commit ed6cd6a1 ("net, neigh: Set lower cap for neigh_managed_work rearming")
fixed a case when DELAY_PROBE_TIME is configured to 0, the processing of the
system work queue hog CPU to 100%, and further more we should introduce
a new option used by periodic probe
Signed-off-by: NYuwei Wang <wangyuweihx@gmail.com>
Signed-off-by: NPaolo Abeni <pabeni@redhat.com>

add proc_dointvec_ms_jiffies_minmax to fit read msecs value to jiffies
with a limited range of values
Signed-off-by: NYuwei Wang <wangyuweihx@gmail.com>
Signed-off-by: NPaolo Abeni <pabeni@redhat.com>

Delete the redundant word 'slot'.
Signed-off-by: NJilin Yuan <yuanjilin@cdjrlc.com>
Link: https://lore.kernel.org/r/20220629081632.54445-1-yuanjilin@cdjrlc.comSigned-off-by: NPaolo Abeni <pabeni@redhat.com>

On failure, devm_platform_ioremap_resource() returns a ERR_PTR() value
and not NULL. Fix return value checking by using IS_ERR() and return
PTR_ERR() as error value.

Fixes: 7dc54d3b ("net: pcs: add Renesas MII converter driver")
Reported-by: NHulk Robot <hulkci@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Reviewed-by: NClément Léger <clement.leger@bootlin.com>
Link: https://lore.kernel.org/r/20220628131259.3109124-1-yangyingliang@huawei.comSigned-off-by: NJakub Kicinski <kuba@kernel.org>

The devm_platform_ioremap_resource() function never returns NULL.
It returns error pointers.
Signed-off-by: NPeng Wu <wupeng58@huawei.com>
Reported-by: NHulk Robot <hulkci@huawei.com>
Reviewed-by: NClément Léger <clement.leger@bootlin.com>
Link: https://lore.kernel.org/r/20220628130920.49493-1-wupeng58@huawei.comSigned-off-by: NJakub Kicinski <kuba@kernel.org>

Add doc build infrastructure for txgbe driver.
Initialize PCI memory space for WangXun 10 Gigabit Ethernet devices.
Signed-off-by: NJiawen Wu <jiawenwu@trustnetic.com>
Link: https://lore.kernel.org/r/20220628095530.889344-1-jiawenwu@trustnetic.comSigned-off-by: NJakub Kicinski <kuba@kernel.org>

There is no need to store the result of the addition back to variable count
after the addition. The store is redundant, replace += with just +

Cleans up clang scan build warning:
warning: Although the value stored to 'count' is used in the enclosing
expression, the value is never actually read from 'count'
Signed-off-by: NColin Ian King <colin.i.king@gmail.com>
Reviewed-by: NDavid Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/20220628145406.183527-1-colin.i.king@gmail.comSigned-off-by: NJakub Kicinski <kuba@kernel.org>

br_switchdev_fdb_notify() creates an on-stack FDB info variable, and
initializes it member by member. As such, newly added fields which are
not initialized by br_switchdev_fdb_notify() will contain junk bytes
from the stack.

Other uses of struct switchdev_notifier_fdb_info have a struct
initializer which should put zeroes in the uninitialized fields.

Add a reminder above the structure for future developers. Recently
discussed during review.

Link: https://patchwork.kernel.org/project/netdevbpf/patch/20220524152144.40527-2-schultz.hans+netdev@gmail.com/#24877698
Link: https://patchwork.kernel.org/project/netdevbpf/patch/20220524152144.40527-3-schultz.hans+netdev@gmail.com/#24912269Signed-off-by: NVladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: NIdo Schimmel <idosch@nvidia.com>
Reviewed-by: NNikolay Aleksandrov <razor@blackwall.org>
Link: https://lore.kernel.org/r/20220628100831.2899434-1-vladimir.oltean@nxp.comSigned-off-by: NJakub Kicinski <kuba@kernel.org>

Oleksij Rempel says:

====================
net: dsa: add pause stats support
====================

Link: https://lore.kernel.org/r/20220628085155.2591201-1-o.rempel@pengutronix.deSigned-off-by: NJakub Kicinski <kuba@kernel.org>

This switch is calculating tx/rx_bytes for all packets including pause.
So, include rx/tx_pause counter to rx/tx_packets to make tx/rx_bytes fit
to rx/tx_packets.

Link: https://lore.kernel.org/all/20220624220317.ckhx6z7cmzegvoqi@skbuf/Signed-off-by: NOleksij Rempel <o.rempel@pengutronix.de>
Reviewed-by: NVladimir Oltean <olteanv@gmail.com>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NJakub Kicinski <kuba@kernel.org>

Add support for pause specific stats.

Tested on ksz9477.
Signed-off-by: NOleksij Rempel <o.rempel@pengutronix.de>
Reviewed-by: NVladimir Oltean <olteanv@gmail.com>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NJakub Kicinski <kuba@kernel.org>

Add support for pause stats.
Signed-off-by: NOleksij Rempel <o.rempel@pengutronix.de>
Reviewed-by: NVladimir Oltean <olteanv@gmail.com>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NJakub Kicinski <kuba@kernel.org>

Add support for pause stats
Signed-off-by: NOleksij Rempel <o.rempel@pengutronix.de>
Reviewed-by: NVladimir Oltean <olteanv@gmail.com>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NJakub Kicinski <kuba@kernel.org>

The following is now supported:

  $ tc qdisc add PORT clsact
  $ tc filter add dev PORT egress ...
Signed-off-by: NMaksym Glubokiy <maksym.glubokiy@plvision.eu>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, this script sets up the test scenario, which is supposed to end
in an inability of the system to negotiate a link. It then waits for a bit,
and verifies that the system can diagnose why the link was not established.

The wait time for the scenario where different link speeds are forced on
the two ends of a loopback cable, was set to 4 seconds, which exactly
covered it. As of a recent mlxsw firmware update, this time gets longer,
and this test starts failing.

The time that selftests currently wait for links to be established is
currently $WAIT_TIMEOUT, or 20 seconds. It seems reasonable that if this is
the time necessary to establish and bring up a link, it should also be
enough to determine that a link cannot be established and why.

Therefore in this patch, convert the sleeps to busywaits, so that if a
failure is established sooner (as is expected), the test runs quicker. And
use $WAIT_TIMEOUT as the time to wait.
Signed-off-by: NPetr Machata <petrm@nvidia.com>
Reviewed-by: NAmit Cohen <amcohen@nvidia.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

A number of spelling and language mistakes in the flower section are
fixed. The spacing between the text inside some comments and the comment
symbols are also updated for consistency.
Signed-off-by: NWalter Heymans <walter.heymans@corigine.com>
Reviewed-by: NLouis Peens <louis.peens@corigine.com>
Signed-off-by: NSimon Horman <simon.horman@corigine.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Arun Ramadoss says:

====================
net: dsa: microchip: use ksz_chip_reg for

This patch series tries to use the same approach as struct ksz8 for
register which has different address for different switch family. It
moves the struct ksz8 from ksz8.h to struct ksz_chip_reg. Based on the
switch detect, the corresponding mask, reg and shifts are assigned.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch moves the broadcast ctrl, multicast ctrl and start control
registers from ksz_chip_dat to ksz_chip_reg.
Signed-off-by: NArun Ramadoss <arun.ramadoss@microchip.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch moves the stp_ctrl_reg from the ksz_chip_data to ksz_chip_reg
structure.
Signed-off-by: NArun Ramadoss <arun.ramadoss@microchip.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The register size for the ksz8 switches is u8 and for ksz9477 series is
u16.  To have common struct for ksz series switches the size of reg is
increased from u8 to u16.
Signed-off-by: NArun Ramadoss <arun.ramadoss@microchip.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch removes the struct ksz8 from ksz8.h which is no longer
needed. The platform bus specific details are now deferenced through
dev->priv.
Signed-off-by: NArun Ramadoss <arun.ramadoss@microchip.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch moves ksz8->shifts from ksz8795.c to ksz_common.c. The shifts
are dereferenced using dev->info->shifts.
Signed-off-by: NArun Ramadoss <arun.ramadoss@microchip.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch moves the ksz8->masks from ksz8795.c to ksz_common.c. The
mask will be dereferenced using dev->info->masks.
Signed-off-by: NArun Ramadoss <arun.ramadoss@microchip.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch moves the ksz8->regs from ksz8795.c to the ksz_common.c. And
the regs is dereferrenced using dev->info->regs.
Signed-off-by: NArun Ramadoss <arun.ramadoss@microchip.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Biao Huang says:

====================
net: add more features for mtk-star-emac

Changes in v4:
1. correct the usage of spin_lock/__napi_schedule.
2. fix coding style as Jakub's comments.

Changes in v3:
1. refractor driver as Jakub's comments in patch
   "net: ethernet: mtk-star-emac: separate tx/rx handling with two NAPIs".
2. add acked-by as Rob's comments.
3. add a new patch for halp-duplex support in driver.

Changes in v2:
1. fix coding style as Bartosz's comments.
2. add reviewed-by as Bartosz's comments.

This series add more features for mtk-star-emac:
1. add reference clock pad selection for RMII;
2. add simple timing adjustment for RMII;
3. add support for MII;
4. add support for new IC MT8365;
5. separate tx/rx interrupt handling.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Current driver doesn't support half duplex correctly.
This patch enable half duplex capability in hardware.
Signed-off-by: NBiao Huang <biao.huang@mediatek.com>
Signed-off-by: NYinghua Pan <ot_yinghua.pan@mediatek.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Current driver may lost tx interrupts under bidirectional test with iperf3,
which leads to some unexpected issues.

This patch let rx/tx interrupt enable/disable separately, and rx/tx are
handled in different NAPIs.
Signed-off-by: NBiao Huang <biao.huang@mediatek.com>
Signed-off-by: NYinghua Pan <ot_yinghua.pan@mediatek.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add support for MII interface.
If user wants to use MII, assign "MII" to "phy-mode" property in dts.
Signed-off-by: NBiao Huang <biao.huang@mediatek.com>
Signed-off-by: NYinghua Pan <ot_yinghua.pan@mediatek.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add description for new properties which will be parsed in driver.
Signed-off-by: NBiao Huang <biao.huang@mediatek.com>
Acked-by: NRob Herring <robh@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add simple clock inversion for timing adjustment in driver.
Add property "mediatek,txc-inverse" or "mediatek,rxc-inverse" to
device node when necessary.
Signed-off-by: NBiao Huang <biao.huang@mediatek.com>
Signed-off-by: NYinghua Pan <ot_yinghua.pan@mediatek.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch add a new dts property named "mediatek,rmii-rxc" parsing
in driver, which will configure MAC to select which pin the RMII reference
clock is connected to, TXC or RXC.

TXC pad is the default reference clock pin. If user wants to use RXC pad
instead, add "mediatek,rmii-rxc" to corresponding device node.
Signed-off-by: NBiao Huang <biao.huang@mediatek.com>
Signed-off-by: NYinghua Pan <ot_yinghua.pan@mediatek.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add binding document for Ethernet on MT8365.
Signed-off-by: NBiao Huang <biao.huang@mediatek.com>
Reviewed-by: NBartosz Golaszewski <brgl@bgdev.pl>
Acked-by: NRob Herring <robh@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add Ethernet driver support for MT8365 SoC.
Signed-off-by: NBiao Huang <biao.huang@mediatek.com>
Signed-off-by: NYinghua Pan <ot_yinghua.pan@mediatek.com>
Signed-off-by: NFabien Parent <fparent@baylibre.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If the flags in request_irq() is IRQF_TRIGGER_NONE, the trigger method
is determined by "interrupt" property in dts.
So, modify the flag from IRQF_TRIGGER_FALLING to IRQF_TRIGGER_NONE.
Signed-off-by: NBiao Huang <biao.huang@mediatek.com>
Signed-off-by: NYinghua Pan <ot_yinghua.pan@mediatek.com>
Reviewed-by: NBartosz Golaszewski <brgl@bgdev.pl>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Not all the SoC are using the same clock divider. Move the divider into
a compat structure specific to the SoCs.
Signed-off-by: NBiao Huang <biao.huang@mediatek.com>
Signed-off-by: NFabien Parent <fparent@baylibre.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Martin Habets says:

====================
sfc: Add extra states for VDPA

For EF100 VDPA support we need to enhance the sfc driver's load and
unload functionality so that it can probe and then unregister its
network device, so that VDPA can use services such as MCDI to initialise
VDPA resources.

v2:
- Fix checkpatch errors.
- Correct signoffs.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The netdev probe will be used when moving from vDPA to EF100 BAR config.
The netdev remove will be used when moving from EF100 to vDPA BAR config.

In the process, change several log messages to pci_ instead of netif_
to remove the "(unregistered net_device)" text.
Signed-off-by: NJonathan Cooper <jonathan.s.cooper@amd.com>
Co-developed-by: NMartin Habets <habetsm.xilinx@gmail.com>
Signed-off-by: NMartin Habets <habetsm.xilinx@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Minor fix to existing code to make later patch checkpatch clean.
Signed-off-by: NJonathan Cooper <jonathan.s.cooper@amd.com>
Acked-by: NMartin Habets <habetsm.xilinx@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Minor fix to existing code to later patch checkpatch clean.
Signed-off-by: NJonathan Cooper <jonathan.s.cooper@amd.com>
Acked-by: NMartin Habets <habetsm.xilinx@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This avoids a forward declaration in a subsequent patch.
Signed-off-by: NJonathan Cooper <jonathan.s.cooper@amd.com>
Co-developed-by: NMartin Habets <habetsm.xilinx@gmail.com>
Signed-off-by: NMartin Habets <habetsm.xilinx@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>