In addition to the source/destination IP which are already hashed.
Only for unicast traffic for now.
Signed-off-by: NAchiad Shochat <achiad@mellanox.com>
Signed-off-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

No logical change in this commit.
Signed-off-by: NAchiad Shochat <achiad@mellanox.com>
Signed-off-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

A regular TX WQE execution involves two or more DMA reads -
one to fetch the WQE, and another one per WQE gather entry.

These DMA reads obviously increase the TX latency.
There are two mlx5 mechanisms to bypass these DMA reads:
1) Inline WQE
2) Blue Flame (BF)

An inline WQE contains a whole packet, thus saves the DMA read/s
of the regular WQE gather entry/s. Inline WQE support was already
added in the previous commit.

A BF WQE is written directly to the device I/O mapped memory, thus
enables saving the DMA read that fetches the WQE.

The BF WQE I/O write must be in cache line granularity, thus uses
the CPU write combining mechanism.
A BF WQE I/O write acts also as a TX doorbell for notifying the
device of new TX WQEs.
A BF WQE is written to the same I/O mapped address as the regular TX
doorbell, thus this address is being mapped twice - once by ioremap()
and once by io_mapping_map_wc().

While both mechanisms reduce the TX latency, they both consume more CPU
cycles than a regular WQE:
- A BF WQE must still be written to host memory, in addition to being
  written directly to the device I/O mapped memory.
- An inline WQE involves copying the SKB data into it.

To handle this tradeoff, we introduce here a heuristic algorithm that
strives to avoid using these two mechanisms in case the TX queue is
being back-pressured by the device, and limit their usage rate otherwise.

An inline WQE will always be "Blue Flamed" (written directly to the
device I/O mapped memory) while a BF WQE may not be inlined (may contain
gather entries).

Preliminary testing using netperf UDP_RR shows that the latency goes down
from 17.5us to 16.9us, while the message rate (tested with pktgen) stays
the same.
Signed-off-by: NAchiad Shochat <achiad@mellanox.com>
Signed-off-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

AKA inline WQE.
A TX latency optimization to save data gather DMA reads.
Controlled by ETHTOOL_TX_COPYBREAK.
Signed-off-by: NAchiad Shochat <achiad@mellanox.com>
Signed-off-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

By affinity hints and XPS, each mlx5e channel is assigned a CPU
core.

Channel DMA coherent memory that is written by the NIC and read
by SW (e.g CQ buffer) is allocated on the NUMA node of the CPU
core assigned for the channel.

Channel DMA coherent memory that is written by SW and read by the
NIC (e.g SQ/RQ buffer) is allocated on the NUMA node of the NIC.

Doorbell record (written by SW and read by the NIC) is an
exception since it is accessed by SW more frequently.
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The ConnectX-4 HW implements inverted XOR8.
To make it act as XOR we re-order the HW RSS indirection table.

Set XOR to be the default RSS hash function and add ethtool API to
control it.
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds the missing statistics for the host
and slave ports of the CPSW on K2L and K2E platforms.
Signed-off-by: NWingMan Kwok <w-kwok2@ti.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In certain applications it's beneficial to allow the CPSW h/w
stats counters to continue to increment even while the kernel
polls them. This patch implements this behavior for both 1G
and 10G ethernet subsystem modules.
Signed-off-by: NWingMan Kwok <w-kwok2@ti.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Different Keystone2 platforms have different number and
layouts of hw statistics modules.  This patch consolidates
the statistics processing of different Keystone2 platforms
for easy maintenance.
Signed-off-by: NWingMan Kwok <w-kwok2@ti.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The CPSW driver keeps internally some, but not all, of
the statistics available in the hw statistics modules.  Furthermore,
some of the locations in the hw statistics modules are reserved and
contain no useful information.  Prior to this patch, the driver
allocates memory of the size of the the whole hw statistics modules,
instead of the size of statistics-entries-interested-in (i.e. et_stats),
for internal storage.  This patch fixes that.
Signed-off-by: NWingMan Kwok <w-kwok2@ti.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch fixes error in the setting of the hw statistics
module base for K2HK platform.  In K2HK although there are
4 hw statistics modules, but only 2 are visible at a time.
Thus when setting up the pointers to the base of the
corresponding hw statistics modules, modules 0 and 2 should
point to one base, while modules 1 and 3 should point to the
other.
Signed-off-by: NWingMan Kwok <w-kwok2@ti.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch fixes a bug in which the timer routine synchronized
against the ethtool-triggered statistics updates with spin_lock_bh().
A timer function is itself a bottom-half, so this should be
spin_lock().
Signed-off-by: NWingMan Kwok <w-kwok2@ti.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

VF driver was reading incorrect freelist congestion notification threshold
for FLM queues when packing is enabled for T5 and T6 adapter. Fixing it
now.
Signed-off-by: NHariprasad Shenai <hariprasad@chelsio.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

JUMBO and NO_GIGABIT_HALF have the same capability masks.
Change one of them.
Signed-off-by: NHarini Katakam <harinik@xilinx.com>
Acked-by: NNicolas Ferre <nicolas.ferre@atmel.com>
Acked-by: NAlexandre Belloni <alexandre.belloni@free-electrons.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 8d88c6eb ("net: bcmgenet: enable MoCA link state change
detection") added a fixed PHY link_update callback for MoCA PHYs when
registered using platform_data exclusively, this change is also
applicable to systems using Device Tree as their primary configuration
interface.

In order for this to work, move the link_update assignment into
bcmgenet_moca_phy_setup() where we know for sure that we are running on
a MoCA GENET instance, and do not override phydev->link since this is:

- properly taken care of by the PHY library by getting the link UP/DOWN
  interrupts
- this now runs everytime we call bcmgenet_open(), so we need to
  preserve whatever we detected before we went administratively DOWN and
  then UP
- we need to make sure that MoCA PHYs start with a link DOWN during
  probe in order to force a link transition to occur

To avoid a forward declaration, move bcmgenet_fixed_phy_link_update()
above its caller.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Instead of multiplying the number of checks for IS_ERR(priv->clk),
simply NULLify the 'struct clk' pointer which is something the Linux
common clock framework perfectly deals with and does early return for
each and every single clk_* API functions.

Having every single function check for !IS_ERR(priv->clk) is both
redundant and error prone, as it turns out, we were doing it for the
main GENET clock: priv->clk, but not for the Wake-on-LAN or EEE clock,
so let's just be consistent here.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Acked-by: NPetri Gynther <pgynther@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The current code returns from probe without waiting for the proper handling
of subchannels that may be requested. If the netvsc driver were to be rapidly
loaded/unloaded, we can  trigger a panic as the unload will be tearing
down state that may not have been fully setup yet. We fix this issue by making
sure that we return from the probe call only after ensuring that the
sub-channel offers in flight are properly handled.

Reviewed-and-tested-by: Haiyang Zhang <haiyangz@microsoft.com
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Adapter can go for a toss, if cxgb4 is loaded as slave and we try to
upgrade the firmware. So add a check for the same before flashing
firmware using ethtool.
Signed-off-by: NHariprasad Shenai <hariprasad@chelsio.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Silences the following sparse warnings:
drivers/net/vxlan.c:1818:21: warning: incorrect type in assignment (different base types)
drivers/net/vxlan.c:1818:21:    expected restricted __be32 [usertype] vx_vni
drivers/net/vxlan.c:1818:21:    got unsigned int [unsigned] [usertype] vni
drivers/net/vxlan.c:2014:58: warning: incorrect type in argument 11 (different base types)
drivers/net/vxlan.c:2014:58:    expected unsigned int [unsigned] [usertype] vni
drivers/net/vxlan.c:2014:58:    got restricted __be32 [usertype] <noident>

Fixes: 614732ea ("openvswitch: Use regular VXLAN net_device device")
Reported-by: Nkbuild test robot <fengguang.wu@intel.com>
Signed-off-by: NThomas Graf <tgraf@suug.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add be_get_phys_port_id() function to report physical port id. The port id
should be unique across different be2net devices in the system. We use the
chip serial number along with the physical port number for this.
Signed-off-by: NSriharsha Basavapatna <sriharsha.basavapatna@avagotech.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use macros for abstracting (1 << foo) to BIT(foo)
and (1ULL << foo64) to BIT_ULL(foo64) in order to match
better with kernel requirements.

NOTE: the adminq_cmd.h file was not modified on purpose because
of the dependency upon firmware for that file.

Change-ID: I73ee2e48c880d671948aad19bd53ca6b2ac558fc
Signed-off-by: NJesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: NCatherine Sullivan <catherine.sullivan@intel.com>
Tested-by: NJim Young <james.m.young@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Clean up a little confusion in reporting error status in phy and fcoe
setup error reports by separating the return status from the AQ error.

Add two decoder functions to make this easier.

Change-ID: I960bcdeef3978a15fec1cdb5eff781d5cbae42fb
Signed-off-by: NShannon Nelson <shannon.nelson@intel.com>
Tested-by: NJim Young <james.m.young@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

This driver fully supports VF drivers using both the 1.0 and 1.1
versions of the virtual channel API. However, VF drivers using
version 1.0 get upset if we provide them with a version other than
that, and refuse to play with us.

Correct this by checking the VFs API version at the time that we
store it off, and provide the correct version number back to the VF
so we can all get along.

Change-ID: I86dfe02e67b2bef336b4b49a1bb072f3e7229abc
Signed-off-by: NMitch Williams <mitch.a.williams@intel.com>
Tested-by: NJim Young <james.m.young@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Store off the PF's API version, then use it to determine whether or not
to send it our capabilities. Change the version checking to allow for PF
drivers with lower API versions than our current version, so we can
still talk to PF drivers over the 1.0 API.

Change-ID: I8edc55d1229c7decf0ed3f285a63032694007c2e
Signed-off-by: NMitch Williams <mitch.a.williams@intel.com>
Tested-by: NJim young <james.m.young@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

The most common type of reset that the VF will encounter is a PF reset
that cascades down into a VF reset for each VF. In this case, the VF
will always be assigned the same VSI and recovery is fairly simple.

However, in the case of 'bigger' resets, such as a Core or EMP reset,
when the device is reinitialized, it's probable that the VF will NOT get
the same VSI. When this happens, the VF will not be able to recover, as
it will continue to request resources for its original VSI.

Add an extra state to the admin queue state machine so that the driver
can re-request its configuration information at runtime. During reset
recovery, set this bit in the aq_required field, and fetch the (possibly
new) configuration information before attempting to bring the driver
back up. Since the driver doesn't know what kind of reset it has
encountered, this step is done even for a PF reset, but it doesn't hurt
anything - it just gets the same VSI back.

Change-ID: I915d59ffb40375215117362f4ac7a37811aba748
Signed-off-by: NMitch Williams <mitch.a.williams@intel.com>
Tested-by: NJim Young <james.m.young@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Store off the VF API version for use when figuring out the VF driver
capabilities. Add support for the VF driver handing its capabilities to
the PF driver and then use this information when sending VF resource
information back to the VF driver.

Change-ID: Ic00d0eeeb5b8118085e12f068ef857089a8f7c2d
Signed-off-by: NMitch Williams <mitch.a.williams@intel.com>
Tested-by: NJim Young <james.m.young@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Now that we've rolled the virtual channel API version to 1.1, add some
macros to test what version is being used by our partner in crime. For the
VF, add some macros to determine what our device capabilities are.

Change-ID: I79f6683d4c23bd76a8ad9fd492776fcc1208e1dc
Signed-off-by: NMitch Williams <mitch.a.williams@intel.com>
Tested-by: NJim Young <james.m.young@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

To prepare for the changes coming up in the X722 device and future
devices, the virtual channel interface has to change slightly. The VF
driver can now report what its capable of supporting, which then informs
the PF driver when it sends the configuration information back to the
VF.

A 1.1 VF driver on a 1.0 PF driver should not send its capabilities.
Likewise, a 1.1 PF driver controlling a 1.0 VF driver should not expect
or depend upon receiving the VF capabilities.

All other aspects of the API are unchanged.

Change-ID: I530cc55f107edd1ee8bdf95830aa90b87854058a
Signed-off-by: NMitch Williams <mitch.a.williams@intel.com>
Acked-by: NShannon Nelson <shannon.nelson@intel.com>
Acked-by: NAnjali Singhai <anjali.singhai@intel.com>
Tested-by: NJim Young <james.m.young@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

With a little work we can clean up some unnecessary logic jumping and
drop a variable.
Signed-off-by: NShannon Nelson <shannon.nelson@intel.com>
Cc: Laurent Navet <laurent.navet@gmail.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

This patch changes the switch statement for dynamic interrupt throttling
and adds a default case. With this patch, we check the latency setting
instead of the current ITR settings and the included refactor improves
performance.

Without this patch, the ITR setting would never change dynamically, and
there was no default.

Change-ID: Idb5a8a14c7109ec47c90f6e94bd43baa17d7ee37
Signed-off-by: NCarolyn Wyborny <carolyn.wyborny@intel.com>
Signed-off-by: NAnjali Singhai Jain <anjali.singhai@intel.com>
Tested-by: NJim Young <james.m.young@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Signed-off-by: NTodd Fujinaka <todd.fujinaka@intel.com>
Tested-by: NAaron Brown <aaron.f.brown@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Use the ARRAY_SIZE macro rather than calculating sizeof(a)/sizeof(a[0]).
Also directly replace the code rather than using an unnecessary define.
Reported-by: NManinder Singh <maninder1.s@samsung.com>
Reported-by: NJoe Perches <joe@perches.com>
Signed-off-by: NTodd Fujinaka <todd.fujinaka@intel.com>
Tested-by: NAaron Brown <aaron.f.brown@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

There are many settings possible using ethtool -C/--coalesce, but not
all of them are supported in igb. Report failure when an unsupported
option is set.
Signed-off-by: NTodd Fujinaka <todd.fujinaka@intel.com>
Tested-by: NAaron Brown <aaron.f.brown@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

The driver lacks pm_qos_remove_request in error handling (err_req_irq) of
e1000_open, and qos request inserted by pm_qos_add_request is not removed.
This patch add pm_qos_remove_request in error handling to fix it.
Signed-off-by: NJia-Ju Bai <baijiaju1990@163.com>
Tested-by: NAaron Brown <aaron.f.brown@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

This patch extends the reporting of the RSS key and hash table by
adding support for X550 VFs. The difference is that X550 VFs have
their own registers for RSS key and indirection table, so there is
no need to query the PF.

The RSS key and indirection table are stored in the adapter structure
during the configuration of VFRSSRK and VFRETA which in turn can be
used in ethtool for reporting.

The logic for writing VFRETA is also changed to make sure that the
indirection table is reported correctly.

In addition this patch adds defines for the VFRETA entries and number
of VFRSSRK registers as well as some whitespace cleanups.
Reported-by: NVlad Zolotarov <vladz@cloudius-systems.com>
Signed-off-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Tested-by: NPhil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Support the new USB gigabit ethernet.
Signed-off-by: NHayes Wang <hayeswang@realtek.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NAriel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Devices with up-to-date management FW will be able to store register dumps
on their persistent storage - in case management FW identifies a fatal
error it would gather and store such dumps, which could later be retrieved
using specific debug tools.

This patch adds the necessary part in the driver in order to make the
feature operational, as well as update users [under debug] during load
in case their device contains a dump of a previous crash.
Signed-off-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NAriel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This adds support to a new multi-function mode, enabling driver to
initialize such devices and correctly interacting with management FW
for fully utilizing their features.
Signed-off-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NAriel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This adds support to a new copper phy.
Signed-off-by: NYaniv Rosner <Yaniv.Rosner@qlogic.com>
Signed-off-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>