Pull kvm fixes from Paolo Bonzini:
 "Two serious ARM fixes:

   - Plug a buffer overflow due to the use of the user-provided register
     width for firmware regs. Outright reject accesses where the user
     register width does not match the kernel representation.

   - Protect non-atomic RMW operations on vCPU flags against preemption,
     as an update to the flags by an intervening preemption could be
     lost"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  KVM: arm64: Fix buffer overflow in kvm_arm_set_fw_reg()
  KVM: arm64: Make vcpu flag updates non-preemptible

KVM/arm64 fixes for 6.3, part #4

 - Plug a buffer overflow due to the use of the user-provided register
   width for firmware regs. Outright reject accesses where the
   user register width does not match the kernel representation.

 - Protect non-atomic RMW operations on vCPU flags against preemption,
   as an update to the flags by an intervening preemption could be lost.

The KVM_REG_SIZE() comes from the ioctl and it can be a power of two
between 0-32768 but if it is more than sizeof(long) this will corrupt
memory.

Fixes: 99adb567632b ("KVM: arm/arm64: Add save/restore support for firmware workaround state")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Reviewed-by: Steven Price <steven.price@arm.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/4efbab8c-640f-43b2-8ac6-6d68e08280fe@kili.mountain
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>

Per-vcpu flags are updated using a non-atomic RMW operation.
Which means it is possible to get preempted between the read and
write operations.

Another interesting thing to note is that preemption also updates
flags, as we have some flag manipulation in both the load and put
operations.

It is thus possible to lose information communicated by either
load or put, as the preempted flag update will overwrite the flags
when the thread is resumed. This is specially critical if either
load or put has stored information which depends on the physical
CPU the vcpu runs on.

This results in really elusive bugs, and kudos must be given to
Mostafa for the long hours of debugging, and finally spotting
the problem.

Fix it by disabling preemption during the RMW operation, which
ensures that the state stays consistent. Also upgrade vcpu_get_flag
path to use READ_ONCE() to make sure the field is always atomically
accessed.

Fixes: e87abb73e594 ("KVM: arm64: Add helpers to manipulate vcpu flags among a set")
Reported-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20230418125737.2327972-1-maz@kernel.org
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>

Pull ARM SoC fixes from Arnd Bergmann:
 "There are a number of updates for devicetree files for Qualcomm,
  Rockchips, and NXP i.MX platforms, addressing mistakes in the DT
  contents:

   - Wrong GPIO polarity on some boards

   - Lower SD card interface speed for better stability

   - Incorrect power supply, clock, pmic, cache properties

   - Disable broken hbr3 on sc7280-herobrine

   - Devicetree warning fixes

  The only other changes are:

   - A regression fix for the Amlogic performance monitoring unit
     driver, along with two related DT changes.

   - imx_v6_v7_defconfig enables PCI support again.

   - Trivial fixes for tee, optee and psci firmware drivers, addressing
     compiler warning and error output"

* tag 'arm-fixes-6.3-3' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc: (32 commits)
  firmware/psci: demote suspend-mode warning to info level
  arm64: dts: qcom: sc7280: remove hbr3 support on herobrine boards
  ARM: imx_v6_v7_defconfig: Fix unintentional disablement of PCI
  arm64: dts: rockchip: correct panel supplies on some rk3326 boards
  arm64: dts: rockchip: use just "port" in panel on RockPro64
  arm64: dts: rockchip: use just "port" in panel on Pinebook Pro
  ARM: dts: imx6ull-colibri: Remove unnecessary #address-cells/#size-cells
  ARM: dts: imx7d-remarkable2: Remove unnecessary #address-cells/#size-cells
  arm64: dts: imx8mp-verdin: correct off-on-delay
  arm64: dts: imx8mm-verdin: correct off-on-delay
  arm64: dts: imx8mm-evk: correct pmic clock source
  arm64: dts: qcom: sc8280xp-pmics: fix pon compatible and registers
  arm64: dts: rockchip: Remove non-existing pwm-delay-us property
  arm64: dts: rockchip: Add clk_rtc_32k to Anbernic xx3 Devices
  tee: Pass a pointer to virt_to_page()
  perf/amlogic: adjust register offsets
  arm64: dts: meson-g12-common: resolve conflict between canvas & pmu
  arm64: dts: meson-g12-common: specify full DMC range
  arm64: dts: imx8mp: fix address length for LCDIF2
  riscv: dts: canaan: drop invalid spi-max-frequency
  ...

A few more Qualcomm ARM64 DeviceTree fixes for 6.3

The GPIO polarity of the WSA881x shutdown GPIO was inconsistent and had
to be corrected in the driver, this fixes the polarity in the DeviceTree
for QRB5165 RB5, SM8250 MTP, Samsung Galaxy Book 2 and Lenovo Yoga C630.

The recent rearrangement of nodes among the IPQ8074 accidentally enabled
the PCIe PHYs, rather than the PCIe controllers. This is being
corrected, to restore PCIe functionality.

PMK8280 PON node has the wrong compatible, which recently caused the
driver to stop probing. This is corrected and the required "pbs" region
is added.

With support for HBR3 introduced, it's noted that SC7280 Herobrine
devices are having trouble running at this rate. This drops the claim
that it's supported, until further analysis can be done.

* tag 'qcom-arm64-fixes-for-6.3-2' of https://git.kernel.org/pub/scm/linux/kernel/git/qcom/linux:
  arm64: dts: qcom: sc7280: remove hbr3 support on herobrine boards
  arm64: dts: qcom: sc8280xp-pmics: fix pon compatible and registers
  arm64: dts: qcom: ipq8074-hk10: enable QMP device, not the PHY node
  arm64: dts: qcom: ipq8074-hk01: enable QMP device, not the PHY node
  arm64: dts: qcom: qrb5165-rb5: Use proper WSA881x shutdown GPIO polarity
  arm64: dts: qcom: sm8250-mtp: Use proper WSA881x shutdown GPIO polarity
  arm64: dts: qcom: sdm850-samsung-w737: Use proper WSA881x shutdown GPIO polarity
  arm64: dts: qcom: sdm850-lenovo-yoga-c630: Use proper WSA881x shutdown GPIO polarity

Link: https://lore.kernel.org/r/20230410153850.4752-1-andersson@kernel.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Lower sd card speeds for two boards to make them run more reliable,
missing 32k clock definition for Anbric xx3 devices, missing cache-levels
for rk3588, fixed rk3326-board display supplies and more dt-schema fixes.

* tag 'v6.3-rockchip-dtsfixes1' of git://git.kernel.org/pub/scm/linux/kernel/git/mmind/linux-rockchip:
  arm64: dts: rockchip: correct panel supplies on some rk3326 boards
  arm64: dts: rockchip: use just "port" in panel on RockPro64
  arm64: dts: rockchip: use just "port" in panel on Pinebook Pro
  arm64: dts: rockchip: Remove non-existing pwm-delay-us property
  arm64: dts: rockchip: Add clk_rtc_32k to Anbernic xx3 Devices
  arm64: dts: rockchip: add rk3588 cache level information
  arm64: dts: rockchip: Lower SD card speed on rk3399 Pinebook Pro
  arm64: dts: rockchip: Lower sd speed on rk3566-soquartz
  ARM: dts: rockchip: fix a typo error for rk3288 spdif node
  arm64: dts: rockchip: Fix rk3399 GICv3 ITS node name

Link: https://lore.kernel.org/r/10559306.CDJkKcVGEf@phil
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Pull networking fixes from Jakub Kicinski:
 "Including fixes from bpf, and bluetooth.

  Not all that quiet given spring celebrations, but "current" fixes are
  thinning out, which is encouraging. One outstanding regression in the
  mlx5 driver when using old FW, not blocking but we're pushing for a
  fix.

  Current release - new code bugs:

   - eth: enetc: workaround for unresponsive pMAC after receiving
     express traffic

  Previous releases - regressions:

   - rtnetlink: restore RTM_NEW/DELLINK notification behavior, keep the
     pid/seq fields 0 for backward compatibility

  Previous releases - always broken:

   - sctp: fix a potential overflow in sctp_ifwdtsn_skip

   - mptcp:
      - use mptcp_schedule_work instead of open-coding it and make the
        worker check stricter, to avoid scheduling work on closed
        sockets
      - fix NULL pointer dereference on fastopen early fallback

   - skbuff: fix memory corruption due to a race between skb coalescing
     and releasing clones confusing page_pool reference counting

   - bonding: fix neighbor solicitation validation on backup slaves

   - bpf: tcp: use sock_gen_put instead of sock_put in bpf_iter_tcp

   - bpf: arm64: fixed a BTI error on returning to patched function

   - openvswitch: fix race on port output leading to inf loop

   - sfp: initialize sfp->i2c_block_size at sfp allocation to avoid
     returning a different errno than expected

   - phy: nxp-c45-tja11xx: unregister PTP, purge queues on remove

   - Bluetooth: fix printing errors if LE Connection times out

   - Bluetooth: assorted UaF, deadlock and data race fixes

   - eth: macb: fix memory corruption in extended buffer descriptor mode

  Misc:

   - adjust the XDP Rx flow hash API to also include the protocol layers
     over which the hash was computed"

* tag 'net-6.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (50 commits)
  selftests/bpf: Adjust bpf_xdp_metadata_rx_hash for new arg
  mlx4: bpf_xdp_metadata_rx_hash add xdp rss hash type
  veth: bpf_xdp_metadata_rx_hash add xdp rss hash type
  mlx5: bpf_xdp_metadata_rx_hash add xdp rss hash type
  xdp: rss hash types representation
  selftests/bpf: xdp_hw_metadata remove bpf_printk and add counters
  skbuff: Fix a race between coalescing and releasing SKBs
  net: macb: fix a memory corruption in extended buffer descriptor mode
  selftests: add the missing CONFIG_IP_SCTP in net config
  udp6: fix potential access to stale information
  selftests: openvswitch: adjust datapath NL message declaration
  selftests: mptcp: userspace pm: uniform verify events
  mptcp: fix NULL pointer dereference on fastopen early fallback
  mptcp: stricter state check in mptcp_worker
  mptcp: use mptcp_schedule_work instead of open-coding it
  net: enetc: workaround for unresponsive pMAC after receiving express traffic
  sctp: fix a potential overflow in sctp_ifwdtsn_skip
  net: qrtr: Fix an uninit variable access bug in qrtr_tx_resume()
  rtnetlink: Restore RTM_NEW/DELLINK notification behavior
  net: ti/cpsw: Add explicit platform_device.h and of_platform.h includes
  ...

Pull kvm fixes from Paolo Bonzini:
 "Two ARM fixes:

   - Ensure the guest PMU context is restored before the first KVM_RUN,
     fixing an issue where EL0 event counting is broken after vCPU
     save/restore

   - Actually initialize ID_AA64PFR0_EL1.{CSV2,CSV3} based on the
     sanitized, system-wide values for protected VMs"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  KVM: arm64: Advertise ID_AA64PFR0_EL1.CSV2/3 to protected VMs
  KVM: arm64: PMU: Restore the guest's EL0 event counting after migration

Daniel Borkmann says:

====================
pull-request: bpf 2023-04-08

We've added 4 non-merge commits during the last 11 day(s) which contain
a total of 5 files changed, 39 insertions(+), 6 deletions(-).

The main changes are:

1) Fix BPF TCP socket iterator to use correct helper for dropping
   socket's refcount, that is, sock_gen_put instead of sock_put,
   from Martin KaFai Lau.

2) Fix a BTI exception splat in BPF trampoline-generated code on arm64,
   from Xu Kuohai.

3) Fix a LongArch JIT error from missing BPF_NOSPEC no-op, from George Guo.

4) Fix dynamic XDP feature detection of veth in xdp_redirect selftest,
   from Lorenzo Bianconi.

* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
  selftests/bpf: fix xdp_redirect xdp-features selftest for veth driver
  bpf, arm64: Fixed a BTI error on returning to patched function
  LoongArch, bpf: Fix jit to skip speculation barrier opcode
  bpf: tcp: Use sock_gen_put instead of sock_put in bpf_iter_tcp
====================

Link: https://lore.kernel.org/r/20230407224642.30906-1-daniel@iogearbox.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>