkaraju <Raju.Lakkaraju@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: lan743x: Add support for PTP-IO Event Input External Timestamp (extts)
Raju Lakkaraju [Thu, 17 Mar 2022 10:43:09 +0000 (16:13 +0530)]
net: lan743x: Add support for PTP-IO Event Input External Timestamp (extts)

PTP-IOs block provides for time stamping PTP-IO input events.
PTP-IOs are numbered from 0 to 11.
When a PTP-IO is enabled by the corresponding bit in the PTP-IO
Capture Configuration Register, a rising or falling edge,
respectively, will capture the 1588 Local Time Counter

Signed-off-by: Raju Lakkaraju <Raju.Lakkaraju@microchip.com>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: lan743x: Add support for OTP
Raju Lakkaraju [Thu, 17 Mar 2022 10:43:08 +0000 (16:13 +0530)]
net: lan743x: Add support for OTP

Add new the OTP read and write access functions for PCI11010/PCI11414 chips
PCI11010/PCI11414 OTP module register offsets are different from
LAN743x OTP module

Signed-off-by: Raju Lakkaraju <Raju.Lakkaraju@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: lan743x: Add support for EEPROM
Raju Lakkaraju [Thu, 17 Mar 2022 10:43:07 +0000 (16:13 +0530)]
net: lan743x: Add support for EEPROM

Add new the EEPROM read and write access functions and system lock
protection to access by devices for PCI11010/PCI11414 chips

Signed-off-by: Raju Lakkaraju <Raju.Lakkaraju@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: lan743x: Add support to display Tx Queue statistics
Raju Lakkaraju [Thu, 17 Mar 2022 10:43:06 +0000 (16:13 +0530)]
net: lan743x: Add support to display Tx Queue statistics

Tx 4 queue statistics display through ethtool application

Signed-off-by: Raju Lakkaraju <Raju.Lakkaraju@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agorfkill: make new event layout opt-in
Johannes Berg [Wed, 16 Mar 2022 20:27:51 +0000 (21:27 +0100)]
rfkill: make new event layout opt-in

Again new complaints surfaced that we had broken the ABI here,
although previously all the userspace tools had agreed that it
was their mistake and fixed it. Yet now there are cases (e.g.
RHEL) that want to run old userspace with newer kernels, and
thus are broken.

Since this is a bit of a whack-a-mole thing, change the whole
extensibility scheme of rfkill to no longer just rely on the
message lengths, but instead require userspace to opt in via a
new ioctl to a given maximum event size that it is willing to
understand.

By default, set that to RFKILL_EVENT_SIZE_V1 (8), so that the
behaviour for userspace not calling the ioctl will look as if
it's just running on an older kernel.

Fixes: 14486c82612a ("rfkill: add a reason to the HW rfkill state")
Cc: stable@vger.kernel.org # 5.11+
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220316212749.16491491b270.Ifcb1950998330a596f29a2a162e00b7546a1d6d0@changeid
2 years agoMerge tag 'mlx5-updates-2022-03-17' of git://git.kernel.org/pub/scm/linux/kernel...
David S. Miller [Fri, 18 Mar 2022 10:30:00 +0000 (10:30 +0000)]
Merge tag 'mlx5-updates-2022-03-17' of git://git./linux/kernel/git/saeed/linux

Saeed Mahameed says:

====================
mlx5-updates-2022-03-17

1) From Maxim Mikityanskiy,
   Datapath improvements in preparation for XDP multi buffer

   This series contains general improvements for the datapath that are
   useful for the upcoming XDP multi buffer support:

   a. Non-linear legacy RQ: validate MTU for robustness, build the linear
      part of SKB over the first hardware fragment (instead of copying the
      packet headers), adjust headroom calculations to allow enabling headroom
      in the non-linear mode (useful for XDP multi buffer).

   b. XDP: do the XDP program test before function call, optimize
      parameters of mlx5e_xdp_handle.

2) From Rongwei Liu, DR, reduce steering memory usage
   Currently, mlx5 driver uses mlx5_htbl/chunk/ste to organize
   steering logic. However there is a little memory waste.

   This update targets to reduce steering memory footprint by:
   a. Adjust struct member layout.
   b. Remove duplicated indicator by using simple functions call.

   With 500k TX rules(3 ste) plus 500k RX rules(6 stes), these patches
   can save around 17% memory.

3) Three cleanup commits at the end of this series.
===================

Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoMerge branch 'mirroring-for-ocelot-switches'
Jakub Kicinski [Fri, 18 Mar 2022 00:42:50 +0000 (17:42 -0700)]
Merge branch 'mirroring-for-ocelot-switches'

Vladimir Oltean says:

====================
Mirroring for Ocelot switches

This series adds support for tc-matchall (port-based) and tc-flower
(flow-based) offloading of the tc-mirred action. Support has been added
for both the ocelot switchdev driver and felix DSA driver.
====================

Link: https://lore.kernel.org/r/20220316204144.2679277-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: dsa: felix: add port mirroring support
Vladimir Oltean [Wed, 16 Mar 2022 20:41:44 +0000 (22:41 +0200)]
net: dsa: felix: add port mirroring support

Gain support for port mirroring using tc-matchall by forwarding the
calls to the ocelot switch library.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: dsa: pass extack to dsa_switch_ops :: port_mirror_add()
Vladimir Oltean [Wed, 16 Mar 2022 20:41:43 +0000 (22:41 +0200)]
net: dsa: pass extack to dsa_switch_ops :: port_mirror_add()

Drivers might have error messages to propagate to user space, most
common being that they support a single mirror port.

Propagate the netlink extack so that they can inform user space in a
verbal way of their limitations.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: mscc: ocelot: offload per-flow mirroring using tc-mirred and VCAP IS2
Vladimir Oltean [Wed, 16 Mar 2022 20:41:42 +0000 (22:41 +0200)]
net: mscc: ocelot: offload per-flow mirroring using tc-mirred and VCAP IS2

Per-flow mirroring with the VCAP IS2 TCAM (in itself handled as an
offload for tc-flower) is done by setting the MIRROR_ENA bit from the
action vector of the filter. The packet is mirrored to the port mask
configured in the ANA:ANA:MIRRORPORTS register (the same port mask as
the destinations for port-based mirroring).

Functionality was tested with:

tc qdisc add dev swp3 clsact
tc filter add dev swp3 ingress protocol ip \
flower skip_sw ip_proto icmp \
action mirred egress mirror dev swp1

and pinging through swp3, while seeing that the ICMP replies are
mirrored towards swp1.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: mscc: ocelot: establish functions for handling VCAP aux resources
Vladimir Oltean [Wed, 16 Mar 2022 20:41:41 +0000 (22:41 +0200)]
net: mscc: ocelot: establish functions for handling VCAP aux resources

Some VCAP filters utilize resources which are global to the switch, like
for example VCAP IS2 policers take an index into a global policer pool.

In commit c9a7fe1238e5 ("net: mscc: ocelot: add action of police on
vcap_is2"), Xiaoliang expressed this by hooking into the low-level
ocelot_vcap_filter_add_to_block() and ocelot_vcap_block_remove_filter()
functions, and allocating/freeing the policers from there.

Evaluating the code, there probably isn't a better place, but we'll need
to do something similar for the mirror ports, and the code will start to
look even more hacked up than it is right now.

Create two ocelot_vcap_filter_{add,del}_aux_resources() functions to
contain the madness, and pollute less the body of other functions such
as ocelot_vcap_filter_add_to_block().

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: mscc: ocelot: add port mirroring support using tc-matchall
Vladimir Oltean [Wed, 16 Mar 2022 20:41:40 +0000 (22:41 +0200)]
net: mscc: ocelot: add port mirroring support using tc-matchall

Ocelot switches perform port-based ingress mirroring if
ANA:PORT:PORT_CFG field SRC_MIRROR_ENA is set, and egress mirroring if
the port is in ANA:ANA:EMIRRORPORTS.

Both ingress-mirrored and egress-mirrored frames are copied to the port
mask from ANA:ANA:MIRRORPORTS.

So the choice of limiting to a single mirror port via ocelot_mirror_get()
and ocelot_mirror_put() may seem bizarre, but the hardware model doesn't
map very well to the user space model. If the user wants to mirror the
ingress of swp1 towards swp2 and the ingress of swp3 towards swp4, we'd
have to program ANA:ANA:MIRRORPORTS with BIT(2) | BIT(4), and that would
make swp1 be mirrored towards swp4 too, and swp3 towards swp2. But there
are no tc-matchall rules to describe those actions.

Now, we could offload a matchall rule with multiple mirred actions, one
per desired mirror port, and force the user to stick to the multi-action
rule format for subsequent matchall filters. But both DSA and ocelot
have the flow_offload_has_one_action() check for the matchall offload,
plus the fact that it will get cumbersome to cross-check matchall
mirrors with flower mirrors (which will be added in the next patch).

As a result, we limit the configuration to a single mirror port, with
the possibility of lifting the restriction in the future.

Frames injected from the CPU don't get egress-mirrored, since they are
sent with the BYPASS bit in the injection frame header, and this
bypasses the analyzer module (effectively also the mirroring logic).
I don't know what to do/say about this.

Functionality was tested with:

tc qdisc add dev swp3 clsact
tc filter add dev swp3 ingress \
matchall skip_sw \
action mirred egress mirror dev swp1

and pinging through swp3, while seeing that the ICMP replies are
mirrored towards swp1.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: mscc: ocelot: refactor policer work out of ocelot_setup_tc_cls_matchall
Vladimir Oltean [Wed, 16 Mar 2022 20:41:39 +0000 (22:41 +0200)]
net: mscc: ocelot: refactor policer work out of ocelot_setup_tc_cls_matchall

In preparation for adding port mirroring support to the ocelot driver,
the dispatching function ocelot_setup_tc_cls_matchall() must be free of
action-specific code. Move port policer creation and deletion to
separate functions.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoptp: ocp: Make debugfs variables the correct bitwidth
Jonathan Lemon [Wed, 16 Mar 2022 16:53:47 +0000 (09:53 -0700)]
ptp: ocp: Make debugfs variables the correct bitwidth

An earlier patch mistakenly changed these variables from u32 to u16,
leading to unintended truncation.  Restore the original logic.

Fixes: a509a7c61e3b ("ptp: ocp: Add support for selectable SMA directions.")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Link: https://lore.kernel.org/r/20220316165347.599154-1-jonathan.lemon@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: dsa: microchip: ksz8795: handle eee specif erratum
Oleksij Rempel [Wed, 16 Mar 2022 12:55:29 +0000 (13:55 +0100)]
net: dsa: microchip: ksz8795: handle eee specif erratum

According to erratum described in DS80000687C[1]: "Module 2: Link drops with
some EEE link partners.", we need to "Disable the EEE next page
exchange in EEE Global Register 2"

1 - https://ww1.microchip.com/downloads/en/DeviceDoc/KSZ87xx-Errata-DS80000687C.pdf

Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Link: https://lore.kernel.org/r/20220316125529.1489045-1-o.rempel@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoMerge branch 'net-bridge-multiple-spanning-trees'
Jakub Kicinski [Thu, 17 Mar 2022 23:50:03 +0000 (16:50 -0700)]
Merge branch 'net-bridge-multiple-spanning-trees'

Tobias Waldekranz says:

====================
net: bridge: Multiple Spanning Trees

The bridge has had per-VLAN STP support for a while now, since:

https://lore.kernel.org/netdev/20200124114022.10883-1-nikolay@cumulusnetworks.com/

The current implementation has some problems:

- The mapping from VLAN to STP state is fixed as 1:1, i.e. each VLAN
  is managed independently. This is awkward from an MSTP (802.1Q-2018,
  Clause 13.5) point of view, where the model is that multiple VLANs
  are grouped into MST instances.

  Because of the way that the standard is written, presumably, this is
  also reflected in hardware implementations. It is not uncommon for a
  switch to support the full 4k range of VIDs, but that the pool of
  MST instances is much smaller. Some examples:

  Marvell LinkStreet (mv88e6xxx): 4k VLANs, but only 64 MSTIs
  Marvell Prestera: 4k VLANs, but only 128 MSTIs
  Microchip SparX-5i: 4k VLANs, but only 128 MSTIs

- By default, the feature is enabled, and there is no way to disable
  it. This makes it hard to add offloading in a backwards compatible
  way, since any underlying switchdevs have no way to refuse the
  function if the hardware does not support it

- The port-global STP state has precedence over per-VLAN states. In
  MSTP, as far as I understand it, all VLANs will use the common
  spanning tree (CST) by default - through traffic engineering you can
  then optimize your network to group subsets of VLANs to use
  different trees (MSTI). To my understanding, the way this is
  typically managed in silicon is roughly:

  Incoming packet:
  .----.----.--------------.----.-------------
  | DA | SA | 802.1Q VID=X | ET | Payload ...
  '----'----'--------------'----'-------------
                        |
                        '->|\     .----------------------------.
                           | +--> | VID | Members | ... | MSTI |
                   PVID -->|/     |-----|---------|-----|------|
                                  |   1 | 0001001 | ... |    0 |
                                  |   2 | 0001010 | ... |   10 |
                                  |   3 | 0001100 | ... |   10 |
                                  '----------------------------'
                                                             |
                               .-----------------------------'
                               |  .------------------------.
                               '->| MSTI | Fwding | Lrning |
                                  |------|--------|--------|
                                  |    0 | 111110 | 111110 |
                                  |   10 | 110111 | 110111 |
                                  '------------------------'

  What this is trying to show is that the STP state (whether MSTP is
  used, or ye olde STP) is always accessed via the VLAN table. If STP
  is running, all MSTI pointers in that table will reference the same
  index in the STP stable - if MSTP is running, some VLANs may point
  to other trees (like in this example).

  The fact that in the Linux bridge, the global state (think: index 0
  in most hardware implementations) is supposed to override the
  per-VLAN state, is very awkward to offload. In effect, this means
  that when the global state changes to blocking, drivers will have to
  iterate over all MSTIs in use, and alter them all to match. This
  also means that you have to cache whether the hardware state is
  currently tracking the global state or the per-VLAN state. In the
  first case, you also have to cache the per-VLAN state so that you
  can restore it if the global state transitions back to forwarding.

This series adds a new mst_enable bridge setting (as suggested by Nik)
that can only be changed when no VLANs are configured on the
bridge. Enabling this mode has the following effect:

- The port-global STP state is used to represent the CST (Common
  Spanning Tree) (1/15)

- Ingress STP filtering is deferred until the frame's VLAN has been
  resolved (1/15)

- The preexisting per-VLAN states can no longer be controlled directly
  (1/15). They are instead placed under the MST module's control,
  which is managed using a new netlink interface (described in 3/15)

- VLANs can br mapped to MSTIs in an arbitrary M:N fashion, using a
  new global VLAN option (2/15)

Switchdev notifications are added so that a driver can track:
- MST enabled state
- VID to MSTI mappings
- MST port states

An offloading implementation is this provided for mv88e6xxx.
====================

Link: https://lore.kernel.org/r/20220316150857.2442916-1-tobias@waldekranz.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: dsa: mv88e6xxx: MST Offloading
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:57 +0000 (16:08 +0100)]
net: dsa: mv88e6xxx: MST Offloading

Allocate a SID in the STU for each MSTID in use by a bridge and handle
the mapping of MSTIDs to VLANs using the SID field of each VTU entry.

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: dsa: mv88e6xxx: Export STU as devlink region
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:56 +0000 (16:08 +0100)]
net: dsa: mv88e6xxx: Export STU as devlink region

Export the raw STU data in a devlink region so that it can be
inspected from userspace and compared to the current bridge
configuration.

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: dsa: mv88e6xxx: Disentangle STU from VTU
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:55 +0000 (16:08 +0100)]
net: dsa: mv88e6xxx: Disentangle STU from VTU

In early LinkStreet silicon (e.g. 6095/6185), the per-VLAN STP states
were kept in the VTU - there was no concept of a SID. Later, the
information was split into two tables, where the VTU only tracked
memberships and deferred the STP state tracking to the STU via a
pointer (SID). This meant that a group of VLANs could share the same
STU entry. Most likely, this was done to align with MSTP (802.1Q-2018,
Clause 13), which is built on this principle.

While the VTU is still 4k lines on most devices, the STU is capped at
64 entries. This means that the current stategy, updating STU info
whenever a VTU entry is updated, can not easily support MSTP because:

- The maximum number of VIDs would also be capped at 64, as we would
  have to allocate one SID for every VTU entry - even if many VLANs
  would effectively share the same MST.

- MSTP updates would be unnecessarily slow as you would have to
  iterate over all VLANs that share the same MST.

In order to support MSTP offloading in the future, manage the STU as a
separate entity from the VTU.

Only add support for newer hardware with separate VTU and
STU. VTU-only devices can also be supported, but essentially this
requires a software implementation of an STU (fanning out state
changed to all VLANs tied to the same MST).

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: dsa: Handle MST state changes
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:54 +0000 (16:08 +0100)]
net: dsa: Handle MST state changes

Add the usual trampoline functionality from the generic DSA layer down
to the drivers for MST state changes.

When a state changes to disabled/blocking/listening, make sure to fast
age any dynamic entries in the affected VLANs (those controlled by the
MSTI in question).

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: dsa: Pass VLAN MSTI migration notifications to driver
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:53 +0000 (16:08 +0100)]
net: dsa: Pass VLAN MSTI migration notifications to driver

Add the usual trampoline functionality from the generic DSA layer down
to the drivers for VLAN MSTI migrations.

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: dsa: Validate hardware support for MST
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:52 +0000 (16:08 +0100)]
net: dsa: Validate hardware support for MST

When joining a bridge where MST is enabled, we validate that the
proper offloading support is in place, otherwise we fallback to
software bridging.

When then mode is changed on a bridge in which we are members, we
refuse the change if offloading is not supported.

At the moment we only check for configurable learning, but this will
be further restricted as we support more MST related switchdev events.

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: bridge: mst: Add helper to query a port's MST state
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:51 +0000 (16:08 +0100)]
net: bridge: mst: Add helper to query a port's MST state

This is useful for switchdev drivers who are offloading MST states
into hardware. As an example, a driver may wish to flush the FDB for a
port when it transitions from forwarding to blocking - which means
that the previous state must be discoverable.

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: bridge: mst: Add helper to check if MST is enabled
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:50 +0000 (16:08 +0100)]
net: bridge: mst: Add helper to check if MST is enabled

This is useful for switchdev drivers that might want to refuse to join
a bridge where MST is enabled, if the hardware can't support it.

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: bridge: mst: Add helper to map an MSTI to a VID set
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:49 +0000 (16:08 +0100)]
net: bridge: mst: Add helper to map an MSTI to a VID set

br_mst_get_info answers the question: "On this bridge, which VIDs are
mapped to the given MSTI?"

This is useful in switchdev drivers, which might have to fan-out
operations, relating to an MSTI, per VLAN.

An example: When a port's MST state changes from forwarding to
blocking, a driver may choose to flush the dynamic FDB entries on that
port to get faster reconvergence of the network, but this should only
be done in the VLANs that are managed by the MSTI in question.

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: bridge: mst: Notify switchdev drivers of MST state changes
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:48 +0000 (16:08 +0100)]
net: bridge: mst: Notify switchdev drivers of MST state changes

Generate a switchdev notification whenever an MST state changes. This
notification is keyed by the VLANs MSTI rather than the VID, since
multiple VLANs may share the same MST instance.

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: bridge: mst: Notify switchdev drivers of VLAN MSTI migrations
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:47 +0000 (16:08 +0100)]
net: bridge: mst: Notify switchdev drivers of VLAN MSTI migrations

Whenever a VLAN moves to a new MSTI, send a switchdev notification so
that switchdevs can track a bridge's VID to MSTI mappings.

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: bridge: mst: Notify switchdev drivers of MST mode changes
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:46 +0000 (16:08 +0100)]
net: bridge: mst: Notify switchdev drivers of MST mode changes

Trigger a switchdev event whenever the bridge's MST mode is
enabled/disabled. This allows constituent ports to either perform any
required hardware config, or refuse the change if it not supported.

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: bridge: mst: Support setting and reporting MST port states
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:45 +0000 (16:08 +0100)]
net: bridge: mst: Support setting and reporting MST port states

Make it possible to change the port state in a given MSTI by extending
the bridge port netlink interface (RTM_SETLINK on PF_BRIDGE).The
proposed iproute2 interface would be:

    bridge mst set dev <PORT> msti <MSTI> state <STATE>

Current states in all applicable MSTIs can also be dumped via a
corresponding RTM_GETLINK. The proposed iproute interface looks like
this:

$ bridge mst
port              msti
vb1               0
    state forwarding
  100
    state disabled
vb2               0
    state forwarding
  100
    state forwarding

The preexisting per-VLAN states are still valid in the MST
mode (although they are read-only), and can be queried as usual if one
is interested in knowing a particular VLAN's state without having to
care about the VID to MSTI mapping (in this example VLAN 20 and 30 are
bound to MSTI 100):

$ bridge -d vlan
port              vlan-id
vb1               10
    state forwarding mcast_router 1
  20
    state disabled mcast_router 1
  30
    state disabled mcast_router 1
  40
    state forwarding mcast_router 1
vb2               10
    state forwarding mcast_router 1
  20
    state forwarding mcast_router 1
  30
    state forwarding mcast_router 1
  40
    state forwarding mcast_router 1

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: bridge: mst: Allow changing a VLAN's MSTI
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:44 +0000 (16:08 +0100)]
net: bridge: mst: Allow changing a VLAN's MSTI

Allow a VLAN to move out of the CST (MSTI 0), to an independent tree.

The user manages the VID to MSTI mappings via a global VLAN
setting. The proposed iproute2 interface would be:

    bridge vlan global set dev br0 vid <VID> msti <MSTI>

Changing the state in non-zero MSTIs is still not supported, but will
be addressed in upcoming changes.

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: bridge: mst: Multiple Spanning Tree (MST) mode
Tobias Waldekranz [Wed, 16 Mar 2022 15:08:43 +0000 (16:08 +0100)]
net: bridge: mst: Multiple Spanning Tree (MST) mode

Allow the user to switch from the current per-VLAN STP mode to an MST
mode.

Up to this point, per-VLAN STP states where always isolated from each
other. This is in contrast to the MSTP standard (802.1Q-2018, Clause
13.5), where VLANs are grouped into MST instances (MSTIs), and the
state is managed on a per-MSTI level, rather that at the per-VLAN
level.

Perhaps due to the prevalence of the standard, many switching ASICs
are built after the same model. Therefore, add a corresponding MST
mode to the bridge, which we can later add offloading support for in a
straight-forward way.

For now, all VLANs are fixed to MSTI 0, also called the Common
Spanning Tree (CST). That is, all VLANs will follow the port-global
state.

Upcoming changes will make this actually useful by allowing VLANs to
be mapped to arbitrary MSTIs and allow individual MSTI states to be
changed.

Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agor8169: improve driver unload and system shutdown behavior on DASH-enabled systems
Heiner Kallweit [Wed, 16 Mar 2022 21:31:00 +0000 (22:31 +0100)]
r8169: improve driver unload and system shutdown behavior on DASH-enabled systems

There's a number of systems supporting DASH remote management.
Driver unload and system shutdown can result in the PHY suspending,
thus making DASH unusable. Improve this by handling DASH being enabled
very similar to WoL being enabled.

Tested-by: Yanko Kaneti <yaneti@declera.com>
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://lore.kernel.org/r/1de3b176-c09c-1654-6f00-9785f7a4f954@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoMerge branch '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next...
Jakub Kicinski [Thu, 17 Mar 2022 23:40:32 +0000 (16:40 -0700)]
Merge branch '100GbE' of git://git./linux/kernel/git/tnguy/next-queue

Tony Nguyen says:

====================
100GbE Intel Wired LAN Driver Updates 2022-03-16

This series contains updates to gtp and ice driver.

Wojciech fixes smatch reported inconsistent indenting for gtp and ice.

Yang Yingliang fixes a couple of return value checks for GNSS to IS_PTR
instead of null.

Jacob adds support for trace events on tx timestamps.

* '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue:
  ice: add trace events for tx timestamps
  ice: fix return value check in ice_gnss.c
  ice: Fix inconsistent indenting in ice_switch
  gtp: Fix inconsistent indenting
====================

Link: https://lore.kernel.org/r/20220316204024.3201500-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoethernet: sun: Fix spelling mistake "mis-matched" -> "mismatched"
Colin Ian King [Wed, 16 Mar 2022 23:46:20 +0000 (23:46 +0000)]
ethernet: sun: Fix spelling mistake "mis-matched" -> "mismatched"

There is a spelling mistake in a dev_err message. Fix it.

Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Link: https://lore.kernel.org/r/20220316234620.55885-1-colin.i.king@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: ethernet: ti: Fix spelling mistake and clean up message
Colin Ian King [Wed, 16 Mar 2022 23:34:55 +0000 (23:34 +0000)]
net: ethernet: ti: Fix spelling mistake and clean up message

There is a spelling mistake in a dev_err message and the MAX_SKB_FRAGS
value does not need to be printed between parentheses. Fix this.

Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Link: https://lore.kernel.org/r/20220316233455.54541-1-colin.i.king@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agovlan: use correct format characters
Bill Wendling [Wed, 16 Mar 2022 21:31:25 +0000 (14:31 -0700)]
vlan: use correct format characters

When compiling with -Wformat, clang emits the following warning:

net/8021q/vlanproc.c:284:22: warning: format specifies type 'unsigned
short' but the argument has type 'int' [-Wformat]
                                   mp->priority, ((mp->vlan_qos >> 13) & 0x7));
                                                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~

The types of these arguments are unconditionally defined, so this patch
updates the format character to the correct ones for ints and unsigned
ints.

Link: https://github.com/ClangBuiltLinux/linux/issues/378
Signed-off-by: Bill Wendling <morbo@google.com>
Link: https://lore.kernel.org/r/20220316213125.2353370-1-morbo@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet/fsl: xgmac_mdio: use correct format characters
Bill Wendling [Wed, 16 Mar 2022 21:31:14 +0000 (14:31 -0700)]
net/fsl: xgmac_mdio: use correct format characters

When compiling with -Wformat, clang emits the following warning:

drivers/net/ethernet/freescale/xgmac_mdio.c:243:22: warning: format
specifies type 'unsigned char' but the argument has type 'int'
[-Wformat]
                        phy_id, dev_addr, regnum);
                                          ^~~~~~
./include/linux/dev_printk.h:163:47: note: expanded from macro 'dev_dbg'
                dev_printk(KERN_DEBUG, dev, dev_fmt(fmt), ##__VA_ARGS__); \
                                                    ~~~     ^~~~~~~~~~~
./include/linux/dev_printk.h:129:34: note: expanded from macro 'dev_printk'
                _dev_printk(level, dev, fmt, ##__VA_ARGS__);            \
                                        ~~~    ^~~~~~~~~~~

The types of these arguments are unconditionally defined, so this patch
updates the format character to the correct ones for ints and unsigned
ints.

Link: https://github.com/ClangBuiltLinux/linux/issues/378
Signed-off-by: Bill Wendling <morbo@google.com>
Link: https://lore.kernel.org/r/20220316213114.2352352-1-morbo@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agobnx2x: use correct format characters
Bill Wendling [Wed, 16 Mar 2022 21:31:04 +0000 (14:31 -0700)]
bnx2x: use correct format characters

When compiling with -Wformat, clang emits the following warnings:

drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c:6181:40: warning: format
specifies type 'unsigned short' but the argument has type 'u32'
(aka 'unsigned int') [-Wformat]
        ret = scnprintf(str, *len, "%hx.%hx", num >> 16, num);
                                    ~~~       ^~~~~~~~~
                                    %x
drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c:6181:51: warning: format
specifies type 'unsigned short' but the argument has type 'u32'
(aka 'unsigned int') [-Wformat]
        ret = scnprintf(str, *len, "%hx.%hx", num >> 16, num);
                                        ~~~              ^~~
                                        %x
drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c:6196:47: warning: format
specifies type 'unsigned char' but the argument has type 'u32'
(aka 'unsigned int') [-Wformat]
        ret = scnprintf(str, *len, "%hhx.%hhx.%hhx", num >> 16, num >> 8, num);
                                    ~~~~             ^~~~~~~~~
                                    %x
drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c:6196:58: warning: format
specifies type 'unsigned char' but the argument has type 'u32'
(aka 'unsigned int') [-Wformat]
        ret = scnprintf(str, *len, "%hhx.%hhx.%hhx", num >> 16, num >> 8, num);
                                         ~~~~                   ^~~~~~~~
                                         %x
drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c:6196:68: warning: format
specifies type 'unsigned char' but the argument has type 'u32'
(aka 'unsigned int') [-Wformat]
        ret = scnprintf(str, *len, "%hhx.%hhx.%hhx", num >> 16, num >> 8, num);
                                              ~~~~                        ^~~
                                              %x

The types of these arguments are unconditionally defined, so this patch
updates the format character to the correct ones for ints and unsigned
ints.

Link: https://github.com/ClangBuiltLinux/linux/issues/378
Signed-off-by: Bill Wendling <morbo@google.com>
Link: https://lore.kernel.org/r/20220316213104.2351651-1-morbo@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoenetc: use correct format characters
Bill Wendling [Wed, 16 Mar 2022 21:31:09 +0000 (14:31 -0700)]
enetc: use correct format characters

When compiling with -Wformat, clang emits the following warning:

drivers/net/ethernet/freescale/enetc/enetc_mdio.c:151:22: warning:
format specifies type 'unsigned char' but the argument has type 'int'
[-Wformat]
                        phy_id, dev_addr, regnum);
                                          ^~~~~~
./include/linux/dev_printk.h:163:47: note: expanded from macro 'dev_dbg'
                dev_printk(KERN_DEBUG, dev, dev_fmt(fmt), ##__VA_ARGS__); \
                                                    ~~~     ^~~~~~~~~~~
./include/linux/dev_printk.h:129:34: note: expanded from macro 'dev_printk'
                _dev_printk(level, dev, fmt, ##__VA_ARGS__);            \
                                        ~~~    ^~~~~~~~~~~

The types of these arguments are unconditionally defined, so this patch
updates the format character to the correct ones for ints and unsigned
ints.

Link: https://github.com/ClangBuiltLinux/linux/issues/378
Signed-off-by: Bill Wendling <morbo@google.com>
Reviewed-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Link: https://lore.kernel.org/r/20220316213109.2352015-1-morbo@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Jakub Kicinski [Thu, 17 Mar 2022 20:56:06 +0000 (13:56 -0700)]
Merge git://git./linux/kernel/git/netdev/net

No conflicts.

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoMerge tag 'net-5.17-final' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Linus Torvalds [Thu, 17 Mar 2022 19:55:26 +0000 (12:55 -0700)]
Merge tag 'net-5.17-final' of git://git./linux/kernel/git/netdev/net

Pull networking fixes from Jakub Kicinski:
 "Including fixes from netfilter, ipsec, and wireless.

  A few last minute revert / disable and fix patches came down from our
  sub-trees. We're not waiting for any fixes at this point.

  Current release - regressions:

   - Revert "netfilter: nat: force port remap to prevent shadowing
     well-known ports", restore working conntrack on asymmetric paths

   - Revert "ath10k: drop beacon and probe response which leak from
     other channel", restore working AP and mesh mode on QCA9984

   - eth: intel: fix hang during reboot/shutdown

  Current release - new code bugs:

   - netfilter: nf_tables: disable register tracking, it needs more work
     to cover all corner cases

  Previous releases - regressions:

   - ipv6: fix skb_over_panic in __ip6_append_data when (admin-only)
     extension headers get specified

   - esp6: fix ESP over TCP/UDP, interpret ipv6_skip_exthdr's return
     value more selectively

   - bnx2x: fix driver load failure when FW not present in initrd

  Previous releases - always broken:

   - vsock: stop destroying unrelated sockets in nested virtualization

   - packet: fix slab-out-of-bounds access in packet_recvmsg()

  Misc:

   - add Paolo Abeni to networking maintainers!"

* tag 'net-5.17-final' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (26 commits)
  iavf: Fix hang during reboot/shutdown
  net: mscc: ocelot: fix backwards compatibility with single-chain tc-flower offload
  net: bcmgenet: skip invalid partial checksums
  bnx2x: fix built-in kernel driver load failure
  net: phy: mscc: Add MODULE_FIRMWARE macros
  net: dsa: Add missing of_node_put() in dsa_port_parse_of
  net: handle ARPHRD_PIMREG in dev_is_mac_header_xmit()
  Revert "ath10k: drop beacon and probe response which leak from other channel"
  hv_netvsc: Add check for kvmalloc_array
  iavf: Fix double free in iavf_reset_task
  ice: destroy flow director filter mutex after releasing VSIs
  ice: fix NULL pointer dereference in ice_update_vsi_tx_ring_stats()
  Add Paolo Abeni to networking maintainers
  atm: eni: Add check for dma_map_single
  net/packet: fix slab-out-of-bounds access in packet_recvmsg()
  net: mdio: mscc-miim: fix duplicate debugfs entry
  net: phy: marvell: Fix invalid comparison in the resume and suspend functions
  esp6: fix check on ipv6_skip_exthdr's return value
  net: dsa: microchip: add spi_device_id tables
  netfilter: nf_tables: disable register tracking
  ...

2 years agoMerge tag 'acpi-5.17-rc9' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael...
Linus Torvalds [Thu, 17 Mar 2022 19:40:59 +0000 (12:40 -0700)]
Merge tag 'acpi-5.17-rc9' of git://git./linux/kernel/git/rafael/linux-pm

Pull ACPI fix from Rafael Wysocki:
 "Revert recent commit that caused multiple systems to misbehave due to
  firmware issues"

* tag 'acpi-5.17-rc9' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  Revert "ACPI: scan: Do not add device IDs from _CID if _HID is not valid"

2 years agoMerge branch 'akpm' (patches from Andrew)
Linus Torvalds [Thu, 17 Mar 2022 19:36:47 +0000 (12:36 -0700)]
Merge branch 'akpm' (patches from Andrew)

Merge misc fixes from Andrew Morton:
 "Four patches.

  Subsystems affected by this patch series: mm/swap, kconfig, ocfs2, and
  selftests"

* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
  selftests: vm: fix clang build error multiple output files
  ocfs2: fix crash when initialize filecheck kobj fails
  configs/debug: restore DEBUG_INFO=y for overriding
  mm: swap: get rid of livelock in swapin readahead

2 years agonet/mlx5: Remove unused fill page array API function
Tariq Toukan [Thu, 20 May 2021 12:34:57 +0000 (15:34 +0300)]
net/mlx5: Remove unused fill page array API function

mlx5_fill_page_array API function is not used.
Remove it, reduce the number of exported functions.

Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: Remove unused exported contiguous coherent buffer allocation API
Tariq Toukan [Tue, 27 Apr 2021 13:30:32 +0000 (16:30 +0300)]
net/mlx5: Remove unused exported contiguous coherent buffer allocation API

All WQ types moved to using the fragmented allocation API
for coherent memory. Contiguous API is not used anymore.
Remove it, reduce the number of exported functions.

Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: CT: Remove extra rhashtable remove on tuple entries
Paul Blakey [Wed, 23 Feb 2022 13:37:50 +0000 (15:37 +0200)]
net/mlx5: CT: Remove extra rhashtable remove on tuple entries

On tuple offload del command, tuples are tried to be removed twice
from the hashtable, once directly via mlx5_tc_ct_entry_remove_from_tuples()
and a second time in the following mlx5_tc_ct_entry_put()->
mlx5_tc_ct_entry_del()->mlx5_tc_ct_entry_remove_from_tuples() call.

This doesn't cause any issue since rhashtable first checks if the
removed object exists in the hashtable.

Remove the extra mlx5_tc_ct_entry_remove_from_tuples().

Signed-off-by: Paul Blakey <paulb@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: DR, Remove hw_ste from mlx5dr_ste to reduce memory
Rongwei Liu [Fri, 28 Jan 2022 04:15:12 +0000 (06:15 +0200)]
net/mlx5: DR, Remove hw_ste from mlx5dr_ste to reduce memory

It can be calculated via function mlx5dr_ste_get_hw_ste().
Very simple and lightweight, no need to use a dedicated member.

Reduce 8 bytes from struct mlx5dr_ste and its size is 48 bytes now.

Signed-off-by: Rongwei Liu <rongweil@nvidia.com>
Reviewed-by: Shun Hao <shunh@nvidia.com>
Reviewed-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: DR, Remove 4 members from mlx5dr_ste_htbl to reduce memory
Rongwei Liu [Fri, 28 Jan 2022 04:14:16 +0000 (06:14 +0200)]
net/mlx5: DR, Remove 4 members from mlx5dr_ste_htbl to reduce memory

Remove chunk_size in struct mlx5dr_icm_chunk and use
chunk->size instead.

Remove ste_arr/hw_ste_arr/miss_list since they can be accessed
from htbl->chunk pointer, no need to keep a copy.

This commit reduces 28 bytes from struct mlx5dr_ste_htbl and its
size is 32 bytes now.

Signed-off-by: Rongwei Liu <rongweil@nvidia.com>
Reviewed-by: Shun Hao <shunh@nvidia.com>
Reviewed-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: DR, Remove num_of_entries byte_size from struct mlx5_dr_icm_chunk
Rongwei Liu [Fri, 28 Jan 2022 04:13:07 +0000 (06:13 +0200)]
net/mlx5: DR, Remove num_of_entries byte_size from struct mlx5_dr_icm_chunk

Target to reduce the memory consumption in large scale of flow rules.

They can be calculated quickly from buddy memory pool.
1. num_of_entries calls dr_icm_pool_get_chunk_num_of_entries().
2. byte_size calls dr_icm_pool_get_chunk_byte_size().

Use chunk size in dr_icm_chunk to speed up and the one in dr_ste_htbl
will be removed in the upcoming commit.

This commit reduce 8 bytes from struct mlx5_dr_icm_chunk and its
current size is 56 bytes.

Signed-off-by: Rongwei Liu <rongweil@nvidia.com>
Reviewed-by: Shun Hao <shunh@nvidia.com>
Reviewed-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: DR, Remove icm_addr from mlx5dr_icm_chunk to reduce memory
Rongwei Liu [Fri, 28 Jan 2022 04:12:09 +0000 (06:12 +0200)]
net/mlx5: DR, Remove icm_addr from mlx5dr_icm_chunk to reduce memory

It can be calculated quickly from buddy memory pool by
function mlx5dr_icm_pool_get_chunk_icm_addr().
This function is very lightweight and straightforward.

Reduce 8 bytes and current size of struct mlx5_dr_icm_chunk
is 64 bytes.

Signed-off-by: Rongwei Liu <rongweil@nvidia.com>
Reviewed-by: Shun Hao <shunh@nvidia.com>
Reviewed-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: DR, Remove mr_addr rkey from struct mlx5dr_icm_chunk
Rongwei Liu [Fri, 28 Jan 2022 04:10:54 +0000 (06:10 +0200)]
net/mlx5: DR, Remove mr_addr rkey from struct mlx5dr_icm_chunk

Reduce memory footprint by removing mr_addr and rkey from
mlx5_dr_icm_chunk.
1. mr_addr is calculated by mlx5dr_icm_pool_get_chunk_mr_addr()
2. rkey is calculated by mlx5dr_icm_pool_get_chunk_rkey()
The two new functions are very lightweight and straightforward.

Reduce 8 bytes from struct mlx5_dr_icm_chunk, its current size is
72 bytes.

Signed-off-by: Rongwei Liu <rongweil@nvidia.com>
Reviewed-by: Shun Hao <shunh@nvidia.com>
Reviewed-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: DR, Adjust structure member to reduce memory hole
Rongwei Liu [Fri, 28 Jan 2022 03:56:28 +0000 (05:56 +0200)]
net/mlx5: DR, Adjust structure member to reduce memory hole

Accord to profiling, mlx5dr_ste/mlx5dr_icm_chunk are the two
hot structures. Their memory layout can be optimized by
adjusting member sequences.

Struct mlx5dr_ste size changes from 64 bytes to 56 bytes.

In the upcoming commits, struct mlx5dr_icm_chunk memory layout
will change automatically after removing some members.
Keep it untouched here.

Signed-off-by: Rongwei Liu <rongweil@nvidia.com>
Reviewed-by: Shun Hao <shunh@nvidia.com>
Reviewed-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: Drop cqe_bcnt32 from mlx5e_skb_from_cqe_mpwrq_linear
Maxim Mikityanskiy [Mon, 14 Feb 2022 18:57:56 +0000 (20:57 +0200)]
net/mlx5e: Drop cqe_bcnt32 from mlx5e_skb_from_cqe_mpwrq_linear

The packet size in mlx5e_skb_from_cqe_mpwrq_linear can't overflow u16,
since the maximum packet size in linear striding RQ is 2^13 bytes. Drop
the unneeded u32 variable.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: Drop the len output parameter from mlx5e_xdp_handle
Maxim Mikityanskiy [Thu, 27 Jan 2022 14:42:18 +0000 (16:42 +0200)]
net/mlx5e: Drop the len output parameter from mlx5e_xdp_handle

The len parameter of mlx5e_xdp_handle is used to output the new packet
length after XDP has processed the packet and returned XDP_PASS.
However, this value can be calculated on the caller site, as the caller
knows if it was an XDP_PASS.

This commit drops the len parameter and moves the calculation to the
caller, reducing the number of parameters passed to the function and
preparing for XDP support in non-linear legacy RQ.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: RX, Test the XDP program existence out of the handler
Tariq Toukan [Wed, 19 Jan 2022 19:28:36 +0000 (21:28 +0200)]
net/mlx5e: RX, Test the XDP program existence out of the handler

Instead of early return inside mlx5e_xdp_handle(), let the caller check
if an XDP program is loaded.  This allows saving a few unnecessary
function calls and calculations in case !prog.

Performance test: single core, drop packets in iptables
Before: 3,872,504 pps
After:  3,975,628 pps (+2.66%)

Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: Build SKB in place over the first fragment in non-linear legacy RQ
Maxim Mikityanskiy [Wed, 26 Jan 2022 15:18:26 +0000 (17:18 +0200)]
net/mlx5e: Build SKB in place over the first fragment in non-linear legacy RQ

As a performance optimization and preparation to enabling XDP multi
buffer on non-linear legacy RQ, build the linear part of the SKB over
the first fragment, instead of allocating a new buffer and copying the
first 256 bytes there.

To achieve this, add headroom and tailroom to the first fragment.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: Add headroom only to the first fragment in legacy RQ
Maxim Mikityanskiy [Wed, 26 Jan 2022 13:49:23 +0000 (15:49 +0200)]
net/mlx5e: Add headroom only to the first fragment in legacy RQ

Currently, rq->buff.headroom is applied to all fragments in legacy RQ.
In the linear mode, there is a non-zero headroom, but there is only one
fragment per packet. In the non-linear mode, the headroom is zero.

This commit changes the logic to apply the headroom only to the first
fragment. The current behavior remains the same for both linear and
non-linear modes. However, it allows the next commit to enable headroom
for the non-linear mode, which will be applied only to the first
fragment.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: Validate MTU when building non-linear legacy RQ fragments info
Maxim Mikityanskiy [Mon, 14 Feb 2022 17:59:19 +0000 (19:59 +0200)]
net/mlx5e: Validate MTU when building non-linear legacy RQ fragments info

mlx5e_build_rq_frags_info() assumes that MTU is not bigger than
PAGE_SIZE * MLX5E_MAX_RX_FRAGS, which is 16K for 4K pages. Currently,
the firmware limits MTU to 10K, so the assumption doesn't lead to a bug.

This commits adds an additional driver check for reliability, since the
firmware boundary might be changed.

The calculation is taken to a separate function with a comment
explaining it. It's a preparation for the following patches that
introcuce XDP multi buffer support.

Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agoselftests: vm: fix clang build error multiple output files
Yosry Ahmed [Wed, 16 Mar 2022 23:15:12 +0000 (16:15 -0700)]
selftests: vm: fix clang build error multiple output files

When building the vm selftests using clang, some errors are seen due to
having headers in the compilation command:

  clang -Wall -I ../../../../usr/include  -no-pie    gup_test.c ../../../../mm/gup_test.h -lrt -lpthread -o .../tools/testing/selftests/vm/gup_test
  clang: error: cannot specify -o when generating multiple output files
  make[1]: *** [../lib.mk:146: .../tools/testing/selftests/vm/gup_test] Error 1

Rework to add the header files to LOCAL_HDRS before including ../lib.mk,
since the dependency is evaluated in '$(OUTPUT)/%:%.c $(LOCAL_HDRS)' in
file lib.mk.

Link: https://lkml.kernel.org/r/20220304000645.1888133-1-yosryahmed@google.com
Signed-off-by: Yosry Ahmed <yosryahmed@google.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Nathan Chancellor <nathan@kernel.org>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2 years agoocfs2: fix crash when initialize filecheck kobj fails
Joseph Qi [Wed, 16 Mar 2022 23:15:09 +0000 (16:15 -0700)]
ocfs2: fix crash when initialize filecheck kobj fails

Once s_root is set, genric_shutdown_super() will be called if
fill_super() fails.  That means, we will call ocfs2_dismount_volume()
twice in such case, which can lead to kernel crash.

Fix this issue by initializing filecheck kobj before setting s_root.

Link: https://lkml.kernel.org/r/20220310081930.86305-1-joseph.qi@linux.alibaba.com
Fixes: 5f483c4abb50 ("ocfs2: add kobject for online file check")
Signed-off-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Gang He <ghe@suse.com>
Cc: Jun Piao <piaojun@huawei.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2 years agoconfigs/debug: restore DEBUG_INFO=y for overriding
Qian Cai [Wed, 16 Mar 2022 23:15:06 +0000 (16:15 -0700)]
configs/debug: restore DEBUG_INFO=y for overriding

Previously, I failed to realize that Kees' patch [1] has not been merged
into the mainline yet, and dropped DEBUG_INFO=y too eagerly from the
mainline.  As the results, "make debug.config" won't be able to flip
DEBUG_INFO=n from the existing .config.  This should close the gaps of a
few weeks before Kees' patch is there, and work regardless of their
merging status anyway.

Link: https://lore.kernel.org/all/20220125075126.891825-1-keescook@chromium.org/
Link: https://lkml.kernel.org/r/20220308153524.8618-1-quic_qiancai@quicinc.com
Signed-off-by: Qian Cai <quic_qiancai@quicinc.com>
Reported-by: Daniel Thompson <daniel.thompson@linaro.org>
Reviewed-by: Daniel Thompson <daniel.thompson@linaro.org>
Cc: Kees Cook <keescook@chromium.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2 years agomm: swap: get rid of livelock in swapin readahead
Guo Ziliang [Wed, 16 Mar 2022 23:15:03 +0000 (16:15 -0700)]
mm: swap: get rid of livelock in swapin readahead

In our testing, a livelock task was found.  Through sysrq printing, same
stack was found every time, as follows:

  __swap_duplicate+0x58/0x1a0
  swapcache_prepare+0x24/0x30
  __read_swap_cache_async+0xac/0x220
  read_swap_cache_async+0x58/0xa0
  swapin_readahead+0x24c/0x628
  do_swap_page+0x374/0x8a0
  __handle_mm_fault+0x598/0xd60
  handle_mm_fault+0x114/0x200
  do_page_fault+0x148/0x4d0
  do_translation_fault+0xb0/0xd4
  do_mem_abort+0x50/0xb0

The reason for the livelock is that swapcache_prepare() always returns
EEXIST, indicating that SWAP_HAS_CACHE has not been cleared, so that it
cannot jump out of the loop.  We suspect that the task that clears the
SWAP_HAS_CACHE flag never gets a chance to run.  We try to lower the
priority of the task stuck in a livelock so that the task that clears
the SWAP_HAS_CACHE flag will run.  The results show that the system
returns to normal after the priority is lowered.

In our testing, multiple real-time tasks are bound to the same core, and
the task in the livelock is the highest priority task of the core, so
the livelocked task cannot be preempted.

Although cond_resched() is used by __read_swap_cache_async, it is an
empty function in the preemptive system and cannot achieve the purpose
of releasing the CPU.  A high-priority task cannot release the CPU
unless preempted by a higher-priority task.  But when this task is
already the highest priority task on this core, other tasks will not be
able to be scheduled.  So we think we should replace cond_resched() with
schedule_timeout_uninterruptible(1), schedule_timeout_interruptible will
call set_current_state first to set the task state, so the task will be
removed from the running queue, so as to achieve the purpose of giving
up the CPU and prevent it from running in kernel mode for too long.

(akpm: ugly hack becomes uglier.  But it fixes the issue in a
backportable-to-stable fashion while we hopefully work on something
better)

Link: https://lkml.kernel.org/r/20220221111749.1928222-1-cgel.zte@gmail.com
Signed-off-by: Guo Ziliang <guo.ziliang@zte.com.cn>
Reported-by: Zeal Robot <zealci@zte.com.cn>
Reviewed-by: Ran Xiaokai <ran.xiaokai@zte.com.cn>
Reviewed-by: Jiang Xuexin <jiang.xuexin@zte.com.cn>
Reviewed-by: Yang Yang <yang.yang29@zte.com.cn>
Acked-by: Hugh Dickins <hughd@google.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roger Quadros <rogerq@kernel.org>
Cc: Ziliang Guo <guo.ziliang@zte.com.cn>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2 years agoiavf: Fix hang during reboot/shutdown
Ivan Vecera [Thu, 17 Mar 2022 10:45:24 +0000 (11:45 +0100)]
iavf: Fix hang during reboot/shutdown

Recent commit 974578017fc1 ("iavf: Add waiting so the port is
initialized in remove") adds a wait-loop at the beginning of
iavf_remove() to ensure that port initialization is finished
prior unregistering net device. This causes a regression
in reboot/shutdown scenario because in this case callback
iavf_shutdown() is called and this callback detaches the device,
makes it down if it is running and sets its state to __IAVF_REMOVE.
Later shutdown callback of associated PF driver (e.g. ice_shutdown)
is called. That callback calls among other things sriov_disable()
that calls indirectly iavf_remove() (see stack trace below).
As the adapter state is already __IAVF_REMOVE then the mentioned
loop is end-less and shutdown process hangs.

The patch fixes this by checking adapter's state at the beginning
of iavf_remove() and skips the rest of the function if the adapter
is already in remove state (shutdown is in progress).

Reproducer:
1. Create VF on PF driven by ice or i40e driver
2. Ensure that the VF is bound to iavf driver
3. Reboot

[52625.981294] sysrq: SysRq : Show Blocked State
[52625.988377] task:reboot          state:D stack:    0 pid:17359 ppid:     1 f2
[52625.996732] Call Trace:
[52625.999187]  __schedule+0x2d1/0x830
[52626.007400]  schedule+0x35/0xa0
[52626.010545]  schedule_hrtimeout_range_clock+0x83/0x100
[52626.020046]  usleep_range+0x5b/0x80
[52626.023540]  iavf_remove+0x63/0x5b0 [iavf]
[52626.027645]  pci_device_remove+0x3b/0xc0
[52626.031572]  device_release_driver_internal+0x103/0x1f0
[52626.036805]  pci_stop_bus_device+0x72/0xa0
[52626.040904]  pci_stop_and_remove_bus_device+0xe/0x20
[52626.045870]  pci_iov_remove_virtfn+0xba/0x120
[52626.050232]  sriov_disable+0x2f/0xe0
[52626.053813]  ice_free_vfs+0x7c/0x340 [ice]
[52626.057946]  ice_remove+0x220/0x240 [ice]
[52626.061967]  ice_shutdown+0x16/0x50 [ice]
[52626.065987]  pci_device_shutdown+0x34/0x60
[52626.070086]  device_shutdown+0x165/0x1c5
[52626.074011]  kernel_restart+0xe/0x30
[52626.077593]  __do_sys_reboot+0x1d2/0x210
[52626.093815]  do_syscall_64+0x5b/0x1a0
[52626.097483]  entry_SYSCALL_64_after_hwframe+0x65/0xca

Fixes: 974578017fc1 ("iavf: Add waiting so the port is initialized in remove")
Signed-off-by: Ivan Vecera <ivecera@redhat.com>
Link: https://lore.kernel.org/r/20220317104524.2802848-1-ivecera@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: mscc: ocelot: fix backwards compatibility with single-chain tc-flower offload
Vladimir Oltean [Wed, 16 Mar 2022 19:21:17 +0000 (21:21 +0200)]
net: mscc: ocelot: fix backwards compatibility with single-chain tc-flower offload

ACL rules can be offloaded to VCAP IS2 either through chain 0, or, since
the blamed commit, through a chain index whose number encodes a specific
PAG (Policy Action Group) and lookup number.

The chain number is translated through ocelot_chain_to_pag() into a PAG,
and through ocelot_chain_to_lookup() into a lookup number.

The problem with the blamed commit is that the above 2 functions don't
have special treatment for chain 0. So ocelot_chain_to_pag(0) returns
filter->pag = 224, which is in fact -32, but the "pag" field is an u8.

So we end up programming the hardware with VCAP IS2 entries having a PAG
of 224. But the way in which the PAG works is that it defines a subset
of VCAP IS2 filters which should match on a packet. The default PAG is
0, and previous VCAP IS1 rules (which we offload using 'goto') can
modify it. So basically, we are installing filters with a PAG on which
no packet will ever match. This is the hardware equivalent of adding
filters to a chain which has no 'goto' to it.

Restore the previous functionality by making ACL filters offloaded to
chain 0 go to PAG 0 and lookup number 0. The choice of PAG is clearly
correct, but the choice of lookup number isn't "as before" (which was to
leave the lookup a "don't care"). However, lookup 0 should be fine,
since even though there are ACL actions (policers) which have a
requirement to be used in a specific lookup, that lookup is 0.

Fixes: 226e9cd82a96 ("net: mscc: ocelot: only install TCAM entries into a specific lookup and PAG")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://lore.kernel.org/r/20220316192117.2568261-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: bcmgenet: skip invalid partial checksums
Doug Berger [Thu, 17 Mar 2022 01:28:12 +0000 (18:28 -0700)]
net: bcmgenet: skip invalid partial checksums

The RXCHK block will return a partial checksum of 0 if it encounters
a problem while receiving a packet. Since a 1's complement sum can
only produce this result if no bits are set in the received data
stream it is fair to treat it as an invalid partial checksum and
not pass it up the stack.

Fixes: 810155397890 ("net: bcmgenet: use CHECKSUM_COMPLETE for NETIF_F_RXCSUM")
Signed-off-by: Doug Berger <opendmb@gmail.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/r/20220317012812.1313196-1-opendmb@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agobnx2x: fix built-in kernel driver load failure
Manish Chopra [Wed, 16 Mar 2022 21:46:13 +0000 (14:46 -0700)]
bnx2x: fix built-in kernel driver load failure

Commit b7a49f73059f ("bnx2x: Utilize firmware 7.13.21.0")
added request_firmware() logic in probe() which caused
load failure when firmware file is not present in initrd (below),
as access to firmware file is not feasible during probe.

  Direct firmware load for bnx2x/bnx2x-e2-7.13.15.0.fw failed with error -2
  Direct firmware load for bnx2x/bnx2x-e2-7.13.21.0.fw failed with error -2

This patch fixes this issue by -

1. Removing request_firmware() logic from the probe()
   such that .ndo_open() handle it as it used to handle
   it earlier

2. Given request_firmware() is removed from probe(), so
   driver has to relax FW version comparisons a bit against
   the already loaded FW version (by some other PFs of same
   adapter) to allow different compatible/close enough FWs with which
   multiple PFs may run with (in different environments), as the
   given PF who is in probe flow has no idea now with which firmware
   file version it is going to initialize the device in ndo_open()

Link: https://lore.kernel.org/all/46f2d9d9-ae7f-b332-ddeb-b59802be2bab@molgen.mpg.de/
Reported-by: Paul Menzel <pmenzel@molgen.mpg.de>
Tested-by: Paul Menzel <pmenzel@molgen.mpg.de>
Fixes: b7a49f73059f ("bnx2x: Utilize firmware 7.13.21.0")
Signed-off-by: Manish Chopra <manishc@marvell.com>
Signed-off-by: Ariel Elior <aelior@marvell.com>
Link: https://lore.kernel.org/r/20220316214613.6884-1-manishc@marvell.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: phy: mscc: Add MODULE_FIRMWARE macros
Juerg Haefliger [Wed, 16 Mar 2022 15:18:35 +0000 (16:18 +0100)]
net: phy: mscc: Add MODULE_FIRMWARE macros

The driver requires firmware so define MODULE_FIRMWARE so that modinfo
provides the details.

Fixes: fa164e40c53b ("net: phy: mscc: split the driver into separate files")
Signed-off-by: Juerg Haefliger <juergh@canonical.com>
Link: https://lore.kernel.org/r/20220316151835.88765-1-juergh@canonical.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoMerge tag 'mt76-for-kvalo-2022-03-16' of https://github.com/nbd168/wireless
Kalle Valo [Thu, 17 Mar 2022 14:29:50 +0000 (16:29 +0200)]
Merge tag 'mt76-for-kvalo-2022-03-16' of https://github.com/nbd168/wireless

mt76 patches for 5.18

- bugfixes
- mbssid support for mt7915
- 6 GHz support for mt7915
- mt7921u driver

2 years agortw89: implement stop and resume channels transmission v1
Ping-Ke Shih [Thu, 17 Mar 2022 05:55:43 +0000 (13:55 +0800)]
rtw89: implement stop and resume channels transmission v1

These function is used to stop transmitting when we are going to switch
channels or do some RF calibration. Before these operations, we need to
stop channel transmission and backup setting into parameter tx_en. After
operations are done, resume transmitting by backup parameter tx_en.

Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220317055543.40514-13-pkshih@realtek.com
2 years agortw89: extend mac tx_en bits from 16 to 32
Ping-Ke Shih [Thu, 17 Mar 2022 05:55:42 +0000 (13:55 +0800)]
rtw89: extend mac tx_en bits from 16 to 32

In order to support 8852C that uses 32 bits to control TX types.

This patch doesn't really use 32 bits tx_en yet, but next patch will
use it.

Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220317055543.40514-12-pkshih@realtek.com
2 years agortw89: change value assignment style of rtw89_mac_cfg_gnt()
Ping-Ke Shih [Thu, 17 Mar 2022 05:55:41 +0000 (13:55 +0800)]
rtw89: change value assignment style of rtw89_mac_cfg_gnt()

Use if() style would be clear than "? :", because the else cases are
always 0. The read val from rtw89_mac_read_lte() isn't used, so remove
this statement.

Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220317055543.40514-11-pkshih@realtek.com
2 years agortw89: 8852c: add mac_ctrl_path and mac_cfg_gnt APIs
Chia-Yuan Li [Thu, 17 Mar 2022 05:55:40 +0000 (13:55 +0800)]
rtw89: 8852c: add mac_ctrl_path and mac_cfg_gnt APIs

The BT-coexistence uses these function to control antenna and TDMA, so
implement the variant type to support all chips.

Signed-off-by: Chia-Yuan Li <leo.li@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220317055543.40514-10-pkshih@realtek.com
2 years agoselftests: net: fix array_size.cocci warning
Guo Zhengkui [Wed, 16 Mar 2022 09:28:57 +0000 (17:28 +0800)]
selftests: net: fix array_size.cocci warning

Fix array_size.cocci warning in tools/testing/selftests/net.

Use `ARRAY_SIZE(arr)` instead of forms like `sizeof(arr)/sizeof(arr[0])`.

It has been tested with gcc (Debian 8.3.0-6) 8.3.0.

Signed-off-by: Guo Zhengkui <guozhengkui@vivo.com>
Link: https://lore.kernel.org/r/20220316092858.9398-1-guozhengkui@vivo.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2 years agortw89: disable FW and H2C function if CPU disabled
Chia-Yuan Li [Thu, 17 Mar 2022 05:55:39 +0000 (13:55 +0800)]
rtw89: disable FW and H2C function if CPU disabled

Initialize FW status and disabled H2C function if CPU disabled. Then, it
can reset to initial state.

Signed-off-by: Chia-Yuan Li <leo.li@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220317055543.40514-9-pkshih@realtek.com
2 years agortw89: initialize preload window of D-MAC
Ping-Ke Shih [Thu, 17 Mar 2022 05:55:38 +0000 (13:55 +0800)]
rtw89: initialize preload window of D-MAC

8852C add new hardware feature -- preload window, which is used to load
more data to D-MAC, so it can possibly yield better performance.

This patch is to configure preload and reserved size for next window.

Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220317055543.40514-8-pkshih@realtek.com
2 years agortw89: modify MAC enable functions
Chia-Yuan Li [Thu, 17 Mar 2022 05:55:37 +0000 (13:55 +0800)]
rtw89: modify MAC enable functions

Modify functions that control D-MAC (data MAC) and LDO to support variant
chips.

Signed-off-by: Chia-Yuan Li <leo.li@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220317055543.40514-7-pkshih@realtek.com
2 years agortw89: add config_rf_reg_v1 to configure RF parameter tables
Ping-Ke Shih [Thu, 17 Mar 2022 05:55:36 +0000 (13:55 +0800)]
rtw89: add config_rf_reg_v1 to configure RF parameter tables

The format of RF parameter is changed; it doesn't encode delay parameters
into table, but the delay coding becomes regular pair of register address
and value.

To help firmware to recover RF register settings, we need to download
these parameters to firmware. For v1 format, only download partial
parameters (ignore them with addr < 0x100).

Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220317055543.40514-6-pkshih@realtek.com
2 years agortw89: 8852c: add read/write rf register function
Chung-Hsuan Hung [Thu, 17 Mar 2022 05:55:35 +0000 (13:55 +0800)]
rtw89: 8852c: add read/write rf register function

Using encoded address which BIT(16) is used to discriminate which region is
going to access. Illustrate the calling flow as below

rtw89_phy_write_rf_v1() -+-> rtw89_phy_write_rf()   // old interface
                         +-> rtw89_phy_write_rf_a() // new interface

Signed-off-by: Chung-Hsuan Hung <hsuan8331@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220317055543.40514-5-pkshih@realtek.com
2 years agortw89: 8852c: add setting of TB UL TX power offset
Yuan-Han Zhang [Thu, 17 Mar 2022 05:55:34 +0000 (13:55 +0800)]
rtw89: 8852c: add setting of TB UL TX power offset

Configure this TX power to indicate TX power offset that uses to transmit
TB (trigger base) uplink frames.
Also, shrink the unit of TX power offset changes to suitable type s8.

Signed-off-by: Yuan-Han Zhang <yuanhan1020@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220317055543.40514-4-pkshih@realtek.com
2 years agortw89: 8852c: add write/read crystal function in CFO tracking
Yuan-Han Zhang [Thu, 17 Mar 2022 05:55:33 +0000 (13:55 +0800)]
rtw89: 8852c: add write/read crystal function in CFO tracking

The CFO tracking algorithm is the same, but control methods are different.
Set parameters via xtal serial interfaces (SI).

Signed-off-by: Yuan-Han Zhang <yuanhan1020@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220317055543.40514-3-pkshih@realtek.com
2 years agortw89: modify dcfo_comp to share with chips
Yuan-Han Zhang [Thu, 17 Mar 2022 05:55:32 +0000 (13:55 +0800)]
rtw89: modify dcfo_comp to share with chips

The dcfo_comp is digital CFO (central frequency offset) compensation.
Since the flow can be shared with all chips, add chip parameters to support
variant register address and format.

Signed-off-by: Yuan-Han Zhang <yuanhan1020@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220317055543.40514-2-pkshih@realtek.com
2 years agortw89: Fix spelling mistake "Mis-Match" -> "Mismatch"
Colin Ian King [Wed, 16 Mar 2022 23:42:42 +0000 (23:42 +0000)]
rtw89: Fix spelling mistake "Mis-Match" -> "Mismatch"

There are some spelling mistakes in some literal strings. Fix them.

Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Acked-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220316234242.55515-1-colin.i.king@gmail.com
2 years agobrcmfmac: p2p: Fix spelling mistake "Comback" -> "Comeback"
Colin Ian King [Wed, 16 Mar 2022 23:39:38 +0000 (23:39 +0000)]
brcmfmac: p2p: Fix spelling mistake "Comback" -> "Comeback"

There are some spelling mistakes in comments and brcmf_dbg messages.
Fix these.

Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220316233938.55135-1-colin.i.king@gmail.com
2 years agoiwlwifi: mei: fix building iwlmei
Arnd Bergmann [Wed, 16 Mar 2022 18:36:03 +0000 (19:36 +0100)]
iwlwifi: mei: fix building iwlmei

Building iwlmei without CONFIG_CFG80211 causes a link-time warning:

ld.lld: error: undefined symbol: ieee80211_hdrlen
>>> referenced by net.c
>>>               net/wireless/intel/iwlwifi/mei/net.o:(iwl_mei_tx_copy_to_csme) in archive drivers/built-in.a

Add an explicit dependency to avoid this. In theory it should not
be needed here, but it also seems pointless to allow IWLMEI
for configurations without CFG80211.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Emmanuel Grumbach <Emmanuel.grumbach@intel.com>
Acked-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20220316183617.1470631-1-arnd@kernel.org
2 years agonet: stmmac: clean up impossible condition
Dan Carpenter [Wed, 16 Mar 2022 08:37:44 +0000 (11:37 +0300)]
net: stmmac: clean up impossible condition

This code works but it has a static checker warning:

    drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:1687 init_dma_rx_desc_rings()
    warn: always true condition '(queue >= 0) => (0-u32max >= 0)'

Obviously, it makes no sense to check if an unsigned int is >= 0.  What
prevents this code from being a forever loop is that later there is a
separate check for if (queue == 0).

The "queue" variable is less than MTL_MAX_RX_QUEUES (8) so it can easily
fit in an int type.  Any larger value for "queue" would lead to an array
overflow when we assign "rx_q = &priv->rx_queue[queue]".

Fixes: de0b90e52a11 ("net: stmmac: rearrange RX and TX desc init into per-queue basis")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Link: https://lore.kernel.org/r/20220316083744.GB30941@kili
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2 years agonet: dsa: Add missing of_node_put() in dsa_port_parse_of
Miaoqian Lin [Wed, 16 Mar 2022 08:26:02 +0000 (08:26 +0000)]
net: dsa: Add missing of_node_put() in dsa_port_parse_of

The device_node pointer is returned by of_parse_phandle()  with refcount
incremented. We should use of_node_put() on it when done.

Fixes: 6d4e5c570c2d ("net: dsa: get port type at parse time")
Signed-off-by: Miaoqian Lin <linmq006@gmail.com>
Link: https://lore.kernel.org/r/20220316082602.10785-1-linmq006@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2 years agonet: geneve: support IPv4/IPv6 as inner protocol
Eyal Birger [Wed, 16 Mar 2022 06:15:57 +0000 (08:15 +0200)]
net: geneve: support IPv4/IPv6 as inner protocol

This patch adds support for encapsulating IPv4/IPv6 within GENEVE.

In order to use this, a new IFLA_GENEVE_INNER_PROTO_INHERIT flag needs
to be provided at device creation. This property cannot be changed for
the time being.

In case IP traffic is received on a non-tun device the drop count is
increased.

Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Link: https://lore.kernel.org/r/20220316061557.431872-1-eyal.birger@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2 years agoMerge branch 'net-mvneta-armada-98dx2530-soc'
Paolo Abeni [Thu, 17 Mar 2022 09:49:30 +0000 (10:49 +0100)]
Merge branch 'net-mvneta-armada-98dx2530-soc'

Chris Packham says:

====================
net: mvneta: Armada 98DX2530 SoC

This is split off from [1] to let it go in via net-next rather than waiting for
the rest of the series to land.

[1] - https://lore.kernel.org/lkml/20220314213143.2404162-1-chris.packham@alliedtelesis.co.nz/
====================

Link: https://lore.kernel.org/r/20220315215207.2746793-1-chris.packham@alliedtelesis.co.nz
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2 years agonet: mvneta: Add support for 98DX2530 Ethernet port
Chris Packham [Tue, 15 Mar 2022 21:52:07 +0000 (10:52 +1300)]
net: mvneta: Add support for 98DX2530 Ethernet port

The 98DX2530 SoC is similar to the Armada 3700 except it needs a
different MBUS window configuration. Add a new compatible string to
identify this device and the required MBUS window configuration.

Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2 years agodt-bindings: net: mvneta: Add marvell,armada-ac5-neta
Chris Packham [Tue, 15 Mar 2022 21:52:06 +0000 (10:52 +1300)]
dt-bindings: net: mvneta: Add marvell,armada-ac5-neta

The out of band port on the 98DX2530 SoC is similar to the armada-3700
except it requires a slightly different MBUS window configuration. Add a
new compatible string so this difference can be accounted for.

Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2 years agoptp: ocp: Fix PTP_PF_* verification requests
Jonathan Lemon [Tue, 15 Mar 2022 19:46:26 +0000 (12:46 -0700)]
ptp: ocp: Fix PTP_PF_* verification requests

Update and check functionality for pin configuration requests:

PTP_PF_NONE: requests "IN: None", disabling the pin.

  # testptp -d /dev/ptp3 -L3,0 -i1
  set pin function okay
  # cat sma4
  IN: None

PTP_PF_EXTTS: should configure external timestamps, but since the
timecard can steer inputs to multiple inputs as well as timestamps,
allow the request, but don't change configurations.

  # testptp -d /dev/ptp3 -L3,1 -i1
  set pin function okay

  (no functional or configuration change here yet)

PTP_PF_PEROUT: Channel 0 is the PHC, at 1PPS.  Channels 1-4 are
the programmable frequency generators.

  # fails because period is not 1PPS.
  # testptp -d /dev/ptp3 -L3,2 -i0  -p 500000000
  PTP_PEROUT_REQUEST: Invalid argument

  # testptp -d /dev/ptp3 -L3,2 -i0  -p 1000000000
  periodic output request okay
  # cat sma4
  OUT: PHC

  # testptp -d /dev/ptp3 -L3,2 -i1 -p 500000000 -w 200000000
  periodic output request okay
  # cat sma4
  OUT: GEN1
  # cat gen1/signal
  500000000 40 0 1 2022-03-10T23:55:26 TAI
  # cat gen1/running
  1

  # testptp -d /dev/ptp3 -L3,2 -i1 -p 0
  periodic output request okay
  # cat gen1/running
  0

Signed-off-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Link: https://lore.kernel.org/r/20220315194626.1895-1-jonathan.lemon@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2 years agoMerge branch 'flow_offload-add-tc-vlan-push_eth-and-pop_eth-actions'
Jakub Kicinski [Thu, 17 Mar 2022 02:59:43 +0000 (19:59 -0700)]
Merge branch 'flow_offload-add-tc-vlan-push_eth-and-pop_eth-actions'

Roi Dayan says:

====================
flow_offload: add tc vlan push_eth and pop_eth actions

Offloading vlan push_eth and pop_eth actions is needed in order to
correctly offload MPLSoUDP encap and decap flows, this series extends
the flow offload API to support these actions and updates mlx5 to
parse them.
====================

Link: https://lore.kernel.org/r/20220315110211.1581468-1-roid@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet/mlx5e: MPLSoUDP encap, support action vlan pop_eth explicitly
Maor Dickman [Tue, 15 Mar 2022 11:02:11 +0000 (13:02 +0200)]
net/mlx5e: MPLSoUDP encap, support action vlan pop_eth explicitly

Currently the MPLSoUDP encap offload does the L2 pop implicitly
while adding such action explicitly (vlan eth_push) will cause
the rule to not be offloaded.

Solve it by adding offload support for vlan eth_push in case of
MPLSoUDP decap case.

Flow example:
filter root protocol ip pref 1 flower chain 0
filter root protocol ip pref 1 flower chain 0 handle 0x1
  eth_type ipv4
  dst_ip 2.2.2.22
  src_ip 2.2.2.21
  in_hw in_hw_count 1
        action order 1: vlan  pop_eth pipe
         index 1 ref 1 bind 1
        used_hw_stats delayed

        action order 2: mpls  push protocol mpls_uc label 555 tc 3 ttl 255 pipe
         index 1 ref 1 bind 1
        used_hw_stats delayed

        action order 3: tunnel_key  set
        src_ip 8.8.8.21
        dst_ip 8.8.8.22
        dst_port 6635
        csum
        tos 0x4
        ttl 6 pipe
         index 1 ref 1 bind 1
        used_hw_stats delayed

        action order 4: mirred (Egress Redirect to device bareudp0) stolen
        index 1 ref 1 bind 1
        used_hw_stats delayed

Signed-off-by: Maor Dickman <maord@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet/mlx5e: MPLSoUDP decap, use vlan push_eth instead of pedit
Maor Dickman [Tue, 15 Mar 2022 11:02:10 +0000 (13:02 +0200)]
net/mlx5e: MPLSoUDP decap, use vlan push_eth instead of pedit

Currently action pedit of source and destination MACs is used
to fill the MACs in L2 push step in MPLSoUDP decap offload,
this isn't aligned to tc SW which use vlan eth_push action
to do this.

To fix that, offload support for vlan veth_push action is
added together with mpls pop action, and deprecate the use
of pedit of MACs.

Flow example:
filter protocol mpls_uc pref 1 flower chain 0
filter protocol mpls_uc pref 1 flower chain 0 handle 0x1
  eth_type 8847
  mpls_label 555
  enc_dst_port 6635
  in_hw in_hw_count 1
        action order 1: tunnel_key  unset pipe
         index 2 ref 1 bind 1
        used_hw_stats delayed

        action order 2: mpls  pop protocol ip pipe
         index 2 ref 1 bind 1
        used_hw_stats delayed

        action order 3: vlan  push_eth dst_mac de:a2:ec:d6:69:c8 src_mac de:a2:ec:d6:69:c8 pipe
         index 2 ref 1 bind 1
        used_hw_stats delayed

        action order 4: mirred (Egress Redirect to device enp8s0f0_0) stolen
        index 2 ref 1 bind 1
        used_hw_stats delayed

Signed-off-by: Maor Dickman <maord@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet/sched: add vlan push_eth and pop_eth action to the hardware IR
Maor Dickman [Tue, 15 Mar 2022 11:02:09 +0000 (13:02 +0200)]
net/sched: add vlan push_eth and pop_eth action to the hardware IR

Add vlan push_eth and pop_eth action to the hardware intermediate
representation model which would subsequently allow it to be used
by drivers for offload.

Signed-off-by: Maor Dickman <maord@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: handle ARPHRD_PIMREG in dev_is_mac_header_xmit()
Nicolas Dichtel [Tue, 15 Mar 2022 09:20:08 +0000 (10:20 +0100)]
net: handle ARPHRD_PIMREG in dev_is_mac_header_xmit()

This kind of interface doesn't have a mac header. This patch fixes
bpf_redirect() to a PIM interface.

Fixes: 27b29f63058d ("bpf: add bpf_redirect() helper")
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Link: https://lore.kernel.org/r/20220315092008.31423-1-nicolas.dichtel@6wind.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>