This effectively reverts 4b5f82f6. On a number of systems ASPM L1
causes tx timeouts with RTL8168h, see referenced bug report.

Fixes: 4b5f82f6 ("r8169: enable ASPM L1/L1.1 from RTL8168h")
Cc: stable@vger.kernel.org
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=217814Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

'tp->irq_max' value is either 1 [L16336] or 5 [L16354], as indicated in
tg3_get_invariants(). Therefore, 'i' can't exceed 4 in tg3_init_one()
that makes (i <= 4) always true. Moreover, 'intmbx' value set at the
last iteration is not used later in it's scope.

Found by Linux Verification Center (linuxtesting.org) with SVACE.

Fixes: 78f90dcf ("tg3: Move napi_add calls below tg3_get_invariants")
Signed-off-by: Mikhail Kobuk <m.kobuk@ispras.ru>
Reviewed-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

There was a previous attempt to fix an out-of-bounds access in the DCCP
error handlers, but that fix assumed that the error handlers only want
to access the first 8 bytes of the DCCP header. Actually, they also look
at the DCCP sequence number, which is stored beyond 8 bytes, so an
explicit pskb_may_pull() is required.

Fixes: 6706a97f ("dccp: fix out of bound access in dccp_v4_err()")
Fixes: 1aa9d1a0 ("ipv6: dccp: fix out of bound access in dccp_v6_err()")
Cc: stable@vger.kernel.org
Signed-off-by: Jann Horn <jannh@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fix wrong switch name in compatibility list. 88E6163 switch does not exist
and is in fact 88E6361

Fixes: 9229a948 ("dt-bindings: net: dsa: marvell: add MV88E6361 switch to compatibility list")
Signed-off-by: Alexis Lothoré <alexis.lothore@bootlin.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Acked-by: Conor Dooley <conor.dooley@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Increase the RX buffer size to 3K when the SBP bit is on. The size of
the RX buffer determines the number of pages allocated which may not
be sufficient for receive frames larger than the set MTU size.

Cc: stable@vger.kernel.org
Fixes: 89eaefb6 ("igb: Support RX-ALL feature flag.")
Reported-by: Manfred Rudigier <manfred.rudigier@omicronenergy.com>
Signed-off-by: Radoslaw Tyl <radoslawx.tyl@intel.com>
Tested-by: Arpana Arland <arpanax.arland@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

syzkaller reported null-ptr-deref [0] related to AF_NETROM.
This is another self-accept issue from the strace log. [1]

syz-executor creates an AF_NETROM socket and calls connect(), which
is blocked at that time.  Then, sk->sk_state is TCP_SYN_SENT and
sock->state is SS_CONNECTING.

  [pid  5059] socket(AF_NETROM, SOCK_SEQPACKET, 0) = 4
  [pid  5059] connect(4, {sa_family=AF_NETROM, sa_data="..." <unfinished ...>

Another thread calls connect() concurrently, which finally fails
with -EINVAL.  However, the problem here is the socket state is
reset even while the first connect() is blocked.

  [pid  5060] connect(4, NULL, 0 <unfinished ...>
  [pid  5060] <... connect resumed>)      = -1 EINVAL (Invalid argument)

As sk->state is TCP_CLOSE and sock->state is SS_UNCONNECTED, the
following listen() succeeds.  Then, the first connect() looks up
itself as a listener and puts skb into the queue with skb->sk itself.
As a result, the next accept() gets another FD of itself as 3, and
the first connect() finishes.

  [pid  5060] listen(4, 0 <unfinished ...>
  [pid  5060] <... listen resumed>)       = 0
  [pid  5060] accept(4, NULL, NULL <unfinished ...>
  [pid  5060] <... accept resumed>)       = 3
  [pid  5059] <... connect resumed>)      = 0

Then, accept4() is called but blocked, which causes the general protection
fault later.

  [pid  5059] accept4(4, NULL, 0x20000400, SOCK_NONBLOCK <unfinished ...>

After that, another self-accept occurs by accept() and writev().

  [pid  5060] accept(4, NULL, NULL <unfinished ...>
  [pid  5061] writev(3, [{iov_base=...}] <unfinished ...>
  [pid  5061] <... writev resumed>)       = 99
  [pid  5060] <... accept resumed>)       = 6

Finally, the leader thread close()s all FDs.  Since the three FDs
reference the same socket, nr_release() does the cleanup for it
three times, and the remaining accept4() causes the following fault.

  [pid  5058] close(3)                    = 0
  [pid  5058] close(4)                    = 0
  [pid  5058] close(5)                    = -1 EBADF (Bad file descriptor)
  [pid  5058] close(6)                    = 0
  [pid  5058] <... exit_group resumed>)   = ?
  [   83.456055][ T5059] general protection fault, probably for non-canonical address 0xdffffc0000000003: 0000 [#1] PREEMPT SMP KASAN

To avoid the issue, we need to return an error for connect() if
another connect() is in progress, as done in __inet_stream_connect().

[0]:
general protection fault, probably for non-canonical address 0xdffffc0000000003: 0000 [#1] PREEMPT SMP KASAN
KASAN: null-ptr-deref in range [0x0000000000000018-0x000000000000001f]
CPU: 0 PID: 5059 Comm: syz-executor.0 Not tainted 6.5.0-rc5-syzkaller-00194-gace0ab3a #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 07/26/2023
RIP: 0010:__lock_acquire+0x109/0x5de0 kernel/locking/lockdep.c:5012
Code: 45 85 c9 0f 84 cc 0e 00 00 44 8b 05 11 6e 23 0b 45 85 c0 0f 84 be 0d 00 00 48 ba 00 00 00 00 00 fc ff df 4c 89 d1 48 c1 e9 03 <80> 3c 11 00 0f 85 e8 40 00 00 49 81 3a a0 69 48 90 0f 84 96 0d 00
RSP: 0018:ffffc90003d6f9e0 EFLAGS: 00010006
RAX: ffff8880244c8000 RBX: 1ffff920007adf6c RCX: 0000000000000003
RDX: dffffc0000000000 RSI: 0000000000000000 RDI: 0000000000000018
RBP: 0000000000000001 R08: 0000000000000001 R09: 0000000000000001
R10: 0000000000000018 R11: 0000000000000000 R12: 0000000000000000
R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000
FS:  00007f51d519a6c0(0000) GS:ffff8880b9800000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f51d5158d58 CR3: 000000002943f000 CR4: 00000000003506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 <TASK>
 lock_acquire kernel/locking/lockdep.c:5761 [inline]
 lock_acquire+0x1ae/0x510 kernel/locking/lockdep.c:5726
 __raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:110 [inline]
 _raw_spin_lock_irqsave+0x3a/0x50 kernel/locking/spinlock.c:162
 prepare_to_wait+0x47/0x380 kernel/sched/wait.c:269
 nr_accept+0x20d/0x650 net/netrom/af_netrom.c:798
 do_accept+0x3a6/0x570 net/socket.c:1872
 __sys_accept4_file net/socket.c:1913 [inline]
 __sys_accept4+0x99/0x120 net/socket.c:1943
 __do_sys_accept4 net/socket.c:1954 [inline]
 __se_sys_accept4 net/socket.c:1951 [inline]
 __x64_sys_accept4+0x96/0x100 net/socket.c:1951
 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: 0033:0x7f51d447cae9
Code: 28 00 00 00 75 05 48 83 c4 28 c3 e8 e1 20 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 b0 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f51d519a0c8 EFLAGS: 00000246 ORIG_RAX: 0000000000000120
RAX: ffffffffffffffda RBX: 00007f51d459bf80 RCX: 00007f51d447cae9
RDX: 0000000020000400 RSI: 0000000000000000 RDI: 0000000000000004
RBP: 00007f51d44c847a R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000800 R11: 0000000000000246 R12: 0000000000000000
R13: 000000000000000b R14: 00007f51d459bf80 R15: 00007ffc25c34e48
 </TASK>

Link: https://syzkaller.appspot.com/text?tag=CrashLog&x=152cdb63a80000 [1]
Fixes: 1da177e4 ("Linux-2.6.12-rc2")
Reported-by: syzbot+666c97e4686410e79649@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=666c97e4686410e79649Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Shannon Nelson says:

====================
pds_core: error handling fixes

Some fixes for better handling of broken states.
====================

Link: https://lore.kernel.org/r/20230824161754.34264-1-shannon.nelson@amd.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Don't rely on the PCI memory for the devcmd opcode because we
read a 0xff value if the PCI bus is broken, which can cause us
to report a bogus dev_cmd opcode later.

Fixes: 523847df ("pds_core: add devcmd device interfaces")
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://lore.kernel.org/r/20230824161754.34264-6-shannon.nelson@amd.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Add a check that the wq exists before queuing up work for a
failed devcmd, as the PF is responsible for health and the VF
doesn't have a wq.

Fixes: c2dbb090 ("pds_core: health timer and workqueue")
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://lore.kernel.org/r/20230824161754.34264-5-shannon.nelson@amd.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

The VF doesn't need to send a reset command, and in a PCI reset
scenario it might not have a valid IO space to write to anyway.

Fixes: 523847df ("pds_core: add devcmd device interfaces")
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://lore.kernel.org/r/20230824161754.34264-4-shannon.nelson@amd.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Make sure the health reporter is set up before we use it in
our devlink health updates, especially since the VF doesn't
set up the health reporter.

Fixes: 25b450c0 ("pds_core: add devlink health facilities")
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://lore.kernel.org/r/20230824161754.34264-3-shannon.nelson@amd.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Don't access structs that have been cleared when in the fw_down
state and the various structs have been cleaned and are waiting
to recover.  This caused a panic on rmmod when already in fw_down
and devlink_param_unregister() tried to check the parameters.

Fixes: 40ced894 ("pds_core: devlink params for enabling VIF support")
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://lore.kernel.org/r/20230824161754.34264-2-shannon.nelson@amd.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

HFSC assumes that inner classes have an fsc curve, but it is currently
possible for classes without an fsc curve to become parents. This leads
to bugs including a use-after-free.

Don't allow non-root classes without HFSC_FSC to become parents.

Fixes: 1da177e4 ("Linux-2.6.12-rc2")
Reported-by: Budimir Markovic <markovicbudimir@gmail.com>
Signed-off-by: Budimir Markovic <markovicbudimir@gmail.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Link: https://lore.kernel.org/r/20230824084905.422-1-markovicbudimir@gmail.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Not all firmware variants support RSS filters. Do not fail all PTP
functionality when raw ethernet PTP filters fail to insert.

Fixes: e4616f64 ("sfc: support PTP over Ethernet")
Signed-off-by: Alex Austin <alex.austin@amd.com>
Acked-by: Edward Cree <ecree.xilinx@gmail.com>
Reviewed-by: Pieter Jansen van Vuuren <pieter.jansen-van-vuuren@amd.com>
Link: https://lore.kernel.org/r/20230824164657.42379-1-alex.austin@amd.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Suman Ghosh says:

====================
Fix PFC related issues

This patchset fixes multiple PFC related issues related to Octeon.

Patch #1: octeontx2-pf: Fix PFC TX scheduler free

Patch #2: octeontx2-af: CN10KB: fix PFC configuration

Patch #3: octeonxt2-pf: Fix backpressure config for multiple PFC
priorities to work simultaneously
====================

Link: https://lore.kernel.org/r/20230824081032.436432-1-sumang@marvell.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

MAC (CGX or RPM) asserts backpressure at TL3 or TL2 node of the egress
hierarchical scheduler tree depending on link level config done. If
there are multiple PFC priorities enabled at a time and for all such
flows to backoff, each priority will have to assert backpressure at
different TL3/TL2 scheduler nodes and these flows will need to submit
egress pkts to these nodes.

Current PFC configuration has an issue where in only one backpressure
scheduler node is being allocated which is resulting in only one PFC
priority to work. This patch fixes this issue.

Fixes: 99c969a8 ("octeontx2-pf: Add egress PFC support")
Signed-off-by: Suman Ghosh <sumang@marvell.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://lore.kernel.org/r/20230824081032.436432-4-sumang@marvell.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Suppose user has enabled pfc with prio 0,1 on a PF netdev(eth0)
	dcb pfc set dev eth0 prio-pfc o:on 1:on
later user enabled pfc priorities 2 and 3 on the VF interface(eth1)
	dcb pfc set dev eth1 prio-pfc 2:on 3:on

Instead of enabling pfc on all priorities (0..3), the driver only
enables on priorities 2,3. This patch corrects the issue by using
the proper CSR address.

Fixes: b9d0fedc ("octeontx2-af: cn10kb: Add RPM_USX MAC support")
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://lore.kernel.org/r/20230824081032.436432-3-sumang@marvell.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

During PFC TX schedulers free, flag TXSCHQ_FREE_ALL was being set
which caused free up all schedulers other than the PFC schedulers.
This patch fixes that to free only the PFC Tx schedulers.

Fixes: 99c969a8 ("octeontx2-pf: Add egress PFC support")
Signed-off-by: Suman Ghosh <sumang@marvell.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://lore.kernel.org/r/20230824081032.436432-2-sumang@marvell.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Petr Machata says:

====================
mlxsw: Assorted fixes

This patchset contains several fixes for the mlxsw driver.

Patch #1 - Fixes buffer size in I2C mailbox buffer.
Patch #2 - Sets limitation of chunk size in I2C transaction.
Patch #3 - Fixes module label names based on MTCAP sensor counter
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Transceiver module temperature sensors are indexed after ASIC and
platform sensors. The current label printing method does not take this
into account and simply prints the index of the transceiver module
sensor.

On new systems that have platform sensors this results in incorrect
(shifted) transceiver module labels being printed:

$ sensors
[...]
front panel 002:  +37.0°C  (crit = +70.0°C, emerg = +75.0°C)
front panel 003:  +47.0°C  (crit = +70.0°C, emerg = +75.0°C)
[...]

Fix by taking the sensor count into account. After the fix:

$ sensors
[...]
front panel 001:  +37.0°C  (crit = +70.0°C, emerg = +75.0°C)
front panel 002:  +47.0°C  (crit = +70.0°C, emerg = +75.0°C)
[...]

Fixes: a53779de ("mlxsw: core: Add QSFP module temperature label attribute to hwmon")
Signed-off-by: Vadim Pasternak <vadimp@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Maximum size of buffer is obtained from underlying I2C adapter and in
case adapter allows I2C transaction buffer size greater than 100 bytes,
transaction will fail due to firmware limitation.

As a result driver will fail initialization.

Limit the maximum size of transaction buffer by 100 bytes to fit to
firmware.

Remove unnecessary calculation:
max_t(u16, MLXSW_I2C_BLK_DEF, quirk_size).
This condition can not happened.

Fixes: 3029a693 ("mlxsw: i2c: Allow flexible setting of I2C transactions size")
Signed-off-by: Vadim Pasternak <vadimp@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The driver reads commands output from the output mailbox. If the size
of the output mailbox is not a multiple of the transaction /
block size, then the driver will not issue enough read transactions
to read the entire output, which can result in driver initialization
errors.

Fix by determining the number of transactions using DIV_ROUND_UP().

Fixes: 3029a693 ("mlxsw: i2c: Allow flexible setting of I2C transactions size")
Signed-off-by: Vadim Pasternak <vadimp@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

It is not allowed to call kfree_skb() from hardware interrupt
context or with hardware interrupts being disabled.
So replace kfree_skb() with dev_kfree_skb_irq() under
local_irq_disable(). Compile tested only.

Fixes: 05fcd31c ("arcnet: add err_skb package for package status feedback")
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

octeontx2 driver calls page_pool_create() during driver probe()
and fails if queue size > 32k. Page pool infra uses these buffers
as shock absorbers for burst traffic. These pages are pinned down
over time as working sets varies, due to the recycling nature
of page pool, given page pool (currently) don't have a shrinker
mechanism, the pages remain pinned down in ptr_ring.
Instead of clamping page_pool size to 32k at
most, limit it even more to 2k to avoid wasting memory.

This have been tested on octeontx2 CN10KA hardware.
TCP and UDP tests using iperf shows no performance regressions.

Fixes: b2e3406a ("octeontx2-pf: Add support for page pool")
Suggested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Reviewed-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: Ratheesh Kannoth <rkannoth@marvell.com>
Acked-by: Jesper Dangaard Brouer <hawk@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

The ice hardware has a synchronization mechanism used to drive the
simultaneous application of commands on both PHY ports and the source timer
in the MAC.

When issuing a sync via ice_ptp_exec_tmr_cmd(), the hardware will
simultaneously apply the commands programmed for the main timer and each
PHY port. Neither the main timer command register, nor the PHY port command
registers auto clear on command execution.

During the execution of a timer command intended for a single port on E822
devices, such as those used to configure a PHY during link up, the driver
is not correctly clearing the previous commands.

This results in unintentionally executing the last programmed command on
the main timer and other PHY ports whenever performing reconfiguration on
E822 ports after link up. This results in unintended side effects on other
timers, depending on what command was previously programmed.

To fix this, the driver must ensure that the main timer and all other PHY
ports are properly initialized to perform no action.

The enumeration for timer commands does not include an enumeration value
for doing nothing. Introduce ICE_PTP_NOP for this purpose. When writing a
timer command to hardware, leave the command bits set to zero which
indicates that no operation should be performed on that port.

Modify ice_ptp_one_port_cmd() to always initialize all ports. For all ports
other than the one being configured, write their timer command register to
ICE_PTP_NOP. This ensures that no side effect happens on the timer command.

To fix this for the PHY ports, modify ice_ptp_one_port_cmd() to always
initialize all other ports to ICE_PTP_NOP. This ensures that no side
effects happen on the other ports.

Call ice_ptp_src_cmd() with a command value if ICE_PTP_NOP in
ice_sync_phy_timer_e822() and ice_start_phy_timer_e822().

With both of these changes, the driver should no longer execute a stale
command on the main timer or another PHY port when reconfiguring one of the
PHY ports after link up.

Fixes: 3a749623 ("ice: implement basic E822 PTP support")
Signed-off-by: Siddaraju DH <siddaraju.dh@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Sunitha Mekala <sunithax.d.mekala@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

The goal is to support a bpf_redirect() from an ethernet device (ingress)
to a ppp device (egress).
The l2 header is added automatically by the ppp driver, thus the ethernet
header should be removed.

CC: stable@vger.kernel.org
Fixes: 27b29f63 ("bpf: add bpf_redirect() helper")
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Tested-by: Siwar Zitouni <siwar.zitouni@6wind.com>
Reviewed-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use a dynamic calculation to determine the shift value for the internal
timer cyclecounter that will lead to the highest precision frequency
adjustments. Previously used a constant for the shift value assuming all
devices supported by the driver had a nominal frequency of 1GHz. However,
there are devices that operate at different frequencies. The previous shift
value constant would break the PHC functionality for those devices.
Reported-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Closes: https://lore.kernel.org/netdev/20230815151507.3028503-1-vadfed@meta.com/
Fixes: 6a401092 ("net/mlx5: Update cyclecounter shift value to improve ptp free running mode precision")
Signed-off-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Tested-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Acked-by: Saeed Mahameed <saeedm@nvidia.com>
Link: https://lore.kernel.org/r/20230821230554.236210-1-rrameshbabu@nvidia.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Pull networking fixes from Paolo Abeni:
 "Including fixes from wifi, can and netfilter.

  Fixes to fixes:

   - nf_tables:
       - GC transaction race with abort path
       - defer gc run if previous batch is still pending

  Previous releases - regressions:

   - ipv4: fix data-races around inet->inet_id

   - phy: fix deadlocking in phy_error() invocation

   - mdio: fix C45 read/write protocol

   - ipvlan: fix a reference count leak warning in ipvlan_ns_exit()

   - ice: fix NULL pointer deref during VF reset

   - i40e: fix potential NULL pointer dereferencing of pf->vf in
     i40e_sync_vsi_filters()

   - tg3: use slab_build_skb() when needed

   - mtk_eth_soc: fix NULL pointer on hw reset

  Previous releases - always broken:

   - core: validate veth and vxcan peer ifindexes

   - sched: fix a qdisc modification with ambiguous command request

   - devlink: add missing unregister linecard notification

   - wifi: mac80211: limit reorder_buf_filtered to avoid UBSAN warning

   - batman:
      - do not get eth header before batadv_check_management_packet
      - fix batadv_v_ogm_aggr_send memory leak

   - bonding: fix macvlan over alb bond support

   - mlxsw: set time stamp fields also when its type is MIRROR_UTC"

* tag 'net-6.5-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (54 commits)
  selftests: bonding: add macvlan over bond testing
  selftest: bond: add new topo bond_topo_2d1c.sh
  bonding: fix macvlan over alb bond support
  rtnetlink: Reject negative ifindexes in RTM_NEWLINK
  netfilter: nf_tables: defer gc run if previous batch is still pending
  netfilter: nf_tables: fix out of memory error handling
  netfilter: nf_tables: use correct lock to protect gc_list
  netfilter: nf_tables: GC transaction race with abort path
  netfilter: nf_tables: flush pending destroy work before netlink notifier
  netfilter: nf_tables: validate all pending tables
  ibmveth: Use dcbf rather than dcbfl
  i40e: fix potential NULL pointer dereferencing of pf->vf i40e_sync_vsi_filters()
  net/sched: fix a qdisc modification with ambiguous command request
  igc: Fix the typo in the PTM Control macro
  batman-adv: Hold rtnl lock during MTU update via netlink
  igb: Avoid starting unnecessary workqueues
  can: raw: add missing refcount for memory leak fix
  can: isotp: fix support for transmission of SF without flow control
  bnx2x: new flag for track HW resource allocation
  sfc: allocate a big enough SKB for loopback selftest packet
  ...

Florian Westphal says:

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

This PR contains nf_tables updates for your *net* tree.

First patch fixes table validation, I broke this in 6.4 when tracking
validation state per table, reported by Pablo, fixup from myself.

Second patch makes sure objects waiting for memory release have been
released, this was broken in 6.1, patch from Pablo Neira Ayuso.

Patch three is a fix-for-fix from previous PR: In case a transaction
gets aborted, gc sequence counter needs to be incremented so pending
gc requests are invalidated, from Pablo.

Same for patch 4: gc list needs to use gc list lock, not destroy lock,
also from Pablo.

Patch 5 fixes a UaF in a set backend, but this should only occur when
failslab is enabled for GFP_KERNEL allocations, broken since feature
was added in 5.6, from myself.

Patch 6 fixes a double-free bug that was also added via previous PR:
We must not schedule gc work if the previous batch is still queued.

netfilter pull request 2023-08-23

* tag 'nf-23-08-23' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:
  netfilter: nf_tables: defer gc run if previous batch is still pending
  netfilter: nf_tables: fix out of memory error handling
  netfilter: nf_tables: use correct lock to protect gc_list
  netfilter: nf_tables: GC transaction race with abort path
  netfilter: nf_tables: flush pending destroy work before netlink notifier
  netfilter: nf_tables: validate all pending tables
====================

Link: https://lore.kernel.org/r/20230823152711.15279-1-fw@strlen.deSigned-off-by: Paolo Abeni <pabeni@redhat.com>

Hangbin Liu says:

====================
fix macvlan over alb bond support

Currently, the macvlan over alb bond is broken after commit
14af9963 ("bonding: Support macvlans on top of tlb/rlb mode bonds").
Fix this and add relate tests.
====================

Link: https://lore.kernel.org/r/20230823071907.3027782-1-liuhangbin@gmail.comSigned-off-by: Paolo Abeni <pabeni@redhat.com>

Add a macvlan over bonding test with mode active-backup, balance-tlb
and balance-alb.

]# ./bond_macvlan.sh
TEST: active-backup: IPv4: client->server                           [ OK ]
TEST: active-backup: IPv6: client->server                           [ OK ]
TEST: active-backup: IPv4: client->macvlan_1                        [ OK ]
TEST: active-backup: IPv6: client->macvlan_1                        [ OK ]
TEST: active-backup: IPv4: client->macvlan_2                        [ OK ]
TEST: active-backup: IPv6: client->macvlan_2                        [ OK ]
TEST: active-backup: IPv4: macvlan_1->macvlan_2                     [ OK ]
TEST: active-backup: IPv6: macvlan_1->macvlan_2                     [ OK ]
TEST: active-backup: IPv4: server->client                           [ OK ]
TEST: active-backup: IPv6: server->client                           [ OK ]
TEST: active-backup: IPv4: macvlan_1->client                        [ OK ]
TEST: active-backup: IPv6: macvlan_1->client                        [ OK ]
TEST: active-backup: IPv4: macvlan_2->client                        [ OK ]
TEST: active-backup: IPv6: macvlan_2->client                        [ OK ]
TEST: active-backup: IPv4: macvlan_2->macvlan_2                     [ OK ]
TEST: active-backup: IPv6: macvlan_2->macvlan_2                     [ OK ]
[...]
TEST: balance-alb: IPv4: client->server                             [ OK ]
TEST: balance-alb: IPv6: client->server                             [ OK ]
TEST: balance-alb: IPv4: client->macvlan_1                          [ OK ]
TEST: balance-alb: IPv6: client->macvlan_1                          [ OK ]
TEST: balance-alb: IPv4: client->macvlan_2                          [ OK ]
TEST: balance-alb: IPv6: client->macvlan_2                          [ OK ]
TEST: balance-alb: IPv4: macvlan_1->macvlan_2                       [ OK ]
TEST: balance-alb: IPv6: macvlan_1->macvlan_2                       [ OK ]
TEST: balance-alb: IPv4: server->client                             [ OK ]
TEST: balance-alb: IPv6: server->client                             [ OK ]
TEST: balance-alb: IPv4: macvlan_1->client                          [ OK ]
TEST: balance-alb: IPv6: macvlan_1->client                          [ OK ]
TEST: balance-alb: IPv4: macvlan_2->client                          [ OK ]
TEST: balance-alb: IPv6: macvlan_2->client                          [ OK ]
TEST: balance-alb: IPv4: macvlan_2->macvlan_2                       [ OK ]
TEST: balance-alb: IPv6: macvlan_2->macvlan_2                       [ OK ]
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Acked-by: Jay Vosburgh <jay.vosburgh@canonical.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

Add a new testing topo bond_topo_2d1c.sh which is used more commonly.
Make bond_topo_3d1c.sh just source bond_topo_2d1c.sh and add the
extra link.
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Acked-by: Jay Vosburgh <jay.vosburgh@canonical.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

The commit 14af9963 ("bonding: Support macvlans on top of tlb/rlb mode
bonds") aims to enable the use of macvlans on top of rlb bond mode. However,
the current rlb bond mode only handles ARP packets to update remote neighbor
entries. This causes an issue when a macvlan is on top of the bond, and
remote devices send packets to the macvlan using the bond's MAC address
as the destination. After delivering the packets to the macvlan, the macvlan
will rejects them as the MAC address is incorrect. Consequently, this commit
makes macvlan over bond non-functional.

To address this problem, one potential solution is to check for the presence
of a macvlan port on the bond device using netif_is_macvlan_port(bond->dev)
and return NULL in the rlb_arp_xmit() function. However, this approach
doesn't fully resolve the situation when a VLAN exists between the bond and
macvlan.

So let's just do a partial revert for commit 14af9963 in rlb_arp_xmit().
As the comment said, Don't modify or load balance ARPs that do not originate
locally.

Fixes: 14af9963 ("bonding: Support macvlans on top of tlb/rlb mode bonds")
Reported-by: susan.zheng@veritas.com
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2117816Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Acked-by: Jay Vosburgh <jay.vosburgh@canonical.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

Negative ifindexes are illegal, but the kernel does not validate the
ifindex in the ancillary header of RTM_NEWLINK messages, resulting in
the kernel generating a warning [1] when such an ifindex is specified.

Fix by rejecting negative ifindexes.

[1]
WARNING: CPU: 0 PID: 5031 at net/core/dev.c:9593 dev_index_reserve+0x1a2/0x1c0 net/core/dev.c:9593
[...]
Call Trace:
 <TASK>
 register_netdevice+0x69a/0x1490 net/core/dev.c:10081
 br_dev_newlink+0x27/0x110 net/bridge/br_netlink.c:1552
 rtnl_newlink_create net/core/rtnetlink.c:3471 [inline]
 __rtnl_newlink+0x115e/0x18c0 net/core/rtnetlink.c:3688
 rtnl_newlink+0x67/0xa0 net/core/rtnetlink.c:3701
 rtnetlink_rcv_msg+0x439/0xd30 net/core/rtnetlink.c:6427
 netlink_rcv_skb+0x16b/0x440 net/netlink/af_netlink.c:2545
 netlink_unicast_kernel net/netlink/af_netlink.c:1342 [inline]
 netlink_unicast+0x536/0x810 net/netlink/af_netlink.c:1368
 netlink_sendmsg+0x93c/0xe40 net/netlink/af_netlink.c:1910
 sock_sendmsg_nosec net/socket.c:728 [inline]
 sock_sendmsg+0xd9/0x180 net/socket.c:751
 ____sys_sendmsg+0x6ac/0x940 net/socket.c:2538
 ___sys_sendmsg+0x135/0x1d0 net/socket.c:2592
 __sys_sendmsg+0x117/0x1e0 net/socket.c:2621
 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

Fixes: 38f7b870 ("[RTNETLINK]: Link creation API")
Reported-by: syzbot+5ba06978f34abb058571@syzkaller.appspotmail.com
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20230823064348.2252280-1-idosch@nvidia.comSigned-off-by: Paolo Abeni <pabeni@redhat.com>

Pull ACPI fix from Rafael Wysocki:
 "Make an existing ACPI IRQ override quirk for PCSpecialist Elimina Pro
  16 M work as intended (Hans de Goede)"

* tag 'acpi-6.5-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  ACPI: resource: Fix IRQ override quirk for PCSpecialist Elimina Pro 16 M

Merge tag 'platform-drivers-x86-v6.5-5' of git://git.kernel.org/pub/scm/linux/kernel/git/pdx86/platform-drivers-x86

Pull x86 platform driver fixes from Hans de Goede:
 "Final set of three small fixes for 6.5"

* tag 'platform-drivers-x86-v6.5-5' of git://git.kernel.org/pub/scm/linux/kernel/git/pdx86/platform-drivers-x86:
  platform/mellanox: Fix mlxbf-tmfifo not handling all virtio CONSOLE notifications
  platform/x86: ideapad-laptop: Add support for new hotkeys found on ThinkBook 14s Yoga ITL
  platform/x86: lenovo-ymc: Add Lenovo Yoga 7 14ACN6 to ec_trigger_quirk_dmi_table

rshim console does not show all entries of dmesg.

Fixed by setting MLXBF_TM_TX_LWM_IRQ for every CONSOLE notification.
Signed-off-by: Shih-Yi Chen <shihyic@nvidia.com>
Reviewed-by: Liming Sung <limings@nvidia.com>
Reviewed-by: David Thompson <davthompson@nvidia.com>
Link: https://lore.kernel.org/r/20230821150627.26075-1-shihyic@nvidia.comReviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>

Don't queue more gc work, else we may queue the same elements multiple
times.

If an element is flagged as dead, this can mean that either the previous
gc request was invalidated/discarded by a transaction or that the previous
request is still pending in the system work queue.

The latter will happen if the gc interval is set to a very low value,
e.g. 1ms, and system work queue is backlogged.

The sets refcount is 1 if no previous gc requeusts are queued, so add
a helper for this and skip gc run if old requests are pending.

Add a helper for this and skip the gc run in this case.

Fixes: f6c383b8 ("netfilter: nf_tables: adapt set backend to use GC transaction API")
Signed-off-by: Florian Westphal <fw@strlen.de>
Reviewed-by: Pablo Neira Ayuso <pablo@netfilter.org>

Several instances of pipapo_resize() don't propagate allocation failures,
this causes a crash when fault injection is enabled for gfp_kernel slabs.

Fixes: 3c4287f6 ("nf_tables: Add set type for arbitrary concatenation of ranges")
Signed-off-by: Florian Westphal <fw@strlen.de>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>

Use nf_tables_gc_list_lock spinlock, not nf_tables_destroy_list_lock to
protect the gc list.

Fixes: 5f68718b ("netfilter: nf_tables: GC transaction API to avoid race with control plane")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Florian Westphal <fw@strlen.de>