The header ocelot_dev_gmii.h is unused since the inclusion of the driver.
It is unused, lets just remove it.
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Ido Schimmel says:

====================
mlxsw: Further MC-awareness configuration

Petr says:

Due to an issue in Spectrum chips, when unicast traffic shares the same
queue as BUM traffic, and there is congestion, the BUM traffic is
admitted to the queue anyway, thus pushing out all UC traffic. In order
to give unicast traffic precedence over BUM traffic, multicast-aware
mode is now configured on all ports. Under MC-aware mode, egress TCs
8..15 are used for BUM traffic, which has its own dedicated pool.

This patch set improves the way that the MC pool and the higher-order
TCs are integrated into the system.

In patch #1, shaper at the higher TCs is configured to the same value
that it has by default. It's better to have the corresponding artifact
in the code explicitly.

The 8 following patches gradually extend the devlink handling in mlxsw
to support the extra TCs and the new MC pool.

Patch #2 changes the way that pools are indexed in mlxsw. Instead of
using (FW index, direction) tuple to identify the pool and the
associated cache, mlxsw now uses devlink index. This change is necessary
because the new pool 15 is not contiguously adjacent to the
currently-used pools 0..3, and because it's only relevant on egress.
Using devlink index relaxes the requirement for symmetry and adjacency
imposed by using FW indexing.

In patch #3, the assumption that number of ingress TCs matches that of
egress TCs is relaxed to allow exposition of egress TCs 8..15.

In patches #4, #5 and #6, support for infinite quotas is introduced.
Infinite quotas are reported as taking all the memory in the system, but
actually use a mechanism where the infinity is configured explicitly.

In patches #7 and #8, support for configuring static pool sizes in
introduced. Statically-sized pools have been supported for a while now,
but during initialization, all pools have dynamic size. The patches
allow there to be a mix of by-default static and dynamic pools.

In patches #9 and #10, pool 15 resp. per-priority MC quotas are
explicitly configured to be in sync with the current recommendation for
handling BUM traffic in Spectrum chips.

In the following 3 patches, an mlxsw-specific selftest is added to test
the MC-awareness configuration.

First in patches #11 and #12, lib.sh is extended with functions to
collect ethtool stats, and to manage port MTU.

Then in patch #13 the selftest itself is added.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

A so-called "MC-aware" mode has recently been enabled in mlxsw. In
MC-aware mode, BUM traffic is handled in a special way so that when a
switch is flooded with BUM, UC performance isn't unduly impacted.
Without enablement of this mode, a stream of BUM traffic can cause
sustained UC throughput drop in excess of 99 %.

Add a test for this behavior. Compare how much UC throughput degrades as
a stream of broadcast frames floods the switch. A minimal degradation is
tolerated to cover for glitches in traffic injection performance.
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Some selftests need to tweak MTU of an interface, and naturally should
at teardown restore the MTU back to the original value. Add two
functions to facilitate this MTU handling: mtu_set() to change MTU
value, and mtu_reset() to change it back to what it was before.
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add a new service function to obtain ethtool counters.
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The SBMM register configures shared buffer allocation and settings for
MC packets according to switch priority. The recommended values are no
reserved buffer and alpha of 1/4, which corresponds to buf_max of 6.
Update mlxsw_sp_sb_mms accordingly.
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Pool 15 (indexed as 8) is dedicated to MC traffic. Its configuration has
been kept at default, because the table-based configuration wasn't
expressive enough to allow the explicit configuration.

Now that the configuration of pool 15 can be described, do so. The MC
pool should have infinite size, infinite per-TC quota, and per-port
limit of 90K.
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Some pools configured through the sb_pm entries may have by default
static size. The MC pool is now not explicitly configured, however it
gets configured as static implicitly by 0-initializing sb->prs, and a
follow-up patch adds an explicit configuration to the same effect.

To support this, pass max_buff taken from sb_pm and sb_cm entries
through cell conversion before handing it to mlxsw_sp_sb_pm_write(), if
the pool that the sb_pm entry configures is statically-sized.

To keep current behavior, update mlxsw_sp_sb_cms_egress[] to denote
buffer sizes in bytes (assuming Spectrum 1 cell sizes, which the
original code assumed as well) instead of cells. Note that a follow-up
patch changes this to infinite size.

Also tweak a comment at SBMM configuration to remain true now that
statically-sized pools exist.
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The SBPM register configures the shared buffer allocation and
configuration per port and pool. The min_buff value is the buffer size
dedicated to this single function, and is configured in cells.
Currently, all sb_pm entries have 0 for min_buff, and therefore the
actual unit is immaterial. However, in a follow-up patch we want to add
entries with non-zero minimum.

Therefore pass the min_buff from the sb_pm table through the cell
conversion before handing it over to mlxsw_sp_sb_pm_write().
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The SBCM register configures the shared buffer configuration according
to port and TC. So far all pools have had a dynamic size, where the
infinite size is easy to express by using max_buff of 0xff. However the
MC pool should be configured with static size, and the infinite size
thus needs to be set using the field SBCM.infi_max.

Therefore add the field infi_max to the SBCM register and to
mlxsw_reg_sbcm_pack(). Extend mlxsw_sp_sb_cm_write() to handle infinite
sizes as well. Report infinite pool limits as if the limit actually were
the total shared buffer size.
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The MC pool should have an infinite size (i.e. no quota).

To that end, add infi_size to the SBPR register and extend
mlxsw_reg_sbpr_pack(). Also add MLXSW_SP_SB_INFI to denote
buffers that should have an infinite size.

Change mlxsw_sp_sb_pr_write() to take as parameter byte size,
instead of cell size, and add the special handling of infinite
buffers. Report pools with infinite size as if they actually
take the full shared buffer size.
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Entities of infinite size will be reported as if they had the maximum
size allowed by the chip. To that end, keep track of maximum shared
buffer size in mlxsw_sp->sb.
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Current code assumes that ingress and egress has the same number of
traffic classes. Since the introduction of MC-aware mode that assumption
hasn't held anymore, and there have been 16 TCs on the egress as opposed
to 8 on ingress.

Break the assumption of symmetry by splitting the artifacts related to
shared-buffer TC counting to ingress and egress parts.
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, mlxsw assumes that each ingress pool has its egress
counterpart, and that pool index for purposes of caching matches the
index with which the hardware should be configured. As we want to expose
the MC pool, both of these assumptions break.

Instead, maintain the pool index as long as possible. Unify ingress and
egress caches and use the pool index as cache index as well. Only
translate to FW pool numbering when actually packing the registers. This
simplifies things considerably, as the pool index is the only quantity
necessary to uniquely identify a pool, and the pool/direction split is
not necessary until firmware is talked to.

To support the mapping between pool indices and pool numbers and
directions, which is not neatly mathematical anymore, introduce a pool
descriptor table, indexed by pool index, to facilitate the translation.

Include the MC pool in the descriptor table as well, so that it can be
referenced from mlxsw_sp_sb_cms_egress.
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

With introduction of MC-aware mode to mlxsw, it became necessary to
configure TCs above 7 as well. There is now code in mlxsw to disable ETS
for these higher classes, but disablement of max shaper was neglected.

By default, max shaper is currently disabled to begin with, so the
problem is just cosmetic. However, for symmetry, do like we do for ETS
configuration, and call mlxsw_sp_port_ets_maxrate_set() for both TC i
and i + 8.
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Clang warns when more than one set of parentheses are used for single
conditional statements:

drivers/isdn/hisax/w6692.c:627:30: warning: equality comparison with
extraneous parentheses [-Wparentheses-equality]
                if ((cs->dc.w6692.ph_state == W_L1IND_DRD)) {
                     ~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~
drivers/isdn/hisax/w6692.c:627:30: note: remove extraneous parentheses
around the comparison to silence this warning
                if ((cs->dc.w6692.ph_state == W_L1IND_DRD)) {
                    ~                      ^             ~
drivers/isdn/hisax/w6692.c:627:30: note: use '=' to turn this equality
comparison into an assignment
                if ((cs->dc.w6692.ph_state == W_L1IND_DRD)) {
                                           ^~
                                           =
1 warning generated.

Remove the parentheses to silence this warning.
Reported-by: NNick Desaulniers <ndesaulniers@google.com>
Signed-off-by: NNathan Chancellor <natechancellor@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fixes gcc '-Wunused-but-set-variable' warning:

drivers/net/ethernet/atheros/atlx/atl1.c: In function 'atl1_set_link_ksettings':
drivers/net/ethernet/atheros/atlx/atl1.c:3280:6: warning:
 variable 'advertising' set but not used [-Wunused-but-set-variable]
Signed-off-by: NYueHaibing <yuehaibing@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fixes gcc '-Wunused-but-set-variable' warning:

drivers/net/ethernet/intel/e1000/e1000_main.c: In function 'e1000_watchdog':
drivers/net/ethernet/intel/e1000/e1000_main.c:2436:9: warning:
 variable 'txb2b' set but not used [-Wunused-but-set-variable]
Signed-off-by: NYueHaibing <yuehaibing@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The local variable 'index_specified' is never used after being assigned.
hence it should be redundant adn can be removed.
Signed-off-by: Nzhong jiang <zhongjiang@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

NFP supports fairly enormous ring sizes (up to 256k descriptors).
In commit 46627170 ("nfp: use kvcalloc() to allocate SW buffer
descriptor arrays") we have started using kvcalloc() functions to
make sure the allocation of software state arrays doesn't hit
the MAX_ORDER limit.  Unfortunately, we can't use virtual mappings
for the DMA region holding HW descriptors.  In case this allocation
fails instead of the generic (and fairly scary) warning/splat in
the logs print a helpful message explaining what happened and
suggesting how to fix it.
Signed-off-by: NJakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: NDirk van der Merwe <dirk.vandermerwe@netronome.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Setting register 0x82 to value 01 is done a few lines before for all
chip versions <= 06 anyway. And setting PHY register 0x0b to value 00
is done at the end of rtl8169s_hw_phy_config() already. So we can
remove this.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

PCI_LATENCY_TIMER is ignored on PCIe, therefore we have to do this
for the PCI chips (version <= 06) only. Also we can move setting
PCI_CACHE_LINE_SIZE.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

That local variable are never used after being assigned.
hence it should be redundant and can be removed.
Signed-off-by: Nzhong jiang <zhongjiang@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The local variable 'k' is never used after being assigned.
hence it should be redundant adn can be removed.
Signed-off-by: Nzhong jiang <zhongjiang@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There is no way currently for an IPv6 client connect using a loopback
address in a VRF, whereas for IPv4 the loopback address can be added:

    $ sudo ip addr add dev vrfred 127.0.0.1/8
    $ sudo ip -6 addr add ::1/128 dev vrfred
    RTNETLINK answers: Cannot assign requested address

So allow ::1 to be configured on an L3 master device. In order for
this to be usable ip_route_output_flags needs to not consider ::1 to
be a link scope address (since oif == l3mdev and so it would be
dropped), and ipv6_rcv needs to consider the l3mdev to be a loopback
device so that it doesn't drop the packets.
Signed-off-by: NRobert Shearman <rshearma@vyatta.att-mail.com>
Signed-off-by: NMike Manning <mmanning@vyatta.att-mail.com>
Reviewed-by: NDavid Ahern <dsahern@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Salil Mehta says:

====================
Fixes, cleanups & minor additions to HNS3 driver

This patch-set present some fixes, cleanups to the HNS3 PF and VF driver.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

So far all the places calling hclge_tm_q_to_qs_map_cfg() are assigning
an u16 type value to "q_id", and in the processing of
hclge_tm_q_to_qs_map_cfg(), it also converts the "q_id" to le16.

The max tqp number for pf can be more than 256, we should use "u16" to
store the queue id, instead of "u8", which may cause data lost.

Fixes: 84844054 ("net: hns3: Add support of TX Scheduler & Shaper to HNS3 driver")
Signed-off-by: NJian Shen <shenjian15@huawei.com>
Signed-off-by: NPeng Li <lipeng321@huawei.com>
Signed-off-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When roce is loaded before nic, the roce client will not be initialized
until nic client is initialized, but roce init flag is set before it.
Furthermore, in this case of nic initialized success and roce failed,
the nic init flag is not set, and roce init flag is not cleared.

This patch fixes it by set init flag only after the client is initialized
successfully.

Fixes: e2cb1dec ("net: hns3: Add HNS3 VF HCL(Hardware Compatibility Layer) Support")
Fixes: 46a3df9f ("net: hns3: Add HNS3 Acceleration Engine & Compatibility Layer Support")
Signed-off-by: NJian Shen <shenjian15@huawei.com>
Signed-off-by: NPeng Li <lipeng321@huawei.com>
Signed-off-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If initialize client failed or finish uninitializing client, we should
clear the client pointer. It may cause unexpected result when use
uninitialized client. Meanwhile, we also should check whether client
exist when uninitialize it.

Fixes: 46a3df9f ("net: hns3: Add HNS3 Acceleration Engine & Compatibility Layer Support")
Signed-off-by: NJian Shen <shenjian15@huawei.com>
Signed-off-by: NPeng Li <lipeng321@huawei.com>
Signed-off-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

According to hardware's description, the head pointer register should
be written before the tail pointer register while initializing the vf
command queue. Otherwise, it may trigger an interrupt even though there
is no command received.

Fixes: fedd0c15 ("net: hns3: Add HNS3 VF IMP(Integrated Management Proc) cmd interface")
Signed-off-by: NJian Shen <shenjian15@huawei.com>
Signed-off-by: NPeng Li <lipeng321@huawei.com>
Signed-off-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The function of genphy_read_status is that reading phy information
from HW and using these information to update SW variable. If user
is using ethtool to setting the speed of phy and service task is calling
by hclge_get_mac_phy_link, the result of speed setting is uncertain.
Because ethtool cmd will modified phydev and hclge_get_mac_phy_link also
will modified phydev.

Because phy state machine will update phy link periodically, we can
just use phydev->link to check the link status. This patch removes
function call of genphy_read_status. To ensure accuracy, this patch
adds a phy state check. If phy state is not PHY_RUNNING, we consider
link is down. Because in some scenarios, phydev->link may be link up,
but phy state is not PHY_RUNNING. This is just an intermediate state.
In fact, the link is not ready yet.

Fixes: 46a3df9f ("net: hns3: Add HNS3 Acceleration Engine & Compatibility Layer Support")
Signed-off-by: NFuyun Liang <liangfuyun1@huawei.com>
Signed-off-by: NPeng Li <lipeng321@huawei.com>
Signed-off-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

By default, HW link status is up. If hclge_update_link_status is called
before net up, driver will print "link up". It is not suitable. hdev
state check is needed when getting link status.

Fixes: 46a3df9f ("net: hns3: Add HNS3 Acceleration Engine & Compatibility Layer Support")
Fixes: e2cb1dec ("net: hns3: Add HNS3 VF HCL(Hardware Compatibility Layer) Support")
Signed-off-by: NFuyun Liang <liangfuyun1@huawei.com>
Signed-off-by: NPeng Li <lipeng321@huawei.com>
Signed-off-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We clear STATE_DOWN bit of hdev state when starting net, but do not set
it again when stopping net. It causes that the net is down, but hdev state
is still up. STATE_DOWN bit of hdev state should be set when stopping net.

Fixes: 46a3df9f ("net: hns3: Add HNS3 Acceleration Engine & Compatibility Layer Support")
Fixes: e2cb1dec ("net: hns3: Add HNS3 VF HCL(Hardware Compatibility Layer) Support")
Signed-off-by: NFuyun Liang <liangfuyun1@huawei.com>
Signed-off-by: NPeng Li <lipeng321@huawei.com>
Signed-off-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds the .ndo_do_ioctl net_device_ops operation to support
the PHY MII ioctl for PF driver.

Fixes: 76ad4f0e ("net: hns3: Add support of HNS3 Ethernet Driver for hip08 SoC")
Signed-off-by: NXi Wang <wangxi11@huawei.com>
Signed-off-by: NPeng Li <lipeng321@huawei.com>
Signed-off-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

All pf have permission to read packet statistics of public in hardware,
but the read operation will clear registers which cause statistical
inaccuracy.

This patch removes all packet statistics of public.
Signed-off-by: NJunxin Chen <chenjunxin1@huawei.com>
Signed-off-by: NPeng Li <lipeng321@huawei.com>
Signed-off-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The the actual Tx work is minimal, driver can clean up as more
Tx descriptors as possible in a irq.
Signed-off-by: NPeng Li <lipeng321@huawei.com>
Signed-off-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds unlikely for buf_num check.
Signed-off-by: NPeng Li <lipeng321@huawei.com>
Signed-off-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

All irq will float to cpu0 if do not set irq affinity.
This patch adds default irq affinity in hns3 driver, users can
also change the irq affinity in OS.
Signed-off-by: NPeng Li <lipeng321@huawei.com>
Signed-off-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, but the implementation in this
driver returns an 'int'.

Found by coccinelle.
Signed-off-by: NYueHaibing <yuehaibing@huawei.com>
Acked-by: NShannon Nelson <shannon.nelson@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, so make sure the implementation in
this driver has returns 'netdev_tx_t' value, and change the function
return type to netdev_tx_t.

Found by coccinelle.
Signed-off-by: NYueHaibing <yuehaibing@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>