Add a new interface for mapping data to a given id range (max_id),
and back again. It uses xarray as the id allocator and for finding a
given id. For locking it uses xa_lock (spin_lock) for add()/del(),
and rcu_read_lock for find().

This mapping interface also supports delaying the mapping removal via
a workqueue. This is for cases where we need the mapping to have
some grace period in regards to finding it back again, for example
for packets arriving from hardware that were marked with by a rule
with an old mapping that no longer exists.
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Oz Shlomo <ozsh@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>

Set the starting chain from the tc skb ext chain value. Once we read
the tc skb ext, delete it, so cloned/redirect packets won't inherit it.

In order to lookup a chain by the chain index on the ingress block
at ingress classification, provide a lookup function.
Co-developed-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>

To allow lookup of a block's chain under atomic context.
Co-developed-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>

On ingress and cls_act qdiscs init, save the block on ingress
mini_Qdisc and and pass it on to ingress classification, so it
can be used for the looking up a specified chain index.
Co-developed-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>

TC multi chain configuration can cause offloaded tc chains to miss in
hardware after jumping to some chain. In such cases the software should
continue from the chain that missed in hardware, as the hardware may
have manipulated the packet and updated some counters.

Currently a single tcf classification function serves both ingress and
egress. However, multi chain miss processing (get tc skb extension on
hw miss, set tc skb extension on tc miss) should happen only on
ingress.

Refactor the code to use ingress classification function, and move setting
the tc skb extension from general classification to it, as a prestep
for supporting the hw miss scenario.
Co-developed-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>

Added tests for 'u32' extended match rules for u8 alignment.
Signed-off-by: Roman Mashak <mrv@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Florian Fainelli says:

====================
net: phy: Better support for BCM54810

This patch series updates the broadcom PHY driver to better support the
BCM54810 and allow it to make use of the exiting
bcm54xx_adjust_rxrefclk() as well as fix suspend/resume for it.

Changes in v2:

- added Reviewed-by tags from Andrew for patches #1 and #3
- expanded commit message in #2 to explain the change
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

The BCM54810 PHY can use the standard BMCR Power down suspend, but needs
a custom resume routine which first clear the Power down bit, and then
re-initializes the PHY. While in low-power mode, the PHY only accepts
writes to the BMCR register. The datasheet clearly says it:

Reads or writes to any MII register other than MII Control register
(address 00h) while the device is in the standby power-down mode may
cause unpredictable results.
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bcm54xx_adjust_rxrefclk() already checks for PHY_BRCM_AUTO_PWRDWN_ENABLE
and PHY_BRCM_DIS_TXCRXC_NOENRGY in order to set the appropriate bit. The
situation is a bit more complicated with the flag
PHY_BRCM_RX_REFCLK_UNUSED but essentially amounts to the same situation.

The default setting for the 125MHz clock is to be on for all PHYs and
we still treat BCM50610 and BCM50610M specifically with the polarity of
the bit reversed.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

The function bcm54xx_adjust_rxrefclk() works correctly on the BCM54810
PHY, allow this device ID to proceed through.
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/sfc/efx.c:116:38: warning:
 efx_default_channel_type defined but not used [-Wunused-const-variable=]

commit 83975485 ("sfc: move channel alloc/removal code")
left behind this, remove it.
Reported-by: Hulk Robot <hulkci@huawei.com>
Fixes: 83975485 ("sfc: move channel alloc/removal code")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Martin Habets <mhabets@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Huazhong Tan says:

====================
net: hns3: misc updates for -net-next

This series includes some misc updates for the HNS3
ethernet driver.

[patch 1] modifies an unsuitable print when setting dulex mode.
[patch 2] adds some debugfs info for TC and DWRR.
[patch 3] adds some debugfs info for loopback.
[patch 4] adds a missing help info for QS shaper in debugfs.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

HNS3 driver can dump QS shaper configs via debugfs, but missing
help info in debugfs for this operation.
Signed-off-by: Yonglong Liu <liuyonglong@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The MAC ID and loopback status information are obtained from
the hardware, which will be helpful for debugging. This patch
adds support for these two items in debugfs.
Signed-off-by: Yufeng Mo <moyufeng@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The actual enabled TC numbers and the DWRR weight of each
TC may be helpful for debugging, so adds them into debugfs.
Signed-off-by: Yonglong Liu <liuyonglong@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Currently, if device is in link down status and user uses
'ethtool -s' command to set speed but not specify duplex
mode, the duplex mode passed from ethtool to driver is
unknown value(255), and the fibre port will identify this
value as half duplex mode and print "only copper port
supports half duplex!". This message is confusing.

So for fibre port, only the setting duplex is half, prints
error and returns.
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:

struct foo {
        int stuff;
        struct boo array[];
};

By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.

Also, notice that, dynamic memory allocations won't be affected by
this change:

"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]

This issue was found with the help of Coccinelle.

[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 76497732 ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Tested-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

commit 93c09704 ("net: phy: consider latched link-down status in
polling mode") removed double-read of latched link-state register for
polling mode from genphy_update_link(). This added extra ~1s delay into
sequence link down->up.
Following scenario:
 - After boot link goes up
 - phy_start() is called triggering an aneg restart, hence link goes
   down and link-down info is latched.
 - After aneg has finished link goes up. In phy_state_machine is checked
   link state but it is latched "link is down". The state machine is
   scheduled after one second and there is detected "link is up". This
   extra delay can be avoided when we keep link-state register double read
   in case when link was down previously.

With this solution we don't miss a link-down event in polling mode and
link-up is faster.

Details about this quirky behavior on Realtek phy:
Without patch:
T0:    aneg is started, link goes down, link-down status is latched
T0+3s: state machine runs, up-to-date link-down is read
T0+4s: state machine runs, aneg is finished (BMSR_ANEGCOMPLETE==1),
       here i read link-down (BMSR_LSTATUS==0),
T0+5s: state machine runs, aneg is finished (BMSR_ANEGCOMPLETE==1),
       up-to-date link-up is read (BMSR_LSTATUS==1),
       phydev->link goes up, state change PHY_NOLINK to PHY_RUNNING

With patch:
T0:    aneg is started, link goes down, link-down status is latched
T0+3s: state machine runs, up-to-date link-down is read
T0+4s: state machine runs, aneg is finished (BMSR_ANEGCOMPLETE==1),
       first BMSR read: BMSR_ANEGCOMPLETE==1 and BMSR_LSTATUS==0,
       second BMSR read: BMSR_ANEGCOMPLETE==1 and BMSR_LSTATUS==1,
       phydev->link goes up, state change PHY_NOLINK to PHY_RUNNING
Signed-off-by: Petr Oros <poros@redhat.com>
Reviewed-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Heiner Kallweit says:

====================
net: core: add helper tcp_v6_gso_csum_prep

Several network drivers for chips that support TSO6 share the same code
for preparing the TCP header, so let's factor it out to a helper.
A difference is that some drivers reset the payload_len whilst others
don't do this. This value is overwritten by TSO anyway, therefore
the new helper resets it in general.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Use new helper tcp_v6_gso_csum_prep in additional network drivers.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use new helper tcp_v6_gso_csum_prep in additional network drivers.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Acked-by: Hayes Wang <hayeswang@realtek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use new helper tcp_v6_gso_csum_prep in additional network drivers.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use new helper tcp_v6_gso_csum_prep in additional network drivers.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Acked-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use new helper tcp_v6_gso_csum_prep in additional network drivers.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use new helper tcp_v6_gso_csum_prep in additional network drivers.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Acked-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use new helper tcp_v6_gso_csum_prep in additional network drivers.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use new helper tcp_v6_gso_csum_prep in additional network drivers.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Alexander Duyck <alexander.h.duyck@linux.intel.com>
Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use new helper tcp_v6_gso_csum_prep in additional network drivers.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use new helper tcp_v6_gso_csum_prep in additional network drivers.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use new helper tcp_v6_gso_csum_prep in additional network drivers.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Simplify the code by using the new helper tcp_v6_gso_csum_prep.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Several network drivers for chips that support TSO6 share the same code
for preparing the TCP header, so let's factor it out to a helper.
A difference is that some drivers reset the payload_len whilst others
don't do this. This value is overwritten by TSO anyway, therefore
the new helper resets it in general.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Alexander Duyck <alexander.h.duyck@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

It is currenty possible to switch the TCP congestion control algorithm
in non-initial network namespaces:

unshare -U --map-root --net --fork --pid --mount-proc
echo "reno" > /proc/sys/net/ipv4/tcp_congestion_control

works just fine. But currently non-initial network namespaces have no
way of kowing which congestion algorithms are available or allowed other
than through trial and error by writing the names of the algorithms into
the aforementioned file.
Since we already allow changing the congestion algorithm in non-initial
network namespaces by exposing the tcp_congestion_control file there is
no reason to not also expose the
tcp_{allowed,available}_congestion_control files to non-initial network
namespaces. After this change a container with a separate network
namespace will show:

root@f1:~# ls -al /proc/sys/net/ipv4/tcp_* | grep congestion
-rw-r--r-- 1 root root 0 Feb 19 11:54 /proc/sys/net/ipv4/tcp_allowed_congestion_control
-r--r--r-- 1 root root 0 Feb 19 11:54 /proc/sys/net/ipv4/tcp_available_congestion_control
-rw-r--r-- 1 root root 0 Feb 19 11:54 /proc/sys/net/ipv4/tcp_congestion_control

Link: https://github.com/lxc/lxc/issues/3267Reported-by: Haw Loeung <haw.loeung@canonical.com>
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Introduce "rx" prefix in the name scheme for xdp counters
on rx path.
Differentiate between XDP_TX and ndo_xdp_xmit counters
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Acked-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Sunil Goutham says:

====================
octeontx2-af: Cleanup changes

These patches cleanup AF driver by removing unnecessary function
exports and cleanup repititive logic.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Cleanedup repititive nixlf and blkaddr retrieving logic
is various mailbox handlers throughout the rvu_nix.c file.
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Most of the CGX register config is restricted to mapped RVU PFs,
this patch cleans up these permission checks spread across
the rvu_cgx.c file by moving the checks to a common fn().
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Since CGX driver and AF driver are built into a single module
the export symbols in CGX driver are not needed. This patch
gets rid of them.
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Saeed Mahameed says:

====================
This series adds two moderate updates and some misc small patches to
mlx5 driver.

1) From Aya, Add the missing devlink health dump callbacks support for
both rx and tx health reporters.

First patch of the series is extending devlink API to set binary fmsg
data.

All others patches in the series are adding the mlx5 devlink health
callbacks support and the needed FW commands.

2) Also from Aya, Support for FEC modes based on 50G per lane links.
Part of this series, Aya adds one missing link mode define "FEC_LLRS"
to include/uapi/linux/ethtool.h.

3) From Joe, Use proper logging and tracing line terminations

4) From Christophe, Remove a useless 'drain_workqueue()'
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

'destroy_workqueue()' already calls 'drain_workqueue()', there is no need
to call it explicitly.
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>