VFIO updates for v7.3-rc1

 - Add nv_falcon vfio selftest driver.  The Falcon is a general-purpose
   microcontroller embedded within NVIDIA GPUs, presenting a relatively
   simple DMA programming interface.  This adds another selftest target
   for vfio-pci with real DMA transfers. (Rubin Du, Alex Williamson)

 - Add allocation assertion helpers to vfio selftests and use them to
   avoid variable length arrays and the compiler errors they generate.
   (Alex Mastro)

 - Fix use-after-free hazard where an init path error in MSI support
   leaves a stray pointer that can later be reused or double-freed.
   (Xiang Mei)

 - Fix previous refactor of PCI BAR mappings to honor non_mappable_bars
   flag, which otherwise generates a warning when trying to pci_iomap()
   a 256TiB BAR on ISM devices on s390. (Farhan Ali)

 - Add igb vfio selftest driver.  Like nv_falcon, this provides another
   target for DMA testing with vfio selftests, but importantly this
   driver supports both physical 82576 NICs and the emulation model in
   QEMU.  This therefore enables a vfio selftest vector with no physical
   hardware requirements. (Josh Hilke, Alex Williamson)

 - Mark selftest fixture objects __maybe_unused to accommodate builds
   with clang -Wunused-but-set-global. (David Matlack)

 - Add error recovery for vfio-pci devices on s390x.  This expands
   devices which expose the existing error eventfd and introduces a
   device feature for reporting firmware defined error state
   information to the user, allowing recovery through hypervisor
   channels. (Farhan Ali)

One of the "small improvements all over the place" releases for us.
It's hard to draw any direct comparisons because summer vacations
disrupted our patch processing (and presumably - generation) quite
a bit.

Quick and dirty count suggests we (Paolo and I) merged a very similar
number of net (632) and net-next (648) patches. This is not telling
the full story either because 1/3 to 1/2 of the net-next patches also
*seem* like AI-driven low priority fixes, cleanups and clarifications.

We are completely overwhelmed, of course. The glimmer of hope is that
we secured sufficient LLM budget and access (thank you Meta!) to run
reviews with multiple frontier models on each patch. This eliminates
some hallucinations. That said, in terms of review, the LLMs can only
do so much. The sad truth is that our APIs (especially for rare events
like PCIe errors, timeouts etc.) have always been racy, and now LLMs
don't let us ignore that. I expect our direction for the next release
will be to tweak the reviews a little bit more, but start shifting
focus to letting the LLMs take care of the busy work - managing
patchwork, automating common process complaints, editing commit
messages, and maybe applying patches which already got "reviewed-by"
tags from people we trust...

Enough blabber. Returning to this PR - you will see these conflicts:
 - trivial in MAINTAINERS, remove EHEA
 - trivial in include/linux/pci_ids.h, keep both
 - wifi/mt76 - keep the lockdep and the comment, rest it obvious
   https://lore.kernel.org/aoMaGJ1GcPZOoT_i@sirena.org.uk

Networking changes for 7.3.

Core & protocols

soc: devicetree updates for 7.3

There are two new mystery SoCs with very little public information
about them so far, Qualcomm's Cortex-A78C based "Shikra" and
Altera's Cortex-A720 based Agilex72 SOCFPGA.

We have also gained support for a couple of SoCs from the
2023/2024 timeframe that have been in the making for a while:

 - The Apple platform gains support for M3 Pro, Max and Ultra SoC,
   following the basic M3 support from 7.2.

 - Samsung Exynos 1580 is a high-end mobile phone SoC from 2024

 - Canaan K230 is a RISC-V based 64-bit AI SoC, based on the
   earlier K210 chip

 - Sophgo SG2000 is a mixed Arm/RISC-V chip that was already
   supported using the Arm core but is now also added for RISC-V
   along with several other variants of the cv18xx series

In terms of newly added machines we have reference platforms
for all the chips above, plus

 - only four 32-bit Arm boards: two older phones and two
   older industrial/embedded boards; using Allwinner, Qualcomm and
   Rockchip SoCs

 - Three laptops und three phones using Qualcomm SoCs

 - Ten 64-bit Rockchips based single-board computers, along with one
 NAS box and a game console.

 - Seven industrial/embedded boards and modules using NXP i.MX8/9 SoCs.

 - A Lenovo desktop box based on NVIDIA Jetson Xavier NX

 - A few more single-board computers based on Allwinnner A133,
   Spacemit K1 and TI AM62.

As usual, there is a constant stream of minor cleanups and fixes
towards addressing the 'dt-check-style --mode strict' warnings,
and everyone is expected to address those warnings for new
submissions now.

soc: drivers for 7.3

The SoC driver changes once more consist of many small fixes and
cleanups, that are to a large part the result of automated testing. On
platform specific drivers, this includes SoC specific code for xilinx,
freescale/nxp, qualcomm, TI, aspeed, omap, tegra, samsung, rockchip,
renesas, ixp4xx. In firmware drivers, we see a similar picture for
SCMI and qcomtee.

Aside from these, we see actual new hardware support in a few
areas:

 - The Apple platform gets a new driver for low power states

 - Updates to Qualcomm platform drivers add several new
   hardware specific features and additional SoCs.

 - Amlogic SoC support for A1 and T7 is added

 - The Mediatek MMSYS driver is refactored as a cleanup.

soc: defconfig updates for 7.3

Just the usual updates to the main defconfig files as well as the omap2
specific one, to enable more loadable modules for better default
hardware support.

soc: Arm platform updates for 7.3

The 32-bit Arm platforms are a bit more interesting this time: I refresh=
ed
an earlier series to mark code as deprecated that does have the tendency
of getting in the way of cleanups and new features but has close to zero
users. Among these are:

 - 22 of the remaining 28 legacy board files that predate the current
   devicetree based descriptions, using old chips from Intel and Marvell.
   The remaining six board files are for TI OMAP1 and Samsung s3c64xx
   chips and all still have known users.

 - support for Cortex-M3/M4/M7 and ARM1136r0 CPU cores and the 25
   machines based on these. These all use devicetree but the CPU
   support causes disproportional work. Most of them are just
   reference boards, the notable exceptions being the Nokia N800/N810
   tablet and the Buglabs BUG platform.

 - be8, be32, oabi and iwmmxt userspace binaries, which were mostly
   associated with the platforms now scheduled for removal and
   are increasingly problematic to support with modern toolchains.

Nothing is actually removed at this point, to ensure that any remaining
users continue to have the 7.3-LTS for a while longer. Patches for
removal are currently being tested.

Other updates include a continued work to convert GPIO number based
interfaces to descriptors, a patch to restore little-endian mode
on the one Arm platform (ixp4xx) that only worked in big-endian mode
recently, and some minor cleanups and bugfixes.

This one is again a little over 1000 non-merge commits, from 263
individual authors, with Qualcomm, NXP and Rockchip devicetree
files making up a majority of the changes. The new SoCs this
time are mostly a bit older, these include Apple M3 Pro/Max/Ultra,
Samsung Exynos 1580, Canaan K210, Sophgo SG2000 and the new
Qualcomm Shikra and as well as Altera Agilex72.

A small but noteworthy change is the deprecation of a number of
older 32-bit Arm platforms that were getting in the way.
For this release, nothing is actually being removed, but my
current testing has shown hundreds of now orphaned drivers that
can be removed along with those platforms.

The long-running devicetree cleanups are slowly getting to the
point of fixing known dtc compiler warnings and being able to
enforce this better for new contributions.

The most active contributors this time are:
     73 Krzysztof Kozlowski
     52 Stefano Radaelli
     50 Manivannan Sadhasivam
     46 Mukesh Ojha
     29 Konrad Dybcio
     23 Frank Li
     21 Arnd Bergmann
     20 Dmitry Baryshkov
     19 Sudeep Holla
     13 Jie Gan
     12 Luca Weiss
     12 Komal Bajaj
     12 James Calligeros
     12 Ioana Ciornei
     12 Aurelien Jarno
     11 Pengpeng Hou

Dirstat:
   0.6% Documentation/devicetree/bindings/arm/
   0.5% Documentation/devicetree/bindings/clock/
   0.1% Documentation/devicetree/bindings/gpio/
   0.2% Documentation/devicetree/bindings/soc/
   0.5% Documentation/devicetree/bindings/
   0.1% arch/arm/boot/dts/allwinner/
   0.6% arch/arm/boot/dts/aspeed/
   0.1% arch/arm/boot/dts/marvell/
   0.1% arch/arm/boot/dts/nxp/
   0.9% arch/arm/boot/dts/qcom/
   0.1% arch/arm/boot/dts/renesas/
   0.6% arch/arm/boot/dts/rockchip/
   0.8% arch/arm/boot/dts/ti/omap/
   0.3% arch/arm/boot/dts/vt8500/
   0.2% arch/arm/boot/dts/
   0.1% arch/arm/configs/
   0.1% arch/arm/mach-lpc32xx/
   0.1% arch/arm/mach-orion5x/
   0.3% arch/arm/mach-sa1100/
   0.7% arch/arm/
   0.4% arch/arm64/boot/dts/allwinner/
   0.3% arch/arm64/boot/dts/amd/
   0.2% arch/arm64/boot/dts/amlogic/
   8.3% arch/arm64/boot/dts/apple/
   2.0% arch/arm64/boot/dts/exynos/
  16.4% arch/arm64/boot/dts/freescale/
   0.4% arch/arm64/boot/dts/intel/
   0.5% arch/arm64/boot/dts/mediatek/
   0.7% arch/arm64/boot/dts/nvidia/
  28.0% arch/arm64/boot/dts/qcom/
   2.4% arch/arm64/boot/dts/renesas/
  11.8% arch/arm64/boot/dts/rockchip/
   0.2% arch/arm64/boot/dts/sophgo/
   0.1% arch/arm64/boot/dts/tesla/
   5.3% arch/arm64/boot/dts/ti/
   1.2% arch/arm64/boot/dts/xilinx/
   0.2% arch/arm64/boot/dts/
   0.6% arch/riscv/boot/dts/canaan/
   0.4% arch/riscv/boot/dts/sophgo/
   1.3% arch/riscv/boot/dts/spacemit/
   0.1% arch/riscv/boot/dts/thead/
   0.4% drivers/firmware/arm_scmi/
   2.3% drivers/firmware/qcom/
   0.1% drivers/firmware/xilinx/
   0.2% drivers/memory/tegra/
   0.4% drivers/soc/amlogic/
   0.2% drivers/soc/apple/
   0.3% drivers/soc/fsl/qe/
   0.7% drivers/soc/fsl/
   1.1% drivers/soc/mediatek/
   1.7% drivers/soc/qcom/
   0.1% drivers/soc/tegra/
   0.4% drivers/
   0.7% include/dt-bindings/clock/
   0.1% include/linux/soc/qcom/
   0.1% include/linux/
 1088 files changed, 70562 insertions(+), 7263 deletions(-)

All commits have been in linux-next; two signed tags for "software node
fw_devlink support" and "Rust I/O type generalization and projection" have =
been
shared with other trees.

Expect a few minor conflicts with other trees (DRM, PM, PWM), and a minor
semantic one with the Rust tree [1].

Thanks,
Danilo

[1] https://lore.kernel.org/linux-next/20260807074729.108159-1-ojeda@kernel=
.org/

Driver core changes for 7.3-rc1

- container_of:

  - Apply typeof_member(), remove the local __mptr variable to eliminate
    variable shadowing warnings on nested container_of() calls, and remove
    unnecessary parentheses.

- core:

  - Add driver name to probe debug print for initcall_debug.

  - Avoid repeatedly printing the same 'Fixed dependency cycle' log.

  - Unwind device_add() on attribute creation failure in
    attribute_container_add_class_device().

  - Remove statistics group if encryption group creation fails in
    transport_add_class_device().

- debugfs:

  - Fix lockdown check for mmap_prepare().

  - Warn if file creation failed due to uninitialized debugfs.

- device property:

  - Implement fw_devlink support for software nodes by adding
    software_node_add_links(), which creates fwnode links from DEV_PROP_REF
    properties to enable automatic probe ordering. Add kunit-managed fwnode
    helpers and test coverage.

  - Fix infinite loop in fwnode_for_each_child_node() when the secondary fw=
node
    has more than one child. Add test cases.

  - Fix out-of-bounds access in software_node_get_reference_args() when cal=
led
    with index -1 (UINT_MAX).

  - Refactor to use RAII approach with __free().

  - Add Bartosz Golaszewski as software node reviewer.

- firmware loader:

  - Fix race where a sysfs fallback request can complete before being queue=
d as
    pending, leading to a use-after-free on the next fallback request.

  - Reject 0-size built-in firmware and fail the build on empty firmware fi=
les
    in CONFIG_EXTRA_FIRMWARE.

- kobject:

  - Provide __KOBJ_ATTR() and __KOBJ_ATTR_RO/WO() initialization macros and
    allow the constification of kobject attributes, enabling them to reside=
 in
    read-only memory.

- platform:

  - Provide platform_device_set_of_node(), platform_device_set_fwnode(), an=
d
    platform_device_set_of_node_from_dev() helpers that encapsulate firmwar=
e
    node reference counting for dynamically allocated platform devices. Con=
vert
    all in-tree users that manually assigned dev.of_node or dev.fwnode, fix=
ing a
    pre-existing refcount bug in powermac.  Switch to counting references o=
f all
    firmware node types, not only OF nodes.

  - Unify the release path for dynamically allocated platform devices by
    removing platform_device_release_full(). Amend the fwnode setter API
    contract to warn if a primary software node is overwritten. Add KUnit t=
ests
    for correct software node removal on device unregistration.

- Rust:

  - Auxiliary:

    - Add registration_data_with() closure-based API for invariant ForLt ty=
pes.

  - Debugfs:

    - Migrate BinaryWriter and BinaryReaderMut trait requirements from
      kernel::transmute traits to zerocopy traits.

  - Device:

    - Add BoundInternal device context and InternalBoundContext trait for b=
us
      abstractions that need internal access to a bound device.

    - Make the lifetime on Core and CoreInternal invariant to prevent coerc=
ion
      to shorter lifetimes.

  - Devres:

    - Fix race between concurrent revokers where the losing revoker could r=
eturn
      before the winning revoker finished dropping the inner data, causing
      use-after-free.

    - Ensure revocation is complete before the device finishes unbinding by
      making the synchronization bidirectional.

    - Add DevresLt<F: ForLt>, a wrapper around Devres that shortens 'static=
 back
      to the caller's borrow scope. Implement ForLt and CovariantForLt for =
Bar,
      IoMem, and ExclusiveIoMem.

  - Driver:

    - Switch from index-based to pointer-based device ID info lookup, stori=
ng
      static references in driver_data. Centralize device ID handling in
      device_id.rs, removing the open-coded ACPI/OF matching logic and dupl=
icate
      ID table from driver.rs.

  - I/O:

    - Make I/O regions typed (with a dynamically-sized Region type for the
      existing untyped case), create view types representing subregions of =
a
      mapped I/O region, and add io_project!() for safely creating subviews=
.

    - Split Io into a base trait (IoBase) and an extension trait (Io) with =
a
      blanket implementation, preventing implementers from overriding provi=
ded
      methods that unsafe code relies on.

    - Add a SysMem backend for shared system memory with volatile access, a=
nd
      make Coherent implement Io via an I/O view type.  Add IoSysMap as sum=
 type
      of Mmio and SysMem. Add copying methods (memcpy_{from,to}io()) and
      read_val()/write_val() for typed access.

    - Replace dma_read!()/dma_write!() with io_read!()/io_write!() for
      primitives and copying methods for aggregates; drop the old macros.
      Convert nova-core to use I/O projection.

    - Fix internal shortcut rule dispatch in the register!() macro, remove
      unused rule arguments, and use path fragments for alias destinations.

  - IRQ:

    - Make irq::Registration compatible with lifetime-bound drivers by
      removing the 'static bound on Handler/ThreadedHandler and
      replacing Devres<RegistrationInner> with direct
      request_irq()/free_irq() calls.  Handlers can now directly own
      lifetime-bound device resources.

  - PCI:

    - Convert IrqVectorRegistration to a lifetime-annotated owning type,
      giving drivers explicit control over the allocation lifetime.
      IrqVector embeds a resolved IrqRequest, making the conversion
      infallible. Remove the redundant
      request_irq()/request_threaded_irq() wrappers from pci::Device.

    - Add pci_irq_type() C helper and expose it via irq_type() on
      IrqVectorRegistration and IrqVector, returning PCI_IRQ_MSIX, PCI_IRQ_=
MSI,
      or PCI_IRQ_INTX.

    - Mark pci::Device refcount methods inline.

  - Serdev:

    - Add Rust abstractions for the serial device bus, including serdev::Dr=
iver
      trait, serdev::Device wrapping struct serdev_device, and serdev::Adap=
ter
      implementing RegistrationOps.  Includes a sample driver. Markus Probs=
t
      takes over as serdev maintainer for both C and Rust code.

  - Misc:

    - Split ForLt into a base trait (providing the Of<'a> GAT) and an unsaf=
e
      CovariantForLt subtrait guaranteeing covariance, enabling invariant t=
ypes
      (e.g. those containing Mutex<&'bound T>) to participate in the ForLt
      abstraction.

    - Fix Coherent read past EOF returning -ERANGE instead of zero.

    - Fix firmware example UB by avoiding null-pointer ARef.

- misc:

  - Avoid iattr allocation in kernfs listxattr by using kernfs_iattrs_noall=
oc().
  - Unregister SoC bus on early device registration failure.
  - Remove unused DMA_FENCE_TRACE Kconfig symbol.
  - Fix /sys/module path in comment.
  - Refactor ISA bus init to remove nested blocks.
  - Remove redundant nodemask clears in numa_init().
  - Add kernel-doc for fwnode_operations and sys_soc.h, mark internal prope=
rty
    data as private for kernel-doc, and add property.h/fwnode.h to driver-a=
pi
    infrastructure docs.
  - Add MAINTAINERS entry for sys_soc.h.

please pull the x86/cpu lineup for v7.3-rc1.

This is conflicting with other work from tip, see end of mail for possible
resolution.

Thx.

---

- Get rid of static_cpu_has() - one less API to care about testing CPU
  features

- Unify the handling of CPU core types (performance, efficient, etc) by
  mapping the vendor-specific types to Linux ones

- Continuation of the work of Ahmed Darwish to centralize CPUID leafs
  representation

This pull request contains two fixes for Alpha floating-point exception
handling, two clang-related fixes, the preparatory changes from the
generic-entry series, an interrupt-entry lockdep fix, two Marvel/EV7 IRQ
fixes, an RTC fix, and one header cleanup.

The generic-entry preparation adds regset-based ptrace and core dumps,
ARCH_STACKWALK and lockdep hardirq-state tracking. These changes are
useful independently and enable previously missing debugging facilities
on Alpha.

The final patch switching Alpha to GENERIC_ENTRY is intentionally not
included in this pull request. I am deferring that change to allow
further testing and to reduce the risk of conflicts with ongoing
entry-path work elsewhere in the kernel.

alpha updates for v7.3

- alpha: read $gp and $sp explicitly for clang
- alpha: pass -Wa,-mev6 only when using GNU as
- alpha: annotate hardirqs-off on IPL 7 interrupt entry
- alpha: run the remote RTC access in a worker, not an IPI callback
- alpha: don't leak hardware-fabricated FP exception bits to user space
- alpha: fix ieee_swcr_to_fpcr setting FPCR_DNOD unconditionally
- alpha: enable lockdep hardirq state tracking
- alpha: use raw spinlocks for low-level platform locks
- alpha: provide ftrace return address support for lockdep
- alpha: make irqflags helpers operate on IPL state
- alpha: add ARCH_STACKWALK-based stacktrace support
- alpha: enable regset-based ptrace and core dumps
- alpha: marvel: Fix lock ordering in init_io7_irqs()
- alpha: marvel: Fix irq_set_status_flags to use correct IRQ number
- alpha: remove unnecessary architecture-specific <asm/device.h>

please pull the x86/cache lineup for v7.3-rc1.

Thx.

---

- How refreshing: no new features but a whole pile of fixes to more or less
  serious issues reported by Sashiko along with miscellaneous cleanups all
  over the place. All except one by Reinette Chatre, the one by Tony Luck

There are two things in here of note. First, there are fixes to reduce
the time that the TLB flushing code sits around with interrupts disabled.
It should be well tested, but it's certainly something to keep an eye on.

Second is a little back-and-forth in the interactions between the
set_memory*() locking and DEBUG_PAGEALLOC. There was a slightly painful
re-discovery of why DEBUG_PAGEALLOC has special locking rules, but the
code did end up better for it.

--

 * Fix pkey_alloc() return value when pkeys are not supported
 * Allow preemption during IPI completion waiting to improve wakeup latency
 * Convert more x86 page table code to ptdescs
 * Fixes and cleanups for set_memory*() locking, mostly around large pages

exceptions whenever they do port I/O. The guest then does some
instruction decoding and makes a call up to the host for help. That
decoding had a couple of bugs.

Fix those bugs. Use an existing and now shared KVM helper for one
of them.

--

 * Fix TDX port I/O bugs

pretty benign, just adding some documentation and removing a super
stale comment about an LED driver that was moved long ago.

--

 * Document the intricacies of GS context switching
 * Remove old TODO message about Geode LED driver

pointer unwinder when it encounters FRED stack frames.

--

Fix frame pointer unwinder when it encounters FRED stack frames

Probes updates for v7.3:

- probes: Advanced BTF typecasting and variable fetch enhancements
  . Typecast support across probe events: Extended BTF typecasting syntax
    (e.g., (STRUCT)PARAM->MEMBER) to kprobes, uprobes, and fprobes on
    function entry and return.
  . Nested typecasts: Added support for chaining and nesting typecasts up
    to 3 levels, including casting registers and stack variables.
  . Field specifier option: Added (STRUCT,FIELD) syntax to emulate
    container_of(), allowing retrieval of parent structures from member
    pointers.
  . $current variable support: Introduced $current special variable to
    access the running task_struct via BTF dereferencing.
  . Per-CPU variable access: Added this_cpu_read() and this_cpu_ptr()
    fetcharg methods to trace CPU-local data safely.
  . Fetcharg bytecode dumper: Added CONFIG_PROBE_EVENTS_DUMP_FETCHARG to
    dump the compiled fetcharg bytecode instructions as comments in
    dynamic_events.
  . Extended symbol name handling: Removed the MAX_COMMON_HEAD_LEN limit
    and extended MAX_ARGSTR_LEN to 256 bytes, enabling probing of long
    symbols, mangled Rust symbols and complex BTF expressions.
  . eprobe variable syntax: Allowed eprobes to reference event fields
    directly without requiring a '$' prefix.
  . Cleanup unused parameters, redundant codes, duplicate macros and
    pointer arithmetic.
  . Use a ternary operator for simplifying fetch_type_from_btf_type().

- bootconfig / boottime-trace: Expanded dynamic probe support
  . Added boot-time tracing configuration support for event probes
    (eprobes), function probes (fprobes), and tracepoint probes
    (tprobes).
  . ftrace2bootconfig: Allowed comment lines ('#') in dynamic_events file.

- fprobe / kprobe: Optimization, robustness, and cleanups
  . fprobe: Simplified fprobe_remove_ips() by reusing graph and ftrace
    helpers.
  . fprobe: Removed __packed attribute from struct __fprobe_header to avoid
    unaligned memory access penalties on RISC architectures.
  . kprobe & fprobe: Removed redundant memset() calls in perf event probe
    handlers.
  . kprobes: Replaced legacy __ASSEMBLY__ with __ASSEMBLER__ in header
    files.

- selftests & refactoring:
  . Refactored parse_probe_arg() and parse_probe_vars(), and eliminated
    recursion in probe argument parsing to protect kernel stack depth.
  . Added selftests for BTF typecasts and module probing without module
    prefixes.
  . Forced LC_ALL=C in ftracetest to prevent test failures on localized
    systems.
  . Refactored btf_type_skip_modifiers() to remove ignored id parameter.
  . Sorted ERRORS list in trace_probe.h alphabetically.
  . Fixed typo in fprobe docs, and trace_fprobe function name.
  . Renamed FETCH_OP_DATA to FETCH_OP_IMMSTR.
  . Made file offset error message probe-agnostic.




probes-v7.3

This is the fixed version of the primary PR from the GPIO tree for this merge
window. It's essentially the same except that it contains two additional merge
commits: one pulling in the upstream v7.2 tag in order to address a conflict
between my for-next and your master branches. The second pulls in an immutable
branch from Mark Brown I missed before but which contains a commit on which
some of the regmap changes depend.

There's one new driver for the Realtek DHC RTD1625 GPIO controller. If adding
it, reverting it and then adding it again raises your eyebrow: this was done
becuase the initial implementation - while correct - didn't use the gpio-regmap
abstraction that suited it nicely. The conversion was done after I had already
merged v7.2-rc2 into my tree so instead of rebasing, I went with a revert for
a cleaner git blame output.

The merges you see - other than pulling in your upstream tags with conflicting
fixes - are for a PR from Arnd containing his work on further isolating legacy
GPIO APIs and the fw_devlink support for software nodes that will go upstream
via the driver core tree.

Nothing else really stands out, details are in the signed tag. Please consider
pulling.

Thanks,
Bartosz

gpio updates for v7.3-rc1

GPIO core:
- extend the gpio-regmap abstraction layer with more features allowing users
  to override configuration setting, translate register values and masks and
  enable/disable interrupts
- extend GPIO kunit tests with suites verifying probe ordering by software
  node devlink support and software node hogs
- shrink GPIO kunit initialization code
- coding style updates (remove commas from sentinels where applicable)
- with all users now converted treewide to using real firmware node links for
  software node GPIO lookup: remove the deprecated label-matching mechanism
  from from GPIO core
- drop redundant return value check of nonseekable_open() in gpiolib-cdev
- use IRQ trigger helpers where applicable

Driver updates:
- refactor error paths and logging in gpio-nomadik
- use more modern interfaces for getting resources in gpio-rockchip,
  gpio-bt8xx and gpio-pca9570
- add missing MODULE_DEVICE_TABLE() to gpio-sifive and gpio-vf610
- drop unused FILONOFF macro from gpio-rcar
- extend build coverage of ioport GPIO drivers with COMPILE_TEST=y
- only enable the gpio-rtd driver by default with ARCH_REALTEK=y to avoid
  bloating the build
- refactor coding style in several drivers
- use correct endianess translation in gpio-pcf85x
- add wake-up interrupt support to gpio-mvebu
- apply initial value in direction output setter in gpio-by-pinctrl

Misc:
- replace linux/gpio.h inclusions treewide with linux/gpio/legacy.h which now
  exports all the deprecated APIs
- select GPIOLIB_LEGACY in Kconfig where required treewide
- use software nodes for gpio-keys in MFD drivers

Devicetree bindings:
- describe the realtek rtd1625 GPIO controller
- document new models for gpio-pca95xx and gpio-cadence
- document new property in gpio-rockchip

 
Regards,
Konstantin

Changes for 7.3-rc1

Added:
    load ATTR_BITMAP run extents from $MFT extension records
    initialize err in attr_wof_frame_info
    reserve NUL byte when converting UTF-16 names
    reject restart table growth beyond U16_MAX entries
    validate dirty page table on log replay
    basic support for alternative data streams
    validate ef->size covers the record's name and value

Fixed:
    slab-out-of-bounds write in ni_create_attr_list()
    out-of-bounds read of INDEX_ROOT in reparse/objid init
    boundary check in ntfs_dir_count()
    info-leak in ntfs_rename()
    lseek EINVAL on sparse/compressed files with 64-bit clusters
    info-leak on partial LZNT decompress in ni_read_frame()
    bound page_lcns[] index by the log record
    memory leak in indx_find_sort()
    integer overflow in MFT cluster validation
    reject out-of-range evcn in mi_enum_attr()
    out-of-bounds read in read_log_rec_buf()

Changed:
    widen inode/record number storage to u64
    cosmetic fixes and improvements
    rename 'err' to 'ret' in read paths

This update includes the following changes:

API:

- Add af_alg_restrict sysctl and white list.
- Fix potential suspend/resume races in hwrng.

Algorithms:

- Optimize vli additive operations using compiler builtins in ecc.

Drivers:

- Remove unsafe/deprecated algorithms from qce.
- Mark qce as BROKEN.
- Add runtime PM and interconnect bandwidth scaling support to qce.
- Remove crypto_rng from qcom, sun8i and caam.
- Fix SG list issues in iaa.
- Fix SEV init path bugs in ccp.

please pull a x86/alternatives cleanup for v7.3-rc1.

Thx.

---

- Remove the smp_locks alternatives machinery which was used to patch out lock
  prefixes when running a SMP kernel on a uniprocessor machine

please pull a single ras/core fixlet for v7.3-rc1.

Thx.

---

- Add mce=print_all to the mce= kernel cmdline params documentation

Various ext4 updates for 7.3-rc1, including:

* Improve performance by allowing parallel DIO writes when we were
  previously being overly conservative when checking whether it was safe
  to avoid requiring an exclusive lock
* Improve the performance of ext4_mb_prefetch() used by fallocate() by
  avoiding work when it is not needed
* Remove the unnecessary custom end_io function ext4_end_buffer_io_sync()
* Improve performance when performing an overwrite to an already uptodate folio
* Clean up how we handle deallocating EA inodes to avoid a potential
  lock ordering issue when there is a failed mount while an EA inode is
  still being evicted.
* Use str_plural() instead of a custom macro
* Avoid soft lockups or RCU stalls if there are many busy buffers
  (caused by heavy I/O) while checkpointing
* Use scoped NOFS when starting a handle in nojournal mode
* Align fields in handle structure to optimize setting and getting
  the h_type and h_line_no fields
* Fix documentation of the meta_bg block group layout
* Bug fixes, including
   - Fix a potential out-of-bounds read in ext4_read_inline_dir()
   - Fix a potential deadlock when concurrent xattr operations are racing
     with each other when some of the xattrs are using the ea_inode feature.
   - Fix a spurious warning with data=journal that can be triggered when
     writeback races with remounting the file system read-only
   - Fix a potential deadlock when EXT4_IOC_MIGRATE races with a file system
     freeze operation
   - Make sure all in-flight direct I/O operations are complete before
     falling back to buffered I/O.
   - Handle IOCB_NOWAIT properly when performing a extending DAX write
   - Prevent potentially sleeping on a block allocation when
     IOCB_NOWAIT is set
   - Fix potential races when racing an inline data write with a page fault
   - Propagate errors when adding or removing extent ranges
     during a fast commit replay
   - Avoid trying to expand an inode's extra size when it is being
     evicted to avoid a number of corner case or deadlocks
   - Avoid spurious error when retrying inode extra size expansion
   - Fix corner cases where we underestimate the number of journal
     credits needed
   - Avoid hangs/crashes/WARNINGS caused by maliciously corrupted
     file systems
   - Don't issue spurious orphan clean message on RO file systems
   - Avoid leaving the file system in an inconsistent state after a
     crash when a WRITE_ZEROS in progress converting an unwritten
     extent to a written extent
   - Handle WRITE_ZEROS correctly when there are some partially dirtied
     regions in the page cache
   - Pass errors during zero-rage, truncate, or punch hole to the caller
     if ext4_get_block() fails
   - Wait for writeback to finish when triggered by zero-range or
     zero-range for those devices that require stable writes
   - If the reserved gid superblock field is set, set the reserved gid
     instead of the reserved uid

Could consider these commits for 7.3-rc1?

The most notable enhancement is to allow passing source fds via
fsconfig() for composefs. The others are all various fixes.

All commits have been in -next and no potential merge conflict is
observed.

Thanks,
Gao Xiang

Changes since last update:

 - Allow source fds via fsconfig(), in addition to source paths

 - Use dedicated metadata inodes for file-backed mounts

 - Disallow invalid interlaced ztailpacking pclusters

 - Validate on-disk compression algorithm IDs against supported ones

 - Fix unused pcluster pools on higher page-size platforms

Thanks,
Guenter
------

hwmon updates for v7.3

* New drivers

  - Driver for Kandou KB9002 retimer

  - Driver for the temp/voltage sensor on PolarFire SoC

  - Driver for Eswin EIC7700 PVT sensor

  - PMBus:

    - Driver for Analog Devices MAX16545/MAX16550 and Volterra VT7505

    - Drivers for Monolithic MPQ82D00 and MPQ8646

    - Driver for Silergy SQ24860

* Added support to existing drivers

  - asus-ec-sensors: Support for ROG STRIX Z390-E GAMING,
    ProArt Z690-CREATOR WIFI, ROG STRIX X870E-E GAMING WIFI7 R2,
    ROG CROSSHAIR X870E HERO, and ROG Maximus Z790 Hero

  - asus_rog_ryujin: Siupport for ROG Ryujin III

  - ina2xx: Support for INA232

  - k10temp: Per-CCD temperature monitoring for Zen5 Turin

  - nct6775: List NCT5585D as supported chip

  - nzxt-kraken3: Support for NZXT Kraken 2024 Elite

  - sht3x: Support for GXCAS GXHT30

  - tmp102: Add device IDs for TMP110 and TMP113

  - yogafan: Support for LOQ 15IAX9, XiaoXin Pro 13ARE 2020,
    IdeaPad 3 15ALC6, Legion Pro 7 16AFR10H, Yoga Pro 7 14IAH10,
    Yoga 7 16ARP8, and Lenovo LOQ 15IAX9

  - PMBus:

    - max20830: Support for max20830c and max20840c

    - max34440: Support for MAX34452, and support for newer version
      of max34451

    - adm1275: Support for ROHM BD12780 and BD12790

* Other notable changes

  - Constify various device attributes

  - Remove redundant dev_err() and dev_err_probe() from various drivers

  - applesmc: Convert to hwmon_device_register_with_info

  - adt7470: Add thermal zone sensor support

  - coretemp: Fix core_data leak on CPUs without PTS

  - emc1403: Drop hysteresis for low limit temperature

  - max6621: Fix various over- and underflow problems

  - PMBus:

    - Introduce pmbus_read_smbus_i2c_block_data() and use it in
      various drivers

    - Export and use pmbus_check_and_notify_faults()

    - Let PMBus drivers report the supported PMBus revision

* Various other minor fixes and improvements

i.MX clock changes for 7.3

- Add audio PLL debugfs support for K-divider control.
- Add missing MODULE_DEVICE_TABLE() declarations for i.MX8QXP clock drivers.
- Add PCIe FUNC_OUTPUT_EN clock gate support on i.MX95.
- Configure the i.MX95 PCIe transmitter current reference to fix REFCLK
  rise-fall timing.

aspeed: Update MAINTAINERS for 7.3

Add Ryan and Billy as reviewers for ASPEED SoC patches. I asked both to add
themselves to increase ASPEED's visibility of upstream changes.

aspeed: First batch of driver changes for 7.3

Just the one patch adding MODULE_DEVICE_TABLE() where it was previously mis=
sing.

aspeed: First batch of ARM devicetree changes for 7.3

No significant changes this time around. Instead, several cleanups
for Meta's San Miguel platform, along with some changes enabling
dt-check-style.

nuvoton: First batch of arm64 devicetree changes for 7.3

Similar to the 7.3 ARM PR for Nuvoton we have a couple of arm64 tidy-ups
from Tomer.

nuvoton: First batch of ARM devicetree changes for 7.3

This time around we have a couple of cleanups from Tomer that make way for
improvements in the bindings.

Hi Joerg,

Hi SoC Maintainers,


Yixun Lan

RISC-V SpacemiT DT changes for 7.3

For K3 SoC
- Enable SD card support
- Enable QSPI
- Fix reset delay of ETH PHY
- Fix PHY ID of ETH
- Symbolic PDAM ID
- Add SPI controller support
- Add USB and PHY support
- Add I2S nodes

For K1 SoC
- Symbolic PDAM ID
- Enable cpufreq support
- Fix max CPU core voltage

For boards of K3
- Pico-ITX
  - Enable ALDPS on RTL8211F PHY
  - Disable 125MHz clock on RTL8211F PHY
  - Enable NOR flash
  - Add EEPROM support
- CoM260
  - Add EEPROM support

For boards of K1
- Milk-V Jupiter
  - Enable CPU DVFS
- Banana Pi BPI-CM6
  - Add initial support
- OrangePi R2S
  - Enable PCIe

Hi Stephen,


Yixun Lan

RISC-V SpacemiT clock changes for 7.3

- Add clock for UFS controller
- Add new I2S clock and fixes

Hi Peter,


The major changes are the introduction of preempt_count backed interrupt
disabling level tracking, i.e. the local_interrupt_{enable, disable}(),
along with the user (the Rust SpinLockIrq). This should unblock a few
drivers in Rust. For other changes, please refer the following change
log.

Rust synchronization changes for v7.3:

* Safe SpinLockIrq
  - Introduce refcounted interrupt disabling and corresponding spinlock
    primitives, which tracks the interrupt disabling "level" in
    preempt_count, as a result, this set of primitives can automatically
    restore the interrupt disabling status when the outermost critical
    section exits.

  - Based on the above, a SpinLockIrq is introduced on the Rust side,
    the guard of which will automatically restore the interrupt
    disabling status when dropped. This avoid playing unsafe games when
    using irq disabled spinlocks.

* Add synchronize_rcu() safe abstraction.

* Generic memory barrier: this extends our Rust barrier primitives to
  include more barriers for DMA as well.

* Replace the atomic usage in Revocable with LKMM atomics.
-----BEGIN PGP SIGNATURE-----

iQFhBAABCABLFiEEj5IosQTPz8XU1wRHSXnow7UH+rgFAmptA2MbFIAAAAAABAAO
bWFudTIsMi41KzEuMTIsMCwzERxib3F1bkBrZXJuZWwub3JnAAoJEEl56MO1B/q4
UBAH/108a8dCkDaK3MUuS0SeuqY57EPwLCTE/k4gJK35XlF0fCpEPfFtqfEs6xkG
E0rMhSx3bDmtfccEo16LOkmbcrSx3NSsc9KlSRiYRywptT+CRld5FPqmJqDd2nCh
mYYUsNppEXFd4rUiGfKAcVeLHD+PCio0/ElNN7X3TCfQgtWUuDp4pcB23m9t6ste
Re4idhKoTmqeKTX3N8AuQ0QMsr1FNNLu0ESpUv1n6/DrEB1g8XevF1J2PKYr9fxt
aahCMk2qQF3vDCi4Ydh9o9AWqSkinU2yIW64ZBWRC2S7QBVRIrPAEtX4V/n7d/n+
2irbb1ZuVu5zFD0/CyNkgyQM/og=
=DzLT
-----END PGP SIGNATURE-----


Regards,
Boqun

Steve,

rv changes for v7.3 (for-next)

Summary of changes:

- Switch LTL and DOT parsers to Lark in code generation tool

  The rvgen code generation tool originally parsed DOT files and LTL
  specifications using custom string parsing and Ply, which is no longer
  maintained. The DOT parser was fragile and prone to failure on minor
  format variations. Both LTL and DOT parsers have been rewritten to use
  the Lark parsing library.

- Simplify Hybrid Automata clock variables

  The clock variables in hybrid automata monitors now use a single
  representation of the elapsed time since the clock was reset, rather
  than converting between invariant and guard representations.
  This allows simpler code generation for the newly refactored parser.

- Generate cleanup hook for per-obj monitor

  The code generation scripts now adds a cleanup function to per-obj
  monitors for the user to wire to the appropriate event (e.g.
  sched_process_exit for tasks).

- Reduce read_lock scope during per-task cleanup

  Take the tasklist_lock only when necessary, that is when iterating
  over for_each_process_thread().

- Simplify task monitor slot management

  Only rely on the slot array for per-task slot management to avoid
  inconsistency with the unused counter.

- Improve rvgen code robustness and templates

  Use pathlib in rvgen and improve kernel path discovery. Also improve
  consistency across templates when generating code (e.g. author
  placeholder and monitor struct name).

- Update rtapp sleep monitor

  Simplify the sleep monitor by excluding kernel threads and
  updating the nanosleep check to focus only on CLOCK_REALTIME. Also
  switch to use the sched_exit tracepoint to run in the context of the
  offending (wakee) task.

- Add wakeup monitor

  Add the new rtapp/wakeup monitor to detect when lower-priority tasks
  wake up higher-priority ones, complementing the existing sleep monitor
  by running in the waker context and capturing its stack trace.

- Fix tools/rv exit status on failure

  Ensure the rv tool returns a failure exit code when a monitor fails to
  start because it was already running.

- Add automated selftests for tools/rv and rvgen

  Introduced automated bash selftests to validate rv monitor listing and
  execution under different configurations. Added tests for the rvgen code
  generator, validating generated files against expected output (golden).
  Tests are reachable via make check.

- Add KUnit test coverage for verification monitors

  Added comprehensive KUnit tests to validate the functionality of
  deterministic, hybrid, and LTL monitors by emulating event sequences
  and timing in a mock environment without affecting the running kernel
  while expecting mock reactions to fire. Ensure real RV monitors cannot
  run during KUnit tests to avoid state corruption.

- Mock current in rv monitors

  Mock the call to current in rv monitors when the KUnit tests are built
  to allow them to run the test on dummy tasks. No overhead is expected
  when KUnit tests aren't running.

- Introduce rvgen kunit subcommand

  Added a new 'kunit' subcommand to rvgen to automatically patch an already
  generated monitor with KUnit integration templates by parsing its event
  handlers and creating the required mock structures and initializations.

- Refine kernel verification selftests

  Added new selftests for the deadline and stall monitors and rearranged
  the existing wwnr_printk test to resolve flakiness.
  Additionally, fixed an issue in the selftests framework where negative
  assertion failures were not correctly propagated due to shell rules.

SoCFPGA firmware updates for v7.3
- Document stratix10-rsu for QSPI size and erase size
- Add method to retrieve device info table using RSU
- Add support for hardware monitoring using service driver

Hi Stephen and Michael,

Here's some changes that were posted since June 29th that I feel are ready to
be picked up. These are changes where a clk submaintainer typically won't send
the changes in via a separate pull.

Note that I have a separate outstanding pull for some changes posted during
the last merge window that I feel is also ready for inclusion.

    https://lore.kernel.org/linux-clk/akLzCVzbMbUsgDda@redhat.com/

There is a minor conflict in drivers/clk/clk-si544.c between these two pulls.
You'll need to add the .name and .driver_data name initializers like the
following based on the work from Uwe Kleine-KЖnig:

-        { "si544a", (kernel_ulong_t)&clk_si544a_info },
+        { .name = "si544a", .driver_data = (kernel_ulong_t)&clk_si544a_info },

With the conflict resolved, I verified that both merges can go in together, or
individually.


clk: Clock driver updates for v7.3 (round two)

- New clock controller drivers for the Cix Sky1 audio subsystem (AUDSS),
  UltraRISC DP1000, and MediaTek MT8173 MFG_TOP, along with their devicetree
  bindings. Si549 support was added to the existing si544 driver.

- New clock and reset support for the Aspeed AST2700 PECI controller and Airoha
  EN7523 PCIe PERSTOUT reset lines.

- The clk core gains devm_clk_bulk_get_enable() as the mandatory counterpart to
  the existing optional variant, and exports
  devm_clk_hw_register_composite_pdata() for modular drivers.

- Tegra gets a proper EMC clock implementation for Tegra114, 48MHz pll_p_out1
  support needed for UEFI on Surface2, and the Xilinx clocking-wizard gains PLL
  charge pump/lock parameter programming during dynamic reconfiguration.

- Bug fixes for a NULL pointer dereference from uninitialized clk_init_data in
  the eswin driver, an IO remap leak in the MediaTek pllfh error path, inverted
  gate control for MT8135 devapc_ck, a missing OF node put in tegra124-emc on
  registration failure, a prepare reference leak in the palmas driver,
  unregistered PLLs on MT6735 probe failure, a missing kasprintf NULL check in
  pmc_atom, PCIe warm boot instability in the Airoha EN7523 driver, and a
  clocking-wizard clock difference detection bug.

- Various cleanups across tegra, st, and mvebu providers to stop misusing the
  consumer clock API, along with other minor fixes and improvements.

Signed-off-by: Brian Masney <bmasney@redhat.com>

T-HEAD device tree changes for v7.3

There is just one set of changes that adds the remaining IO expansion
chips on the TH1520-based LicheePi 4a board.

Memory controller drivers for v7.3

1. Few improvements: STM32 OMM, Tegra MC and JZ4780-nemc.

2. Refactor Tegra MC interconnect code to better represent actual
   hardware in multi-socket NUMA cases.

Samsung SoC drivers for v7.3

1. Fix very old leak of OF node in PMU driver.
2. Few cleanups and new bindings for ExynosAutov920 HSI0 syscon.

arm64: realtek: Device tree updates for v7.3

This pull request includes the following Realtek DT updates for
v7.3, please pull the following:

- Add GPIO node support for the RTD1625 SoC.
- Marc adds the missing the EL2 virtual timer interrupt.

Hi Stephen,

Here is a pull for the clk-eyeq driver adding support for Mobileye EyeQ7H
SoC. This pull request is a clean rebase of the version that was posted
during the last release cycle[1] and it was also posted as a patch series[2=
].

Alongside the changes to the clk-eyeq driver, this pull request also
provides the dt-bindings reviewed by Krzysztof Kozlowski, a change to the
reset-eyeq driver acked by Philipp Zabel and a small rework of the
fixed-factor clocks reviewed by Brian Masney.

[1]: https://lore.kernel.org/linux-clk/h_BA0KRKTGSANQ5Y4eqo6A@bootlin.com
[2]: https://patch.msgid.link/20260604-clk-eyeq7-v8-0-53f618174cf6@bootlin.=
com

T-HEAD clock changes for v7.3

There is just one change for this cycle. It allows COMPILE_TEST to
select the th1520 clk driver without having CONFIG_ARCH_THEAD enabled.

Hi, I created my riscv-confg branches to shepard defconfig patches
related to the riscv platforms that I maintain. The branches are in
linux-next. For 7.3, there is just a single patch for the WiFi driver on
the TH1520-based LPi4a.

Thanks,
Drew

RISC-V defconfig changes for 7.3

Enable PCA953X GPIO driver to properly probe Wifi pwrseq driver on
the Sipeed Lichee Pi 4a board which is based on the T-HEAD TH1520 SoC.

Chinese translation docs for 7.3

This is the Chinese translation subtree for 7.3. It includes
the following changes:
         - Add the some rust, module-signing docs translation
         - Fix/update couples Chinese translation

Above patches are tested by 'make htmldocs'

Signed-off-by: Alex Shi <alexs@kernel.org>

ARM: platform and feature deprecation for 7.3

This is attempt to bring the source code to the state that we discussed
back in 2024, regarding a timeline for deprecating and removing features,
in particular platforms and ABIs that are basically unused and are
also getting in the way of maintenance.

The earlier consensus was to remove these features in 2025 or 2026
after the corresponding LTS kernel release. I would suggest pushing
these to early 2027 now, see the patch description for details on
each one:

 - all board files that are still not converted to DT
   (except for OMAP1 and S3C)

 - iWMMXt FPU support

 - ARMv6/ARM1136r0 (not ARMv6K/ARM1136r1/ARM1176) along with the
   OMAP24xx and i.MX31 SoCs using it

 - Cortex-M3/M4/M7 based microcontroller support
   (stm32, imxrt, lpc18xx, samv7)

 - LSI Axxia platform

 - OABI, OABI-compat and NWFPE are confined to StrongARM based
   builds and scheduled for removal once StrongARM is gone.

 - DEPRECATED_PARAM_STRUCT, which was originally scheduled for
   removal in 2006 (!)

The ARCH_S3C64XX platform was scheduled for removal in 2024, but is
still actively used and will get removed once Mark Brown stops using
it. The OMAP1 platform has some remaining boardfiles and is expected to
get converted to DT.

Big-endian ARMv7 CPU support is not scheduled for removal but marked
as broken the same way as ARM64 has, to reduce the need for testing
unused configurations. This may be removed in the future if arm64 does
it as well. Note that big-endian ARMv5 support is can still be
enabled since the patches to revive little-endianm mode are still
fresh.

Regarding actually removing the code, I would suggest linux-7.4 (January
2027), removing both the board files and any drivers that have either
become newly unused or were missed in a previous machine removal series,
followed by removing support for then unused CPUs a release later.

Links to previous discussion:

  https://lore.kernel.org/linux-arm-kernel/20210109174357.GB1551@shell.armlinux.org.uk/
  https://lore.kernel.org/linux-arm-kernel/2831c5a6-cfbf-4fe0-b51c-0396e5b0aeb7@app.fastmail.com/
  https://lore.kernel.org/all/20241204102904.1863796-1-arnd@kernel.org/
  https://lore.kernel.org/all/20260701212353.2196041-1-arnd@kernel.org/

Hey folks,

with for the 7.2 merge window. Some of these patches date from 2024, so I
have elected not to squash or have the submitters of the clock changes etc
resubmit the original k230 series with their stuff squashed in so that
the history of the initial bring-up is retained, particularly given
Yangyu's detailed commit message in "riscv: dts: add initial canmv-k230
and k230-evb dts" that covers the testing done and decisions made about
supporting the standard Svpbmt extension or T-Head custom MAEE stuff etc.

Thanks,
Conor.

RISC-V devicetrees for v7.3 - early k230 material

Canaan:
Support for the k230 SoC that was too late for the 7.2 window.
Consists of some bindings not taken by the various subsystem
maintainers, the original dts patches from Yangyu back in 2024
and the more recent clock, reset and pinctrl additions from various
contributors.

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>

If it's possible to get this into to kvm/next before you disappear on holiday,
please pull a set of selftest changes that (lightly) touch all x86 tests that
use nested virtualization in any way.

If it's too late to get this into kvm/next, I'd say just ignore this pull
request and I'll include them in the normal 7.3 selftests pull requests.  I
don't think there will be any cross-branch conflicts to juggle?  But I've had
these sitting in -next for a few weeks now, and getting them into kvm/next
"now" would all but guarantee I won't end up having to play weird games with
selftests patches.

Thanks!

KVM selftests for 7.3, early edition

 - Automatically allocate a full page for L2 guest stacks on x86 instead of
   requiring test-specific L1 guest code to carve out a portion of the L1
   stack for L2 usage, and to ensure the L2 stack also adheres to the x86-64
   calling convention ABI.

 - Add a selftest to verify {Guest,Host}-Only behavior in x86's mediated PMU.