Handle memory allocation failure from nci_skb_alloc() (calling
alloc_skb()) to avoid possible NULL pointer dereference.

Reported-by: 黄思聪 <huangsicong@iie.ac.cn>
Fixes: 391d8a2d

 ("NFC: Add NCI over SPI receive")
Cc: <stable@vger.kernel.org>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://lore.kernel.org/r/20231013184129.18738-1-krzysztof.kozlowski@linaro.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

rtnl_offload_xstats_get_size_hw_s_info_one() conditionalizes the
size-computation for IFLA_OFFLOAD_XSTATS_HW_S_INFO_USED based on whether
or not the device has offload_xstats enabled.

However, rtnl_offload_xstats_fill_hw_s_info_one() is adding the u8 for
that field uncondtionally.

syzkaller triggered a WARNING in rtnl_stats_get due to this:
------------[ cut here ]------------
WARNING: CPU: 0 PID: 754 at net/core/rtnetlink.c:5982 rtnl_stats_get+0x2f4/0x300
Modules linked in:
CPU: 0 PID: 754 Comm: syz-executor148 Not tainted 6.6.0-rc2-g331b78eb12af #45
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.11.0-2.el7 04/01/2014
RIP: 0010:rtnl_stats_get+0x2f4/0x300 net/core/rtnetlink.c:5982
Code: ff ff 89 ee e8 7d 72 50 ff 83 fd a6 74 17 e8 33 6e 50 ff 4c 89 ef be 02 00 00 00 e8 86 00 fa ff e9 7b fe ff ff e8 1c 6e 50 ff <0f> 0b eb e5 e8 73 79 7b 00 0f 1f 00 90 90 90 90 90 90 90 90 90 90
RSP: 0018:ffffc900006837c0 EFLAGS: 00010293
RAX: ffffffff81cf7f24 RBX: ffff8881015d9000 RCX: ffff888101815a00
RDX: 0000000000000000 RSI: 00000000ffffffa6 RDI: 00000000ffffffa6
RBP: 00000000ffffffa6 R08: ffffffff81cf7f03 R09: 0000000000000001
R10: ffff888101ba47b9 R11: ffff888101815a00 R12: ffff8881017dae00
R13: ffff8881017dad00 R14: ffffc90000683ab8 R15: ffffffff83c1f740
FS:  00007fbc22dbc740(0000) GS:ffff88813bc00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000020000046 CR3: 000000010264e003 CR4: 0000000000170ef0
Call Trace:
 <TASK>
 rtnetlink_rcv_msg+0x677/0x710 net/core/rtnetlink.c:6480
 netlink_rcv_skb+0xea/0x1c0 net/netlink/af_netlink.c:2545
 netlink_unicast+0x430/0x500 net/netlink/af_netlink.c:1342
 netlink_sendmsg+0x4fc/0x620 net/netlink/af_netlink.c:1910
 sock_sendmsg+0xa8/0xd0 net/socket.c:730
 ____sys_sendmsg+0x22a/0x320 net/socket.c:2541
 ___sys_sendmsg+0x143/0x190 net/socket.c:2595
 __x64_sys_sendmsg+0xd8/0x150 net/socket.c:2624
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x47/0xa0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x6e/0xd8
RIP: 0033:0x7fbc22e8d6a9
Code: 5c c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 4f 37 0d 00 f7 d8 64 89 01 48
RSP: 002b:00007ffc4320e778 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
RAX: ffffffffffffffda RBX: 00000000004007d0 RCX: 00007fbc22e8d6a9
RDX: 0000000000000000 RSI: 0000000020000000 RDI: 0000000000000003
RBP: 0000000000000001 R08: 0000000000000000 R09: 00000000004007d0
R10: 0000000000000008 R11: 0000000000000246 R12: 00007ffc4320e898
R13: 00007ffc4320e8a8 R14: 00000000004004a0 R15: 00007fbc22fa5a80
 </TASK>
---[ end trace 0000000000000000 ]---

Which didn't happen prior to commit bf9f1baa ("net: add dedicated
kmem_cache for typical/small skb->head") as the skb always was large
enough.

Fixes: 0e7788fd

 ("net: rtnetlink: Add UAPI for obtaining L3 offload xstats")
Signed-off-by: Christoph Paasch <cpaasch@apple.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Link: https://lore.kernel.org/r/20231013041448.8229-1-cpaasch@apple.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

In the smc_listen_work(), if smc_listen_prfx_check() failed,
the real reason: SMC_CLC_DECL_DIFFPREFIX was dropped, and
SMC_CLC_DECL_NOSMCDEV was returned.

Althrough this is also kind of SMC_CLC_DECL_NOSMCDEV, but return
the real reason is much friendly for debugging.

Fixes: e49300a6

 ("net/smc: add listen processing for SMC-Rv2")
Signed-off-by: Dust Li <dust.li@linux.alibaba.com>
Reviewed-by: Alexandra Winter <wintera@linux.ibm.com>
Reviewed-by: Wenjia Zhang <wenjia@linux.ibm.com>
Link: https://lore.kernel.org/r/20231012123729.29307-1-dust.li@linux.alibaba.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

From: Aaron Conole <aconole@redhat.com>
To: netdev@vger.kernel.org
Cc: dev@openvswitch.org, linux-kselftest@vger.kernel.org,
	linux-kernel@vger.kernel.org, Pravin B Shelar <pshelar@ovn.org>,
	"David S. Miller" <davem@davemloft.net>,
	Eric Dumazet <edumazet@google.com>,
	Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	Adrian Moreno <amorenoz@redhat.com>,
	Eelco Chaudron <echaudro@redhat.com>,
	shuah@kernel.org
Subject: [PATCH net v2 0/4] selftests: openvswitch: Minor fixes for some systems
Date: Wed, 11 Oct 2023 15:49:35 -0400	[thread overview]
Message-ID: <20231011194939.704565-1-aconole@redhat.com> (raw)

A number of corner cases were caught when trying to run the selftests on
older systems.  Missed skip conditions, some error cases, and outdated
python setups would all report failures but the issue would actually be
related to some other condition rather than the selftest suite.

Address these individual cases.
====================

Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

The ct_tuple v4 data structure decode / encode routines were using
the v6 IP address decode and relying on default encode. This could
cause exceptions during encode / decode depending on how a ct4
tuple would appear in a netlink message.

Caught during code review.

Fixes: e52b07aa

 ("selftests: openvswitch: add flow dump support")
Signed-off-by: Aaron Conole <aconole@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Kernels that don't have support for openvswitch drop reasons also
won't have the drop counter reasons, so we should skip the test
completely.  It previously wasn't possible to build a test case
for this without polluting the datapath, so we introduce a mechanism
to clear all the flows from a datapath allowing us to test for
explicit drop actions, and then clear the flows to build the
original test case.

Fixes: 42420291

 ("selftests: openvswitch: add explicit drop testcase")
Signed-off-by: Aaron Conole <aconole@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In case of fatal signal, or early abort at least cleanup the current
test case.

Fixes: 25f16c87

 ("selftests: add openvswitch selftest suite")
Signed-off-by: Aaron Conole <aconole@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Paolo Abeni reports that on some systems the pyroute2 version isn't
new enough to run the test suite.  Ensure that we support a minimum
version of 0.6 for all cases (which does include the existing ones).
The 0.6.1 version was released in May of 2021, so should be
propagated to most installations at this point.

The alternative that Paolo proposed was to only skip when the
add-flow is being run.  This would be okay for most cases, except
if a future test case is added that needs to do flow dump without
an associated add (just guessing).  In that case, it could also be
broken and we would need additional skip logic anyway.  Just draw
a line in the sand now.

Fixes: 25f16c87

 ("selftests: add openvswitch selftest suite")
Reported-by: Paolo Abeni <pabeni@redhat.com>
Closes: https://lore.kernel.org/lkml/8470c431e0930d2ea204a9363a60937289b7fdbe.camel@redhat.com/
Signed-off-by: Aaron Conole <aconole@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Syzbot reported two new paths to hit an internal WARNING using the
new virtio gso type VIRTIO_NET_HDR_GSO_UDP_L4.

    RIP: 0010:skb_checksum_help+0x4a2/0x600 net/core/dev.c:3260
    skb len=64521 gso_size=344
and

    RIP: 0010:skb_warn_bad_offload+0x118/0x240 net/core/dev.c:3262

Older virtio types have historically had loose restrictions, leading
to many entirely impractical fuzzer generated packets causing
problems deep in the kernel stack. Ideally, we would have had strict
validation for all types from the start.

New virtio types can have tighter validation. Limit UDP GSO packets
inserted via virtio to the same limits imposed by the UDP_SEGMENT
socket interface:

1. must use checksum offload
2. checksum offload matches UDP header
3. no more segments than UDP_MAX_SEGMENTS
4. UDP GSO does not take modifier flags, notably SKB_GSO_TCP_ECN

Fixes: 860b7f27

 ("linux/virtio_net.h: Support USO offload in vnet header.")
Reported-by:  <syzbot+01cdbc31e9c0ae9b33ac@syzkaller.appspotmail.com>
Closes: https://lore.kernel.org/netdev/0000000000005039270605eb0b7f@google.com/
Reported-by:  <syzbot+c99d835ff081ca30f986@syzkaller.appspotmail.com>
Closes: https://lore.kernel.org/netdev/0000000000005426680605eb0b9f@google.com/
Signed-off-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Driver allocates the LL2 rx buffers from kmalloc()
area to construct the skb using slab_build_skb()

The required size allocation seems to have overlooked
for accounting both skb_shared_info size and device
placement padding bytes which results into the below
panic when doing skb_put() for a standard MTU sized frame.

skbuff: skb_over_panic: text:ffffffffc0b0225f len:1514 put:1514
head:ff3dabceaf39c000 data:ff3dabceaf39c042 tail:0x62c end:0x566
dev:<NULL>
…
skb_panic+0x48/0x4a
skb_put.cold+0x10/0x10
qed_ll2b_complete_rx_packet+0x14f/0x260 [qed]
qed_ll2_rxq_handle_completion.constprop.0+0x169/0x200 [qed]
qed_ll2_rxq_completion+0xba/0x320 [qed]
qed_int_sp_dpc+0x1a7/0x1e0 [qed]

This patch fixes this by accouting skb_shared_info and device
placement padding size bytes when allocating the buffers.

Cc: David S. Miller <davem@davemloft.net>
Fixes: 0a7fb11c

 ("qed: Add Light L2 support")
Signed-off-by: Manish Chopra <manishc@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Saeed Mahameed says:

====================
mlx5 fixes 2023-10-12

This series provides bug fixes to mlx5 driver.

* tag 'mlx5-fixes-2023-10-12' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux:
  net/mlx5e: Fix VF representors reporting zero counters to "ip -s" command
  net/mlx5e: Don't offload internal port if filter device is out device
  net/mlx5e: Take RTNL lock before triggering netdev notifiers
  net/mlx5e: XDP, Fix XDP_REDIRECT mpwqe page fragment leaks on shutdown
  net/mlx5e: RX, Fix page_pool allocation failure recovery for legacy rq
  net/mlx5e: RX, Fix page_pool allocation failure recovery for striding rq
  net/mlx5: Handle fw tracer change ownership event based on MTRC
  net/mlx5: Bridge, fix peer entry ageing in LAG mode
  net/mlx5: E-switch, register event handler before arming the event
  net/mlx5: Perform DMA operations in the right locations
====================

Link: https://lore.kernel.org/r/20231012195127.129585-1-saeed@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Jacob Keller says:

====================
Intel Wired LAN Driver Updates 2023-10-11 (i40e, ice)

This series contains fixes for the i40e and ice drivers.

Jesse adds handling to the ice driver which resetis the device when loading
on a crash kernel, preventing stale transactions from causing machine check
exceptions which could prevent capturing crash data.

Mateusz fixes a bug in the ice driver 'Safe mode' logic for handling the
device when the DDP is missing.

Michal fixes a crash when probing the i40e driver in the event that HW
registers are reporting invalid/unexpected values.

The following are changes since commit a950a592

:
  net/smc: Fix pos miscalculation in statistics

I'm covering for Tony Nguyen while he's out, and don't have access to create
a pull request branch on his net-queue, so these are sent via mail only.
====================

Link: https://lore.kernel.org/r/20231011233334.336092-1-jacob.e.keller@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

One thing is broken in the safe mode, that is
ice_deinit_features() is being executed even
that ice_init_features() was not causing stack
trace during pci_unregister_driver().

Add check on the top of the function.

Fixes: 5b246e53

 ("ice: split probe into smaller functions")
Signed-off-by: Mateusz Pacuszka <mateuszx.pacuszka@intel.com>
Signed-off-by: Jan Sokolowski <jan.sokolowski@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Link: https://lore.kernel.org/r/20231011233334.336092-4-jacob.e.keller@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

When the system boots into the crash dump kernel after a panic, the ice
networking device may still have pending transactions that can cause errors
or machine checks when the device is re-enabled. This can prevent the crash
dump kernel from loading the driver or collecting the crash data.

To avoid this issue, perform a function level reset (FLR) on the ice device
via PCIe config space before enabling it on the crash kernel. This will
clear any outstanding transactions and stop all queues and interrupts.
Restore the config space after the FLR, otherwise it was found in testing
that the driver wouldn't load successfully.

The following sequence causes the original issue:
- Load the ice driver with modprobe ice
- Enable SR-IOV with 2 VFs: echo 2 > /sys/class/net/eth0/device/sriov_num_vfs
- Trigger a crash with echo c > /proc/sysrq-trigger
- Load the ice driver again (or let it load automatically) with modprobe ice
- The system crashes again during pcim_enable_device()

Fixes: 837f08fd

 ("ice: Add basic driver framework for Intel(R) E800 Series")
Reported-by: Vishal Agrawal <vagrawal@redhat.com>
Reviewed-by: Jay Vosburgh <jay.vosburgh@canonical.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Link: https://lore.kernel.org/r/20231011233334.336092-3-jacob.e.keller@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

The hardware provides the indexes of the first and the last available
queue and VF. From the indexes, the driver calculates the numbers of
queues and VFs. In theory, a faulty device might say the last index is
smaller than the first index. In that case, the driver's calculation
would underflow, it would attempt to write to non-existent registers
outside of the ioremapped range and crash.

I ran into this not by having a faulty device, but by an operator error.
I accidentally ran a QE test meant for i40e devices on an ice device.
The test used 'echo i40e > /sys/...ice PCI device.../driver_override',
bound the driver to the device and crashed in one of the wr32 calls in
i40e_clear_hw.

Add checks to prevent underflows in the calculations of num_queues and
num_vfs. With this fix, the wrong device probing reports errors and
returns a failure without crashing.

Fixes: 838d41d9

 ("i40e: clear all queues and interrupts")
Signed-off-by: Michal Schmidt <mschmidt@redhat.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Link: https://lore.kernel.org/r/20231011233334.336092-2-jacob.e.keller@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Florian Westphal says:

====================
netfilter updates for net

Patch 1, from Pablo Neira Ayuso, fixes a performance regression
(since 6.4) when a large pending set update has to be canceled towards
the end of the transaction.

Patch 2 from myself, silences an incorrect compiler warning reported
with a few (older) compiler toolchains.

Patch 3, from Kees Cook, adds __counted_by annotation to
nft_pipapo set backend type.  I took this for net instead of -next
given infra is already in place and no actual code change is made.

Patch 4, from Pablo Neira Ayso, disables timeout resets on
stateful element reset.  The rest should only affect internal object
state, e.g. reset a quota or counter, but not affect a pending timeout.

Patches 5 and 6 fix NULL dereferences in 'inner header' match,
control plane doesn't test for netlink attribute presence before
accessing them. Broken since feature was added in 6.2, fixes from
Xingyuan Mo.

Last patch, from myself, fixes a bogus rule match when skb has
a 0-length mac header, in this case we'd fetch data from network
header instead of canceling rule evaluation.  This is a day 0 bug,
present since nftables was merged in 3.13.

* tag 'nf-23-10-12' of https://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:
  netfilter: nft_payload: fix wrong mac header matching
  nf_tables: fix NULL pointer dereference in nft_expr_inner_parse()
  nf_tables: fix NULL pointer dereference in nft_inner_init()
  netfilter: nf_tables: do not refresh timeout when resetting element
  netfilter: nf_tables: Annotate struct nft_pipapo_match with __counted_by
  netfilter: nfnetlink_log: silence bogus compiler warning
  netfilter: nf_tables: do not remove elements if set backend implements .abort
====================

Link: https://lore.kernel.org/r/20231012085724.15155-1-fw@strlen.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

ICSSG HW stats on TX side considers 8 preamble bytes as data bytes. Due
to this the tx_bytes of ICSSG interface doesn't match the rx_bytes of the
link partner. There is no public errata available yet.

As a workaround to fix this, decrease tx_bytes by 8 bytes for every tx
frame.

Fixes: c1e10d5d

 ("net: ti: icssg-prueth: Add ICSSG Stats")
Signed-off-by: MD Danish Anwar <danishanwar@ti.com>
Link: https://lore.kernel.org/r/20231012064626.977466-1-danishanwar@ti.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Update the "How are representors identified?" documentation
subchapter. For newer kernels driver should use
SET_NETDEV_DEVLINK_PORT instead of ndo_get_devlink_port()
callback.

Fixes: 7712b3e9

 ("Merge branch 'net-fix-netdev-to-devlink_port-linkage-and-expose-to-user'")
Signed-off-by: Mateusz Polchlopek <mateusz.polchlopek@intel.com>
Reviewed-by: Wojciech Drewek <wojciech.drewek@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Edward Cree <ecree.xilinx@gmail.com>
Link: https://lore.kernel.org/r/20231012123144.15768-1-mateusz.polchlopek@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Make sure that the command values used for replies are correct. This is
only affecting generated userspace helpers, no change on kernel code.

Fixes: 7199c862

 ("netlink: specs: devlink: add commands that do per-instance dump")
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Link: https://lore.kernel.org/r/20231012115811.298129-1-jiri@resnulli.us
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

If the netdevice is within a container and communicates externally
through network technologies such as VxLAN, we won't be able to find
routing information in the init_net namespace. To address this issue,
we need to add a struct net parameter to the smc_ib_find_route function.
This allow us to locate the routing information within the corresponding
net namespace, ensuring the correct completion of the SMC CLC interaction.

Fixes: e5c4744c

 ("net/smc: add SMC-Rv2 connection establishment")
Signed-off-by: Albert Huang <huangjie.albert@bytedance.com>
Reviewed-by: Dust Li <dust.li@linux.alibaba.com>
Reviewed-by: Wenjia Zhang <wenjia@linux.ibm.com>
Link: https://lore.kernel.org/r/20231011074851.95280-1-huangjie.albert@bytedance.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

As reported by Tom, .NET and applications build on top of it rely
on connect(AF_UNSPEC) to async cancel pending I/O operations on TCP
socket.

The blamed commit below caused a regression, as such cancellation
can now fail.

As suggested by Eric, this change addresses the problem explicitly
causing blocking I/O operation to terminate immediately (with an error)
when a concurrent disconnect() is executed.

Instead of tracking the number of threads blocked on a given socket,
track the number of disconnect() issued on such socket. If such counter
changes after a blocking operation releasing and re-acquiring the socket
lock, error out the current operation.

Fixes: 4faeee0c

 ("tcp: deny tcp_disconnect() when threads are waiting")
Reported-by: Tom Deseyn <tdeseyn@redhat.com>
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1886305
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/f3b95e47e3dbed840960548aebaa8d954372db41.1697008693.git.pabeni@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Since the introduction of the ice driver the code has been
double-shifting the RSS enabling field, because the define already has
shifts in it and can't have the regular pattern of "a << shiftval &
mask" applied.

Most places in the code got it right, but one line was still wrong. Fix
this one location for easy backports to stable. An in-progress patch
fixes the defines to "standard" and will be applied as part of the
regular -next process sometime after this one.

Fixes: d76a60ba

 ("ice: Add support for VLANs and offloads")
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
CC: stable@vger.kernel.org
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Link: https://lore.kernel.org/r/20231010203101.406248-1-jacob.e.keller@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

In bcm_sf2_mdio_register(), the class_find_device() will call get_device()
to increment reference count for priv->master_mii_bus->dev if
of_mdio_find_bus() succeeds. If mdiobus_alloc() or mdiobus_register()
fails, it will call get_device() twice without decrement reference count
for the device. And it is the same if bcm_sf2_mdio_register() succeeds but
fails in bcm_sf2_sw_probe(), or if bcm_sf2_sw_probe() succeeds. If the
reference count has not decremented to zero, the dev related resource will
not be freed.

So remove the get_device() in bcm_sf2_mdio_register(), and call
put_device() if mdiobus_alloc() or mdiobus_register() fails and in
bcm_sf2_mdio_unregister() to solve the issue.

And as Simon suggested, unwind from errors for bcm_sf2_mdio_register() and
just return 0 if it succeeds to make it cleaner.

Fixes: 461cd1b0

 ("net: dsa: bcm_sf2: Register our slave MDIO bus")
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
Suggested-by: Simon Horman <horms@kernel.org>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Link: https://lore.kernel.org/r/20231011032419.2423290-1-ruanjinjie@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Ido Schimmel says:

====================
selftests: fib_tests: Fixes for multipath list receive tests

Fix two issues in recently added FIB multipath list receive tests.
====================

Link: https://lore.kernel.org/r/20231010132113.3014691-1-idosch@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

The tests rely on the IPv{4,6} FIB trace points being triggered once for
each forwarded packet. If receive processing is deferred to the
ksoftirqd task these invocations will not be counted and the tests will
fail. Fix by specifying the '-a' flag to avoid perf from filtering on
the mausezahn task.

Before:

 # ./fib_tests.sh -t ipv4_mpath_list

 IPv4 multipath list receive tests
     TEST: Multipath route hit ratio (.68)                               [FAIL]

 # ./fib_tests.sh -t ipv6_mpath_list

 IPv6 multipath list receive tests
     TEST: Multipath route hit ratio (.27)                               [FAIL]

After:

 # ./fib_tests.sh -t ipv4_mpath_list

 IPv4 multipath list receive tests
     TEST: Multipath route hit ratio (1.00)                              [ OK ]

 # ./fib_tests.sh -t ipv6_mpath_list

 IPv6 multipath list receive tests
     TEST: Multipath route hit ratio (.99)                               [ OK ]

Fixes: 8ae9efb8

 ("selftests: fib_tests: Add multipath list receive tests")
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/netdev/202309191658.c00d8b8-oliver.sang@intel.com/
Tested-by: kernel test robot <oliver.sang@intel.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Tested-by: Sriram Yagnaraman <sriram.yagnaraman@est.tech>
Link: https://lore.kernel.org/r/20231010132113.3014691-3-idosch@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

The test relies on the fib:fib_table_lookup trace point being triggered
once for each forwarded packet. If RP filter is not disabled, the trace
point will be triggered twice for each packet (for source validation and
forwarding), potentially masking actual bugs. Fix by explicitly
disabling RP filter.

Before:

 # ./fib_tests.sh -t ipv4_mpath_list

 IPv4 multipath list receive tests
     TEST: Multipath route hit ratio (1.99)                              [ OK ]

After:

 # ./fib_tests.sh -t ipv4_mpath_list

 IPv4 multipath list receive tests
     TEST: Multipath route hit ratio (.99)                               [ OK ]

Fixes: 8ae9efb8

 ("selftests: fib_tests: Add multipath list receive tests")
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/netdev/202309191658.c00d8b8-oliver.sang@intel.com/
Tested-by: kernel test robot <oliver.sang@intel.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Tested-by: Sriram Yagnaraman <sriram.yagnaraman@est.tech>
Link: https://lore.kernel.org/r/20231010132113.3014691-2-idosch@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

syzbot reported a warning [0] introduced by commit c48ef9c4 ("tcp: Fix
bind() regression for v4-mapped-v6 non-wildcard address.").

After the cited commit, a v4 socket's address matches the corresponding
v4-mapped-v6 tb2 in inet_bind2_bucket_match_addr(), not vice versa.

During X.X.X.X -> ::ffff:X.X.X.X order bind()s, the second bind() uses
bhash and conflicts properly without checking bhash2 so that we need not
check if a v4-mapped-v6 sk matches the corresponding v4 address tb2 in
inet_bind2_bucket_match_addr().  However, the repro shows that we need
to check that in a no-conflict case.

The repro bind()s two sockets to the 2-tuples using SO_REUSEPORT and calls
listen() for the first socket:

  from socket import *

  s1 = socket()
  s1.setsockopt(SOL_SOCKET, SO_REUSEPORT, 1)
  s1.bind(('127.0.0.1', 0))

  s2 = socket(AF_INET6)
  s2.setsockopt(SOL_SOCKET, SO_REUSEPORT, 1)
  s2.bind(('::ffff:127.0.0.1', s1.getsockname()[1]))

  s1.listen()

The second socket should belong to the first socket's tb2, but the second
bind() creates another tb2 bucket because inet_bind2_bucket_find() returns
NULL in inet_csk_get_port() as the v4-mapped-v6 sk does not match the
corresponding v4 address tb2.

  bhash2[] -> tb2(::ffff:X.X.X.X) -> tb2(X.X.X.X)

Then, listen() for the first socket calls inet_csk_get_port(), where the
v4 address matches the v4-mapped-v6 tb2 and WARN_ON() is triggered.

To avoid that, we need to check if v4-mapped-v6 sk address matches with
the corresponding v4 address tb2 in inet_bind2_bucket_match().

The same checks are needed in inet_bind2_bucket_addr_match() too, so we
can move all checks there and call it from inet_bind2_bucket_match().

Note that now tb->family is just an address family of tb->(v6_)?rcv_saddr
and not of sockets in the bucket.  This could be refactored later by
defining tb->rcv_saddr as tb->v6_rcv_saddr.s6_addr32[3] and prepending
::ffff: when creating v4 tb2.

[0]:
WARNING: CPU: 0 PID: 5049 at net/ipv4/inet_connection_sock.c:587 inet_csk_get_port+0xf96/0x2350 net/ipv4/inet_connection_sock.c:587
Modules linked in:
CPU: 0 PID: 5049 Comm: syz-executor288 Not tainted 6.6.0-rc2-syzkaller-00018-g2cf0f7156238 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 08/04/2023
RIP: 0010:inet_csk_get_port+0xf96/0x2350 net/ipv4/inet_connection_sock.c:587
Code: 7c 24 08 e8 4c b6 8a 01 31 d2 be 88 01 00 00 48 c7 c7 e0 94 ae 8b e8 59 2e a3 f8 2e 2e 2e 31 c0 e9 04 fe ff ff e8 ca 88 d0 f8 <0f> 0b e9 0f f9 ff ff e8 be 88 d0 f8 49 8d 7e 48 e8 65 ca 5a 00 31
RSP: 0018:ffffc90003abfbf0 EFLAGS: 00010293
RAX: 0000000000000000 RBX: ffff888026429100 RCX: 0000000000000000
RDX: ffff88807edcbb80 RSI: ffffffff88b73d66 RDI: ffff888026c49f38
RBP: ffff888026c49f30 R08: 0000000000000005 R09: 0000000000000000
R10: 0000000000000001 R11: 0000000000000000 R12: ffffffff9260f200
R13: ffff888026c49880 R14: 0000000000000000 R15: ffff888026429100
FS:  00005555557d5380(0000) GS:ffff8880b9800000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 000000000045ad50 CR3: 0000000025754000 CR4: 00000000003506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 <TASK>
 inet_csk_listen_start+0x155/0x360 net/ipv4/inet_connection_sock.c:1256
 __inet_listen_sk+0x1b8/0x5c0 net/ipv4/af_inet.c:217
 inet_listen+0x93/0xd0 net/ipv4/af_inet.c:239
 __sys_listen+0x194/0x270 net/socket.c:1866
 __do_sys_listen net/socket.c:1875 [inline]
 __se_sys_listen net/socket.c:1873 [inline]
 __x64_sys_listen+0x53/0x80 net/socket.c:1873
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x38/0xb0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x63/0xcd
RIP: 0033:0x7f3a5bce3af9
Code: 28 00 00 00 75 05 48 83 c4 28 c3 e8 c1 17 00 00 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b8 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007ffc1a1c79e8 EFLAGS: 00000246 ORIG_RAX: 0000000000000032
RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f3a5bce3af9
RDX: 00007f3a5bce3af9 RSI: 0000000000000000 RDI: 0000000000000003
RBP: 00007f3a5bd565f0 R08: 0000000000000006 R09: 0000000000000006
R10: 0000000000000006 R11: 0000000000000246 R12: 0000000000000001
R13: 431bde82d7b634db R14: 0000000000000001 R15: 0000000000000001
 </TASK>

Fixes: c48ef9c4

 ("tcp: Fix bind() regression for v4-mapped-v6 non-wildcard address.")
Reported-by:  <syzbot+71e724675ba3958edb31@syzkaller.appspotmail.com>
Closes: https://syzkaller.appspot.com/bug?extid=71e724675ba3958edb31
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20231010013814.70571-1-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Since 429e3d12 ("bonding: Fix extraction of ports from the packet
headers"), header offsets used to compute a hash in bond_xmit_hash() are
relative to skb->data and not skb->head. If the tail of the header buffer
of an skb really needs to be advanced and the operation is successful, the
pointer to the data must be returned (and not a pointer to the head of the
buffer).

Fixes: 429e3d12

 ("bonding: Fix extraction of ports from the packet headers")
Signed-off-by: Jiri Wiesner <jwiesner@suse.de>
Acked-by: Jay Vosburgh <jay.vosburgh@canonical.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pull networking fixes from Paolo Abeni:
 "Including fixes from CAN and BPF.

  We have a regression in TC currently under investigation, otherwise
  the things that stand off most are probably the TCP and AF_PACKET
  fixes, with both issues coming from 6.5.

  Previous releases - regressions:

   - af_packet: fix fortified memcpy() without flex array.

   - tcp: fix crashes trying to free half-baked MTU probes

   - xdp: fix zero-size allocation warning in xskq_create()

   - can: sja1000: always restart the tx queue after an overrun

   - eth: mlx5e: again mutually exclude RX-FCS and RX-port-timestamp

   - eth: nfp: avoid rmmod nfp crash issues

   - eth: octeontx2-pf: fix page pool frag allocation warning

  Previous releases - always broken:

   - mctp: perform route lookups under a RCU read-side lock

   - bpf: s390: fix clobbering the caller's backchain in the trampoline

   - phy: lynx-28g: cancel the CDR check work item on the remove path

   - dsa: qca8k: fix qca8k driver for Turris 1.x

   - eth: ravb: fix use-after-free issue in ravb_tx_timeout_work()

   - eth: ixgbe: fix crash with empty VF macvlan list"

* tag 'net-6.6-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (54 commits)
  rswitch: Fix imbalance phy_power_off() calling
  rswitch: Fix renesas_eth_sw_remove() implementation
  octeontx2-pf: Fix page pool frag allocation warning
  nfc: nci: assert requested protocol is valid
  af_packet: Fix fortified memcpy() without flex array.
  net: tcp: fix crashes trying to free half-baked MTU probes
  net/smc: Fix pos miscalculation in statistics
  nfp: flower: avoid rmmod nfp crash issues
  net: usb: dm9601: fix uninitialized variable use in dm9601_mdio_read
  ethtool: Fix mod state of verbose no_mask bitset
  net: nfc: fix races in nfc_llcp_sock_get() and nfc_llcp_sock_get_sn()
  mctp: perform route lookups under a RCU read-side lock
  net: skbuff: fix kernel-doc typos
  s390/bpf: Fix unwinding past the trampoline
  s390/bpf: Fix clobbering the caller's backchain in the trampoline
  net/mlx5e: Again mutually exclude RX-FCS and RX-port-timestamp
  net/smc: Fix dependency of SMC on ISM
  ixgbe: fix crash with empty VF macvlan list
  net/mlx5e: macsec: use update_pn flag instead of PN comparation
  net: phy: mscc: macsec: reject PN update requests
  ...

Pull ARM SoC fixes from Arnd Bergmann:
 "AngeloGioacchino Del Regno is stepping in as co-maintainer for the
  MediaTek SoC platform and starts by sending some dts fixes for the
  mt8195 platform that had been pending for a while.

  On the ixp4xx platform, Krzysztof Halasa steps down as co-maintainer,
  reflecting that Linus Walleij has been handling this on his own for
  the past few years.

  Generic RISC-V kernels are now marked as incompatible with the RZ/Five
  platform that requires custom hacks both for managing its DMA bounce
  buffers and for addressing low virtual memory.

 Finally, there is one bugfix for the AMDTEE firmware driver to prevent
 a use-after-free bug"

* tag 'soc-fixes-6.6-2' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc:
  IXP4xx MAINTAINERS entries
  arm64: dts: mediatek: mt8195: Set DSU PMU status to fail
  arm64: dts: mediatek: fix t-phy unit name
  arm64: dts: mediatek: mt8195-demo: update and reorder reserved memory regions
  arm64: dts: mediatek: mt8195-demo: fix the memory size to 8GB
  MAINTAINERS: Add Angelo as MediaTek SoC co-maintainer
  soc: renesas: Make ARCH_R9A07G043 (riscv version) depend on NONPORTABLE
  tee: amdtee: fix use-after-free vulnerability in amdtee_close_session

Pull pmdomain fix from Ulf Hansson:

 - imx: scu-pd: Correct the DMA2 channel

* tag 'pmdomain-v6.6-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/linux-pm:
  pmdomain: imx: scu-pd: correct DMA2 channel

Although vf_vport entry of struct mlx5e_stats is never updated, its
values are mistakenly copied to the caller structure in the VF
representor .ndo_get_stat_64 callback mlx5e_rep_get_stats(). Remove
redundant entry and use the updated one, rep_stats, instead.

Fixes: 64b68e36

 ("net/mlx5: Refactor and expand rep vport stat group")
Reviewed-by: Patrisious Haddad <phaddad@nvidia.com>
Signed-off-by: Amir Tzin <amirtz@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>

In the cited commit, if the routing device is ovs internal port, the
out device is set to uplink, and packets go out after encapsulation.

If filter device is uplink, it can trigger the following syndrome:
mlx5_core 0000:08:00.0: mlx5_cmd_out_err:803:(pid 3966): SET_FLOW_TABLE_ENTRY(0x936) op_mod(0x0) failed, status bad parameter(0x3), syndrome (0xcdb051), err(-22)

Fix this issue by not offloading internal port if filter device is out
device. In this case, packets are not forwarded to the root table to
be processed, the termination table is used instead to forward them
from uplink to uplink.

Fixes: 100ad4e2

 ("net/mlx5e: Offload internal port as encap route device")
Signed-off-by: Jianbo Liu <jianbol@nvidia.com>
Reviewed-by: Ariel Levkovich <lariel@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>

Hold RTNL lock when calling xdp_set_features() with a registered netdev,
as the call triggers the netdev notifiers. This could happen when
switching from nic profile to uplink representor for example.

Similar logic which fixed a similar scenario was previously introduced in
the following commit:
commit 72cc6549 net/mlx5e: Take RTNL lock when needed before calling
xdp_set_features().

This fixes the following assertion and warning call trace:

RTNL: assertion failed at net/core/dev.c (1961)
WARNING: CPU: 13 PID: 2529 at net/core/dev.c:1961
call_netdevice_notifiers_info+0x7c/0x80
Modules linked in: rpcrdma rdma_ucm ib_iser libiscsi
scsi_transport_iscsi ib_umad rdma_cm ib_ipoib iw_cm ib_cm mlx5_ib
ib_uverbs ib_core xt_conntrack xt_MASQUERADE nf_conntrack_netlink
nfnetlink xt_addrtype iptable_nat nf_nat br_netfilter rpcsec_gss_krb5
auth_rpcgss oid_registry overlay mlx5_core zram zsmalloc fuse
CPU: 13 PID: 2529 Comm: devlink Not tainted
6.5.0_for_upstream_min_debug_2023_09_07_20_04 #1
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS
rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014
RIP: 0010:call_netdevice_notifiers_info+0x7c/0x80
Code: 8f ff 80 3d 77 0d 16 01 00 75 c5 ba a9 07 00 00 48
c7 c6 c4 bb 0d 82 48 c7 c7 18 c8 06 82 c6 05 5b 0d 16 01 01 e8 44 f6 8c
ff <0f> 0b eb a2 0f 1f 44 00 00 55 48 89 e5 41 54 48 83 e4 f0 48 83 ec
RSP: 0018:ffff88819930f7f0 EFLAGS: 00010282
RAX: 0000000000000000 RBX: ffffffff8309f740 RCX: 0000000000000027
RDX: ffff88885fb5b5c8 RSI: 0000000000000001 RDI: ffff88885fb5b5c0
RBP: 0000000000000028 R08: ffff88887ffabaa8 R09: 0000000000000003
R10: ffff88887fecbac0 R11: ffff88887ff7bac0 R12: ffff88819930f810
R13: ffff88810b7fea40 R14: ffff8881154e8fd8 R15: ffff888107e881a0
FS:  00007f3ad248f800(0000) GS:ffff88885fb40000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000563b85f164e0 CR3: 0000000113b5c006 CR4: 0000000000370ea0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 <TASK>
 ? __warn+0x79/0x120
 ? call_netdevice_notifiers_info+0x7c/0x80
 ? report_bug+0x17c/0x190
 ? handle_bug+0x3c/0x60
 ? exc_invalid_op+0x14/0x70
 ? asm_exc_invalid_op+0x16/0x20
 ? call_netdevice_notifiers_info+0x7c/0x80
 call_netdevice_notifiers+0x2e/0x50
 mlx5e_set_xdp_feature+0x21/0x50 [mlx5_core]
 mlx5e_build_rep_params+0x97/0x130 [mlx5_core]
 mlx5e_init_ul_rep+0x9f/0x100 [mlx5_core]
 mlx5e_netdev_init_profile+0x76/0x110 [mlx5_core]
 mlx5e_netdev_attach_profile+0x1f/0x90 [mlx5_core]
 mlx5e_netdev_change_profile+0x92/0x160 [mlx5_core]
 mlx5e_vport_rep_load+0x329/0x4a0 [mlx5_core]
 mlx5_esw_offloads_rep_load+0x9e/0xf0 [mlx5_core]
 esw_offloads_enable+0x4bc/0xe90 [mlx5_core]
 mlx5_eswitch_enable_locked+0x3c8/0x570 [mlx5_core]
 ? kmalloc_trace+0x25/0x80
 mlx5_devlink_eswitch_mode_set+0x224/0x680 [mlx5_core]
 ? devlink_get_from_attrs_lock+0x9e/0x110
 devlink_nl_cmd_eswitch_set_doit+0x60/0xe0
 genl_family_rcv_msg_doit+0xd0/0x120
 genl_rcv_msg+0x180/0x2b0
 ? devlink_get_from_attrs_lock+0x110/0x110
 ? devlink_nl_cmd_eswitch_get_doit+0x290/0x290
 ? devlink_pernet_pre_exit+0xf0/0xf0
 ? genl_family_rcv_msg_dumpit+0xf0/0xf0
 netlink_rcv_skb+0x54/0x100
 genl_rcv+0x24/0x40
 netlink_unicast+0x1fc/0x2c0
 netlink_sendmsg+0x232/0x4a0
 sock_sendmsg+0x38/0x60
 ? _copy_from_user+0x2a/0x60
 __sys_sendto+0x110/0x160
 ? handle_mm_fault+0x161/0x260
 ? do_user_addr_fault+0x276/0x620
 __x64_sys_sendto+0x20/0x30
 do_syscall_64+0x3d/0x90
 entry_SYSCALL_64_after_hwframe+0x46/0xb0
RIP: 0033:0x7f3ad231340a
Code: d8 64 89 02 48 c7 c0 ff ff ff ff eb b8 0f 1f 00 f3
0f 1e fa 41 89 ca 64 8b 04 25 18 00 00 00 85 c0 75 15 b8 2c 00 00 00 0f
05 <48> 3d 00 f0 ff ff 77 7e c3 0f 1f 44 00 00 41 54 48 83 ec 30 44 89
RSP: 002b:00007ffd70aad4b8 EFLAGS: 00000246 ORIG_RAX: 000000000000002c
RAX: ffffffffffffffda RBX: 0000000000c36b00 RCX:00007f3ad231340a
RDX: 0000000000000038 RSI: 0000000000c36b00 RDI: 0000000000000003
RBP: 0000000000c36910 R08: 00007f3ad2625200 R09: 000000000000000c
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000001
 </TASK>
---[ end trace 0000000000000000 ]---
------------[ cut here ]------------

Fixes: 4d5ab0ad

 ("net/mlx5e: take into account device reconfiguration for xdp_features flag")
Signed-off-by: Lama Kayal <lkayal@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>

When mlx5e_xdp_xmit is called without the XDP_XMIT_FLUSH set it is
possible that it leaves a mpwqe session open. That is ok during runtime:
the session will be closed on the next call to mlx5e_xdp_xmit. But
having a mpwqe session still open at XDP sq close time is problematic:
the pc counter is not updated before flushing the contents of the
xdpi_fifo. This results in leaking page fragments.

The fix is to always close the mpwqe session at the end of
mlx5e_xdp_xmit, regardless of the XDP_XMIT_FLUSH flag being set or not.

Fixes: 5e0d2eef

 ("net/mlx5e: XDP, Support Enhanced Multi-Packet TX WQE")
Signed-off-by: Dragos Tatulea <dtatulea@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>

When a page allocation fails during refill in mlx5e_refill_rx_wqes, the
page will be released again on the next refill call. This triggers the
page_pool negative page fragment count warning below:

 [  338.326070] WARNING: CPU: 4 PID: 0 at include/net/page_pool/helpers.h:130 mlx5e_page_release_fragmented.isra.0+0x42/0x50 [mlx5_core]
  ...
 [  338.328993] RIP: 0010:mlx5e_page_release_fragmented.isra.0+0x42/0x50 [mlx5_core]
 [  338.329094] Call Trace:
 [  338.329097]  <IRQ>
 [  338.329100]  ? __warn+0x7d/0x120
 [  338.329105]  ? mlx5e_page_release_fragmented.isra.0+0x42/0x50 [mlx5_core]
 [  338.329173]  ? report_bug+0x155/0x180
 [  338.329179]  ? handle_bug+0x3c/0x60
 [  338.329183]  ? exc_invalid_op+0x13/0x60
 [  338.329187]  ? asm_exc_invalid_op+0x16/0x20
 [  338.329192]  ? mlx5e_page_release_fragmented.isra.0+0x42/0x50 [mlx5_core]
 [  338.329259]  mlx5e_post_rx_wqes+0x210/0x5a0 [mlx5_core]
 [  338.329327]  ? mlx5e_poll_rx_cq+0x88/0x6f0 [mlx5_core]
 [  338.329394]  mlx5e_napi_poll+0x127/0x6b0 [mlx5_core]
 [  338.329461]  __napi_poll+0x25/0x1a0
 [  338.329465]  net_rx_action+0x28a/0x300
 [  338.329468]  __do_softirq+0xcd/0x279
 [  338.329473]  irq_exit_rcu+0x6a/0x90
 [  338.329477]  common_interrupt+0x82/0xa0
 [  338.329482]  </IRQ>

This patch fixes the legacy rq case by releasing all allocated fragments
and then setting the skip flag on all released fragments. It is
important to note that the number of released fragments will be higher
than the number of allocated fragments when an allocation error occurs.

Fixes: 3f93f829

 ("net/mlx5e: RX, Defer page release in legacy rq for better recycling")
Tested-by: Chris Mason <clm@fb.com>
Reported-by: Chris Mason <clm@fb.com>
Closes: https://lore.kernel.org/netdev/117FF31A-7BE0-4050-B2BB-E41F224FF72F@meta.com
Signed-off-by: Dragos Tatulea <dtatulea@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>

When a page allocation fails during refill in mlx5e_post_rx_mpwqes, the
page will be released again on the next refill call. This triggers the
page_pool negative page fragment count warning below:

 [ 2436.447717] WARNING: CPU: 1 PID: 2419 at include/net/page_pool/helpers.h:130 mlx5e_page_release_fragmented.isra.0+0x42/0x50 [mlx5_core]
 ...
 [ 2436.447895] RIP: 0010:mlx5e_page_release_fragmented.isra.0+0x42/0x50 [mlx5_core]
 [ 2436.447991] Call Trace:
 [ 2436.447975]  mlx5e_post_rx_mpwqes+0x1d5/0xcf0 [mlx5_core]
 [ 2436.447994]  <IRQ>
 [ 2436.447996]  ? __warn+0x7d/0x120
 [ 2436.448009]  ? mlx5e_handle_rx_cqe_mpwrq+0x109/0x1d0 [mlx5_core]
 [ 2436.448002]  ? mlx5e_page_release_fragmented.isra.0+0x42/0x50 [mlx5_core]
 [ 2436.448044]  ? mlx5e_poll_rx_cq+0x87/0x6e0 [mlx5_core]
 [ 2436.448061]  ? report_bug+0x155/0x180
 [ 2436.448065]  ? handle_bug+0x36/0x70
 [ 2436.448067]  ? exc_invalid_op+0x13/0x60
 [ 2436.448070]  ? asm_exc_invalid_op+0x16/0x20
 [ 2436.448079]  mlx5e_napi_poll+0x122/0x6b0 [mlx5_core]
 [ 2436.448077]  ? mlx5e_page_release_fragmented.isra.0+0x42/0x50 [mlx5_core]
 [ 2436.448113]  ? generic_exec_single+0x35/0x100
 [ 2436.448117]  __napi_poll+0x25/0x1a0
 [ 2436.448120]  net_rx_action+0x28a/0x300
 [ 2436.448122]  __do_softirq+0xcd/0x279
 [ 2436.448126]  irq_exit_rcu+0x6a/0x90
 [ 2436.448128]  sysvec_apic_timer_interrupt+0x6e/0x90
 [ 2436.448130]  </IRQ>

This patch fixes the striding rq case by setting the skip flag on all
the wqe pages that were expected to have new pages allocated.

Fixes: 4c2a1323

 ("net/mlx5e: RX, Defer page release in striding rq for better recycling")
Tested-by: Chris Mason <clm@fb.com>
Reported-by: Chris Mason <clm@fb.com>
Closes: https://lore.kernel.org/netdev/117FF31A-7BE0-4050-B2BB-E41F224FF72F@meta.com
Signed-off-by: Dragos Tatulea <dtatulea@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>

Currently, whenever fw issues a change ownership event, the PF that owns
the fw tracer drops its ownership directly and the other PFs try to pick
up the ownership via what MTRC register suggests.

In some cases, driver releases the ownership of the tracer and reacquires
it later on. Whenever the driver releases ownership of the tracer, fw
issues a change ownership event. This event can be delayed and come after
driver has reacquired ownership of the tracer. Thus the late event will
trigger the tracer owner PF to release the ownership again and lead to a
scenario where no PF is owning the tracer.

To prevent the scenario described above, when handling a change
ownership event, do not drop ownership of the tracer directly, instead
read the fw MTRC register to retrieve the up-to-date owner of the tracer
and set it accordingly in driver level.

Fixes: f53aaa31

 ("net/mlx5: FW tracer, implement tracer logic")
Signed-off-by: Maher Sanalla <msanalla@nvidia.com>
Reviewed-by: Shay Drory <shayd@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>

With current implementation in single FDB LAG mode all packets are
processed by eswitch 0 rules. As such, 'peer' FDB entries receive the
packets for rules of other eswitches and are responsible for updating the
main entry by sending SWITCHDEV_FDB_ADD_TO_BRIDGE notification from their
background update wq task. However, this introduces a race condition when
non-zero eswitch instance decides to delete a FDB entry, sends
SWITCHDEV_FDB_DEL_TO_BRIDGE notification, but another eswitch's update task
refreshes the same entry concurrently while its async delete work is still
pending on the workque. In such case another SWITCHDEV_FDB_ADD_TO_BRIDGE
event may be generated and entry will remain stuck in FDB marked as
'offloaded' since no more SWITCHDEV_FDB_DEL_TO_BRIDGE notifications are
sent for deleting the peer entries.

Fix the issue by synchronously marking deleted entries with
MLX5_ESW_BRIDGE_FLAG_DELETED flag and skipping them in background update
job.

Signed-off-by: Vlad Buslov <vladbu@nvidia.com>
Reviewed-by: Jianbo Liu <jianbol@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>

Currently, mlx5 is registering event handler for vport context change
event some time after arming the event. this can lead to missing an
event, which will result in wrong rules in the FDB.
Hence, register the event handler before arming the event.

This solution is valid since FW is sending vport context change event
only on vports which SW armed, and SW arming the vport when enabling
it, which is done after the FDB has been created.

Fixes: 6933a937

 ("net/mlx5: E-Switch, Use async events chain")
Signed-off-by: Shay Drory <shayd@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>