Put the current logic in bnxt_open() to re-enable SRIOV after detecting
firmware reset into a new function bnxt_reenable_sriov().  This call
needs to be invoked in the firmware reset path also in the next patch.
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NJakub Kicinski <kuba@kernel.org>

Display the following information via devlink info command:
  - Driver name
  - Board id
  - Broad revision
  - Board Serial number
  - Board FW version
  - FW parameter set version
  - FW App version
  - FW management version
  - FW RoCE version

  Standard output example:
  $ devlink dev info pci/0000:3b:00.0
  pci/0000:3b:00.0:
  driver bnxt_en
  serial_number 00-10-18-FF-FE-AD-05-00
  versions:
      fixed:
        asic.id D802
        asic.rev 1
      running:
        fw 216.1.124.0
        fw.psid 0.0.0
        fw.app 216.1.122.0
        fw.mgmt 864.0.32.0
        fw.roce 216.1.15.0

[ This version has incorporated changes suggested by Jakub Kicinski to
  use generic devlink version tags. ]

v2: Use fw.psid

Cc: Jiri Pirko <jiri@mellanox.com>
Cc: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: NVasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Instead of switch_id, renaming it to dsn will be more meaningful
so that it can be used to display device serial number in follow up
patch via devlink_info command.
Signed-off-by: NVasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds status notification to devlink flash update
while flashing is in progress.

$ devlink dev flash pci/0000:05:00.0 file 103.pkg
Preparing to flash
Flashing done

Cc: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: NVasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Latest kernels get the phys_port_name via devlink, if
ndo_get_phys_port_name is not defined. To provide the phys_port_name
correctly, register devlink before registering netdev.

Also call devlink_port_type_eth_set() after registering netdev as
devlink port updates the netdev structure and notifies user.

Cc: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: NVasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This will allow to register for devlink port and use port features.
Also register params only if firmware spec version is at least 0x10600
which will support reading/setting numbered variables in NVRAM.
Signed-off-by: NVasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Define bnxt_dl_params_register() and bnxt_dl_params_unregister()
functions and move params register/unregister code to these newly
defined functions. This patch is in preparation to register
devlink irrespective of firmware spec. version in the next patch.
Signed-off-by: NVasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The hardware bug has been fixed on B0 and newer chips, so disable the
workaround on these chips.
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently the only time we check and remove expired filters is
when we are inserting new filters.
Improving the aRFS expiry handling by adding code to do the above
work periodically.
Signed-off-by: NPavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In bnxt_rx_flow_steer(), if the dissected packet is a fragment, do not
proceed to create the ntuple filter and return error instead.  Otherwise
we would create a filter with 0 source and destination ports because
the dissected ports would not be available for fragments.
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

575XX (P5) chips have the same UDP RSS hashing capability as P4 chips,
so we can enable it on P5 chips.
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The dev_port is meant to distinguish the network ports belonging to
the same PCI function.  Our devices only have one network port
associated with each PCI function and so we should not set it for
correctness.
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If the 2nd parameter fw_dflt is not set, we are calling bnxt_probe_phy()
after the firmware has reset.  There is no need to query the current
PHY settings from firmware as these settings may be different from
the ethtool settings that the driver will re-establish later.  So
return earlier in bnxt_probe_phy() to save one firmware call.
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In bnxt_update_phy_setting(), ethtool_get_link_ksettings() and
bnxt_disable_an_for_lpbk(), we inconsistently use netif_carrier_ok()
to determine link.  Instead, we should use bp->link_info.link_up
which has the true link state.  The netif_carrier state may be off
during self-test and while the device is being reset and may not always
reflect the true link state.

By always using bp->link_info.link_up, the code is now more
consistent and more correct.  Some unnecessary link toggles are
now prevented with this patch.
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Before commit 7587935c ("net: bcmgenet: move NAPI initialization to
ring initialization") moved the code, this used to be
netif_tx_napi_add(), but we lost that small semantic change in the
process, restore that.

Fixes: 7587935c ("net: bcmgenet: move NAPI initialization to ring initialization")
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Acked-by: NDoug Berger <opendmb@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Convert suitable drivers to use new helper phy_do_ioctl_running.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Acked-by: NTimur Tabi <timur@kernel.org>
Acked-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

DSN read can fail, for example on a kdump kernel without PCIe extended
config space support.  If DSN read fails, don't set the
BNXT_FLAG_DSN_VALID flag and continue loading.  Check the flag
to see if the stored DSN is valid before using it.  Only VF reps
creation should fail without valid DSN.

Fixes: 03213a99 ("bnxt: move bp->switch_id initialization to PF probe")
Reported-by: NMarc Smith <msmith626@gmail.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix bnxt_fltr_match() to match ipv6 source and destination addresses.
The function currently only checks ipv4 addresses and will not work
corrently on ipv6 filters.

Fixes: c0c050c5 ("bnxt_en: New Broadcom ethernet driver.")
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The NTUPLE related firmware commands are sent to the wrong firmware
channel, causing all these commands to fail on new firmware that
supports the new firmware channel.  Fix it by excluding the 3
NTUPLE firmware commands from the list for the new firmware channel.

Fixes: 760b6d33 ("bnxt_en: Add support for 2nd firmware message channel.")
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We would not be transmitting using the correct SYSTEMPORT transmit queue
during ndo_select_queue() which looks up the internal TX ring map
because while establishing the mapping we would be off by 4, so for
instance, when we populate switch port mappings we would be doing:

switch port 0, queue 0 -> ring index #0
switch port 0, queue 1 -> ring index #1
...
switch port 0, queue 3 -> ring index #3
switch port 1, queue 0 -> ring index #8 (4 + 4 * 1)
...

instead of using ring index #4. This would cause our ndo_select_queue()
to use the fallback queue mechanism which would pick up an incorrect
ring for that switch port. Fix this by using the correct switch queue
number instead of SYSTEMPORT queue number.

Fixes: 25c44070 ("net: systemport: Simplify queue mapping logic")
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When running in XDP mode, pages come from the page pool, and should
be freed back to the same pool or specifically detached.  Currently,
when the driver re-initializes, the page pool destruction is delayed
forever since it thinks there are oustanding pages.

Fixes: 322b87ca ("bnxt_en: add page_pool support")
Signed-off-by: NJonathan Lemon <jonathan.lemon@gmail.com>
Reviewed-by: NAndy Gospodarek <gospo@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Return EINPROGRESS to devlink health reporter recover as we are not yet
done and call devlink_health_reporter_recovery_done once reset is
successfully completed from workqueue context.
Signed-off-by: NVikas Gupta <vikas.gupta@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is a straight-forward conversion case for the new function, and
while we're at it, we can remove a null write to skb->next by replacing
it with skb_mark_not_on_list.
Signed-off-by: NJason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

ioremap has provided non-cached semantics by default since the Linux 2.6
days, so remove the additional ioremap_nocache interface.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Acked-by: NArnd Bergmann <arnd@arndb.de>

Atomic operations that span cache lines are super-expensive on x86
(not just to the current processor, but also to other processes as all
memory operations are blocked until the operation completes). Upcoming
x86 processors have a switch to cause such operations to generate a #AC
trap. It is expected that some real time systems will enable this mode
in BIOS.

In preparation for this, it is necessary to fix code that may execute
atomic instructions with operands that cross cachelines because the #AC
trap will crash the kernel.

Since "pwol_mask" is local and never exposed to concurrency, there is
no need to set bits in pwol_mask using atomic operations.

Directly operate on the byte which contains the bit instead of using
__set_bit() to avoid any big endian concern due to type cast to
unsigned long in __set_bit().
Suggested-by: NPeter Zijlstra <peterz@infradead.org>
Signed-off-by: NFenghua Yu <fenghua.yu@intel.com>
Signed-off-by: NTony Luck <tony.luck@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

While testing max vlan configuration on the PF, firmware gets
assert as driver was configuring number of vlans more than what
is supported per port/engine, it was figured out that there is an
implicit vlan (hidden default vlan consuming hardware cam entry resource)
which is configured default for all the clients (PF/VFs) on client_init
ramrod by the adapter implicitly, so when allocating resources among the
PFs this implicit vlan should be considered or total vlan entries should
be reduced by one to accommodate that default/implicit vlan entry.
Signed-off-by: NManish Chopra <manishc@marvell.com>
Signed-off-by: NAriel Elior <aelior@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Although it has same value as MAX_MAC_CREDIT_E2,
use MAX_VLAN_CREDIT_E2 appropriately.
Signed-off-by: NManish Chopra <manishc@marvell.com>
Signed-off-by: NAriel Elior <aelior@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The 1588 standard defines one step operation for both Sync and
PDelay_Resp messages.  Up until now, hardware with P2P one step has
been rare, and kernel support was lacking.  This patch adds support of
the mode in anticipation of new hardware developments.
Signed-off-by: NRichard Cochran <richardcochran@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When inserting the TSB, keep track of how many times we had to do
it and if there was a failure in doing so, this helps profile the
driver for possibly incorrect headroom settings.
Signed-off-by: NDoug Berger <opendmb@gmail.com>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

During bcmgenet_put_tx_csum() make sure we differentiate a SKB
headroom re-allocation failure from the normal swap and replace
path.
Signed-off-by: NDoug Berger <opendmb@gmail.com>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We can turn on the RX/TX checksum offloads and the scatter/gather
features by default and make sure that those are properly reflected
back to e.g: stacked devices such as VLAN.
Signed-off-by: NDoug Berger <opendmb@gmail.com>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

During driver resume and open, the HW may have lost its context/state,
utilize bcmgenet_set_features() to make sure we do restore the correct
set of features that were previously configured.
Signed-off-by: NDoug Berger <opendmb@gmail.com>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In preparation for unconditionally enabling TX and RX checksum
offloads, refactor bcmgenet_set_features() a bit such that
__netdev_update_features() during register_netdev() can make sure
that features are correctly programmed during network device
registration.

Since we can now be called during register_netdev() with clocks
gated, we need to temporarily turn them on/off in order to have a
successful register programming.

We also move the CRC forward setting read into
bcmgenet_set_features() since priv->crc_fwd_en matters while
turning on RX checksum offload, that way we are guaranteed they
are in sync in case we ever add support for NETIF_F_RXFCS at some
point in the future.
Signed-off-by: NDoug Berger <opendmb@gmail.com>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This commit updates the Rx checksum offload behavior of the driver
to use the more generic CHECKSUM_COMPLETE method that supports all
protocols over the CHECKSUM_UNNECESSARY method that only applies
to some protocols known by the hardware.

This behavior is perceived to be superior.
Signed-off-by: NDoug Berger <opendmb@gmail.com>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The GENET hardware should be capable of generating IP checksums
using the NETIF_F_HW_CSUM feature, so switch to using that feature
instead of the depricated NETIF_F_IP_CSUM and NETIF_F_IPV6_CSUM.
Signed-off-by: NDoug Berger <opendmb@gmail.com>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This commit configures the DMA masks for the GENET driver and
sets the NETIF_F_HIGHDMA flag to report support of the feature.
Signed-off-by: NDoug Berger <opendmb@gmail.com>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

SYSTEMPORT is capabable of doing up to 40-bit of physical addresses, set
an appropriate DMA mask to permit that.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Driver doesn't calculate total number of PFs configured on a
given engine correctly which messed up resources in the PFs
loaded on that engine, leading driver to exceed configuration
of resources (like vlan filters etc.) beyond the limit per
engine, which ended up with asserts from the firmware.
Signed-off-by: NManish Chopra <manishc@marvell.com>
Signed-off-by: NAriel Elior <aelior@marvell.com>
Signed-off-by: NJakub Kicinski <jakub.kicinski@netronome.com>

Parity error from the hardware will cause PF to lose the state
of their VFs due to PF's internal reload and hardware reset following
the parity error. Restrict any configuration request from the VFs after
the parity as it could cause unexpected hardware behavior, only way
for VFs to recover would be to trigger FLR on VFs and reload them.
Signed-off-by: NManish Chopra <manishc@marvell.com>
Signed-off-by: NAriel Elior <aelior@marvell.com>
Signed-off-by: NJakub Kicinski <jakub.kicinski@netronome.com>

After executing "ethtool -C eth0 rx-usecs-irq 0", the box becomes
unresponsive, likely due to interrupt livelock.  It appears that
a minimum clamp value for the irq timer is computed, but is never
applied.

Fix by applying the corrected clamp value.

Fixes: 74706afa ("bnxt_en: Update interrupt coalescing logic.")
Signed-off-by: NJonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NJakub Kicinski <jakub.kicinski@netronome.com>