Add a hook in rt6_flush_exceptions, rt6_remove_exception_rt,
rt6_update_exception_stamp_rt, and rt6_age_exceptions to handle
nexthop struct in a fib6_info.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add a hook in fib6_info_uses_dev to handle nexthop struct in a fib6_info.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add a hook in rt6_nlmsg_size to handle nexthop struct in a fib6_info.
rt6_nh_nlmsg_size is used to sum the space needed for all nexthops in
the fib entry.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add a hook in __find_rr_leaf to handle nexthop struct in a fib6_info.
nexthop_for_each_fib6_nh is used to walk each fib6_nh in a nexthop and
call find_match. On a match, use the fib6_nh saved in the callback arg
to setup fib6_result.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add a hook in rt6_device_match to handle nexthop struct in a fib6_info.
The new rt6_nh_dev_match uses nexthop_for_each_fib6_nh to walk each
fib6_nh in a nexthop and call __rt6_device_match. On match,
rt6_nh_dev_match returns the fib6_nh and rt6_device_match uses it to
setup fib6_result.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use nexthop_for_each_fib6_nh to walk all fib6_nh in a nexthop when
dropping 'from' reference in pcpu routes.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

IPv6 has traditionally had a single fib6_nh per fib6_info. With
nexthops we can have multiple fib6_nh associated with a fib6_info.
Add a nexthop helper to invoke a callback for each fib6_nh in a
'struct nexthop'. If the callback returns non-0, the loop is
stopped and the return value passed to the caller.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fix sparse warning:

net/ipv4/tcp_fastopen.c:75:29: warning:
 symbol 'tcp_fastopen_alloc_ctx' was not declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Jason Baron <jbaron@akamai.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Heiner Kallweit says:

====================
r8169: improve handling of chip-specific configuration

This series improves and simplifies handling of chip-specific
configuration.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Simplify the code by removing struct rtl_cfg_info. Only info we need
per PCI ID is whether it supports GBit or not.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

To prepare removal of struct rtl_cfg_info, set the coalesce
config based on the chip version number.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

After the latest changes we don't need separate functions
rtl_hw_start_8168 and rtl_hw_start_8101 any longer. This allows us to
simplify the code. For this change we need to move rtl_hw_start() and
rtl_hw_start_8169(). rtl_hw_start_8169() is unchanged.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

CPCMD_QUIRK_MASK isn't specific to certain chip versions. The vendor
driver applies this mask to all 8168 versions. Therefore remove QUIRK
from the mask name and apply it on all chip versions.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

So far several places in the code deal with setting the interrupt mask
for the respective chip versions. Improve this by having one function
for this only. In addition don't set RxFIFOOver for all 8101 chip
versions like in the vendor driver.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Page pods are used for direct data placement, this patch
enables eDRAM page pods if firmware supports this feature.
Signed-off-by: Varun Prakash <varun@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Maxime Chevallier says:

====================
net: mvpp2: Add extra ethtool stats

This series adds support for more ethtool counters in PPv2 :
 - Per port counters, including one indicating the classifier drops
 - Per RXQ and per TXQ counters

The first 2 patches perform some light rework and renaming, and the 3rd
adds the extra counters.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Besides the MIB counters, some other useful counters can be exposed to
the user. This commit adds support for :

 - Per-port counters, that indicate FIFO drops and classifier drops,
 - Per-rxq counters,
 - Per-txq counters
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Since we'll be adding support for other kind of internal counters, make
clear that the currently supported counters are the MIB counters.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

When first configuring a port on PPv2, we want to clear the internal
counters so that we don't get values from previous boot stages.

However, we can't really clear these counters when resetting the MAC,
since there are valid reasons to do so while the port is being used,
such as when reconfiguring the interface mode with the PHY.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixes gcc '-Wunused-but-set-variable' warning:

drivers/net/ethernet/mscc/ocelot_ace.c: In function ‘vcap_cmd’:
drivers/net/ethernet/mscc/ocelot_ace.c:108:6: warning: variable ‘rc’ set
but not used [-Wunused-but-set-variable]
  int rc;
      ^
It's never used since introduction in commit b5962294 ("net: mscc:
ocelot: Add support for tcam")
Signed-off-by: Mao Wenan <maowenan@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In case autoflowlabel is in action, skb_get_hash_flowi6()
derives a non zero skb->hash to the flowlabel.

If skb->hash is zero, a flow dissection is performed.

Since all TCP skbs sent from ESTABLISH state inherit their
skb->hash from sk->sk_txhash, we better keep a copy
of sk->sk_txhash into the TIME_WAIT socket.

After this patch, ACK or RST packets sent on behalf of
a TIME_WAIT socket have the flowlabel that was previously
used by the flow.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Vladimir Oltean says:

====================
RGMII delays for SJA1105 DSA driver

This patchset configures the Tunable Delay Lines of the SJA1105 P/Q/R/S
switches. These add a programmable phase offset on the RGMII RX and TX
clock signals and get used by the driver for fixed-link interfaces that
use the rgmii-id, rgmii-txid or rgmii-rxid phy-modes.

Tested on a board where RGMII delays were already set up, by adding
MAC-side delays on the RGMII interface towards a BCM5464R PHY and
noticing that the MAC now reports SFD, preamble, FCS etc. errors.

Conflicts trivially in drivers/net/dsa/sja1105/sja1105_spi.c with
https://patchwork.ozlabs.org/project/netdev/list/?series=112614&state=*
which must be applied first.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

As per the DT phy-mode specification, RGMII delays are applied by the
MAC when there is no PHY present on the link.
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

The pad_mii_tx registers point to the same memory region but were
unused. So convert to using these for RGMII I/O cell configuration, as
they bear a shorter name.
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This puts the quad PHY ports in power-down mode when the PHY transitions
to the PHY_HALTED state.  It is likely that all the other PHYs support
the BMCR_PDOWN bit, but I only have the BCM5464R to test.
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Vladimir Oltean says:

====================
Rethink PHYLINK callbacks for SJA1105 DSA

This patchset implements phylink_mac_link_up and phylink_mac_link_down,
while also removing the code that was modifying the EGRESS and INGRESS
MAC settings for STP and replacing them with the "inhibit TX"
functionality.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

The first fact that needs to be stated is that the per-MAC settings in
SJA1105 called EGRESS and INGRESS do *not* disable egress and ingress on
the MAC. They only prevent non-link-local traffic from being
sent/received on this port.

So instead of having .phylink_mac_config essentially mess with the STP
state and force it to DISABLED/BLOCKING (which also brings useless
complications in sja1105_static_config_reload), simply add the
.phylink_mac_link_down and .phylink_mac_link_up callbacks which inhibit
TX at the MAC level, while leaving RX essentially enabled.

Also stop from trying to put the link down in .phylink_mac_config, which
is incorrect.
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This will be used to stop egress traffic in .phylink_mac_link_up.
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Since the driver is now using PHYLINK exclusively, it makes sense to
remove all references to it and replace them with PHYLINK.
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This is a cosmetic patch that replaces the link speed numbers used in
the driver with the corresponding ethtool macros.
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Suggested-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

fix below warnings reported by coccicheck

net/key/af_key.c:932:2-5: WARNING: Use BUG_ON instead of if condition
followed by BUG.
net/key/af_key.c:948:2-5: WARNING: Use BUG_ON instead of if condition
followed by BUG.
Signed-off-by: Hariprasad Kelam <hariprasad.kelam@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

syzbot found a crash in tcp_v6_send_reset() caused by my latest
change.

Problem is that if an skb has been queued to socket prequeue,
skb_dst(skb)->dev can not anymore point to the device.

Fortunately in this case the socket pointer is not NULL.

A similar issue has been fixed in commit 0f85feae ("tcp: fix
more NULL deref after prequeue changes"), I should have known better.

Fixes: 323a53c4 ("ipv6: tcp: enable flowlabel reflection in some RST packets")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Sebastian Andrzej says:

====================
Avoid local_irq_save() and use napi_alloc_frag() where possible

The first two patches remove local_irq_save() around
`netdev_alloc_cache' which does not work on -RT. Besides helping -RT it
whould benefit the users of the function since they can avoid disabling
interrupts and save a few cycles.
The remaining patches are from a time when I tried to remove
`netdev_alloc_cache' but then noticed that we still have non-NAPI
drivers using netdev_alloc_skb() and I dropped that idea. Using
napi_alloc_frag() over netdev_alloc_frag() would skip the not required
local_bh_disable() around the allocation.

v1…v2:
  - 1/7 + 2/7 use now "(in_irq() || irqs_disabled())" instead just
    "irqs_disabled()" to align with __dev_kfree_skb_any(). Pointed out
    by Eric Dumazet.

  - 6/7 has a typo less. Pointed out by Sergei Shtylyov.

  - 3/7 + 4/7 added acks from Ioana Radulescu.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Based on review, `lock' is only acquired in hwbm_pool_add() which is
invoked via ->probe(), ->resume() and ->ndo_change_mtu(). Based on this
the lock can become a mutex and there is no need to disable interrupts
during the procedure.
Now that the lock is a mutex, hwbm_pool_add() no longer invokes
hwbm_pool_refill() in an atomic context so we can pass GFP_KERNEL to
hwbm_pool_refill() and remove the `gfp' argument from hwbm_pool_add().

Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

tg3_alloc_rx_data() uses netdev_alloc_frag() for skb allocation. All
callers of tg3_alloc_rx_data() either hold tp->lock (which is held with
BH disabled) or run in NAPI context.

Use napi_alloc_frag() for skb allocations.

Cc: Siva Reddy Kallam <siva.kallam@broadcom.com>
Cc: Prashant Sreedharan <prashant@broadcom.com>
Cc: Michael Chan <mchan@broadcom.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

SKB allocation via bnx2x_frag_alloc() is always performed in NAPI
context. Preemptible context passes GFP_KERNEL and bnx2x_frag_alloc()
uses then __get_free_page() for the allocation.

Use napi_alloc_frag() for memory allocation.

Cc: Ariel Elior <aelior@marvell.com>
Cc: Sudarsana Kalluru <skalluru@marvell.com>
Cc: GR-everest-linux-l2@marvell.com
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

The driver is using netdev_alloc_frag() for allocation in the
->ndo_start_xmit() path. That one is always invoked in a BH disabled
region so we could also use napi_alloc_frag().

Use napi_alloc_frag() for skb allocation.

Cc: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Acked-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

According to the comment, the preempt_disable() statement is required
due to synchronisation in napi_alloc_frag(). The awful truth is that
local_bh_disable() is required because otherwise the NAPI poll callback
can be invoked while the open function setup buffers. This isn't
unlikely since the dpaa2 provides multiple devices.

The usage of napi_alloc_frag() has been removed in commit

 27c87486 ("dpaa2-eth: Use a single page per Rx buffer")

which means that the comment is not accurate and the preempt_disable()
statement is not required.

Remove the outdated comment and the no longer required
preempt_disable().

Cc: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Acked-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

__netdev_alloc_skb() can be used from any context and is used by NAPI
and non-NAPI drivers. Non-NAPI drivers use it in interrupt context and
NAPI drivers use it during initial allocation (->ndo_open() or
->ndo_change_mtu()). Some NAPI drivers share the same function for the
initial allocation and the allocation in their NAPI callback.

The interrupts are disabled in order to ensure locked access from every
context to `netdev_alloc_cache'.

Let __netdev_alloc_skb() check if interrupts are disabled. If they are, use
`netdev_alloc_cache'. Otherwise disable BH and use `napi_alloc_cache.page'.
The IRQ check is cheaper compared to disabling & enabling interrupts and
memory allocation with disabled interrupts does not work on -RT.
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

netdev_alloc_frag() can be used from any context and is used by NAPI
and non-NAPI drivers. Non-NAPI drivers use it in interrupt context
and NAPI drivers use it during initial allocation (->ndo_open() or
->ndo_change_mtu()). Some NAPI drivers share the same function for the
initial allocation and the allocation in their NAPI callback.

The interrupts are disabled in order to ensure locked access from every
context to `netdev_alloc_cache'.

Let netdev_alloc_frag() check if interrupts are disabled. If they are,
use `netdev_alloc_cache' otherwise disable BH and invoke
__napi_alloc_frag() for the allocation. The IRQ check is cheaper
compared to disabling & enabling interrupts and memory allocation with
disabled interrupts does not work on -RT.
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>