For only PF driver can configure flow director rule, it's
better to call hclge_del_all_fd_entries() directly in hclge
layer, rather than call hns3_del_all_fd_entries() in hns3
layer. Then the ae_algo->ops.del_all_fd_entries can be removed.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Currently, the flow director rule of aRFS is configured in
the IO path. It's time-consuming. So move out the configuration,
and configure it asynchronously. And keep ethtool and tc flower
rule using synchronous way, otherwise the application maybe
unable to know the rule is installed or pending.

Add a state member for each flow director rule to indicate the
rule state. There are 4 states:
TO_ADD: the rule is waiting to add to hardware
TO_DEL: the rule is waiting to remove from hardware
DELETED: the rule has been removed from hardware. It's a middle
        state, used to remove the rule node in the fd_rule_list.
ACTIVE: the rule is already added in hardware

For asynchronous way, when receive a new request to add or delete
flow director rule by aRFS, update the rule list, then request to
schedule the service task to finish the configuration.

For synchronous way, when receive a new request to add or delete
flow director rule by ethtool or tc flower, configure hardware
directly, then update the rule list if success.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The hardware supports to parse and match the traffic class field
of IPv6 packet for flow director, uses the same tuple as ip tos.
So removes the limitation of configure 'tclass' by driver.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Currently, there are too many branches for hclge_fd_convert_tuple().
And it may be more when add new tuples. Refactor it by sorting the
tuples according to their length. So it only needs several KEY_OPT
now, and being flexible to add new tuples.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The process of function hclge_fd_get_tuple() is complex and
prolix. To make it more readable, extract the process of each
flow-type tuple to a single function.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The process of function hclge_add_fd_entry() is complex and
prolix. To make it more readable, extract the process of
fs->ring_cookie to a single function.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

s/unrequired/"not required"/
s/consme/consume/ .....two different places
s/accros/across/
Signed-off-by: Bhaskar Chowdhury <unixbhaskar@gmail.com>
Acked-by: Igor Russkikh <irusskikh@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

s/subsytem/subsystem/
Signed-off-by: Bhaskar Chowdhury <unixbhaskar@gmail.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

s/packaet/packet/
Signed-off-by: Bhaskar Chowdhury <unixbhaskar@gmail.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Cristian Ciocaltea says:

====================
Add support for Actions Semi Owl Ethernet MAC

This patch series adds support for the Ethernet MAC found on the Actions
Semi Owl family of SoCs.

For the moment I have only tested the driver on RoseapplePi SBC, which is
based on the S500 SoC variant. It might work on S900 as well, but I cannot
tell for sure since the S900 datasheet I currently have doesn't provide
any information regarding the MAC registers - so I couldn't check the
compatibility with S500.

Similar story for S700: the datasheet I own is incomplete, but it seems
the MAC is advertised with Gigabit capabilities. For that reason most
probably we need to extend the current implementation in order to support
this SoC variant as well.

Please note that for testing the driver it is also necessary to update the
S500 clock subsystem:

https://lore.kernel.org/lkml/cover.1615221459.git.cristian.ciocaltea@gmail.com/

The DTS changes for the S500 SBCs will be provided separately.

Thanks,
Cristi

Changes in v3:
 - Dropped the 'debug' module parameter and passed the default NETIF_MSG flags
to netif_msg_init(), according to David's review

 - Removed the owl_emac_generate_mac_addr() function and the related
OWL_EMAC_GEN_ADDR_SYS_SN config option until a portable solution to get
the system serial number is found - when building on arm64 the following
error is thrown (as reported by Rob's kernel bot):
 '[...]/owl-emac.c:9:10: fatal error: asm/system_info.h: No such file or directory'

 - Rebased patchset on v5.12-rc4

Changes in v2:
* According to Philipp's review
 - Requested exclusive control over serial line via
   devm_reset_control_get_exclusive()
 - Optimized error handling by using dev_err_probe()

* According to Andrew's review
 - Dropped the inline keywords
 - Applied Reverse Christmas Tree format to local variable declarations
 - Renamed owl_emac_phy_config() to owl_emac_update_link_state()
 - Documented the purpose of the special descriptor used in the context of
   owl_emac_setup_frame_xmit()
 - Updated comment inside owl_emac_mdio_clock_enable() regarding the MDC
   clock divider setup
 - Indicated MAC support for symmetric pause via phy_set_sym_pause()
   in owl_emac_phy_init()
 - Changed the MAC addr generation algorithm in owl_emac_generate_mac_addr()
   by setting the locally administered bit in byte 0 and replacing bytes 1 & 2
   with additional entries from enc_sn
 - Moved devm_add_action_or_reset() before clk_set_rate() in owl_emac_probe()

* Other
 - Added SMII interface support: updated owl_emac_core_sw_reset(), added
   owl_emac_clk_set_rate(), updated description in the YAML binding
 - Changed OWL_EMAC_TX_TIMEOUT from 0.05*HZ to 2*HZ
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Add entries for Actions Semi Owl Ethernet MAC binding and driver.
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add new driver for the Ethernet MAC used on the Actions Semi Owl
family of SoCs.

Currently this has been tested only on the Actions Semi S500 SoC
variant.
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add devicetree binding for the Ethernet MAC present on the Actions
Semi Owl family of SoCs.

For the moment advertise only the support for the Actions Semi S500 SoC
variant.
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The existing code is functionally correct: iproute2 parses the ip_flags
argument for tc-flower and really packs it as big endian into the
TCA_FLOWER_KEY_FLAGS netlink attribute. But there is a problem in the
fact that W=1 builds complain:

net/sched/cls_flower.c:1047:15: warning: cast to restricted __be32

This is because we should use the dedicated helper for obtaining a
__be32 pointer to the netlink attribute, not a u32 one. This ensures
type correctness for be32_to_cpu.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

A make W=1 build complains that:

net/sched/cls_flower.c:214:20: warning: cast from restricted __be16
net/sched/cls_flower.c:214:20: warning: incorrect type in argument 1 (different base types)
net/sched/cls_flower.c:214:20:    expected unsigned short [usertype] val
net/sched/cls_flower.c:214:20:    got restricted __be16 [usertype] dst

This is because we use htons on struct flow_dissector_key_ports members
src and dst, which are defined as __be16, so they are already in network
byte order, not host. The byte swap function for the other direction
should have been used.

Because htons and ntohs do the same thing (either both swap, or none
does), this change has no functional effect except to silence the
warnings.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add a function to set the dynamic queue limit minimum value.

Some specific drivers might have legitimate reasons to configure
dql.min_limit to a given value. Typically, this is the case when the
PDU of the protocol is smaller than the packet size to used to
carry those frames to the device.

Concrete example: a CAN (Control Area Network) device with an USB 2.0
interface.  The PDU of classical CAN protocol are roughly 16 bytes but
the USB packet size (which is used to carry the CAN frames to the
device) might be up to 512 bytes.  Wen small traffic burst occurs, BQL
algorithm is not able to immediately adjust and this would result in
having to send many small USB packets (i.e packet of 16 bytes for each
CAN frame). Filling up the USB packet with CAN frames is relatively
fast (small latency issue) but the gain of not having to send several
small USB packets is huge (big throughput increase). In this case,
forcing dql.min_limit to a given value that would allow to stuff the
USB packet is always a win.

This function is to be used by network drivers which are able to prove
through a rationale and through empirical tests on several environment
(with other applications, heavy context switching, virtualization...),
that they constantly reach better performances with a specific
predefined dql.min_limit value with no noticeable latency impact.
Signed-off-by: Vincent Mailhol <mailhol.vincent@wanadoo.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>

1. Remove CONFIG_HAVE_NET_DSA.

CONFIG_HAVE_NET_DSA is a legacy leftover from the times when drivers
should have selected CONFIG_NET_DSA manually.
Currently, all drivers has explicit 'depends on NET_DSA', so this is
no more needed.

2. CONFIG_HAVE_NET_DSA dependencies became CONFIG_NET_DSA's ones.

 - dropped !S390 dependency which was introduced to be sure NET_DSA
   can select CONFIG_PHYLIB. DSA migrated to Phylink almost 3 years
   ago and the PHY library itself doesn't depend on !S390 since
   commit 870a2b5e ("phylib: remove !S390 dependeny from Kconfig");
 - INET dependency is kept to be sure we can select NET_SWITCHDEV;
 - NETDEVICES dependency is kept to be sure we can select PHYLINK.

3. DSA drivers menu now depends on NET_DSA.

Instead on 'depends on NET_DSA' on every single driver, the entire
menu now depends on it. This eliminates a lot of duplicated lines
from Kconfig with no loss (when CONFIG_NET_DSA=m, drivers also can
be only m or n).
This also has a nice side effect that there's no more empty menu on
configurations without DSA.

4. Kbuild will now descend into 'drivers/net/dsa' only when
   CONFIG_NET_DSA is y or m.

This is safe since no objects inside this folder can be built without
DSA core, as well as when CONFIG_NET_DSA=m, no objects can be
built-in.
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This reverts commit e9bf9694.

The topic of the reverted patch is the support for switches with global
VLAN filtering, added by commit 061f6a50 ("net: dsa: Add
ndo_vlan_rx_{add, kill}_vid implementation"). Be there a switch with 4
ports swp0 -> swp3, and the following setup:

ip link add br0 type bridge vlan_filtering 1
ip link set swp0 master br0
ip link set swp1 master br0

What would happen with VLAN-tagged traffic received on standalone ports
swp2 and swp3? Well, it would get dropped, were it not for the
.ndo_vlan_rx_add_vid and .ndo_vlan_rx_kill_vid implementations (called
from vlan_vid_add and vlan_vid_del respectively). Basically, for DSA
switches where VLAN filtering is a global attribute, we enforce the
standalone ports to have 'rx-vlan-filter: off [fixed]' in their ethtool
features, which lets the user know that all VLAN-tagged packets that are
not explicitly added in the RX filtering list are dropped.

As for the sja1105 driver, at the time of the reverted patch, it was
operating in a pretty handicapped mode when it had ports under a bridge
with vlan_filtering=1. Specifically, it was unable to terminate traffic
through the CPU port (for further explanation see "Traffic support" in
Documentation/networking/dsa/sja1105.rst).

However, since then, the sja1105 driver has made considerable progress,
and that limitation is no longer as severe now. Specifically, since
commit 2cafa72e ("net: dsa: sja1105: add a new
best_effort_vlan_filtering devlink parameter"), the driver is able to
perform CPU termination even when some ports are under bridges with
vlan_filtering=1. Then, since commit 8841f6e6 ("net: dsa: sja1105:
make devlink property best_effort_vlan_filtering true by default"), this
even became the default operating mode.

So we can now take advantage of the logic in the DSA core.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support for the ethtool get_ringparam operation.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Alex Elder says:

====================
net: ipa: more configuration data updates

This series starts with two patches that should have been included
in an earlier series.  With these in place, QSB settings are
programmed from information found in the data files rather than
being embedded in code.  Support is then added for reprenting
another QSB property (supported for IPA v4.0+).

The third patch updates the definition of the sequencer type used
for an endpoint.  Previously a set of 2-byte symbols with fairly
long names defined the sequencer type, but now those are broken into
1-byte halves whose names are a little more informative.

The fourth patch moves the sequencer type definition so it only
applies to TX endpoints (they aren't valid for RX endpoints).  And
the last makes some minor documentation updates.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Fix/expand some comments in "ipa_data.h".
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

We only program the sequencer type for TX endpoints.  So move the
definition of the sequencer type fields into the TX-specific portion
of the endpoint configuration data.  There's no need to maintain
this in the IPA structure; we can extract it from the configuration
data it points to in the one spot it's needed.

We previously specified the sequencer type for RX endpoints with
INVALID values.  These are no longer needed, so get rid of them.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

An IPA endpoint has a sequencer that must be configured based on how
the endpoint is to be used.  Currently the IPA code programs the
sequencer type by splitting a value into four 4-bit nibbles.  Doing
that doesn't really add much value, and regardless, a better way of
splitting the sequencer type is into two halves--the lower byte
describing how normal packet processing is handled, and the next
byte describing information about processing replicas.

So split the sequencer type into two sub-parts:  the sequencer type
and the replication sequencer type.  Define the values supported for
the "main" sequencer type, and define the values supported for the
replication part separately.

In addition, the sequencer type names are quite verbose, encoding
what the type includes, but also what it *excludes*.  Rename the
sequencer types in a way that mainly describes the number of passes
that a packet takes through the IPA processing pipeline, and how
many of those passes end by supplying the processed packet to the
microprocessor.

The result expands the supported types beyond what is required for
now, but simplifies the way these are defined.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Starting with IPA v4.0, a limit is placed on the number of bytes
outstanding in a transaction, to reduce latency.  The limit is
imposed only if this value is non-zero.

We don't use a non-zero value for SC7180, but newer versions of IPA
do.  Prepare for that by allowing a programmed value to be specified
in the platform configuration data.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use the QSB configuration data in ipa_hardware_config_qsb(), rather
than determining in code what values to use based on IPA version.
Pass configuration data to ipa_hardware_config() so it can be passed
to ipa_hardware_config_qsb().
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Made changes to coding style as suggested by checkpatch.pl
changes are of the type:
	open brace '{' following struct go on the same line
	do not use assignment in if condition
Signed-off-by: Sai Kalyaan Palla <saikalyaan63@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Tony Nguyen says:

====================
1GbE Intel Wired LAN Driver Updates 2021-03-19

This series contains updates to igc and e1000e drivers.

Sasha removes unused defines in igc driver.

Jiapeng Zhong changes bool assignments from 0/1 to false/true for igc.

Wei Yongjun marks e1000e_pm_prepare() as __maybe_unused to resolve a
defined but not used warning under certain configurations.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

The function e1000e_pm_prepare() may have no callers depending
on configuration, so it must be marked __maybe_unused to avoid
harmless warning:

drivers/net/ethernet/intel/e1000e/netdev.c:6926:12:
 warning: 'e1000e_pm_prepare' defined but not used [-Wunused-function]
 6926 | static int e1000e_pm_prepare(struct device *dev)
      |            ^~~~~~~~~~~~~~~~~

Fixes: ccf8b940 ("e1000e: Leverage direct_complete to speed up s2ram")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>

Fix the following coccicheck warnings:

./drivers/net/ethernet/intel/igc/igc_main.c:4961:2-14: WARNING:
Assignment of 0/1 to bool variable.

./drivers/net/ethernet/intel/igc/igc_main.c:4955:2-14: WARNING:
Assignment of 0/1 to bool variable.

./drivers/net/ethernet/intel/igc/igc_main.c:4933:1-13: WARNING:
Assignment of 0/1 to bool variable.

./drivers/net/ethernet/intel/igc/igc_main.c:4592:1-24: WARNING:
Assignment of 0/1 to bool variable.

./drivers/net/ethernet/intel/igc/igc_main.c:4438:2-25: WARNING:
Assignment of 0/1 to bool variable.

./drivers/net/ethernet/intel/igc/igc_main.c:4396:2-25: WARNING:
Assignment of 0/1 to bool variable.

./drivers/net/ethernet/intel/igc/igc_main.c:4018:2-25: WARNING:
Assignment of 0/1 to bool variable.
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Jiapeng Zhong <abaci-bugfix@linux.alibaba.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>

MII_CR_LOOPBACK masks not in use in i225 device and can be removed.
Signed-off-by: Sasha Neftin <sasha.neftin@intel.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>

Force PHY speed not supported for i225 devices.
MII_CR_SPEED masks not in use in i225 device and can be removed.
Signed-off-by: Sasha Neftin <sasha.neftin@intel.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>

I was working on a syzbot issue, claiming one device could not be
dismantled because its refcount was -1

unregister_netdevice: waiting for sit0 to become free. Usage count = -1

It would be nice if syzbot could trigger a warning at the time
this reference count became negative.

This patch adds CONFIG_PCPU_DEV_REFCNT options which defaults
to per cpu variables (as before this patch) on SMP builds.

v2: free_dev label in alloc_netdev_mqs() is moved to avoid
    a compiler warning (-Wunused-label), as reported
    by kernel test robot <lkp@intel.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Alex Elder says:

====================
net: ipa: update configuration data

Each IPA version has a "data" file defining how various things are
configured.  This series gathers a few updates to this information:
  - The first patch makes all configuration data constant
  - The second fixes an incorrect (but seemingly harmless) value
  - The third simplifies things a bit by using implicit zero
    initialization for memory regions that are empty
  - The fourth adds definitions for memory regions that exist but
    are not yet used
  - The fifth use configuration data rather than conditional code to
    set some bus parameters
====================

Define the maximum number of reads and writes to configure for the
QSB masters used for IPA in configuration data.

We don't use these values yet; the next commit takes care of that.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

There are several memory regions that are defined starting with IPA
v4.0, but which were not used for the SC7180 SoC (IPA v4.2).  Even
though they're not used (yet), define them so they are ready to be
used for SoCs when they become supported.

There are two QUOTA statistics memory regions, one for the modem and
one for the AP.  Define distinct names for these regions, and get
rid of the definition of IPA_MEM_STATS_QUOTA.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

The AP_HEADER memory region for both the SDM845 and SC7180 SoCs has
zero size, and has no canaries.  Defining an offset for such a
zero-length region is not meaningful, so it's better not to define
it at all.  The size of this region is used in the code, but its
value will still be zero because the memory regions are defined in
statically initialized memory.

For the SC7180, the STATS_DROP memory region has a zero size and no
canaries as well.

These regions are the only place where a zero-sized region is
defined despite having no canaries.  Remove them.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

There should be no canary values written before the beginning of the
UC_INFO memory region.  This was correct for SDM845, but somehow was
committed with the wrong value for SC7180.

This bug seems to cause no harm, so we'll just correct it without
back-porting.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

All of the platform configuration data should be constant, but
that isn't the case for the memory regions, interconnects, and
clocks.  Fix this.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

MII_CR_RESET mask not in use in i225 device and can be removed
Signed-off-by: Sasha Neftin <sasha.neftin@intel.com>
Tested-by: Dvora Fuxbrumer <dvorax.fuxbrumer@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>

Bjarni Jonasson says:

====================
Fixes applied to VCS8584 family

Three different fixes applied to VSC8584 family:
1. LCPLL reset
2. Serdes calibration
3. Coma mode disabled

The same fixes has already been applied to VSC8514
and most of the functionality can be reused for the VSC8584.

v1 -> v2:
  Preserved reversed christmas tree
====================
Signed-off-by: David S. Miller <davem@davemloft.net>