A few of the static queues are enabled along with the bcast
STA. Make sure they are removed along with it, rather than
waiting for the mac ctxt release.

This is needed because we sometimes have a STA being removed
and then added again (either with the same sta_id or a
different one). If we wait for the mac ctxt release we will
try to allocate the queues again (as this is currently done
in the STA allocation and not in the MAC init) although
they weren't freed, and even if the sta_id of the STA has
changed.
Signed-off-by: NLiad Kaufman <liad.kaufman@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

Instead of setting the tx_cmd length in the mvm code, which is
complicated by the fact that DQA may want to temporarily store
the SKB on the side, adjust the length in the PCIe code which
also knows about this since it's responsible for duplicating
all those headers that are account for in this code.

As the PCIe code already relies on the tx_cmd->len field, this
doesn't really introduce any new dependencies.

To make this possible we need to move the memcpy() of the TX
command until after it was updated.

This does even simplify the code though, since the PCIe code
already does a lot of manipulations to build A-MSDUs correctly
and changing the length becomes a simple operation to see how
much was added/removed, rather than predicting it.

Fixes: 24afba76 ("iwlwifi: mvm: support bss dynamic alloc/dealloc of queues")
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

We don't really need clear the skb's status area nor store the
dev_cmd into it until we really commit to the frame by handing
it to the transport - defer those operations until just before
we do that.

This doesn't entirely fix the bug with frames not getting sent
out after having been deferred due to DQA, because it doesn't
restore the info->driver_data[0] place that was already set to
zero (or another value) by the A-MSDU logic.

Fixes: 24afba76 ("iwlwifi: mvm: support bss dynamic alloc/dealloc of queues")
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

The firmware will soon actually look at the AID field, and
when it does that it'll try to ensure that the AID is never
changing. Due to the way the station is added, it may start
with an invalid AID before it's associated, so to ensure a
constant AID (once it becomes non-zero), track the station
state and set the AID only when the station is associated
and when it disassociates.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

This fixes a long-standing bug that was introduced when this code
was introduced: cfg80211 passes a pointer, but we treat it as if
it was passing a value. The result is that we pass the pointer to
the firmware, instead of the value. It's not clear how this could
ever have worked, unless the firmware is ignoring this value.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

When driver needs to access the contents of a streaming DMA buffer
without unmapping it it should call dma_sync_single_for_cpu().
Once the call has been made, the CPU "owns" the DMA buffer and can
work with it as needed.
Before the device accesses the buffer, however, ownership should be
transferred back to it with dma_sync_single_for_device().
Both calls weren't performed by the driver, resulting with odd paging
errors on some platforms. Fix it.

Fixes: a6c4fb44 ("iwlwifi: mvm: Add FW paging mechanism for the UMAC on PCI")
Signed-off-by: NSara Sharon <sara.sharon@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

Support getting alive from two LMACs and dumping debug
data from both.
Deprecate older alive notifications no one is using.
Signed-off-by: NSara Sharon <sara.sharon@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

There's already a variable with the result in scope, use that.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

If iwl_mvm_find_free_queue() doesn't find a free queue, it will
return an inactive one.
However, not all the call paths free this queue before reassigning
it, which is a bug.
Check it in other paths and act accordingly.

Fixes: 9794c64f ("iwlwifi: mvm: support dqa queue inactivation upon timeout")
Signed-off-by: NSara Sharon <sara.sharon@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

For CDB devices we will want to configure scan parameters
per band.
Support the new scan API for now. Logic per band will be
added in the future.
Signed-off-by: NSara Sharon <sara.sharon@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

For CDB arch there is another auxiliary mac.
Support statistics APIs that were changed to reflect that.
Signed-off-by: NSara Sharon <sara.sharon@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

Currently we have up to 3 phy contexts - defined by NUM_PHY_CTX.
However - some code paths validate the ID by using MAX_PHYS define
which is set to 4.
While there is no harm it is incorrect - since the maximum is 3.
Remove the define and use the correct one.
Cleanup the code a bit while at it.
Signed-off-by: NSara Sharon <sara.sharon@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

Since offchannel activity doesn't always require a BSS, e.g. ANQP
sessions, offchannel frames should not use the BSS queue, because it
might not be initialized.
Use the auxilary queue instead

Fixes: e3118ad7 ("iwlwifi: mvm: support tdls in dqa mode")
Signed-off-by: NBeni Lev <beni.lev@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

Add to the v1 of the mfuart loading notification, the size
of the mfuart image, and write it to dmesg once the notification
is received.
Signed-off-by: NGolan Ben-Ami <golan.ben.ami@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

For a000 devices the INIT and RT images are unified to one
image.
The changes in the flow are the following:
* Driver load only RT firmware - meaning that the nvm access
  command will be done in the RT image load flow.
* A new command (NVM_ACCESS_COMPLETE) now signals to the FW that
  the driver is done accessing the NVM and FW can proceed with phy
  calibrations.
* Phy DB is no longer sent from INIT FW to be restored by driver
  for the RT FW - all the phy DB is now internal to the FW.
  INIT complete will now follow the NVM access command, without
  phy DB calls before.
* Paging command is sent earlier in the flow before NVM access
  to enable a complete load of FW.
* caution must be care when restart is called since we may have
  not completed init flow even though we are in RT firmware.
Signed-off-by: NSara Sharon <sara.sharon@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

We added the uAPSD enabled ACs and that made the
firmware choose to pull frames with uAPSD trigger
frames instead of PS-Poll.
Fix that.
Signed-off-by: NEmmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

It's important to inform user so he knows things went wrong. He may also
want to get memory dump for further debugging purposes.
Signed-off-by: NRafał Miłecki <rafal@milecki.pl>
Acked-by: NArend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

This method may be unsupported (see: USB bus) or may just fail (see:
SDIO bus).
While at it rework logic in brcmf_sdio_bus_get_memdump function to avoid
too many conditional code nesting levels.
Signed-off-by: NRafał Miłecki <rafal@milecki.pl>
Acked-by: NArend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: NKalle Valo <kvalo@codeaurora.org>

This extra BCM54612E code in PHY driver isn't really aneg specific. Even
without it aneg works OK but the problem is no packets pass through PHY.

Moreover putting this code inside config_aneg callback didn't allow
resuming PHY correctly. When driver called phy_stop and phy_start it was
putting PHY machine into RESUMING state. After that machine was
switching into AN and NOLINK without ever calling phy_start_aneg. This
prevented this extra setup from being called and PHY didn't work.

This change has been verified to fix network on BCM47186B0 SoC device
with BCM54612E.
Signed-off-by: NRafał Miłecki <rafal@milecki.pl>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Yotam Gigi says:

====================
net/sched: act_sample: Little fixes

Little fixes in sample tc action.
====================
Acked-by: NCong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The ASSERT_RTNL is not necessary in the init function, as it does not
touch any rtnl protected structures, as opposed to the mirred action which
does have to hold a net device.
Reported-by: NCong Wang <xiyou.wangcong@gmail.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NYotam Gigi <yotamg@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix error path of in sample init, by releasing the tc hash in case of
failure in psample_group creation.

Fixes: 5c5670fa ("net/sched: Introduce sample tc action")
Reported-by: NCong Wang <xiyou.wangcong@gmail.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NYotam Gigi <yotamg@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Niklas Söderlund says:

====================
sh_eth: fixes for MagicPacket handling

This series contain two fixes for MagicPacket handling. It's based on
top of net-next and is tested on Koelsch.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If a link change interrupt happens along side the MagicPacket interrupt
and the link change interrupt is ignored the interrupt handler will
return and the wakeup event is not registered. Fix this by moving the
MagicPacket check before the link change check.
Signed-off-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reported-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

To be consistent with the rest of the driver when setting bits using
sh_eth_modify() the same bit should also be cleared. This have no
functional change and should have been done from the start.
Signed-off-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Suggested-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Merge tag 'wireless-drivers-next-for-davem-2017-02-01' of git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/wireless-drivers-next

Kalle Valo says:

====================
wireless-drivers-next patches for 4.11

It's nice to see rt2x00 development has becoming active, for example
adding support for a new chip version. Also wcn36xx has been converted
to use the recently merged QCOM_SMD subsystem. Otherwise new features
and fixes it lots of drivers.

Major changes:

iwlwifi

* some more work in preparation for A000 family support
* add support for radiotap timestamps
* some work on our firmware debugging capabilities

wcn36xx

* convert to a proper QCOM_SMD driver (from the platform_driver interface)

ath10k

* VHT160 support
* dump Copy Engine registers during firmware crash
* search board file extension from SMBIOS

wil6210

* add disable_ap_sme module parameter

rt2x00

* support RT3352 with external PA
* support for RT3352 with 20MHz crystal
* add support for RT5350 WiSoC

brcmfmac

* add support for BCM43455 sdio device

rtl8xxxu

* add support for D-Link DWA-131 rev E1, TP-Link TL-WN822N v4 and others
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

dma_mapping_error() returns 1 if there is an error and 0 if not.

Fixes: 018423e9 ("net: ethernet: aquantia: Add ring support code")
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Flush the mmio writes before releasing spin locks.
if the maintained counts get too high > 2M force
writeback of the counts to clear them
Signed-off-by: NSatanand Burla <satananda.burla@cavium.com>
Signed-off-by: NFelix Manlunas <felix.manlunas@cavium.com>
Signed-off-by: NRaghu Vatsavayi <raghu.vatsavayi@cavium.com>
Signed-off-by: NDerek Chickles <derek.chickles@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

IPv6 does not set the NLM_F_APPEND flag in notifications to signal that
a NEWROUTE is an append versus a new route or a replaced one. Add the
flag if the request has it.
Signed-off-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In switch mode on struct cpsw_slave->ndev field will be initialized with
proper value only for the one cpsw slave port, as result
cpsw_get_usage_count() will generate "Unable to handle kernel NULL pointer
dereference" exception when first ethernet interface is opening
cpsw_ndo_open(). This issue causes boot regression on AM335x EVM and
reproducible on am57xx-evm (switch mode).
Fix it by adding additional check for !cpsw->slaves[i].ndev in
cpsw_get_usage_count().

Cc: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Fixes: 03fd01ad ("net: ethernet: ti: cpsw: don't duplicate ndev_running")
Signed-off-by: NGrygorii Strashko <grygorii.strashko@ti.com>
Reviewed-by: NIvan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Dmitry reported warnings occurring in __skb_gso_segment() [1]

All SKB_GSO_DODGY producers can allow user space to feed
packets that trigger the current check.

We could prevent them from doing so, rejecting packets, but
this might add regressions to existing programs.

It turns out our SKB_GSO_DODGY handlers properly set up checksum
information that is needed anyway when packets needs to be segmented.

By checking again skb_needs_check() after skb_mac_gso_segment(),
we should remove these pesky warnings, at a very minor cost.

With help from Willem de Bruijn

[1]
WARNING: CPU: 1 PID: 6768 at net/core/dev.c:2439 skb_warn_bad_offload+0x2af/0x390 net/core/dev.c:2434
lo: caps=(0x000000a2803b7c69, 0x0000000000000000) len=138 data_len=0 gso_size=15883 gso_type=4 ip_summed=0
Kernel panic - not syncing: panic_on_warn set ...

CPU: 1 PID: 6768 Comm: syz-executor1 Not tainted 4.9.0 #5
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
 ffff8801c063ecd8 ffffffff82346bdf ffffffff00000001 1ffff100380c7d2e
 ffffed00380c7d26 0000000041b58ab3 ffffffff84b37e38 ffffffff823468f1
 ffffffff84820740 ffffffff84f289c0 dffffc0000000000 ffff8801c063ee20
Call Trace:
 [<ffffffff82346bdf>] __dump_stack lib/dump_stack.c:15 [inline]
 [<ffffffff82346bdf>] dump_stack+0x2ee/0x3ef lib/dump_stack.c:51
 [<ffffffff81827e34>] panic+0x1fb/0x412 kernel/panic.c:179
 [<ffffffff8141f704>] __warn+0x1c4/0x1e0 kernel/panic.c:542
 [<ffffffff8141f7e5>] warn_slowpath_fmt+0xc5/0x100 kernel/panic.c:565
 [<ffffffff8356cbaf>] skb_warn_bad_offload+0x2af/0x390 net/core/dev.c:2434
 [<ffffffff83585cd2>] __skb_gso_segment+0x482/0x780 net/core/dev.c:2706
 [<ffffffff83586f19>] skb_gso_segment include/linux/netdevice.h:3985 [inline]
 [<ffffffff83586f19>] validate_xmit_skb+0x5c9/0xc20 net/core/dev.c:2969
 [<ffffffff835892bb>] __dev_queue_xmit+0xe6b/0x1e70 net/core/dev.c:3383
 [<ffffffff8358a2d7>] dev_queue_xmit+0x17/0x20 net/core/dev.c:3424
 [<ffffffff83ad161d>] packet_snd net/packet/af_packet.c:2930 [inline]
 [<ffffffff83ad161d>] packet_sendmsg+0x32ed/0x4d30 net/packet/af_packet.c:2955
 [<ffffffff834f0aaa>] sock_sendmsg_nosec net/socket.c:621 [inline]
 [<ffffffff834f0aaa>] sock_sendmsg+0xca/0x110 net/socket.c:631
 [<ffffffff834f329a>] ___sys_sendmsg+0x8fa/0x9f0 net/socket.c:1954
 [<ffffffff834f5e58>] __sys_sendmsg+0x138/0x300 net/socket.c:1988
 [<ffffffff834f604d>] SYSC_sendmsg net/socket.c:1999 [inline]
 [<ffffffff834f604d>] SyS_sendmsg+0x2d/0x50 net/socket.c:1995
 [<ffffffff84371941>] entry_SYSCALL_64_fastpath+0x1f/0xc2
Signed-off-by: NEric Dumazet <edumazet@google.com>
Reported-by: NDmitry Vyukov  <dvyukov@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Allow a master interface to be specified as one of the parameters when
creating a new interface via rtnl_newlink.  Previously this would
require invoking interface creation, waiting for it to complete, and
then separately binding that new interface to a master.

In particular, this is used when creating a macvlan child interface for
VRRP in a VRF configuration, allowing the interface creator to specify
directly what master interface should be inherited by the child,
without having to deal with asynchronous complications and potential
race conditions.
Signed-off-by: NTheuns Verwoerd <theuns.verwoerd@alliedtelesis.co.nz>
Acked-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Steffen Klassert says:

====================
pull request (net-next): ipsec-next 2017-02-01

1) Some typo fixes, from Alexander Alemayhu.

2) Don't acquire state lock in get_mtu functions.
   The only rece against a dead state does not matter.
   From Florian Westphal.

3) Remove xfrm4_state_fini, it is unused for more than
   10 years. From Florian Westphal.

4) Various rcu usage improvements. From Florian Westphal.

5) Properly handle crypto arrors in ah4/ah6.
   From Gilad Ben-Yossef.

6) Try to avoid skb linearization in esp4 and esp6.

7) The esp trailer is now set up in different places,
   add a helper for this.

8) With the upcomming usage of gro_cells in IPsec,
   a gro merged skb can have a secpath. Drop it
   before freeing or reusing the skb.

9) Add a xfrm dummy network device for napi. With
   this we can use gro_cells from within xfrm,
   it allows IPsec GRO without impact on the generic
   networking code.

Please pull or let me know if there are problems.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Rafał Miłecki says:

====================
net-next: use one struct bgmac & add PHY support

This patchset adds support for initializing PHY using PHY subsystem.
It's required e.g. for wireless access point devices that use bgmac
supported Ethernet device connected to some external PHY.

Implementing this required accessing phydev in bcma specific code which
wasn't possible with core code allocating struct bgmac on its own. This
is why I needed to modify alloc_etherdev usage first.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This adds support for using bgmac with PHYs supported by standalone PHY
drivers. Having any PHY initialization in bgmac is hacky and shouldn't
be extended but rather removed if anyone has hardware to test it.
Signed-off-by: NRafał Miłecki <rafal@milecki.pl>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Adding struct bcma_mdio was a workaround for bcma code not having access
to the struct bgmac used in the core code. Now we don't duplicate this
struct we can just use it internally in bcma code.

This simplifies code & allows access to all bgmac driver details from
all places in bcma code.
Signed-off-by: NRafał Miłecki <rafal@milecki.pl>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

So far were were allocating struct bgmac in 3 places: platform code,
bcma code and shared bgmac_enet_probe function. The reason for this was
bgmac_enet_probe:
1) Requiring early-filled struct bgmac
2) Calling alloc_etherdev on its own in order to use netdev_priv later

This solution got few drawbacks:
1) Was duplicating allocating code
2) Required copying early-filled struct
3) Resulted in platform/bcma code having access only to unused struct

Solve this situation by simply extracting some probe code into the new
bgmac_alloc function.
Signed-off-by: NRafał Miłecki <rafal@milecki.pl>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 7318166c ("net: dsa: bcm_sf2: Add support for
ethtool::rxnfc") added a new object to build: bcm_sf2_cfp.o, but in
doing so, we essentially just built this object and no longer bcm_sf2.o.

Fix this by creating a module named bcm-sf2.ko which links in bcm_sf2.o
and bcm_sf2_cfp.o.

Fixes: 7318166c ("net: dsa: bcm_sf2: Add support for ethtool::rxnfc")
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Change t4fw_version.h to update latest firmware version
number 1.16.26.0.
Signed-off-by: NGanesh Goudar <ganeshgr@chelsio.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

under memory pressure 'ethtool -S' command may warn:
[ 2374.385195] ethtool: page allocation failure: order:4, mode:0x242c0c0
[ 2374.405573] CPU: 12 PID: 40211 Comm: ethtool Not tainted
[ 2374.423071] Call Trace:
[ 2374.423076]  [<ffffffff8148cb29>] dump_stack+0x4d/0x64
[ 2374.423080]  [<ffffffff811667cb>] warn_alloc_failed+0xeb/0x150
[ 2374.423082]  [<ffffffff81169cd3>] ? __alloc_pages_direct_compact+0x43/0xf0
[ 2374.423084]  [<ffffffff8116a25c>] __alloc_pages_nodemask+0x4dc/0xbf0
[ 2374.423091]  [<ffffffffa0023dc2>] ? cmd_exec+0x722/0xcd0 [mlx5_core]
[ 2374.423095]  [<ffffffff811b3dcc>] alloc_pages_current+0x8c/0x110
[ 2374.423097]  [<ffffffff81168859>] alloc_kmem_pages+0x19/0x90
[ 2374.423099]  [<ffffffff81186e5e>] kmalloc_order_trace+0x2e/0xe0
[ 2374.423101]  [<ffffffff811c0084>] __kmalloc+0x204/0x220
[ 2374.423105]  [<ffffffff816c269e>] dev_ethtool+0xe4e/0x1f80
[ 2374.423106]  [<ffffffff816b967e>] ? dev_get_by_name_rcu+0x5e/0x80
[ 2374.423108]  [<ffffffff816d6926>] dev_ioctl+0x156/0x560
[ 2374.423111]  [<ffffffff811d4c68>] ? mem_cgroup_commit_charge+0x78/0x3c0
[ 2374.423117]  [<ffffffff8169d542>] sock_do_ioctl+0x42/0x50
[ 2374.423119]  [<ffffffff8169d9c3>] sock_ioctl+0x1b3/0x250
[ 2374.423121]  [<ffffffff811f0f42>] do_vfs_ioctl+0x92/0x580
[ 2374.423123]  [<ffffffff8100222b>] ? do_audit_syscall_entry+0x4b/0x70
[ 2374.423124]  [<ffffffff8100287c>] ? syscall_trace_enter_phase1+0xfc/0x120
[ 2374.423126]  [<ffffffff811f14a9>] SyS_ioctl+0x79/0x90
[ 2374.423127]  [<ffffffff81002bb0>] do_syscall_64+0x50/0xa0
[ 2374.423129]  [<ffffffff817e19bc>] entry_SYSCALL64_slow_path+0x25/0x25

~1160 mlx5 counters ~= order 4 allocation which is unlikely to succeed
under memory pressure. Convert them to vzalloc() as ethtool_get_regs() does.
Also take care of drivers without counters similar to
commit 67ae7cf1 ("ethtool: Allow zero-length register dumps again")
and reduce warn_on to warn_on_once.
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>