Add the RDMA cookie and RX timestamp to the usercopy whitelist.

After the introduction of hardened usercopy whitelisting
(https://lwn.net/Articles/727322/), a warning is displayed when the
RDMA cookie or RX timestamp is copied to userspace:

kernel: WARNING: CPU: 3 PID: 5750 at
mm/usercopy.c:81 usercopy_warn+0x8e/0xa6
[...]
kernel: Call Trace:
kernel: __check_heap_object+0xb8/0x11b
kernel: __check_object_size+0xe3/0x1bc
kernel: put_cmsg+0x95/0x115
kernel: rds_recvmsg+0x43d/0x620 [rds]
kernel: sock_recvmsg+0x43/0x4a
kernel: ___sys_recvmsg+0xda/0x1e6
kernel: ? __handle_mm_fault+0xcae/0xf79
kernel: __sys_recvmsg+0x51/0x8a
kernel: SyS_recvmsg+0x12/0x1c
kernel: do_syscall_64+0x79/0x1ae

When the whitelisting feature was introduced, the memory for the RDMA
cookie and RX timestamp in RDS was not added to the whitelist, causing
the warning above.
Signed-off-by: Dag Moxnes <dag.moxnes@oracle.com>
Tested-by: Jenny <jenny.x.xu@oracle.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Return code value could be non deterministic in case of wrong size read.
With this patch, if such error occurs, set rc to be -EIO.

In addition, mlx5_hv_config_common() supports reading of
HV_CONFIG_BLOCK_SIZE_MAX bytes only, fix to early return error with
bad input.

Fixes: 913d14e8 ("net/mlx5: Add wrappers for HyperV PCIe operations")
Reported-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

We need a similar fix for ipv6 as Commit 0761680d ("net: ipv4: fix
listify ip_rcv_finish in case of forwarding") does for ipv4.

This issue can be reprocuded by syzbot since Commit 323ebb61 ("net:
use listified RX for handling GRO_NORMAL skbs") on net-next. The call
trace was:

  kernel BUG at include/linux/skbuff.h:2225!
  RIP: 0010:__skb_pull include/linux/skbuff.h:2225 [inline]
  RIP: 0010:skb_pull+0xea/0x110 net/core/skbuff.c:1902
  Call Trace:
    sctp_inq_pop+0x2f1/0xd80 net/sctp/inqueue.c:202
    sctp_endpoint_bh_rcv+0x184/0x8d0 net/sctp/endpointola.c:385
    sctp_inq_push+0x1e4/0x280 net/sctp/inqueue.c:80
    sctp_rcv+0x2807/0x3590 net/sctp/input.c:256
    sctp6_rcv+0x17/0x30 net/sctp/ipv6.c:1049
    ip6_protocol_deliver_rcu+0x2fe/0x1660 net/ipv6/ip6_input.c:397
    ip6_input_finish+0x84/0x170 net/ipv6/ip6_input.c:438
    NF_HOOK include/linux/netfilter.h:305 [inline]
    NF_HOOK include/linux/netfilter.h:299 [inline]
    ip6_input+0xe4/0x3f0 net/ipv6/ip6_input.c:447
    dst_input include/net/dst.h:442 [inline]
    ip6_sublist_rcv_finish+0x98/0x1e0 net/ipv6/ip6_input.c:84
    ip6_list_rcv_finish net/ipv6/ip6_input.c:118 [inline]
    ip6_sublist_rcv+0x80c/0xcf0 net/ipv6/ip6_input.c:282
    ipv6_list_rcv+0x373/0x4b0 net/ipv6/ip6_input.c:316
    __netif_receive_skb_list_ptype net/core/dev.c:5049 [inline]
    __netif_receive_skb_list_core+0x5fc/0x9d0 net/core/dev.c:5097
    __netif_receive_skb_list net/core/dev.c:5149 [inline]
    netif_receive_skb_list_internal+0x7eb/0xe60 net/core/dev.c:5244
    gro_normal_list.part.0+0x1e/0xb0 net/core/dev.c:5757
    gro_normal_list net/core/dev.c:5755 [inline]
    gro_normal_one net/core/dev.c:5769 [inline]
    napi_frags_finish net/core/dev.c:5782 [inline]
    napi_gro_frags+0xa6a/0xea0 net/core/dev.c:5855
    tun_get_user+0x2e98/0x3fa0 drivers/net/tun.c:1974
    tun_chr_write_iter+0xbd/0x156 drivers/net/tun.c:2020

Fixes: d8269e2c ("net: ipv6: listify ipv6_rcv() and ip6_rcv_finish()")
Fixes: 323ebb61 ("net: use listified RX for handling GRO_NORMAL skbs")
Reported-by: syzbot+eb349eeee854e389c36d@syzkaller.appspotmail.com
Reported-by: syzbot+4a0643a653ac375612d1@syzkaller.appspotmail.com
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Hayes Wang says:

====================
r8152: save EEE

v4:
For patch #2, remove redundant calling of "ocp_reg_write(tp, OCP_EEE_ADV, 0)".

v3:
For patch #2, fix the mistake caused by copying and pasting.

v2:
Adjust patch #1. The EEE has been disabled in the beginning of
r8153_hw_phy_cfg() and r8153b_hw_phy_cfg(), so only check if
it is necessary to enable EEE.

Add the patch #2 for the helper function.

v1:
Saving the settings of EEE to avoid they become the default settings
after reset_resume().
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Add a helper function "rtl_eee_enable" for setting EEE. Besides, I
move r8153_eee_en() and r8153b_eee_en(). And, I remove r8152b_enable_eee(),
r8153_set_eee(), and r8153b_set_eee().
Signed-off-by: Hayes Wang <hayeswang@realtek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Saving the settings of EEE to avoid they become the default settings
after reset_resume().
Signed-off-by: Hayes Wang <hayeswang@realtek.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Update response packet length for the following commands per NC-SI spec
- Get Controller Packet Statistics
- Get NC-SI Statistics
- Get NC-SI Pass-through Statistics command
Signed-off-by: Ben Wei <benwei@fb.com>
Reviewed-by: Justin Lee <justin.lee1@dell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The dev_kfree_skb() function performs also input parameter validation.
Thus the test around the shown calls is not needed.

This issue was detected by using the Coccinelle software.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

The dev_kfree_skb() function performs also input parameter validation.
Thus the test around the shown calls is not needed.

This issue was detected by using the Coccinelle software.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

The dev_kfree_skb() function performs also input parameter validation.
Thus the test around the shown calls is not needed.

This issue was detected by using the Coccinelle software.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

The dev_kfree_skb() function performs also input parameter validation.
Thus the test around the shown calls is not needed.

This issue was detected by using the Coccinelle software.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Acked-by: Sean Nyekjaer <sean@geanix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The dev_kfree_skb() function performs also input parameter validation.
Thus the test around the shown calls is not needed.

This issue was detected by using the Coccinelle software.
Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

IEEE 802.3ae clause 45 defines a modified MDIO protocol that uses a two
staged access model in order to increase the address space.

This patch adds support for C45 MDIO read and write accesses, which are
used whenever the MII_ADDR_C45 flag in the regnum argument is set.
In case it is not set, C22 accesses are used as before.
Signed-off-by: Marco Hartmann <marco.hartmann@nxp.com>
Acked-by: Fugang Duan <fugang.duan@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Neil Armstrong says:

====================
dt-bindings: net: meson-dwmac: convert to yaml

This patchsets converts the Amlogic Meson DWMAC glue bindings over to
YAML schemas using the already converted dwmac bindings.

The first patch is needed because the Amlogic glue needs a supplementary
reg cell to access the DWMAC glue registers.

Changes since v3:
- Specified net-next target tree

Changes since v2:
- Added review tags
- Updated allwinner,sun7i-a20-gmac.yaml reg maxItems
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Now that we have the DT validation in place, let's convert the device tree
bindings for the Synopsys DWMAC Glue for Amlogic SoCs over to a YAML schemas.
Reviewed-by: Rob Herring <robh@kernel.org>
Reviewed-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The Amlogic Meson DWMAC glue bindings needs a second reg cells for the
glue registers, thus update the reg minItems/maxItems to allow more
than a single reg cell.

Also update the allwinner,sun7i-a20-gmac.yaml derivative schema to specify
maxItems to 1.
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Acked-by: Rob Herring <robh@kernel.org>
Acked-by: Maxime Ripard <maxime.ripard@bootlin.com>
Reviewed-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Jeff Kirsher says:

====================
40GbE Intel Wired LAN Driver Updates 2019-08-22

This series contains updates to i40e driver only.

Arnd Bergmann reduces the stack usage which was causing warnings on
32-bit architectures due to large structure sizes for 2 functions
getting inlined, so use noinline_for_stack to prevent the compilers from
combining the 2 functions.

Mauro S. M. Rodrigues fixes an issue when reading an EEPROM from SFP
modules that comply with SFF-8472 but do not implement the Digital
Diagnostic Monitoring (DDM) interface for i40e.

Huhai found we were not checking the return value for configuring the
transmit ring and continuing with XDP configuration of the transmit
ring.

Beilei fixes an issue of shifting signed 32-bit integers.

Sylwia adds support for "packet drop mode" to the MAC configuration for
admin queue command.  This bit controls the behavior when a no-drop
packet is blocking a TC queue.  Adds support for persistent LLDP by
checking the LLDP flag and reading the LLDP from the NVM when enabled.

Adrian fixes the "recovery mode" check to take into account which device
we are on, since x710 devices have 4 register values to check for status
and x722 devices only have 2 register values to check.

Piotr Azarewicz bumps the supported firmware API version to 1.9 which
extends the PHY access admin queue command support.

Jake makes sure the traffic class stats for a VEB are reset when the VEB
stats are reset.

Slawomir fixes a NULL pointer dereference where the VSI pointer was not
updated before passing it to the i40e_set_vf_mac() when the VF is in a
reset state, so wait for the reset to complete.

Grzegorz removes the i40e_update_dcb_config() which was not using the
correct NVM reads, so call i40e_init_dcb() in its place to correctly
update the DCB configuration.

Piotr Kwapulinski expands the scope of i40e_set_mac_type() since this is
needed during probe to determine if we are in recovery mode.  Fixed the
driver reset path when in recovery mode.

Marcin fixed an issue where we were breaking out of a loop too early
when trying to get the PHY capabilities.

v2: Combined patch 7 & 9 in the original series, since both patches
    bumped firmware API version.  Also combined patches 12 & 13 in the
    original series, since one increased the scope of checking for MAC
    and the follow-on patch made use of function within the new scope.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Fixed a bug where driver was breaking out of the loop and
reporting an error without retrying first.
Signed-off-by: Marcin Formela <marcin.formela@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

This patch adds a function to read NVM module data and uses it to
read current LLDP agent configuration from NVM API version 1.8.
Signed-off-by: Sylwia Wnuczko <sylwia.wnuczko@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Driver waits after issuing a reset. When a reset takes too long a driver
gives up. Implemented by invoking PF reset in a loop. After defined
number of unsuccessful PF reset trials it returns error.
Without this patch PF reset fails when NIC is in recovery mode.

So make i40e_set_mac_type() public. i40e driver requires i40e_set_mac_type()
to be public. It is required for recovery mode handling. Without this patch
recovery mode could not be detected in i40e_probe().
Signed-off-by: Piotr Kwapulinski <piotr.kwapulinski@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

This patch removes function i40e_update_dcb_config(). Instead of
i40e_update_dcb_config() we use i40e_init_dcb(), which implements the
correct NVM read.
Signed-off-by: Grzegorz Siwik <grzegorz.siwik@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Add update to the VSI pointer passed to the i40e_set_vf_mac function.
If VF is in reset state the driver waits in i40e_set_vf_mac function
for the reset to be complete, yet after reset the vsi pointer
that was passed into this function is no longer valid.

The patch updates local VSI pointer directly from pf->vsi array,
by using the id stored in VF pointer (lan_vsi_idx).

Without this commit the driver might occasionally invoke general
protection fault in kernel and disable the OS entirely.
Signed-off-by: Slawomir Laba <slawomirx.laba@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

The stats structure for the VEB switch statistics is reset periodically,
but the tc_stats are not reset at the same time.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Upcoming FW increment API version to 1.9 due to Extend PHY access AQ
command support. SW is ready for that support as well.
Signed-off-by: Piotr Azarewicz <piotr.azarewicz@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Function check_recovery_mode had wrong if statement.
Now we check proper FWS1B register values, which are responsible for
the recovery mode. Recovery mode has 4 values for x710 and 2 for x722.
That's why we need 6 different flags which are defined in the code.
Now in the if statement, we recognize type of mac address
and register value.
Without those changes driver could show wrong state.
Signed-off-by: Adrian Podlawski <adrian.podlawski@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

This patch adds "drop mode" parameter to set mac config AQ command.
This bit controls the behavior when a no-drop packet is blocking a TC
queue.
0 – The PF driver is notified.
1 – The blocking packet is dropped and then the PF driver is notified.
Signed-off-by: Sylwia Wnuczko <sylwia.wnuczko@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

This patch fixes following error reported by cppcheck:
(error) Shifting signed 32-bit value by 31 bits is undefined behaviour
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

When i40e_configure_tx_ring(vsi->tx_rings[i]) returns an error, we should
exit from i40e_vsi_configure_tx and return the error, instead of continuing
to check whether xdp is enable, and configure the xdp transmit ring.
Signed-off-by: huhai <huhai@kylinos.cn>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Similar to the ixgbe issue fixed in:
655c9141 ("ixgbe: Check DDM existence in transceiver before access)

i40e has the same issue when reading eeprom from SFP's module that comply
with SFF-8472 but not implement the Digital Diagnostic Monitoring (DDM)
interface described in it. The existence of such area is specified by bit
6 of byte 92, set to 1 if implemented.

Without this patch, due to not checking this bit i40e fails to read SFP
module's eeprom with the follow message:

ethtool -m enP51p1s0f0
Cannot get Module EEPROM data: Input/output error

Because it fails to read the additional 256 bytes in which it was assumed
to exist the DDM data.
Signed-off-by: "Mauro S. M. Rodrigues" <maurosr@linux.vnet.ibm.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

The functions i40e_aq_get_phy_abilities_resp() and i40e_set_fc() both
have giant structure on the stack, which makes each one use stack frames
larger than 500 bytes.

As clang decides one function into the other, we get a warning for
exceeding the frame size limit on 32-bit architectures:

drivers/net/ethernet/intel/i40e/i40e_common.c:1654:23: error: stack frame size of 1116 bytes in function 'i40e_set_fc' [-Werror,-Wframe-larger-than=]

When building with gcc, the inlining does not happen, but i40e_set_fc()
calls i40e_aq_get_phy_abilities_resp() anyway, so they add up on the
kernel stack just as much.

The parts that actually use large stacks don't overlap, so make sure
each one is a separate function, and mark them as noinline_for_stack to
prevent the compilers from combining them again.

Fixes: 0a862b43 ("i40e/i40evf: Add module_types and update_link_info")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Variable err is initialized to a value that is never read and it is
re-assigned later. The initialization is redundant and can be removed.

Addresses-Coverity: ("Unused Value")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Haiyang Zhang says:

====================
Add software backchannel and mlx5e HV VHCA stats

This patch set adds paravirtual backchannel in software in pci_hyperv,
which is required by the mlx5e driver HV VHCA stats agent.

The stats agent is responsible on running a periodic rx/tx packets/bytes
stats update.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

HV VHCA stats agent is responsible on running a preiodic rx/tx
packets/bytes stats update. Currently the supported format is version
MLX5_HV_VHCA_STATS_VERSION. Block ID 1 is dedicated for statistics data
transfer from the VF to the PF.

The reporter fetch the statistics data from all opened channels, fill it
in a buffer and send it to mlx5_hv_vhca_write_agent.

As the stats layer should include some metadata per block (sequence and
offset), the HV VHCA layer shall modify the buffer before actually send it
over block 1.
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Control agent is responsible over of the control block (ID 0). It should
update the PF via this block about every capability change. In addition,
upon block 0 invalidate, it should activate all other supported agents
with data requests from the PF.

Upon agent create/destroy, the invalidate callback of the control agent
is being called in order to update the PF driver about this change.

The control agent is an integral part of HV VHCA and will be created
and destroy as part of the HV VHCA init/cleanup flow.
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

HV VHCA is a layer which provides PF to VF communication channel based on
HyperV PCI config channel. It implements Mellanox's Inter VHCA control
communication protocol. The protocol contains control block in order to
pass messages between the PF and VF drivers, and data blocks in order to
pass actual data.

The infrastructure is agent based. Each agent will be responsible of
contiguous buffer blocks in the VHCA config space. This infrastructure will
bind agents to their blocks, and those agents can only access read/write
the buffer blocks assigned to them. Each agent will provide three
callbacks (control, invalidate, cleanup). Control will be invoked when
block-0 is invalidated with a command that concerns this agent. Invalidate
callback will be invoked if one of the blocks assigned to this agent was
invalidated. Cleanup will be invoked before the agent is being freed in
order to clean all of its open resources or deferred works.

Block-0 serves as the control block. All execution commands from the PF
will be written by the PF over this block. VF will ack on those by
writing on block-0 as well. Its format is described by struct
mlx5_hv_vhca_control_block layout.
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add wrapper functions for HyperV PCIe read / write /
block_invalidate_register operations.  This will be used as an
infrastructure in the downstream patch for software communication.

This will be enabled by default if CONFIG_PCI_HYPERV_INTERFACE is set.
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This interface driver is a helper driver allows other drivers to
have a common interface with the Hyper-V PCI frontend driver.
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Windows SR-IOV provides a backchannel mechanism in software for communication
between a VF driver and a PF driver.  These "configuration blocks" are
similar in concept to PCI configuration space, but instead of doing reads and
writes in 32-bit chunks through a very slow path, packets of up to 128 bytes
can be sent or received asynchronously.

Nearly every SR-IOV device contains just such a communications channel in
hardware, so using this one in software is usually optional.  Using the
software channel, however, allows driver implementers to leverage software
tools that fuzz the communications channel looking for vulnerabilities.

The usage model for these packets puts the responsibility for reading or
writing on the VF driver.  The VF driver sends a read or a write packet,
indicating which "block" is being referred to by number.

If the PF driver wishes to initiate communication, it can "invalidate" one or
more of the first 64 blocks.  This invalidation is delivered via a callback
supplied by the VF driver by this driver.

No protocol is implied, except that supplied by the PF and VF drivers.
Signed-off-by: Jake Oshins <jakeo@microsoft.com>
Signed-off-by: Dexuan Cui <decui@microsoft.com>
Cc: Haiyang Zhang <haiyangz@microsoft.com>
Cc: K. Y. Srinivasan <kys@microsoft.com>
Cc: Stephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Saeed Mahameed says:

====================
mlx5 tc flow handling for concurrent execution (Part 3)

This series includes updates to mlx5 ethernet and core driver:

Vlad submits part 3 of 3 part series to allow TC flow handling
for concurrent execution.

Vlad says:
==========

Structure mlx5e_neigh_hash_entry code that uses it are refactored in
following ways:

- Extend neigh_hash_entry with rcu and modify its users to always take
  reference to the structure when using it (neigh_hash_entry has already
  had atomic reference counter which was only used when scheduling neigh
  update on workqueue from atomic context of neigh update netevent).

- Always use mlx5e_neigh_update_table->encap_lock when modifying neigh
  update hash table and list. Originally, this lock was only used to
  synchronize with netevent handler function, which is called from bh
  context and cannot use rtnl lock for synchronization. Use rcu read lock
  instead of encap_lock to lookup nhe in atomic context of netevent even
  handler function. Convert encap_lock to mutex to allow creating new
  neigh hash entries while holding it, which is safe to do because the
  lock is no longer used in atomic context.

- Rcu-ify mlx5e_neigh_hash_entry->encap_list by changing operations on
  encap list to their rcu counterparts and extending encap structure
  with rcu_head to free the encap instances after rcu grace period. This
  allows fast traversal of list of encaps attached to nhe under rcu read
  lock protection.

- Take encap_table_lock when accessing encap entries in neigh update and
  neigh stats update code to protect from concurrent encap entry
  insertion or removal.

This approach leads to potential race condition when neigh update and
neigh stats update code can access encap and flow entries that are not
fully initialized or are being destroyed, or neigh can change state
without updating encaps that are created concurrently. Prevent these
issues by following changes in flow and encap initialization:

- Extend mlx5e_tc_flow with 'init_done' completion. Modify neigh update
  to wait for both encap and flow completions to prevent concurrent
  access to a structure that is being initialized by tc.

- Skip structures that failed during initialization: encaps with
  encap_id<0 and flows that don't have OFFLOADED flag set.

- To ensure that no new flows are added to encap when it is being
  accessed by neigh update or neigh stats update, take encap_table_lock
  mutex.

- To prevent concurrent deletion by tc, ensure that neigh update and
  neigh stats update hold references to encap and flow instances while
  using them.

With changes presented in this patch set it is now safe to execute tc
concurrently with neigh update and neigh stats update. However, these
two workqueue tasks modify same flow "tmp_list" field to store flows
with reference taken in temporary list to release the references after
update operation finishes and should not be executed concurrently with
each other.

Last 3 patches of this series provide 3 new mlx5 trace points to track
mlx5 tc requests and mlx5 neigh updates.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Allow tracing neigh state during neigh update task that is executed on
workqueue and is scheduled by neigh state change event.

Usage example:
 ># cd /sys/kernel/debug/tracing
 ># echo mlx5:mlx5e_rep_neigh_update >> set_event
 ># cat trace
    ...
    kworker/u48:7-2221  [009] ...1  1475.387435: mlx5e_rep_neigh_update:
netdev: ens1f0 MAC: 24:8a:07:9a:17:9a IPv4: 1.1.1.10 IPv6: ::ffff:1.1.1.10 neigh_connected=1

Added corresponding documentation in
    Documentation/networking/device-driver/mellanox/mlx5.rst
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Reviewed-by: Dmytro Linkin <dmitrolin@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>