bp->sriov_cfg is not defined when CONFIG_BNXT_SRIOV is not set.  Fix
it by adding a helper function bnxt_sriov_cfg() to handle the logic
with or without the config option.

Fixes: 46d08f55 ("bnxt_en: extend RTNL to VF check in devlink driver_reinit")
Reported-by: Nkernel test robot <lkp@intel.com>
Reviewed-by: NEdwin Peer <edwin.peer@broadcom.com>
Reviewed-by: NAndy Gospodarek <gospo@broadcom.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/1637090770-22835-1-git-send-email-michael.chan@broadcom.comSigned-off-by: NJakub Kicinski <kuba@kernel.org>

The kernel test robot reported a following issue:

>> drivers/net/ethernet/marvell/mvmdio.c:426:36: warning:
unused variable 'orion_mdio_acpi_match' [-Wunused-const-variable]
   static const struct acpi_device_id orion_mdio_acpi_match[] = {
                                      ^
   1 warning generated.

Fix that by surrounding the variable by appropriate ifdef.

Fixes: c54da4c1 ("net: mvmdio: add ACPI support")
Reported-by: Nkernel test robot <lkp@intel.com>
Signed-off-by: NMarcin Wojtas <mw@semihalf.com>
Reviewed-by: NAndrew Lunn <andrew@lunn.ch>
Link: https://lore.kernel.org/r/20211115153024.209083-1-mw@semihalf.comSigned-off-by: NJakub Kicinski <kuba@kernel.org>

Restore VLAN filters after the link is brought down, and up - since all
filters are deleted from HW during the netdev link down routine.

Fixes: ed1f5b58 ("i40evf: remove VLAN filters on close")
Signed-off-by: NAkeem G Abodunrin <akeem.g.abodunrin@intel.com>
Tested-by: NGeorge Kuruvinakunnel <george.kuruvinakunnel@intel.com>
Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com>

Now setting combine to 0 will be rejected with the
appropriate error code.
This has been implemented by adding a condition that checks
the value of combine equal to zero.
Without this patch, when the user requested it, no error was
returned and combine was set to the default value for VF.

Fixes: 5520deb1 ("iavf: Enable support for up to 16 queues")
Signed-off-by: NGrzegorz Szczurek <grzegorzx.szczurek@intel.com>
Tested-by: NTony Brelinski <tony.brelinski@intel.com>
Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com>

While issuing VF Reset from the guest OS, the VF driver prints
logs about critical / Overflow error detection. This is not an
actual error since the VF_MBX_ARQLEN register is set to all FF's
for a short period of time and the VF would catch the bits set if
it was reading the register during that spike of time.
This patch introduces an additional check to ignore this condition
since the VF is in reset.

Fixes: 19b73d8e ("i40evf: Add additional check for reset")
Signed-off-by: NSurabhi Boob <surabhi.boob@intel.com>
Tested-by: NTony Brelinski <tony.brelinski@intel.com>
Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com>

In some cases, the ethtool get_rxfh handler may be called with a null
key or indir parameter. So check these pointers, or you will have a very
bad day.

Fixes: 43a3d9ba ("i40evf: Allow PF driver to configure RSS")
Signed-off-by: NMitch Williams <mitch.a.williams@intel.com>
Tested-by: NTony Brelinski <tony.brelinski@intel.com>
Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com>

In iavf_config_clsflower, the filter structure could be accidentally
released at the end, if iavf_parse_cls_flower or iavf_handle_tclass ever
return a non-zero but positive value.

In this case, the function continues through to the end, and will call
kfree() on the filter structure even though it has been added to the
linked list.

This can actually happen because iavf_parse_cls_flower will return
a positive IAVF_ERR_CONFIG value instead of the traditional negative
error codes.

Fix this by ensuring that the kfree() check and error checks are
similar. Use the more idiomatic "if (err)" to catch all non-zero error
codes.

Fixes: 0075fa0f ("i40evf: Add support to apply cloud filters")
Signed-off-by: NJacob Keller <jacob.e.keller@intel.com>
Tested-by: NTony Brelinski <tony.brelinski@intel.com>
Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com>

The driver could only quit allmulti when allmulti and promisc modes are
turn on at the same time. If promisc had been off there was no way to turn
off allmulti mode.
The patch corrects this behavior. Switching allmulti does not depends on
promisc state mode anymore

Fixes: f42a5c74 ("i40e: Add allmulti support for the VF")
Signed-off-by: NPiotr Marczak <piotr.marczak@intel.com>
Tested-by: NTony Brelinski <tony.brelinski@intel.com>
Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com>

In iavf_configure_clsflower() the function will bail out if it is unable
to obtain the crit_section lock in a reasonable time. However, it will
clear the lock when exiting, so fix this.

Fixes: 640a8af5 ("i40evf: Reorder configure_clsflower to avoid deadlock on error")
Signed-off-by: NNicholas Nunley <nicholas.d.nunley@intel.com>
Tested-by: NTony Brelinski <tony.brelinski@intel.com>
Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com>

iavf_free_queues() clears adapter->num_active_queues, which
iavf_free_q_vectors() relies on, so swap the order of these two function
calls in iavf_disable_vf(). This resolves a panic encountered when the
interface is disabled and then later brought up again after PF
communication is restored.

Fixes: 65c7006f ("i40evf: assign num_active_queues inside i40evf_alloc_queues")
Signed-off-by: NNicholas Nunley <nicholas.d.nunley@intel.com>
Tested-by: NTony Brelinski <tony.brelinski@intel.com>
Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com>

If the driver has lost contact with the PF then it enters a disabled state
and frees adapter->vf_res. However, ndo_fix_features can still be called on
the interface, so we need to check for this condition first. Since we have
no information on the features at this time simply leave them unmodified
and return.

Fixes: c4445aed ("i40evf: Fix VLAN features")
Signed-off-by: NNicholas Nunley <nicholas.d.nunley@intel.com>
Tested-by: NTony Brelinski <tony.brelinski@intel.com>
Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com>

Fixed return correct code from set the new channel count.
Implemented by check if reset is done in appropriate time.
This solution give a extra time to pf for reset vf in case
when user want set new channel count for all vfs.
Without this patch it is possible to return misleading output
code to user and vf reset not to be correctly performed by pf.

Fixes: 5520deb1 ("iavf: Enable support for up to 16 queues")
Signed-off-by: NGrzegorz Szczurek <grzegorzx.szczurek@intel.com>
Signed-off-by: NMateusz Palczewski <mateusz.palczewski@intel.com>
Tested-by: NKonrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com>

According to upstream commit 5ec55823("net: stmmac:
add clocks management for gmac driver"), it improve clocks
management for stmmac driver. So, it is necessary to implement
the runtime callback in dwmac-socfpga driver because it doesn't
use the common stmmac_pltfr_pm_ops instance. Otherwise, clocks
are not disabled when system enters suspend status.

Fixes: 5ec55823 ("net: stmmac: add clocks management for gmac driver")
Cc: stable@vger.kernel.org
Signed-off-by: NMeng Li <Meng.Li@windriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The driver does not check if hw-tc-offload is enabled for the device
before offloading a flow in the context of indirect block callback.
Fix this by checking NETIF_F_HW_TC in the features flag and rejecting
the offload request.  This will avoid unnecessary dmesg error logs when
hw-tc-offload is disabled, such as these:

bnxt_en 0000:19:00.1 eno2np1: dev(ifindex=294) not on same switch
bnxt_en 0000:19:00.1 eno2np1: Error: bnxt_tc_add_flow: cookie=0xffff8dace1c88000 error=-22
bnxt_en 0000:19:00.0 eno1np0: dev(ifindex=294) not on same switch
bnxt_en 0000:19:00.0 eno1np0: Error: bnxt_tc_add_flow: cookie=0xffff8dace1c88000 error=-22
Reported-by: NMarcelo Ricardo Leitner <mleitner@redhat.com>
Fixes: 627c89d0 ("bnxt_en: flow_offload: offload tunnel decap rules via indirect callbacks")
Signed-off-by: NSriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This fixes type mismatch warning.
Reported-by: Nkernel test robot <lkp@intel.com>
Fixes: 3c415339 ("bnxt_en: implement firmware live patching")
Signed-off-by: NEdwin Peer <edwin.peer@broadcom.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The fixes the race condition between configuring SR-IOV and devlink
reload.  The SR-IOV configure logic already takes the RTNL lock,
setting sriov_cfg under the lock while changes are underway. Extend
the lock scope in devlink driver_reinit to cover the VF check and
don't run concurrently with SR-IOV configure.
Reported-by: NLeon Romanovsky <leon@kernel.org>
Fixes: 228ea8c1 ("bnxt_en: implement devlink dev reload driver_reinit")
Cc: Leon Romanovsky <leon@kernel.org>
Reviewed-by: NSomnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: NPavan Chebbi <pavan.chebbi@broadcom.com>
Reviewed-by: NAndy Gospodarek <gospo@broadcom.com>
Signed-off-by: NEdwin Peer <edwin.peer@broadcom.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix build error and warnings reported by kernel test robot:

drivers/net/ethernet/lantiq_etop.c: In function 'ltq_etop_probe':
drivers/net/ethernet/lantiq_etop.c:673:15: error: implicit declaration of function 'device_property_read_u32' [-Werror=implicit-function-declaration]
     673 |         err = device_property_read_u32(&pdev->dev, "lantiq,tx-burst-length", &priv->tx_burst_len);

   drivers/net/ethernet/lantiq_etop.c: At top level:
   drivers/net/ethernet/lantiq_etop.c:730:1: warning: no previous prototype for 'init_ltq_etop' [-Wmissing-prototypes]
     730 | init_ltq_etop(void)

   drivers/net/ethernet/lantiq_etop.c: In function 'ltq_etop_hw_init':
   drivers/net/ethernet/lantiq_etop.c:276:25: warning: ignoring return value of 'request_irq' declared with attribute 'warn_unused_result' [-Wunused-result]
     276 |                         request_irq(irq, ltq_etop_dma_irq, 0, "etop_tx", priv);
   drivers/net/ethernet/lantiq_etop.c:284:25: warning: ignoring return value of 'request_irq' declared with attribute 'warn_unused_result' [-Wunused-result]
     284 |                         request_irq(irq, ltq_etop_dma_irq, 0, "etop_rx", priv);

Fixes: 14d4e308 ("net: lantiq: configure the burst length in ethernet drivers")
Fixes: dddb29e4 ("net: lantiq_etop: remove deprecated IRQF_DISABLED")
Fixes: 504d4721 ("MIPS: Lantiq: Add ethernet driver")
Signed-off-by: NRandy Dunlap <rdunlap@infradead.org>
Reported-by: Nkernel test robot <lkp@intel.com>
Link: lore.kernel.org/r/202111090621.yjr9xuVj-lkp@intel.com
To: netdev@vger.kernel.org
Cc: Aleksander Jan Bajkowski <olek2@wp.pl>
Cc: Hauke Mehrtens <hauke@hauke-m.de>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: John Crispin <john@phrozen.org>
Cc: linux-mips@vger.kernel.org
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Michael Opdenacker <michael.opdenacker@bootlin.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This bug report shows up when running our research tools. The
reports is SOOB read, but it seems SOOB write is also possible
a few lines below.

In details, fw.len and sw.len are inputs coming from io. A len
over the size of self->rpc triggers SOOB. The patch fixes the
bugs by adding sanity checks.

The bugs are triggerable with compromised/malfunctioning devices.
They are potentially exploitable given they first leak up to
0xffff bytes and able to overwrite the region later.

The patch is tested with QEMU emulater.
This is NOT tested with a real device.

Attached is the log we found by fuzzing.

BUG: KASAN: slab-out-of-bounds in
	hw_atl_utils_fw_upload_dwords+0x393/0x3c0 [atlantic]
Read of size 4 at addr ffff888016260b08 by task modprobe/213
CPU: 0 PID: 213 Comm: modprobe Not tainted 5.6.0 #1
Call Trace:
 dump_stack+0x76/0xa0
 print_address_description.constprop.0+0x16/0x200
 ? hw_atl_utils_fw_upload_dwords+0x393/0x3c0 [atlantic]
 ? hw_atl_utils_fw_upload_dwords+0x393/0x3c0 [atlantic]
 __kasan_report.cold+0x37/0x7c
 ? aq_hw_read_reg_bit+0x60/0x70 [atlantic]
 ? hw_atl_utils_fw_upload_dwords+0x393/0x3c0 [atlantic]
 kasan_report+0xe/0x20
 hw_atl_utils_fw_upload_dwords+0x393/0x3c0 [atlantic]
 hw_atl_utils_fw_rpc_call+0x95/0x130 [atlantic]
 hw_atl_utils_fw_rpc_wait+0x176/0x210 [atlantic]
 hw_atl_utils_mpi_create+0x229/0x2e0 [atlantic]
 ? hw_atl_utils_fw_rpc_wait+0x210/0x210 [atlantic]
 ? hw_atl_utils_initfw+0x9f/0x1c8 [atlantic]
 hw_atl_utils_initfw+0x12a/0x1c8 [atlantic]
 aq_nic_ndev_register+0x88/0x650 [atlantic]
 ? aq_nic_ndev_init+0x235/0x3c0 [atlantic]
 aq_pci_probe+0x731/0x9b0 [atlantic]
 ? aq_pci_func_init+0xc0/0xc0 [atlantic]
 local_pci_probe+0xd3/0x160
 pci_device_probe+0x23f/0x3e0
Reported-by: NBrendan Dolan-Gavitt <brendandg@nyu.edu>
Signed-off-by: NZekun Shen <bruceshenzk@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Smatch says:
	bnx2x_init_ops.h:640 bnx2x_ilt_client_mem_op()
	warn: variable dereferenced before check 'ilt' (see line 638)

Move ilt_cli variable initialization _after_ ilt validation, because
it's unsafe to deref the pointer before validation check.

Fixes: 523224a3 ("bnx2x, cnic, bnx2i: use new FW/HSI")
Signed-off-by: NPavel Skripkin <paskripkin@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The head-of-line blocking timer should only be modified when
head-of-line drop is disabled.

One of the steps in recovering from a modem crash is to enable
dropping of packets with timeout of 0 (immediate).  We don't know
how the modem configured its endpoints, so before we program the
timer, we need to ensure HOL_BLOCK is disabled.

Fixes: 84f9bd12 ("soc: qcom: ipa: IPA endpoints")
Signed-off-by: NAlex Elder <elder@linaro.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Starting with IPA v4.5, the HOL_BLOCK_EN register must be written
twice when enabling head-of-line blocking avoidance.

Fixes: 84f9bd12 ("soc: qcom: ipa: IPA endpoints")
Signed-off-by: NAlex Elder <elder@linaro.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

As stated in the bonding doc, trans_start must be set manually for drivers
using NETIF_F_LLTX:
 Drivers that use NETIF_F_LLTX flag must also update
 netdev_queue->trans_start. If they do not, then the ARP monitor will
 immediately fail any slaves using that driver, and those slaves will stay
 down.

Link: https://www.kernel.org/doc/html/v5.15/networking/bonding.html#arp-monitor-operationSigned-off-by: NNicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

A space has snuck in.
Reported-by: Nkernel test robot <lkp@intel.com>
Fixes: 74fad215 ("ethernet: sis900: use eth_hw_addr_set()")
Link: https://lore.kernel.org/r/20211111210824.676201-1-kuba@kernel.orgSigned-off-by: NJakub Kicinski <kuba@kernel.org>

The source group count was mistakenly assigned to both dst and src loops.
Fix it to make IPA probe and work again.

Fixes: 4fd704b3 ("net: ipa: record number of groups in data")
Acked-by: NAngeloGioacchino Del Regno <angelogioacchino.delregno@somainline.org>
Reviewed-by: NMarijn Suijten <marijn.suijten@somainline.org>
Signed-off-by: NKonrad Dybcio <konrad.dybcio@somainline.org>
Reviewed-by: NAlex Elder <elder@linaro.org>
Link: https://lore.kernel.org/r/20211111183724.593478-1-konrad.dybcio@somainline.orgSigned-off-by: NJakub Kicinski <kuba@kernel.org>

The former patch "defer 6pack kfree after unregister_netdev" reorders
the kfree of two buffer after the unregister_netdev to prevent the race
condition. It also adds free_netdev() function in sixpack_close(), which
is a direct copy from the similar code in mkiss_close().

However, in sixpack driver, the flag needs_free_netdev is set to true in
sp_setup(), hence the unregister_netdev() will free the netdev
automatically. Therefore, as the sp is netdev_priv, use-after-free
occurs.

This patch removes the needs_free_netdev = true and just let the
free_netdev to finish this deallocation task.

Fixes: 0b911192 ("hamradio: defer 6pack kfree after unregister_netdev")
Signed-off-by: NLin Ma <linma@zju.edu.cn>
Link: https://lore.kernel.org/r/20211111141402.7551-1-linma@zju.edu.cnSigned-off-by: NJakub Kicinski <kuba@kernel.org>

curr_phase is unused. Removed the dead code.

Fixes: 8d9be063 ("net: wwan: iosm: transport layer support for fw flashing/cd")
Reported-by: Nkernel test robot <lkp@intel.com>
Signed-off-by: NM Chetan Kumar <m.chetan.kumar@linux.intel.com>
Reviewed-by: NLoic Poulain <loic.poulain@linaro.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Ensure diagnostics monitoring support is implemented for the SFF 8472
compliant port module and set the correct length for ethtool port
module eeprom read.

Fixes: f56ec676 ("cxgb4: Add support for ethtool i2c dump")
Signed-off-by: NManoj Malviya <manojmalviya@chelsio.com>
Signed-off-by: NRahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There is a minus character missing from ERR_PTR(ENOMEM) so if this
allocation fails it will lead to an Oops in the caller.

Fixes: dc48516e ("net/mlx5: Lag, add support to create definers for LAG")
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: NLeon Romanovsky <leonro@nvidia.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The u64_stats_update_end() call is supposed to be inside the curly
braces so it pairs with the u64_stats_update_begin().

Fixes: 37149e93 ("gve: Implement packet continuation for RX.")
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This fixes the error detected when compiling the driver.

Fixes: 14d4e308 ("net: lantiq: configure the burst length in ethernet drivers")
Reported-by: Nkernel test robot <lkp@intel.com>
Signed-off-by: NAleksander Jan Bajkowski <olek2@wp.pl>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit bfe301eb ("net: mvpp2: convert to use
mac_prepare()/mac_finish()") introduced a bug wherein it leaves the MAC
RESET register asserted after mac_finish(), due to wrong order of
function calls.

Before it was:
  .mac_config()
    mvpp22_mode_reconfigure()
      assert reset
    mvpp2_xlg_config()
      deassert reset

Now it is:
  .mac_prepare()
  .mac_config()
    mvpp2_xlg_config()
      deassert reset
  .mac_finish()
    mvpp2_xlg_config()
      assert reset

Obviously this is wrong.

This bug is triggered when phylink tries to change the PHY interface
mode from a GMAC mode (sgmii, 1000base-x, 2500base-x) to XLG mode
(10gbase-r, xaui). The XLG mode does not work since reset is left
asserted. Only after
  ifconfig down && ifconfig up
is called will the XLG mode work.

Move the call to mvpp22_mode_reconfigure() to .mac_prepare()
implementation. Since some of the subsequent functions need to know
whether the interface is being changed, we unfortunately also need to
pass around the new interface mode before setting port->phy_interface.

Fixes: bfe301eb ("net: mvpp2: convert to use mac_prepare()/mac_finish()")
Signed-off-by: NMarek Behún <kabel@kernel.org>
Signed-off-by: NRussell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

It is spurious to allocate a bitmap without initializing it.
So, better safe than sorry, initialize it to 0 at least to have some known
values.

While at it, switch to the devm_bitmap_ API which is less verbose.

Fixes: 4b41d343 ("net: ethernet: ti: cpsw: allow untagged traffic on host port")
Signed-off-by: NChristophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit fe28c53e ("net: stmmac: fix taprio configuration when
base_time is in the past") allowed some base time values in the past,
but apparently not all, the base-time value of 0 (Jan 1st 1970) is still
explicitly denied by the driver.

Remove the bogus check.

Fixes: b60189e0 ("net: stmmac: Integrate EST with TAPRIO scheduler API")
Signed-off-by: NVladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: NKurt Kanzenbach <kurt@linutronix.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, driver only allow configuring ETS bandwidth of TCs according
to the max TC number queried from firmware. However, the hardware actually
supports 8 TCs and users may need to configure ETS bandwidth of all TCs,
so remove the restriction.

Fixes: 330baff5 ("net: hns3: add ETS TC weight setting in SSU module")
Signed-off-by: NGuangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If users set unicast mac address for VFs by PF, they need to guarantee all
VFs' address is different. This patch removes the check mac address exist
of VFs, for usrs can refresh mac addresses of all VFs directly without
need to modify the exist mac address to other value firstly.

Fixes: 8e6de441 ("net: hns3: add support for configuring VF MAC from the host")
Signed-off-by: NGuangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When driver queries the register number of mac statistics from firmware,
the old firmware runs in device version V2 only returns number of valid
registers, not include number of three reserved registers among of them.
It cause driver doesn't record the last three data when query mac
statistics.

To fix this problem, driver never query register number in device version
V2 and set it to a fixed value which include three reserved registers.

Fixes: c8af2887 ("net: hns3: add support pause/pfc durations for mac statistics")
Signed-off-by: NGuangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When fully configure VLANs for a VF, then unload the VF while
triggering a reset to PF, will cause a kernel crash because the
irq is already uninit.

[ 293.177579] ------------[ cut here ]------------
[ 293.183502] kernel BUG at drivers/pci/msi.c:352!
[ 293.189547] Internal error: Oops - BUG: 0 [#1] SMP
......
[ 293.390124] Workqueue: hclgevf hclgevf_service_task [hclgevf]
[ 293.402627] pstate: 80c00009 (Nzcv daif +PAN +UAO)
[ 293.414324] pc : free_msi_irqs+0x19c/0x1b8
[ 293.425429] lr : free_msi_irqs+0x18c/0x1b8
[ 293.436545] sp : ffff00002716fbb0
[ 293.446950] x29: ffff00002716fbb0 x28: 0000000000000000
[ 293.459519] x27: 0000000000000000 x26: ffff45b91ea16b00
[ 293.472183] x25: 0000000000000000 x24: ffffa587b08f4700
[ 293.484717] x23: ffffc591ac30e000 x22: ffffa587b08f8428
[ 293.497190] x21: ffffc591ac30e300 x20: 0000000000000000
[ 293.509594] x19: ffffa58a062a8300 x18: 0000000000000000
[ 293.521949] x17: 0000000000000000 x16: ffff45b91dcc3f48
[ 293.534013] x15: 0000000000000000 x14: 0000000000000000
[ 293.545883] x13: 0000000000000040 x12: 0000000000000228
[ 293.557508] x11: 0000000000000020 x10: 0000000000000040
[ 293.568889] x9 : ffff45b91ea1e190 x8 : ffffc591802d0000
[ 293.580123] x7 : ffffc591802d0148 x6 : 0000000000000120
[ 293.591190] x5 : ffffc591802d0000 x4 : 0000000000000000
[ 293.602015] x3 : 0000000000000000 x2 : 0000000000000000
[ 293.612624] x1 : 00000000000004a4 x0 : ffffa58a1e0c6b80
[ 293.623028] Call trace:
[ 293.630340] free_msi_irqs+0x19c/0x1b8
[ 293.638849] pci_disable_msix+0x118/0x140
[ 293.647452] pci_free_irq_vectors+0x20/0x38
[ 293.656081] hclgevf_uninit_msi+0x44/0x58 [hclgevf]
[ 293.665309] hclgevf_reset_rebuild+0x1ac/0x2e0 [hclgevf]
[ 293.674866] hclgevf_reset+0x358/0x400 [hclgevf]
[ 293.683545] hclgevf_reset_service_task+0xd0/0x1b0 [hclgevf]
[ 293.693325] hclgevf_service_task+0x4c/0x2e8 [hclgevf]
[ 293.702307] process_one_work+0x1b0/0x448
[ 293.710034] worker_thread+0x54/0x468
[ 293.717331] kthread+0x134/0x138
[ 293.724114] ret_from_fork+0x10/0x18
[ 293.731324] Code: f940b000 b4ffff00 a903e7b8 f90017b6 (d4210000)

This patch fixes the problem by waiting for the VF reset done
while unloading the VF.

Fixes: e2cb1dec ("net: hns3: Add HNS3 VF HCL(Hardware Compatibility Layer) Support")
Signed-off-by: NYufeng Mo <moyufeng@huawei.com>
Signed-off-by: NGuangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When the driver processes rx packets, the head pointer is updated only
after the number of received packets reaches 16. However, hardware
relies on the head pointer to calculate the number of FBDs. As a result,
the hardware calculates the FBD incorrectly. Therefore, the driver
proactively updates the head pointer in each common poll to ensure that
the number of FBDs calculated by the hardware is correct.

Fixes: 68752b24 ("net: hns3: schedule the polling again when allocation fails")
Signed-off-by: NYufeng Mo <moyufeng@huawei.com>
Signed-off-by: NGuangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, driver will send command to firmware to query pfc packet number
when user uses dcb tool to get pfc parameters. However, the periodic
service task will also periodically query and record MAC statistics,
including pfc packet number.

As the hardware registers of statistics is cleared after reading, it will
cause pfc packet number of MAC statistics are not correct after using dcb
tool to get pfc parameters.

To fix this problem, when user uses dcb tool to get pfc parameters, driver
updates MAC statistics firstly and then get pfc packet number from MAC
statistics.

Fixes: 64fd2300 ("net: hns3: add support for querying pfc puase packets statistic")
Signed-off-by: NJie Wang <wangjie125@huawei.com>
Signed-off-by: NGuangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, NIC init ROCE interrupt vector with MSIX interrupt. But ROCE use
pci_irq_vector() to get interrupt vector, which adds the relative interrupt
vector again and gets wrong interrupt vector.

So fixes it by assign relative interrupt vector to ROCE instead of MSIX
interrupt vector and delete the unused struct member base_msi_vector
declaration of hclgevf_dev.

Fixes: 46a3df9f ("net: hns3: Add HNS3 Acceleration Engine & Compatibility Layer Support")
Signed-off-by: NJie Wang <wangjie125@huawei.com>
Signed-off-by: NGuangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>