This queueing discipline implements the shaper algorithm defined by
the 802.1Q-2014 Section 8.6.8.2 and detailed in Annex L.

It's primary usage is to apply some bandwidth reservation to user
defined traffic classes, which are mapped to different queues via the
mqprio qdisc.

Only a simple software implementation is added for now.

Signed-off-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: Jesus Sanchez-Palencia <jesus.sanchez-palencia@intel.com>
Tested-by: Henrik Austad <henrik@austad.us>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

When replacing a child qdisc from mqprio, tc_modify_qdisc() must fetch
the netdev_queue pointer that the current child qdisc is associated
with before creating the new qdisc.

Currently, when using mqprio as root qdisc, the kernel will end up
getting the queue #0 pointer from the mqprio (root qdisc), which leaves
any new child qdisc with a possibly wrong netdev_queue pointer.

Implementing the Qdisc_class_ops select_queue() on mqprio fixes this
issue and avoid an inconsistent state when child qdiscs are replaced.

Signed-off-by: Jesus Sanchez-Palencia <jesus.sanchez-palencia@intel.com>
Tested-by: Henrik Austad <henrik@austad.us>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Currently, the class_ops select_queue() implementation on sch_mq
returns a pointer to netdev_queue #0 when it receives and invalid
qdisc id. That can be misleading since all of mq's inner qdiscs are
attached to a valid netdev_queue.

Here we fix that by returning NULL when a qdisc id is invalid. This is
aligned with how select_queue() is implemented for sch_mqprio in the
next patch on this series, keeping a consistent behavior between these
two qdiscs.

Signed-off-by: Jesus Sanchez-Palencia <jesus.sanchez-palencia@intel.com>
Tested-by: Henrik Austad <henrik@austad.us>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

In qdisc_alloc() the dev_queue pointer was used without any checks
being performed. If qdisc_create() gets a null dev_queue pointer, it
just passes it along to qdisc_alloc(), leading to a crash. That
happens if a root qdisc implements select_queue() and returns a null
dev_queue pointer for an "invalid handle", for example, or if the
dev_queue associated with the parent qdisc is null.

This patch is in preparation for the next in this series, where
select_queue() is being added to mqprio and as it may return a null
dev_queue.

Signed-off-by: Jesus Sanchez-Palencia <jesus.sanchez-palencia@intel.com>
Tested-by: Henrik Austad <henrik@austad.us>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Defer ringing the Tx doorbell if skb->xmit_more is set unless the Tx queue
is full or stopped.  To keep latency low, use a deferral limit of 8
packets.  We chose 8 because Octeon can fetch at most 8 packets in a single
PCI read, and our tests show that 8 results in low latency.

Signed-off-by: Intiyaz Basha <intiyaz.basha@cavium.com>
Signed-off-by: Satanand Burla <satananda.burla@cavium.com>
Signed-off-by: Felix Manlunas <felix.manlunas@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

John Allen says:

====================
ibmvnic: Tunable parameter support

This series implements support for changing tunable parameters such as the
mtu, number of tx/rx queues, and number of buffers per queue via ethtool
and ifconfig.

v2: -Fix conflict with Tom's recently applied TSO/SG patches
v3: -Initialize rc in __ibmvnic_reset fixing build warning
    -Fix buggy behavior with pending mac changes. Use boolean flag
     to track if mac change is needed on open rather than relying on
     the desired->mac pointer.
    -Directly include tunable structs in the adapter struct rather
     than keeping pointers, eliminating the need to directly allocate
     them.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

For all non-fatal reset conditions, the hypervisor will send a failover when
we attempt to initialize the crq and the vnic client is expected to handle
that failover instead of the existing non-fatal reset. To handle this, we
need to return from init with a return code that indicates that we have hit
this case.

Signed-off-by: John Allen <jallen@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Update ibmvnic reset infrastructure to include a new reset option that will
allow changing of tunable parameters. There currently is no way to request
different capabilities from the vnic server on the fly so this patch
achieves this by resetting the driver and attempting to log in with the
requested changes. If the reset operation fails, the old values of the
tunable parameters are stored in the "fallback" struct and we attempt to
login with the fallback values.

Signed-off-by: John Allen <jallen@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Subash Abhinov Kasiviswanathan says:

====================
net: qualcomm: rmnet: Add 64 bit stats and GRO

This series adds support for 64 bit per cpu stats and GRO

Patches 1-2 are cleanups of return code and a redundant condition
Patch 3 adds support for 64 bit per cpu stats
Patch 4 adds support for GRO using GRO cells

v1->v2: Since gro_cells_init() could potentially fail, move it from device
setup to ndo_init() as mentioned by Eric.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

Add gro_cells so that rmnet devices can call gro_cells_receive
instead of netif_receive_skb.

Signed-off-by: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Implement 64 bit per cpu stats.

Signed-off-by: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

The rmnet device needs to assigned for all packets in the
deaggregation path based on the mux id, so the check is not needed.

Signed-off-by: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Since packet is always consumed by rmnet_rx_handler(), we always
return RX_HANDLER_CONSUMED. There is no need to pass on this
value through multiple functions.

Signed-off-by: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Michael Chan says:

====================
bnxt_en: Updates for net-next.

This series includes firmware interface update, some optimizations,
some new PCI IDs, new MTU checks, ethtool reset method, interrupt coalescing
code cleanup, and TC flower offload for vxlan encap/decap from Sathya
Perla.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

TC flower is not enabled on VFs and when there's no FW support.
Alloc the tc_info{} struct at init time only when TC flower is being
enabled.

Signed-off-by: Sathya Perla <sathya.perla@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch implements periodic querying of cfa flow stats
in batches to compute the 'lastused' attribute of TC flow stats.

Signed-off-by: Sathya Perla <sathya.perla@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add routines for issuing the hwrm_cfa_encap_record_alloc/free
and hwrm_cfa_decap_filter_alloc/free FW cmds needed for
supporting vxlan encap/decap offload.

Signed-off-by: Sathya Perla <sathya.perla@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch adds IPv4 vxlan encap/decap action support to TC-flower
offload.

For vxlan encap, the driver maintains a tunnel encap hash-table.
When a new flow with a tunnel encap action arrives, this table
is looked up; if an encap entry exists, it uses the already
programmed encap_record_handle as the tunnel_handle in the
hwrm_cfa_flow_alloc cmd. Else, a new encap node is added and the
L2 header fields are queried via a route lookup.
hwrm_cfa_encap_record_alloc cmd is used to create a new encap
record and the encap_record_handle is used as the tunnel_handle
while adding the flow.

For vxlan decap, the driver maintains a tunnel decap hash-table.
When a new flow with a tunnel decap action arrives, this table
is looked up; if a decap entry exists, it uses the already
programmed decap_filter_handle as the tunnel_handle in the
hwrm_cfa_flow_alloc cmd. Else, a new decap node is added and
a decap_filter_handle is alloc'd via the hwrm_cfa_decap_filter_alloc
cmd. This handle is used as the tunnel_handle while adding the flow.

The code to issue the HWRM FW cmds is introduced in a follow-up patch.

Signed-off-by: Sathya Perla <sathya.perla@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The mapping of the ethtool coalescing parameters to hardware parameters
is now done in bnxt_hwrm_set_coal_params().  The same function can
handle both RX and TX settings.  The code is now more clear.  Some
adjustments have been made to get better hardware settings.  The
coal_frames setting is now accurately set in hardware.  The max_timer
is set to coal_ticks value.

Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The current IRQ coalescing logic is a little messy.  The ethtool
parameters are mapped to hardware parameters in a way that is difficult
to understand.  The first step is to better organize the parameters
by adding the new structure bnxt_coal.  The structure is used by both
the RX and TX sets of coalescing parameters.

Adjust the default coal_ticks to 14 us and 28 us for RX and TX.

Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This is a firmware internal reset after driver is unloaded.

Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Some NICs have a firmware enforced maximum MTU setting by management
firmware.  Set up netdev->max_mtu accordingly.

Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

No need to call bnxt_approve_mac() which will send a message to the
PF if the MAC address hasn't changed.

Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The current code retrieves the firmware package version from firmware
everytime ethtool -i is run.  There is no reason to do that as the
firmware will not change while the driver is loaded.  Get the version
once at init time.

Also, display the full 4-part firmware version string and remove the
less useful interface spec version.

Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Return -EINVAL if the length is zero and not proceed to do essentially
nothing.

Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Rob Miller <rmiller@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add new PCIe device ID and chip number for bcm58804

Signed-off-by: Ray Jui <ray.jui@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Vxlan encap/decap filters are added to this firmware spec.

Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Vivien Didelot says:

====================
net: dsa: define port types

The DSA code currently has 3 bitmaps in the dsa_switch structure:
cpu_port_mask, dsa_port_mask and enabled_port_mask.

They are used to store the type of each switch port. This dates back
from when DSA didn't have a dsa_port structure to hold port-specific
data.

The dsa_switch structure is mainly used to communicate with DSA drivers
and must not contain such static data parsed from DTS or pdata, which
belongs the DSA core structures, such as dsa_switch_tree and dsa_port.

Also the enabled_port_mask is misleading, often misinterpreted as the
complement of disabled ports (thus including DSA and CPU ports), while
in fact it only masks the user ports.

A port can be of 3 types when it is not unused: "cpu" (interfacing with
a master device), "dsa" (interconnecting with another "dsa" port from
another switch chip), or "user" (user-facing port.)

This patchset first fixes the usage of DSA port type helpers, then
defines the DSA_PORT_TYPE_UNUSED, DSA_PORT_TYPE_CPU, DSA_PORT_TYPE_DSA,
and DSA_PORT_TYPE_USER port types, and finally removes the misleading
port bitmaps.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

Now that DSA core provides port types, there is no need to keep this
information at the switch level. This is a static information that is
part of a DSA core dsa_port structure. Remove them.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Now that DSA exposes an enumerated type for the ports, we can use them
directly instead of checking bitmaps, which is more consistent.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Introduce an enumerated type for ports, which will be way more explicit
to identify a port type instead of digging into switch port masks.

A port can be of type CPU, DSA, user, or unused by default. This is a
static parsed information that cannot be changed at runtime.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Introduce a dsa_user_ports() helper to return the ds->enabled_port_mask
mask which is more explicit. This will also minimize diffs when touching
this internal mask.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Most of the DSA code still check ds->enabled_port_mask directly to
inspect a given port type instead of using the provided dsa_is_user_port
helper. Change this.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch renames dsa_is_normal_port to dsa_is_user_port because "user"
is the correct term in the DSA terminology, not "normal".

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In order to know if a port is of type user, dsa_is_normal_port checks
that the given port is not of type DSA nor CPU. This is not enough
because a port can be unused.

Without the previous fix, this caused the unused mv88e6xxx ports to be
configured in normal mode.

The ds->enabled_port_mask reports the user ports, so check this instead.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The unused ports are currently configured in normal mode. This does not
prevent the switch from being functional, but it is unnecessary. Skip
unused ports.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

As the comment above the chunk states, the b53 driver attempts to
disable the unused ports. But using ds->enabled_port_mask is misleading,
because this mask reports in fact the user ports.

To avoid confusion and fix this, this patch introduces an explicit
dsa_is_unused_port helper which ensures the corresponding bit is not
masked in any of the switch port masks.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Notice that in this particular case unlikely() is already being called
inside BUG_ON macro.

This issue was detected with the help of Coccinelle.

Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>