linux-block.git
6 weeks agodrm: nova-drm: add initial driver skeleton
Danilo Krummrich [Thu, 24 Apr 2025 16:02:50 +0000 (18:02 +0200)]
drm: nova-drm: add initial driver skeleton

Add the initial nova-drm driver skeleton.

nova-drm is connected to nova-core through the auxiliary bus and
implements the DRM parts of the nova driver stack.

For now, it implements the fundamental DRM abstractions, i.e. creates a
DRM device and registers it, exposing a three sample IOCTLs.

  DRM_IOCTL_NOVA_GETPARAM
    - provides the PCI bar size from the bar that maps the GPUs VRAM
      from nova-core

  DRM_IOCTL_NOVA_GEM_CREATE
    - creates a new dummy DRM GEM object and returns a handle

  DRM_IOCTL_NOVA_GEM_INFO
    - provides metadata for the DRM GEM object behind a given handle

I implemented a small userspace test suite [1] that utilizes this
interface.

Link: https://gitlab.freedesktop.org/dakr/drm-test
Reviewed-by: Maxime Ripard <mripard@kernel.org>
Acked-by: Dave Airlie <airlied@redhat.com>
Link: https://lore.kernel.org/r/20250424160452.8070-3-dakr@kernel.org
[ Kconfig: depend on DRM=y rather than just DRM. - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
6 weeks agogpu: nova-core: register auxiliary device for nova-drm
Danilo Krummrich [Thu, 24 Apr 2025 16:02:49 +0000 (18:02 +0200)]
gpu: nova-core: register auxiliary device for nova-drm

Register an auxiliary device for nova-drm.

For now always use zero for the auxiliary device's ID; we don't use it
yet anyways. However, once it lands, we should switch to XArray.

Acked-by: Dave Airlie <airlied@redhat.com>
Link: https://lore.kernel.org/r/20250424160452.8070-2-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
6 weeks agorust: devres: fix doctest build under `!CONFIG_PCI`
Miguel Ojeda [Sun, 11 May 2025 18:25:33 +0000 (20:25 +0200)]
rust: devres: fix doctest build under `!CONFIG_PCI`

The doctest requires `CONFIG_PCI`:

    error[E0432]: unresolved import `kernel::pci`
        --> rust/doctests_kernel_generated.rs:2689:44
         |
    2689 | use kernel::{device::Core, devres::Devres, pci};
         |                                            ^^^ no `pci` in the root
         |
    note: found an item that was configured out
        --> rust/kernel/lib.rs:96:9
    note: the item is gated here
        --> rust/kernel/lib.rs:95:1

Thus conditionally compile it (which still checks the syntax).

Fixes: f301cb978c06 ("rust: devres: implement Devres::access()")
Signed-off-by: Miguel Ojeda <ojeda@kernel.org>
Link: https://lore.kernel.org/r/20250511182533.1016163-1-ojeda@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
7 weeks agosamples: rust: pci: take advantage of Devres::access()
Danilo Krummrich [Mon, 28 Apr 2025 14:00:29 +0000 (16:00 +0200)]
samples: rust: pci: take advantage of Devres::access()

For the I/O operations executed from the probe() method, take advantage
of Devres::access(), avoiding the atomic check and RCU read lock required
otherwise entirely.

Reviewed-by: Alexandre Courbot <acourbot@nvidia.com>
Acked-by: Boqun Feng <boqun.feng@gmail.com>
Reviewed-by: Joel Fernandes <joelagnelf@nvidia.com>
Link: https://lore.kernel.org/r/20250428140137.468709-4-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
7 weeks agorust: devres: implement Devres::access()
Danilo Krummrich [Mon, 28 Apr 2025 14:00:28 +0000 (16:00 +0200)]
rust: devres: implement Devres::access()

Implement a direct accessor for the data stored within the Devres for
cases where we can prove that we own a reference to a Device<Bound>
(i.e. a bound device) of the same device that was used to create the
corresponding Devres container.

Usually, when accessing the data stored within a Devres container, it is
not clear whether the data has been revoked already due to the device
being unbound and, hence, we have to try whether the access is possible
and subsequently keep holding the RCU read lock for the duration of the
access.

However, when we can prove that we hold a reference to Device<Bound>
matching the device the Devres container has been created with, we can
guarantee that the device is not unbound for the duration of the
lifetime of the Device<Bound> reference and, hence, it is not possible
for the data within the Devres container to be revoked.

Therefore, in this case, we can bypass the atomic check and the RCU read
lock, which is a great optimization and simplification for drivers.

Reviewed-by: Christian Schrefl <chrisi.schrefl@gmail.com>
Reviewed-by: Alexandre Courbot <acourbot@nvidia.com>
Acked-by: Boqun Feng <boqun.feng@gmail.com>
Reviewed-by: Joel Fernandes <joelagnelf@nvidia.com>
Link: https://lore.kernel.org/r/20250428140137.468709-3-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
7 weeks agorust: revocable: implement Revocable::access()
Danilo Krummrich [Mon, 28 Apr 2025 14:00:27 +0000 (16:00 +0200)]
rust: revocable: implement Revocable::access()

Implement an unsafe direct accessor for the data stored within the
Revocable.

This is useful for cases where we can prove that the data stored within
the Revocable is not and cannot be revoked for the duration of the
lifetime of the returned reference.

Reviewed-by: Christian Schrefl <chrisi.schrefl@gmail.com>
Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Acked-by: Miguel Ojeda <ojeda@kernel.org>
Reviewed-by: Alexandre Courbot <acourbot@nvidia.com>
Acked-by: Boqun Feng <boqun.feng@gmail.com>
Reviewed-by: Joel Fernandes <joelagnelf@nvidia.com>
Link: https://lore.kernel.org/r/20250428140137.468709-2-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
8 weeks agorust: device: conditionally expect `dead_code` for `parent()`
Miguel Ojeda [Tue, 29 Apr 2025 21:06:29 +0000 (23:06 +0200)]
rust: device: conditionally expect `dead_code` for `parent()`

When `CONFIG_AUXILIARY_BUS` is disabled, `parent()` is still dead code:

    error: method `parent` is never used
      --> rust/kernel/device.rs:71:19
       |
    64 | impl<Ctx: DeviceContext> Device<Ctx> {
       | ------------------------------------ method in this implementation
    ...
    71 |     pub(crate) fn parent(&self) -> Option<&Self> {
       |                   ^^^^^^
       |
       = note: `-D dead-code` implied by `-D warnings`
       = help: to override `-D warnings` add `#[allow(dead_code)]`

Thus reintroduce the `expect`, but now as a conditional one. Do so as
`dead_code` since that is narrower.

An `allow` would also be possible, but Danilo wants to catch new users
in the future [1].

Link: https://lore.kernel.org/rust-for-linux/aBE8qQrpXOfru_K3@pollux/
Fixes: ce735e73dd59 ("rust: auxiliary: add auxiliary device / driver abstractions")
Signed-off-by: Miguel Ojeda <ojeda@kernel.org>
Link: https://lore.kernel.org/r/20250429210629.513521-1-ojeda@kernel.org
[ Adjust commit subject to "rust: device: conditionally expect
  `dead_code` for `parent()`". - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
8 weeks agoMAINTAINERS: add DRM Rust source files to DRM DRIVERS
Danilo Krummrich [Thu, 10 Apr 2025 23:55:27 +0000 (01:55 +0200)]
MAINTAINERS: add DRM Rust source files to DRM DRIVERS

Add the DRM Rust source files to the DRM DRIVERS maintainers entry.

Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Lyude Paul <lyude@redhat.com>
Link: https://lore.kernel.org/r/20250410235546.43736-9-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
8 weeks agorust: drm: gem: Add GEM object abstraction
Asahi Lina [Thu, 10 Apr 2025 23:55:26 +0000 (01:55 +0200)]
rust: drm: gem: Add GEM object abstraction

DRM GEM is the DRM memory management subsystem used by most modern
drivers; add a Rust abstraction for DRM GEM.

This includes the BaseObject trait, which contains operations shared by
all GEM object classes.

Signed-off-by: Asahi Lina <lina@asahilina.net>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Lyude Paul <lyude@redhat.com>
Link: https://lore.kernel.org/r/20250410235546.43736-8-dakr@kernel.org
[ Rework of GEM object abstractions
    * switch to the Opaque<T> type
    * fix (mutable) references to struct drm_gem_object (which in this
      context is UB)
    * drop all custom reference types in favor of AlwaysRefCounted
    * bunch of minor changes and simplifications (e.g. IntoGEMObject
      trait)
    * write and fix safety and invariant comments
    * remove necessity for and convert 'as' casts
    * original source archive: https://archive.is/dD5SL

  - Danilo ]
[ Fix missing CONFIG_DRM guards in rust/helpers/drm.c. - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: drm: file: Add File abstraction
Asahi Lina [Thu, 10 Apr 2025 23:55:25 +0000 (01:55 +0200)]
rust: drm: file: Add File abstraction

A DRM File is the DRM counterpart to a kernel file structure,
representing an open DRM file descriptor.

Add a Rust abstraction to allow drivers to implement their own File types
that implement the DriverFile trait.

Reviewed-by: Maxime Ripard <mripard@kernel.org>
Signed-off-by: Asahi Lina <lina@asahilina.net>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Lyude Paul <lyude@redhat.com>
Link: https://lore.kernel.org/r/20250410235546.43736-7-dakr@kernel.org
[ Rework of drm::File
    * switch to the Opaque<T> type
    * fix (mutable) references to struct drm_file (which in this context
      is UB)
    * restructure and rename functions to align with common kernel
      schemes
    * write and fix safety and invariant comments
    * remove necessity for and convert 'as' casts
    * original source archive: https://archive.is/GH8oy

  - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: drm: add DRM driver registration
Asahi Lina [Thu, 10 Apr 2025 23:55:24 +0000 (01:55 +0200)]
rust: drm: add DRM driver registration

Implement the DRM driver `Registration`.

The `Registration` structure is responsible to register and unregister a
DRM driver. It makes use of the `Devres` container in order to allow the
`Registration` to be owned by devres, such that it is automatically
dropped (and the DRM driver unregistered) once the parent device is
unbound.

Signed-off-by: Asahi Lina <lina@asahilina.net>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Lyude Paul <lyude@redhat.com>
Link: https://lore.kernel.org/r/20250410235546.43736-6-dakr@kernel.org
[ Rework of drm::Registration
    * move VTABLE to drm::Device to prevent use-after-free bugs; VTABLE
      needs to be bound to the lifetime of drm::Device, not the
      drm::Registration
    * combine new() and register() to get rid of the registered boolean
    * remove file_operations
    * move struct drm_device creation to drm::Device
    * introduce Devres
    * original source archive: https://archive.is/Pl9ys

  - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: drm: add device abstraction
Asahi Lina [Thu, 10 Apr 2025 23:55:23 +0000 (01:55 +0200)]
rust: drm: add device abstraction

Implement the abstraction for a `struct drm_device`.

A `drm::Device` creates a static const `struct drm_driver` filled with
the data from the `drm::Driver` trait implementation of the actual
driver creating the `drm::Device`.

Reviewed-by: Maxime Ripard <mripard@kernel.org>
Signed-off-by: Asahi Lina <lina@asahilina.net>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Lyude Paul <lyude@redhat.com>
Link: https://lore.kernel.org/r/20250410235546.43736-5-dakr@kernel.org
[ Rewrite of drm::Device
    * full rewrite of the drm::Device abstraction using the subclassing
      pattern
    * original source archive: http://archive.today/5NxBo

  - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: drm: add driver abstractions
Asahi Lina [Thu, 10 Apr 2025 23:55:22 +0000 (01:55 +0200)]
rust: drm: add driver abstractions

Implement the DRM driver abstractions.

The `Driver` trait provides the interface to the actual driver to fill
in the driver specific data, such as the `DriverInfo`, driver features
and IOCTLs.

Reviewed-by: Maxime Ripard <mripard@kernel.org>
Signed-off-by: Asahi Lina <lina@asahilina.net>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Lyude Paul <lyude@redhat.com>
Link: https://lore.kernel.org/r/20250410235546.43736-4-dakr@kernel.org
[ MISC changes
    * remove unnecessary DRM features; make remaining ones crate private
    * add #[expect(unused)] to avoid warnings
    * add sealed trait
    * remove shmem::Object references
    * original source archive: https://archive.is/Pl9ys

  - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: drm: ioctl: Add DRM ioctl abstraction
Asahi Lina [Thu, 10 Apr 2025 23:55:21 +0000 (01:55 +0200)]
rust: drm: ioctl: Add DRM ioctl abstraction

DRM drivers need to be able to declare which driver-specific ioctls they
support. Add an abstraction implementing the required types and a helper
macro to generate the ioctl definition inside the DRM driver.

Note that this macro is not usable until further bits of the abstraction
are in place (but it will not fail to compile on its own, if not called).

Signed-off-by: Asahi Lina <lina@asahilina.net>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Lyude Paul <lyude@redhat.com>
Link: https://lore.kernel.org/r/20250410235546.43736-3-dakr@kernel.org
[ MISC fixes
    * wrap raw_data in Opaque to avoid UB when creating a reference
    * fix IOCTL sample declaration
    * fix safety comment of IOCTL argument
    * original source archive: https://archive.is/LqHDQ

  - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agodrm: drv: implement __drm_dev_alloc()
Danilo Krummrich [Thu, 10 Apr 2025 23:55:20 +0000 (01:55 +0200)]
drm: drv: implement __drm_dev_alloc()

In the Rust DRM device abstraction we need to allocate a struct
drm_device.

Currently, there are two options, the deprecated drm_dev_alloc() (which
does not support subclassing) and devm_drm_dev_alloc(). The latter
supports subclassing, but also manages the initial reference through
devres for the parent device.

In Rust we want to conform with the subclassing pattern, but do not want
to get the initial reference managed for us, since Rust has its own,
idiomatic ways to properly deal with it.

There are two options to achieve this.

  1) Allocate the memory ourselves with a KBox.
  2) Implement __drm_dev_alloc(), which supports subclassing, but is
     unmanged.

While (1) would be possible, it would be cumbersome, since it would
require exporting drm_dev_init() and drmm_add_final_kfree().

Hence, go with option (2) and implement __drm_dev_alloc().

Reviewed-by: Maxime Ripard <mripard@kernel.org>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Lyude Paul <lyude@redhat.com>
Link: https://lore.kernel.org/r/20250410235546.43736-2-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agosamples: rust: convert PCI rust sample driver to use try_access_with()
Alexandre Courbot [Fri, 11 Apr 2025 12:09:39 +0000 (21:09 +0900)]
samples: rust: convert PCI rust sample driver to use try_access_with()

This method limits the scope of the revocable guard and is considered
safer to use for most cases, so let's showcase it here.

Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Acked-by: Miguel Ojeda <ojeda@kernel.org>
Link: https://lore.kernel.org/r/20250411-try_with-v4-2-f470ac79e2e2@nvidia.com
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust/revocable: add try_access_with() convenience method
Alexandre Courbot [Fri, 11 Apr 2025 12:09:38 +0000 (21:09 +0900)]
rust/revocable: add try_access_with() convenience method

Revocable::try_access() returns a guard through which the wrapped object
can be accessed. Code that can sleep is not allowed while the guard is
held; thus, it is common for the caller to explicitly drop it before
running sleepable code, e.g:

    let b = bar.try_access()?;
    let reg = b.readl(...);

    // Don't forget this or things could go wrong!
    drop(b);

    something_that_might_sleep();

    let b = bar.try_access()?;
    let reg2 = b.readl(...);

This is arguably error-prone. try_access_with() provides an arguably
safer alternative, by taking a closure that is run while the guard is
held, and by dropping the guard automatically after the closure
completes. This way, code can be organized more clearly around the
critical sections and the risk of forgetting to release the guard when
needed is considerably reduced:

    let reg = bar.try_access_with(|b| b.readl(...))?;

    something_that_might_sleep();

    let reg2 = bar.try_access_with(|b| b.readl(...))?;

The closure can return nothing, or any value including a Result which is
then wrapped inside the Option returned by try_access_with. Error
management is driver-specific, so users are encouraged to create their
own macros that map and flatten the returned values to something
appropriate for the code they are working on.

Suggested-by: Danilo Krummrich <dakr@kernel.org>
Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Reviewed-by: Joel Fernandes <joelagnelf@nvidia.com>
Acked-by: Miguel Ojeda <ojeda@kernel.org>
Link: https://lore.kernel.org/r/20250411-try_with-v4-1-f470ac79e2e2@nvidia.com
[ Link `None`, `Some`, `Option` in doc-comment. - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agosamples: rust: add Rust auxiliary driver sample
Danilo Krummrich [Mon, 14 Apr 2025 13:18:08 +0000 (15:18 +0200)]
samples: rust: add Rust auxiliary driver sample

Add a sample Rust auxiliary driver based on a PCI driver for QEMU's
"pci-testdev" device.

The PCI driver only registers an auxiliary device, in order to make the
corresponding auxiliary driver probe.

Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://lore.kernel.org/r/20250414131934.28418-6-dakr@kernel.org
[ Use `ok_or()` when accessing auxiliary::Device::parent(). - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: auxiliary: add auxiliary registration
Danilo Krummrich [Mon, 14 Apr 2025 13:18:07 +0000 (15:18 +0200)]
rust: auxiliary: add auxiliary registration

Implement the `auxiliary::Registration` type, which provides an API to
create and register new auxiliary devices in the system.

Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Link: https://lore.kernel.org/r/20250414131934.28418-5-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: auxiliary: add auxiliary device / driver abstractions
Danilo Krummrich [Mon, 14 Apr 2025 13:18:06 +0000 (15:18 +0200)]
rust: auxiliary: add auxiliary device / driver abstractions

Implement the basic auxiliary abstractions required to implement a
driver matching an auxiliary device.

The design and implementation is analogous to PCI and platform and is
based on the generic device / driver abstractions.

Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Link: https://lore.kernel.org/r/20250414131934.28418-4-dakr@kernel.org
[ Fix typos, `let _ =` => `drop()`, use `kernel::ffi`. - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: device: implement Device::parent()
Danilo Krummrich [Mon, 14 Apr 2025 13:18:05 +0000 (15:18 +0200)]
rust: device: implement Device::parent()

Device::parent() returns a reference to the device' parent device, if
any.

Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Link: https://lore.kernel.org/r/20250414131934.28418-3-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: types: add `Opaque::zeroed`
Danilo Krummrich [Mon, 14 Apr 2025 13:18:04 +0000 (15:18 +0200)]
rust: types: add `Opaque::zeroed`

Analogous to `Opaque::uninit` add `Opaque::zeroed`, which sets the
corresponding memory to zero. In contrast to `Opaque::uninit`, the
corresponding value, depending on its type, may be initialized.

Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Acked-by: Miguel Ojeda <ojeda@kernel.org>
Link: https://lore.kernel.org/r/20250414131934.28418-2-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: platform: impl TryFrom<&Device> for &platform::Device
Danilo Krummrich [Fri, 21 Mar 2025 21:47:55 +0000 (22:47 +0100)]
rust: platform: impl TryFrom<&Device> for &platform::Device

Implement TryFrom<&device::Device> for &Device.

This allows us to get a &platform::Device from a generic &Device in a safe
way; the conversion fails if the device' bus type does not match with
the platform bus type.

Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Link: https://lore.kernel.org/r/20250321214826.140946-4-dakr@kernel.org
[ Support device context types, use dev_is_platform() helper. - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: pci: impl TryFrom<&Device> for &pci::Device
Danilo Krummrich [Fri, 21 Mar 2025 21:47:54 +0000 (22:47 +0100)]
rust: pci: impl TryFrom<&Device> for &pci::Device

Implement TryFrom<&device::Device> for &Device.

This allows us to get a &pci::Device from a generic &Device in a safe
way; the conversion fails if the device' bus type does not match with
the PCI bus type.

Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Link: https://lore.kernel.org/r/20250321214826.140946-3-dakr@kernel.org
[ Support device context types, use dev_is_pci() helper. - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agoMerge tag 'topic/device-context-2025-04-17' into nova-next
Danilo Krummrich [Sat, 19 Apr 2025 08:10:40 +0000 (10:10 +0200)]
Merge tag 'topic/device-context-2025-04-17' into nova-next

Introduce "Bound" device context

Introduce the "Bound" device context, such that it can be ensured to only
ever pass a bound device to APIs that require this precondition. [1]

This tag exists to share the patches from [1] between multiple trees.

[1] https://lore.kernel.org/lkml/20250413173758.12068-1-dakr@kernel.org/

2 months agorust: dma: require a bound device
Danilo Krummrich [Sun, 13 Apr 2025 17:37:04 +0000 (19:37 +0200)]
rust: dma: require a bound device

Require the Bound device context to be able to create new
dma::CoherentAllocation instances.

DMA memory allocations are only valid to be created for bound devices.

Link: https://lore.kernel.org/r/20250413173758.12068-10-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: devres: require a bound device
Danilo Krummrich [Sun, 13 Apr 2025 17:37:03 +0000 (19:37 +0200)]
rust: devres: require a bound device

Require the Bound device context to be able to a new Devres container.
This ensures that we can't register devres callbacks for unbound
devices.

Link: https://lore.kernel.org/r/20250413173758.12068-9-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: pci: move iomap_region() to impl Device<Bound>
Danilo Krummrich [Sun, 13 Apr 2025 17:37:02 +0000 (19:37 +0200)]
rust: pci: move iomap_region() to impl Device<Bound>

Require the Bound device context to be able to call iomap_region() and
iomap_region_sized(). Creating I/O mapping requires the device to be
bound.

Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Link: https://lore.kernel.org/r/20250413173758.12068-8-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: device: implement Bound device context
Danilo Krummrich [Sun, 13 Apr 2025 17:37:01 +0000 (19:37 +0200)]
rust: device: implement Bound device context

The Bound device context indicates that a device is bound to a driver.
It must be used for APIs that require the device to be bound, such as
Devres or dma::CoherentAllocation.

Implement Bound and add the corresponding Deref hierarchy, as well as the
corresponding ARef conversion for this device context.

Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Link: https://lore.kernel.org/r/20250413173758.12068-7-dakr@kernel.org
[ Add missing `::` prefix in macros. - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: pci: preserve device context in AsRef
Danilo Krummrich [Sun, 13 Apr 2025 17:37:00 +0000 (19:37 +0200)]
rust: pci: preserve device context in AsRef

Since device::Device has a generic over its context, preserve this
device context in AsRef.

For instance, when calling pci::Device<Core> the new AsRef implementation
returns device::Device<Core>.

Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Link: https://lore.kernel.org/r/20250413173758.12068-6-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: platform: preserve device context in AsRef
Danilo Krummrich [Sun, 13 Apr 2025 17:36:59 +0000 (19:36 +0200)]
rust: platform: preserve device context in AsRef

Since device::Device has a generic over its context, preserve this
device context in AsRef.

For instance, when calling platform::Device<Core> the new AsRef
implementation returns device::Device<Core>.

Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Link: https://lore.kernel.org/r/20250413173758.12068-5-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: device: implement device context for Device
Danilo Krummrich [Sun, 13 Apr 2025 17:36:58 +0000 (19:36 +0200)]
rust: device: implement device context for Device

Analogous to bus specific device, implement the DeviceContext generic
for generic devices.

This is used for APIs that work with generic devices (such as Devres) to
evaluate the device's context.

Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Link: https://lore.kernel.org/r/20250413173758.12068-4-dakr@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: device: implement impl_device_context_into_aref!
Danilo Krummrich [Sun, 13 Apr 2025 17:36:57 +0000 (19:36 +0200)]
rust: device: implement impl_device_context_into_aref!

Implement a macro to implement all From conversions of a certain device
to ARef<Device>.

This avoids unnecessary boiler plate code for every device
implementation.

Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Link: https://lore.kernel.org/r/20250413173758.12068-3-dakr@kernel.org
[ Add missing `::` prefix in macros. - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agorust: device: implement impl_device_context_deref!
Danilo Krummrich [Sun, 13 Apr 2025 17:36:56 +0000 (19:36 +0200)]
rust: device: implement impl_device_context_deref!

The Deref hierarchy for device context generics is the same for every
(bus specific) device.

Implement those with a generic macro to avoid duplicated boiler plate
code and ensure the correct Deref hierarchy for every device
implementation.

Co-developed-by: Benno Lossin <benno.lossin@proton.me>
Signed-off-by: Benno Lossin <benno.lossin@proton.me>
Reviewed-by: Christian Schrefl <chrisi.schrefl@gmail.com>
Link: https://lore.kernel.org/r/20250413173758.12068-2-dakr@kernel.org
[ Add missing `::` prefix in macros. - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agogpu: nova-core: remove completed Vec extentions from task list
Andrew Ballance [Sun, 16 Mar 2025 11:16:44 +0000 (06:16 -0500)]
gpu: nova-core: remove completed Vec extentions from task list

The requested Vec methods have been implemented thus, remove
the completed item from the nova task list.

Link: https://lore.kernel.org/r/20250316111644.154602-4-andrewjballance@gmail.com
Signed-off-by: Andrew Ballance <andrewjballance@gmail.com>
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2 months agoLinux 6.15-rc1 v6.15-rc1
Linus Torvalds [Sun, 6 Apr 2025 20:11:33 +0000 (13:11 -0700)]
Linux 6.15-rc1

2 months agotools/include: make uapi/linux/types.h usable from assembly
Thomas Weißschuh [Wed, 2 Apr 2025 20:21:57 +0000 (21:21 +0100)]
tools/include: make uapi/linux/types.h usable from assembly

The "real" linux/types.h UAPI header gracefully degrades to a NOOP when
included from assembly code.

Mirror this behaviour in the tools/ variant.

Test for __ASSEMBLER__ over __ASSEMBLY__ as the former is provided by the
toolchain automatically.

Reported-by: Mark Brown <broonie@kernel.org>
Closes: https://lore.kernel.org/lkml/af553c62-ca2f-4956-932c-dd6e3a126f58@sirena.org.uk/
Fixes: c9fbaa879508 ("selftests: vDSO: parse_vdso: Use UAPI headers instead of libc headers")
Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
Link: https://patch.msgid.link/20250321-uapi-consistency-v1-1-439070118dc0@linutronix.de
Signed-off-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2 months agoMerge tag 'turbostat-2025.05.06' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Sun, 6 Apr 2025 19:32:43 +0000 (12:32 -0700)]
Merge tag 'turbostat-2025.05.06' of git://git./linux/kernel/git/lenb/linux

Pull turbostat updates from Len Brown:

 - support up to 8192 processors

 - add cpuidle governor debug telemetry, disabled by default

 - update default output to exclude cpuidle invocation counts

 - bug fixes

* tag 'turbostat-2025.05.06' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux:
  tools/power turbostat: v2025.05.06
  tools/power turbostat: disable "cpuidle" invocation counters, by default
  tools/power turbostat: re-factor sysfs code
  tools/power turbostat: Restore GFX sysfs fflush() call
  tools/power turbostat: Document GNR UncMHz domain convention
  tools/power turbostat: report CoreThr per measurement interval
  tools/power turbostat: Increase CPU_SUBSET_MAXCPUS to 8192
  tools/power turbostat: Add idle governor statistics reporting
  tools/power turbostat: Fix names matching
  tools/power turbostat: Allow Zero return value for some RAPL registers
  tools/power turbostat: Clustered Uncore MHz counters should honor show/hide options

2 months agoMerge tag 'soundwire-6.15-rc1-fixes' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 6 Apr 2025 19:04:53 +0000 (12:04 -0700)]
Merge tag 'soundwire-6.15-rc1-fixes' of git://git./linux/kernel/git/vkoul/soundwire

Pull soundwire fix from Vinod Koul:

 - add missing config symbol CONFIG_SND_HDA_EXT_CORE required for asoc
   driver CONFIG_SND_SOF_SOF_HDA_SDW_BPT

* tag 'soundwire-6.15-rc1-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/soundwire:
  ASoC: SOF: Intel: Let SND_SOF_SOF_HDA_SDW_BPT select SND_HDA_EXT_CORE

2 months agotools/power turbostat: v2025.05.06
Len Brown [Sun, 6 Apr 2025 18:49:20 +0000 (14:49 -0400)]
tools/power turbostat: v2025.05.06

Support up to 8192 processors
Add cpuidle governor debug telemetry, disabled by default
Update default output to exclude cpuidle invocation counts
Bug fixes

Signed-off-by: Len Brown <len.brown@intel.com>
2 months agotools/power turbostat: disable "cpuidle" invocation counters, by default
Len Brown [Sun, 6 Apr 2025 18:29:57 +0000 (14:29 -0400)]
tools/power turbostat: disable "cpuidle" invocation counters, by default

Create "pct_idle" counter group, the sofware notion of residency
so it can now be singled out, independent of other counter groups.

Create "cpuidle" group, the cpuidle invocation counts.
Disable "cpuidle", by default.

Create "swidle" = "cpuidle" + "pct_idle".
Undocument "sysfs", the old name for "swidle", but keep it working
for backwards compatibilty.

Create "hwidle", all the HW idle counters

Modify "idle", enabled by default
"idle" = "hwidle" + "pct_idle" (and now excludes "cpuidle")

Signed-off-by: Len Brown <len.brown@intel.com>
2 months agoMerge tag 'perf-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Sun, 6 Apr 2025 17:48:12 +0000 (10:48 -0700)]
Merge tag 'perf-urgent-2025-04-06' of git://git./linux/kernel/git/tip/tip

Pull perf event fix from Ingo Molnar:
 "Fix a perf events time accounting bug"

* tag 'perf-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf/core: Fix child_total_time_enabled accounting bug at task exit

2 months agoMerge tag 'sched-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 6 Apr 2025 17:44:58 +0000 (10:44 -0700)]
Merge tag 'sched-urgent-2025-04-06' of git://git./linux/kernel/git/tip/tip

Pull scheduler fixes from Ingo Molnar:

 - Fix a nonsensical Kconfig combination

 - Remove an unnecessary rseq-notification

* tag 'sched-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  rseq: Eliminate useless task_work on execve
  sched/isolation: Make CONFIG_CPU_ISOLATION depend on CONFIG_SMP

2 months agoDisable SLUB_TINY for build testing
Linus Torvalds [Sun, 6 Apr 2025 17:00:04 +0000 (10:00 -0700)]
Disable SLUB_TINY for build testing

... and don't error out so hard on missing module descriptions.

Before commit 6c6c1fc09de3 ("modpost: require a MODULE_DESCRIPTION()")
we used to warn about missing module descriptions, but only when
building with extra warnigns (ie 'W=1').

After that commit the warning became an unconditional hard error.

And it turns out not all modules have been converted despite the claims
to the contrary.  As reported by Damian Tometzki, the slub KUnit test
didn't have a module description, and apparently nobody ever really
noticed.

The reason nobody noticed seems to be that the slub KUnit tests get
disabled by SLUB_TINY, which also ends up disabling a lot of other code,
both in tests and in slub itself.  And so anybody doing full build tests
didn't actually see this failre.

So let's disable SLUB_TINY for build-only tests, since it clearly ends
up limiting build coverage.  Also turn the missing module descriptions
error back into a warning, but let's keep it around for non-'W=1'
builds.

Reported-by: Damian Tometzki <damian@riscv-rocks.de>
Link: https://lore.kernel.org/all/01070196099fd059-e8463438-7b1b-4ec8-816d-173874be9966-000000@eu-central-1.amazonses.com/
Cc: Masahiro Yamada <masahiroy@kernel.org>
Cc: Jeff Johnson <jeff.johnson@oss.qualcomm.com>
Fixes: 6c6c1fc09de3 ("modpost: require a MODULE_DESCRIPTION()")
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2 months agotools/power turbostat: re-factor sysfs code
Len Brown [Sun, 6 Apr 2025 16:53:18 +0000 (12:53 -0400)]
tools/power turbostat: re-factor sysfs code

Probe cpuidle "sysfs" residency and counts separately,
since soon we will make one disabled on, and the
other disabled off.

Clarify that some BIC (build-in-counters) are actually "groups".
since we're about to re-name some of those groups.

no functional change.

Signed-off-by: Len Brown <len.brown@intel.com>
2 months agotools/power turbostat: Restore GFX sysfs fflush() call
Zhang Rui [Wed, 19 Mar 2025 00:53:07 +0000 (08:53 +0800)]
tools/power turbostat: Restore GFX sysfs fflush() call

Do fflush() to discard the buffered data, before each read of the
graphics sysfs knobs.

Fixes: ba99a4fc8c24 ("tools/power turbostat: Remove unnecessary fflush() call")
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
2 months agotools/power turbostat: Document GNR UncMHz domain convention
Len Brown [Sun, 6 Apr 2025 16:23:22 +0000 (12:23 -0400)]
tools/power turbostat: Document GNR UncMHz domain convention

Document that on Intel Granite Rapids Systems,
Uncore domains 0-2 are CPU domains, and
uncore domains 3-4 are IO domains.

Signed-off-by: Len Brown <len.brown@intel.com>
2 months agotools/power turbostat: report CoreThr per measurement interval
Len Brown [Sun, 6 Apr 2025 15:18:39 +0000 (11:18 -0400)]
tools/power turbostat: report CoreThr per measurement interval

The CoreThr column displays total thermal throttling events
since boot time.

Change it to report events during the measurement interval.

This is more useful for showing a user the current conditions.
Total events since boot time are still available to the user via
/sys/devices/system/cpu/cpu*/thermal_throttle/*

Document CoreThr on turbostat.8

Fixes: eae97e053fe30 ("turbostat: Support thermal throttle count print")
Reported-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Cc: Chen Yu <yu.c.chen@intel.com>
2 months agotools/power turbostat: Increase CPU_SUBSET_MAXCPUS to 8192
Justin Ernst [Wed, 19 Mar 2025 20:27:31 +0000 (15:27 -0500)]
tools/power turbostat: Increase CPU_SUBSET_MAXCPUS to 8192

On systems with >= 1024 cpus (in my case 1152), turbostat fails with the error output:
"turbostat: /sys/fs/cgroup/cpuset.cpus.effective: cpu str malformat 0-1151"

A similar error appears with the use of turbostat --cpu when the inputted cpu
range contains a cpu number >= 1024:
# turbostat -c 1100-1151
"--cpu 1100-1151" malformed
...

Both errors are caused by parse_cpu_str() reaching its limit of CPU_SUBSET_MAXCPUS.

It's a good idea to limit the maximum cpu number being parsed, but 1024 is too low.
For a small increase in compute and allocated memory, increasing CPU_SUBSET_MAXCPUS
brings support for parsing cpu numbers >= 1024.

Increase CPU_SUBSET_MAXCPUS to 8192, a common setting for CONFIG_NR_CPUS on x86_64.

Signed-off-by: Justin Ernst <justin.ernst@hpe.com>
Signed-off-by: Len Brown <len.brown@intel.com>
2 months agoMerge tag 'timers-cleanups-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 6 Apr 2025 15:35:37 +0000 (08:35 -0700)]
Merge tag 'timers-cleanups-2025-04-06' of git://git./linux/kernel/git/tip/tip

Pull timer cleanups from Thomas Gleixner:
 "A set of final cleanups for the timer subsystem:

   - Convert all del_timer[_sync]() instances over to the new
     timer_delete[_sync]() API and remove the legacy wrappers.

     Conversion was done with coccinelle plus some manual fixups as
     coccinelle chokes on scoped_guard().

   - The final cleanup of the hrtimer_init() to hrtimer_setup()
     conversion.

     This has been delayed to the end of the merge window, so that all
     patches which have been merged through other trees are in mainline
     and all new users are catched.

  Doing this right before rc1 ensures that new code which is merged post
  rc1 is not introducing new instances of the original functionality"

* tag 'timers-cleanups-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  tracing/timers: Rename the hrtimer_init event to hrtimer_setup
  hrtimers: Rename debug_init_on_stack() to debug_setup_on_stack()
  hrtimers: Rename debug_init() to debug_setup()
  hrtimers: Rename __hrtimer_init_sleeper() to __hrtimer_setup_sleeper()
  hrtimers: Remove unnecessary NULL check in hrtimer_start_range_ns()
  hrtimers: Make callback function pointer private
  hrtimers: Merge __hrtimer_init() into __hrtimer_setup()
  hrtimers: Switch to use __htimer_setup()
  hrtimers: Delete hrtimer_init()
  treewide: Convert new and leftover hrtimer_init() users
  treewide: Switch/rename to timer_delete[_sync]()

2 months agoMerge tag 'irq-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Sun, 6 Apr 2025 15:17:43 +0000 (08:17 -0700)]
Merge tag 'irq-urgent-2025-04-06' of git://git./linux/kernel/git/tip/tip

Pull more irq updates from Thomas Gleixner:
 "A set of updates for the interrupt subsystem:

   - A treewide cleanup for the irq_domain code, which makes the naming
     consistent and gets rid of the original oddity of naming domains
     'host'.

     This is a trivial mechanical change and is done late to ensure that
     all instances have been catched and new code merged post rc1 wont
     reintroduce new instances.

   - A trivial consistency fix in the migration code

     The recent introduction of irq_force_complete_move() in the core
     code, causes a problem for the nostalgia crowd who maintains ia64
     out of tree.

     The code assumes that hierarchical interrupt domains are enabled
     and dereferences irq_data::parent_data unconditionally. That works
     in mainline because both architectures which enable that code have
     hierarchical domains enabled. Though it breaks the ia64 build,
     which enables the functionality, but does not have hierarchical
     domains.

     While it's not really a problem for mainline today, this
     unconditional dereference is inconsistent and trivially fixable by
     using the existing helper function irqd_get_parent_data(), which
     has the appropriate #ifdeffery in place"

* tag 'irq-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  genirq/migration: Use irqd_get_parent_data() in irq_force_complete_move()
  irqdomain: Stop using 'host' for domain
  irqdomain: Rename irq_get_default_host() to irq_get_default_domain()
  irqdomain: Rename irq_set_default_host() to irq_set_default_domain()

2 months agoMerge tag 'timers-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 6 Apr 2025 15:13:16 +0000 (08:13 -0700)]
Merge tag 'timers-urgent-2025-04-06' of git://git./linux/kernel/git/tip/tip

Pull timer fix from Thomas Gleixner:
 "A revert to fix a adjtimex() regression:

  The recent change to prevent that time goes backwards for the coarse
  time getters due to immediate multiplier adjustments via adjtimex(),
  changed the way how the timekeeping core treats that.

  That change result in a regression on the adjtimex() side, which is
  user space visible:

   1) The forwarding of the base time moves the update out of the
      original period and establishes a new one. That's changing the
      behaviour of the [PF]LL control, which user space expects to be
      applied periodically.

   2) The clearing of the accumulated NTP error due to #1, changes the
      behaviour as well.

  An attempt to delay the multiplier/frequency update to the next tick
  did not solve the problem as userspace expects that the multiplier or
  frequency updates are in effect, when the syscall returns.

  There is a different solution for the coarse time problem available,
  so revert the offending commit to restore the existing adjtimex()
  behaviour"

* tag 'timers-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  Revert "timekeeping: Fix possible inconsistencies in _COARSE clockids"

2 months agoMerge tag 'sh-for-v6.15-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/glaubi...
Linus Torvalds [Sun, 6 Apr 2025 15:10:45 +0000 (08:10 -0700)]
Merge tag 'sh-for-v6.15-tag1' of git://git./linux/kernel/git/glaubitz/sh-linux

Pull sh updates from John Paul Adrian Glaubitz:
 "One important fix and one small configuration update.

  The first patch by Artur Rojek fixes an issue with the J2 firmware
  loader not being able to find the location of the device tree blob due
  to insufficient alignment of the .bss section which rendered J2 boards
  unbootable.

  The second patch by Johan Korsnes updates the defconfigs on sh to drop
  the CONFIG_NET_CLS_TCINDEX configuration option which became obsolete
  after 8c710f75256b ("net/sched: Retire tcindex classifier").

  Summary:

   - sh: defconfig: Drop obsolete CONFIG_NET_CLS_TCINDEX

   - sh: Align .bss section padding to 8-byte boundary"

* tag 'sh-for-v6.15-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/glaubitz/sh-linux:
  sh: defconfig: Drop obsolete CONFIG_NET_CLS_TCINDEX
  sh: Align .bss section padding to 8-byte boundary

2 months agoMerge tag 'kbuild-v6.15' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy...
Linus Torvalds [Sat, 5 Apr 2025 22:46:50 +0000 (15:46 -0700)]
Merge tag 'kbuild-v6.15' of git://git./linux/kernel/git/masahiroy/linux-kbuild

Pull Kbuild updates from Masahiro Yamada:

 - Improve performance in gendwarfksyms

 - Remove deprecated EXTRA_*FLAGS and KBUILD_ENABLE_EXTRA_GCC_CHECKS

 - Support CONFIG_HEADERS_INSTALL for ARCH=um

 - Use more relative paths to sources files for better reproducibility

 - Support the loong64 Debian architecture

 - Add Kbuild bash completion

 - Introduce intermediate vmlinux.unstripped for architectures that need
   static relocations to be stripped from the final vmlinux

 - Fix versioning in Debian packages for -rc releases

 - Treat missing MODULE_DESCRIPTION() as an error

 - Convert Nios2 Makefiles to use the generic rule for built-in DTB

 - Add debuginfo support to the RPM package

* tag 'kbuild-v6.15' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild: (40 commits)
  kbuild: rpm-pkg: build a debuginfo RPM
  kconfig: merge_config: use an empty file as initfile
  nios2: migrate to the generic rule for built-in DTB
  rust: kbuild: skip `--remap-path-prefix` for `rustdoc`
  kbuild: pacman-pkg: hardcode module installation path
  kbuild: deb-pkg: don't set KBUILD_BUILD_VERSION unconditionally
  modpost: require a MODULE_DESCRIPTION()
  kbuild: make all file references relative to source root
  x86: drop unnecessary prefix map configuration
  kbuild: deb-pkg: add comment about future removal of KDEB_COMPRESS
  kbuild: Add a help message for "headers"
  kbuild: deb-pkg: remove "version" variable in mkdebian
  kbuild: deb-pkg: fix versioning for -rc releases
  Documentation/kbuild: Fix indentation in modules.rst example
  x86: Get rid of Makefile.postlink
  kbuild: Create intermediate vmlinux build with relocations preserved
  kbuild: Introduce Kconfig symbol for linking vmlinux with relocations
  kbuild: link-vmlinux.sh: Make output file name configurable
  kbuild: do not generate .tmp_vmlinux*.map when CONFIG_VMLINUX_MAP=y
  Revert "kheaders: Ignore silly-rename files"
  ...

2 months agoMerge tag 'drm-next-2025-04-05' of https://gitlab.freedesktop.org/drm/kernel
Linus Torvalds [Sat, 5 Apr 2025 22:35:11 +0000 (15:35 -0700)]
Merge tag 'drm-next-2025-04-05' of https://gitlab.freedesktop.org/drm/kernel

Pull drm fixes from Dave Airlie:
 "Weekly fixes, mostly from the end of last week, this week was very
  quiet, maybe you scared everyone away. It's mostly amdgpu, and xe,
  with some i915, adp and bridge bits, since I think this is overly
  quiet I'd expect rc2 to be a bit more lively.

  bridge:
   - tda998x: Select CONFIG_DRM_KMS_HELPER

  amdgpu:
   - Guard against potential division by 0 in fan code
   - Zero RPM support for SMU 14.0.2
   - Properly handle SI and CIK support being disabled
   - PSR fixes
   - DML2 fixes
   - DP Link training fix
   - Vblank fixes
   - RAS fixes
   - Partitioning fix
   - SDMA fix
   - SMU 13.0.x fixes
   - Rom fetching fix
   - MES fixes
   - Queue reset fix

  xe:
   - Fix NULL pointer dereference on error path
   - Add missing HW workaround for BMG
   - Fix survivability mode not triggering
   - Fix build warning when DRM_FBDEV_EMULATION is not set

  i915:
   - Bounds check for scalers in DSC prefill latency computation
   - Fix build by adding a missing include

  adp:
   - Fix error handling in plane setup"

  # -----BEGIN PGP SIGNATURE-----

* tag 'drm-next-2025-04-05' of https://gitlab.freedesktop.org/drm/kernel: (34 commits)
  drm/i2c: tda998x: select CONFIG_DRM_KMS_HELPER
  drm/amdgpu/gfx12: fix num_mec
  drm/amdgpu/gfx11: fix num_mec
  drm/amd/pm: Add gpu_metrics_v1_8
  drm/amdgpu: Prefer shadow rom when available
  drm/amd/pm: Update smu metrics table for smu_v13_0_6
  drm/amd/pm: Remove host limit metrics support
  Remove unnecessary firmware version check for gc v9_4_2
  drm/amdgpu: stop unmapping MQD for kernel queues v3
  Revert "drm/amdgpu/sdma_v4_4_2: update VM flush implementation for SDMA"
  drm/amdgpu: Parse all deferred errors with UMC aca handle
  drm/amdgpu: Update ta ras block
  drm/amdgpu: Add NPS2 to DPX compatible mode
  drm/amdgpu: Use correct gfx deferred error count
  drm/amd/display: Actually do immediate vblank disable
  drm/amd/display: prevent hang on link training fail
  Revert "drm/amd/display: dml2 soc dscclk use DPM table clk setting"
  drm/amd/display: Increase vblank offdelay for PSR panels
  drm/amd: Handle being compiled without SI or CIK support better
  drm/amd/pm: Add zero RPM enabled OD setting support for SMU14.0.2
  ...

2 months agokbuild: rpm-pkg: build a debuginfo RPM
Uday Shankar [Mon, 31 Mar 2025 22:46:32 +0000 (16:46 -0600)]
kbuild: rpm-pkg: build a debuginfo RPM

The rpm-pkg make target currently suffers from a few issues related to
debuginfo:
1. debuginfo for things built into the kernel (vmlinux) is not available
   in any RPM produced by make rpm-pkg. This makes using tools like
   systemtap against a make rpm-pkg kernel impossible.
2. debug source for the kernel is not available. This means that
   commands like 'disas /s' in gdb, which display source intermixed with
   assembly, can only print file names/line numbers which then must be
   painstakingly resolved to actual source in a separate editor.
3. debuginfo for modules is available, but it remains bundled with the
   .ko files that contain module code, in the main kernel RPM. This is a
   waste of space for users who do not need to debug the kernel (i.e.
   most users).

Address all of these issues by additionally building a debuginfo RPM
when the kernel configuration allows for it, in line with standard
patterns followed by RPM distributors. With these changes:
1. systemtap now works (when these changes are backported to 6.11, since
   systemtap lags a bit behind in compatibility), as verified by the
   following simple test script:

   # stap -e 'probe kernel.function("do_sys_open").call { printf("%s\n", $$parms); }'
   dfd=0xffffffffffffff9c filename=0x7fe18800b160 flags=0x88800 mode=0x0
   ...

2. disas /s works correctly in gdb, with source and disassembly
   interspersed:

   # gdb vmlinux --batch -ex 'disas /s blk_op_str'
   Dump of assembler code for function blk_op_str:
   block/blk-core.c:
   125     {
      0xffffffff814c8740 <+0>:     endbr64

   127
   128             if (op < ARRAY_SIZE(blk_op_name) && blk_op_name[op])
      0xffffffff814c8744 <+4>:     mov    $0xffffffff824a7378,%rax
      0xffffffff814c874b <+11>:    cmp    $0x23,%edi
      0xffffffff814c874e <+14>:    ja     0xffffffff814c8768 <blk_op_str+40>
      0xffffffff814c8750 <+16>:    mov    %edi,%edi

   126             const char *op_str = "UNKNOWN";
      0xffffffff814c8752 <+18>:    mov    $0xffffffff824a7378,%rdx

   127
   128             if (op < ARRAY_SIZE(blk_op_name) && blk_op_name[op])
      0xffffffff814c8759 <+25>:    mov    -0x7dfa0160(,%rdi,8),%rax

   126             const char *op_str = "UNKNOWN";
      0xffffffff814c8761 <+33>:    test   %rax,%rax
      0xffffffff814c8764 <+36>:    cmove  %rdx,%rax

   129                     op_str = blk_op_name[op];
   130
   131             return op_str;
   132     }
      0xffffffff814c8768 <+40>:    jmp    0xffffffff81d01360 <__x86_return_thunk>
   End of assembler dump.

3. The size of the main kernel package goes down substantially,
   especially if many modules are built (quite typical). Here is a
   comparison of installed size of the kernel package (configured with
   allmodconfig, dwarf4 debuginfo, and module compression turned off)
   before and after this patch:

   # rpm -qi kernel-6.13* | grep -E '^(Version|Size)'
   Version     : 6.13.0postpatch+
   Size        : 1382874089
   Version     : 6.13.0prepatch+
   Size        : 17870795887

   This is a ~92% size reduction.

Note that a debuginfo package can only be produced if the following
configs are set:
- CONFIG_DEBUG_INFO=y
- CONFIG_MODULE_COMPRESS=n
- CONFIG_DEBUG_INFO_SPLIT=n

The first of these is obvious - we can't produce debuginfo if the build
does not generate it. The second two requirements can in principle be
removed, but doing so is difficult with the current approach, which uses
a generic rpmbuild script find-debuginfo.sh that processes all packaged
executables. If we want to remove those requirements the best path
forward is likely to add some debuginfo extraction/installation logic to
the modules_install target (controllable by flags). That way, it's
easier to operate on modules before they're compressed, and the logic
can be reused by all packaging targets.

Signed-off-by: Uday Shankar <ushankar@purestorage.com>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
2 months agokconfig: merge_config: use an empty file as initfile
Daniel Gomez [Fri, 28 Mar 2025 14:28:37 +0000 (14:28 +0000)]
kconfig: merge_config: use an empty file as initfile

The scripts/kconfig/merge_config.sh script requires an existing
$INITFILE (or the $1 argument) as a base file for merging Kconfig
fragments. However, an empty $INITFILE can serve as an initial starting
point, later referenced by the KCONFIG_ALLCONFIG Makefile variable
if -m is not used. This variable can point to any configuration file
containing preset config symbols (the merged output) as stated in
Documentation/kbuild/kconfig.rst. When -m is used $INITFILE will
contain just the merge output requiring the user to run make (i.e.
KCONFIG_ALLCONFIG=<$INITFILE> make <allnoconfig/alldefconfig> or make
olddefconfig).

Instead of failing when `$INITFILE` is missing, create an empty file and
use it as the starting point for merges.

Signed-off-by: Daniel Gomez <da.gomez@samsung.com>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
2 months agonios2: migrate to the generic rule for built-in DTB
Masahiro Yamada [Sun, 22 Dec 2024 00:30:53 +0000 (09:30 +0900)]
nios2: migrate to the generic rule for built-in DTB

Commit 654102df2ac2 ("kbuild: add generic support for built-in boot
DTBs") introduced generic support for built-in DTBs.

Select GENERIC_BUILTIN_DTB when built-in DTB support is enabled.

To keep consistency across architectures, this commit also renames
CONFIG_NIOS2_DTB_SOURCE_BOOL to CONFIG_BUILTIN_DTB, and
CONFIG_NIOS2_DTB_SOURCE to CONFIG_BUILTIN_DTB_NAME.

Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
2 months agosh: defconfig: Drop obsolete CONFIG_NET_CLS_TCINDEX
Johan Korsnes [Sun, 23 Mar 2025 19:13:30 +0000 (20:13 +0100)]
sh: defconfig: Drop obsolete CONFIG_NET_CLS_TCINDEX

This option was removed from Kconfig in 8c710f75256b ("net/sched:
Retire tcindex classifier") but from the defconfigs.

Fixes: 8c710f75256b ("net/sched: Retire tcindex classifier")
Signed-off-by: Johan Korsnes <johan.korsnes@gmail.com>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Rich Felker <dalias@libc.org>
Cc: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Signed-off-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
2 months agosh: Align .bss section padding to 8-byte boundary
Artur Rojek [Sun, 16 Feb 2025 17:55:44 +0000 (18:55 +0100)]
sh: Align .bss section padding to 8-byte boundary

J2-based devices expect to find a device tree blob at the end of the
.bss section. As of a77725a9a3c5 ("scripts/dtc: Update to upstream
version v1.6.1-19-g0a3a9d3449c8"), libfdt enforces 8-byte alignment
for the DTB, causing J2 devices to fail early in sh_fdt_init().

As the J2 loader firmware calculates the DTB location based on the kernel
image .bss section size rather than the __bss_stop symbol offset, the
required alignment can't be enforced with BSS_SECTION(0, PAGE_SIZE, 8).

To fix this, inline a modified version of the above macro which grows
.bss by the required size. While this change affects all existing SH
boards, it should be benign on platforms which don't need this alignment.

Signed-off-by: Artur Rojek <contact@artur-rojek.eu>
Reviewed-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Tested-by: Rob Landley <rob@landley.net>
Signed-off-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
2 months agoMerge tag 'input-for-v6.15-rc0' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Sat, 5 Apr 2025 16:20:39 +0000 (09:20 -0700)]
Merge tag 'input-for-v6.15-rc0' of git://git./linux/kernel/git/dtor/input

Pull input updates from Dmitry Torokhov:

 - a brand new driver for touchpads and touchbars in newer Apple devices

 - support for Berlin-A series in goodix-berlin touchscreen driver

 - improvements to matrix_keypad driver to better handle GPIOs toggling

 - assorted small cleanups in other input drivers

* tag 'input-for-v6.15-rc0' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
  Input: goodix_berlin - add support for Berlin-A series
  dt-bindings: input: goodix,gt9916: Document gt9897 compatible
  dt-bindings: input: matrix_keypad - add wakeup-source property
  dt-bindings: input: matrix_keypad - add missing property
  Input: pm8941-pwrkey - fix dev_dbg() output in pm8941_pwrkey_irq()
  Input: synaptics - hide unused smbus_pnp_ids[] array
  Input: apple_z2 - fix potential confusion in Kconfig
  Input: matrix_keypad - use fsleep for delays after activating columns
  Input: matrix_keypad - add settle time after enabling all columns
  dt-bindings: input: matrix_keypad: add settle time after enabling all columns
  dt-bindings: input: matrix_keypad: convert to YAML
  dt-bindings: input: Correct indentation and style in DTS example
  MAINTAINERS: Add entries for Apple Z2 touchscreen driver
  Input: apple_z2 - add a driver for Apple Z2 touchscreens
  dt-bindings: input: touchscreen: Add Z2 controller
  Input: Switch to use hrtimer_setup()
  Input: drop vb2_ops_wait_prepare/finish

2 months agotracing/timers: Rename the hrtimer_init event to hrtimer_setup
Nam Cao [Wed, 5 Feb 2025 10:55:21 +0000 (11:55 +0100)]
tracing/timers: Rename the hrtimer_init event to hrtimer_setup

The function hrtimer_init() doesn't exist anymore. It was replaced by
hrtimer_setup().

Thus, rename the hrtimer_init trace event to hrtimer_setup to keep it
consistent.

Signed-off-by: Nam Cao <namcao@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/all/cba84c3d853c5258aa3a262363a6eac08e2c7afc.1738746927.git.namcao@linutronix.de
2 months agohrtimers: Rename debug_init_on_stack() to debug_setup_on_stack()
Nam Cao [Wed, 5 Feb 2025 10:55:20 +0000 (11:55 +0100)]
hrtimers: Rename debug_init_on_stack() to debug_setup_on_stack()

All the hrtimer_init*() functions have been renamed to hrtimer_setup*().
Rename debug_init_on_stack() to debug_setup_on_stack() as well, to keep the
names consistent.

Signed-off-by: Nam Cao <namcao@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/all/073cf6162779a2f5b12624677d4c49ee7eccc1ed.1738746927.git.namcao@linutronix.de
2 months agohrtimers: Rename debug_init() to debug_setup()
Nam Cao [Wed, 5 Feb 2025 10:55:19 +0000 (11:55 +0100)]
hrtimers: Rename debug_init() to debug_setup()

All the hrtimer_init*() functions have been renamed to hrtimer_setup*().
Rename debug_init() to debug_setup() as well, to keep the names consistent.

Signed-off-by: Nam Cao <namcao@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/all/4b730c1f79648b16a1c5413f928fdc2e138dfc43.1738746927.git.namcao@linutronix.de
2 months agohrtimers: Rename __hrtimer_init_sleeper() to __hrtimer_setup_sleeper()
Nam Cao [Wed, 5 Feb 2025 10:55:18 +0000 (11:55 +0100)]
hrtimers: Rename __hrtimer_init_sleeper() to __hrtimer_setup_sleeper()

All the hrtimer_init*() functions have been renamed to hrtimer_setup*().
Rename __hrtimer_init_sleeper() to __hrtimer_setup_sleeper() as well, to
keep the names consistent.

Signed-off-by: Nam Cao <namcao@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/all/807694aedad9353421c4a7347629a30c5c31026f.1738746927.git.namcao@linutronix.de
2 months agohrtimers: Remove unnecessary NULL check in hrtimer_start_range_ns()
Nam Cao [Wed, 5 Feb 2025 10:55:17 +0000 (11:55 +0100)]
hrtimers: Remove unnecessary NULL check in hrtimer_start_range_ns()

The struct hrtimer::function field can only be changed using
hrtimer_setup*() or hrtimer_update_function(), and both already null-check
'function'. Therefore, null-checking 'function' in hrtimer_start_range_ns()
is not necessary.

Signed-off-by: Nam Cao <namcao@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/all/4661c571ee87980c340ccc318fc1a473c0c8f6bc.1738746927.git.namcao@linutronix.de
2 months agohrtimers: Make callback function pointer private
Nam Cao [Wed, 5 Feb 2025 10:55:16 +0000 (11:55 +0100)]
hrtimers: Make callback function pointer private

Make the struct hrtimer::function field private, to prevent users from
changing this field in an unsafe way. hrtimer_update_function() should be
used if the callback function needs to be changed.

Signed-off-by: Nam Cao <namcao@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/all/7d0e6e0c5c59a64a9bea940051aac05d750bc0c2.1738746927.git.namcao@linutronix.de
2 months agohrtimers: Merge __hrtimer_init() into __hrtimer_setup()
Nam Cao [Wed, 5 Feb 2025 10:55:12 +0000 (11:55 +0100)]
hrtimers: Merge __hrtimer_init() into __hrtimer_setup()

__hrtimer_init() is only called by __hrtimer_setup(). Simplify by merging
__hrtimer_init() into __hrtimer_setup().

Signed-off-by: Nam Cao <namcao@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/all/8a0a847a35f711f66b2d05b57255aa44e7e61279.1738746927.git.namcao@linutronix.de
2 months agohrtimers: Switch to use __htimer_setup()
Nam Cao [Wed, 5 Feb 2025 10:55:11 +0000 (11:55 +0100)]
hrtimers: Switch to use __htimer_setup()

__hrtimer_init_sleeper() calls __hrtimer_init() and also sets up the
callback function. But there is already __hrtimer_setup() which does both
actions.

Switch to use __hrtimer_setup() to simplify the code.

Signed-off-by: Nam Cao <namcao@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/all/d9a45a51b6a8aa0045310d63f73753bf6b33f385.1738746927.git.namcao@linutronix.de
2 months agohrtimers: Delete hrtimer_init()
Nam Cao [Wed, 5 Feb 2025 10:55:10 +0000 (11:55 +0100)]
hrtimers: Delete hrtimer_init()

hrtimer_init() is now unused. Delete it.

Signed-off-by: Nam Cao <namcao@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/all/003722f60c7a2a4f8d4ed24fb741aa313b7e5136.1738746927.git.namcao@linutronix.de
2 months agotreewide: Convert new and leftover hrtimer_init() users
Thomas Gleixner [Fri, 4 Apr 2025 17:31:15 +0000 (19:31 +0200)]
treewide: Convert new and leftover hrtimer_init() users

hrtimer_setup() takes the callback function pointer as argument and
initializes the timer completely.

Replace hrtimer_init() and the open coded initialization of
hrtimer::function with the new setup mechanism.

Coccinelle scripted cleanup.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2 months agotreewide: Switch/rename to timer_delete[_sync]()
Thomas Gleixner [Sat, 5 Apr 2025 08:17:26 +0000 (10:17 +0200)]
treewide: Switch/rename to timer_delete[_sync]()

timer_delete[_sync]() replaces del_timer[_sync](). Convert the whole tree
over and remove the historical wrapper inlines.

Conversion was done with coccinelle plus manual fixups where necessary.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2 months agoMerge branch 'next' into for-linus
Dmitry Torokhov [Sat, 5 Apr 2025 06:04:35 +0000 (23:04 -0700)]
Merge branch 'next' into for-linus

Prepare input updates for 6.15 merge window.

2 months agoMerge tag 'v6.15-p3' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
Linus Torvalds [Sat, 5 Apr 2025 02:34:38 +0000 (19:34 -0700)]
Merge tag 'v6.15-p3' of git://git./linux/kernel/git/herbert/crypto-2.6

Pull crypto fix from Herbert Xu:
 "This fixes a race condition in the newly added eip93 driver"

* tag 'v6.15-p3' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
  crypto: inside-secure/eip93 - acquire lock on eip93_put_descriptor hash

2 months agoMerge tag 's390-6.15-2' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux
Linus Torvalds [Fri, 4 Apr 2025 23:58:34 +0000 (16:58 -0700)]
Merge tag 's390-6.15-2' of git://git./linux/kernel/git/s390/linux

Pull more s390 updates from Vasily Gorbik:

 - Fix machine check handler _CIF_MCCK_GUEST bit setting by adding the
   missing base register for relocated lowcore address

 - Fix build failure on older linkers by conditionally adding the
   -no-pie linker option only when it is supported

 - Fix inaccurate kernel messages in vfio-ap by providing descriptive
   error notifications for AP queue sharing violations

 - Fix PCI isolation logic by ensuring non-VF devices correctly return
   false in zpci_bus_is_isolated_vf()

 - Fix PCI DMA range map setup by using dma_direct_set_offset() to add a
   proper sentinel element, preventing potential overruns and
   translation errors

 - Cleanup header dependency problems with asm-offsets.c

 - Add fault info for unexpected low-address protection faults in user
   mode

 - Add support for HOTPLUG_SMT, replacing the arch-specific "nosmt"
   handling with common code handling

 - Use bitop functions to implement CPU flag helper functions to ensure
   that bits cannot get lost if modified in different contexts on a CPU

 - Remove unused machine_flags for the lowcore

* tag 's390-6.15-2' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
  s390/vfio-ap: Fix no AP queue sharing allowed message written to kernel log
  s390/pci: Fix dev.dma_range_map missing sentinel element
  s390/mm: Dump fault info in case of low address protection fault
  s390/smp: Add support for HOTPLUG_SMT
  s390: Fix linker error when -no-pie option is unavailable
  s390/processor: Use bitop functions for cpu flag helper functions
  s390/asm-offsets: Remove ASM_OFFSETS_C
  s390/asm-offsets: Include ftrace_regs.h instead of ftrace.h
  s390/kvm: Split kvm_host header file
  s390/pci: Fix zpci_bus_is_isolated_vf() for non-VFs
  s390/lowcore: Remove unused machine_flags
  s390/entry: Fix setting _CIF_MCCK_GUEST with lowcore relocation

2 months agoMerge tag '6.15-rc-part2-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6
Linus Torvalds [Fri, 4 Apr 2025 22:27:43 +0000 (15:27 -0700)]
Merge tag '6.15-rc-part2-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6

Pull more smb client updates from Steve French:

 - reconnect fixes: three for updating rsize/wsize and an SMB1 reconnect
   fix

 - RFC1001 fixes: fixing connections to nonstandard ports, and negprot
   retries

 - fix mfsymlinks to old servers

 - make mapping of open flags for SMB1 more accurate

 - permission fixes: adding retry on open for write, and one for stat to
   workaround unexpected access denied

 - add two new xattrs, one for retrieving SACL and one for retrieving
   owner (without having to retrieve the whole ACL)

 - fix mount parm validation for echo_interval

 - minor cleanup (including removing now unneeded cifs_truncate_page)

* tag '6.15-rc-part2-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
  cifs: update internal version number
  cifs: Implement is_network_name_deleted for SMB1
  cifs: Remove cifs_truncate_page() as it should be superfluous
  cifs: Do not add FILE_READ_ATTRIBUTES when using GENERIC_READ/EXECUTE/ALL
  cifs: Improve SMB2+ stat() to work also without FILE_READ_ATTRIBUTES
  cifs: Add fallback for SMB2 CREATE without FILE_READ_ATTRIBUTES
  cifs: Fix querying and creating MF symlinks over SMB1
  cifs: Fix access_flags_to_smbopen_mode
  cifs: Fix negotiate retry functionality
  cifs: Improve handling of NetBIOS packets
  cifs: Allow to disable or force initialization of NetBIOS session
  cifs: Add a new xattr system.smb3_ntsd_owner for getting or setting owner
  cifs: Add a new xattr system.smb3_ntsd_sacl for getting or setting SACLs
  smb: client: Update IO sizes after reconnection
  smb: client: Store original IO parameters and prevent zero IO sizes
  smb:client: smb: client: Add reverse mapping from tcon to superblocks
  cifs: remove unreachable code in cifs_get_tcp_session()
  cifs: fix integer overflow in match_server()

2 months agoMerge tag 'ntb-6.15' of https://github.com/jonmason/ntb
Linus Torvalds [Fri, 4 Apr 2025 21:23:07 +0000 (14:23 -0700)]
Merge tag 'ntb-6.15' of https://github.com/jonmason/ntb

Pull ntb fixes from Jon Mason:
 "Bug fixes for NTB Switchtec driver mw negative shift, Intel NTB link
  status db, ntb_perf double unmap (in error case), and MSI 64bit
  arithmetic.

  Also, add new AMD NTB PCI IDs, update AMD NTB maintainer, and pull in
  patch to reduce the stack usage in IDT driver"

* tag 'ntb-6.15' of https://github.com/jonmason/ntb:
  ntb_hw_amd: Add NTB PCI ID for new gen CPU
  ntb: reduce stack usage in idt_scan_mws
  ntb: use 64-bit arithmetic for the MSI doorbell mask
  MAINTAINERS: Update AMD NTB maintainers
  ntb_perf: Delete duplicate dmaengine_unmap_put() call in perf_copy_chunk()
  ntb: intel: Fix using link status DB's
  ntb_hw_switchtec: Fix shift-out-of-bounds in switchtec_ntb_mw_set_trans

2 months agoMerge tag 'drm-misc-next-fixes-2025-04-04' of https://gitlab.freedesktop.org/drm...
Dave Airlie [Fri, 4 Apr 2025 20:27:56 +0000 (06:27 +1000)]
Merge tag 'drm-misc-next-fixes-2025-04-04' of https://gitlab.freedesktop.org/drm/misc/kernel into drm-next

Short summary of fixes pull:

bridge:
- tda998x: Select CONFIG_DRM_KMS_HELPER

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://lore.kernel.org/r/20250404065105.GA27699@linux.fritz.box
2 months agoRevert "timekeeping: Fix possible inconsistencies in _COARSE clockids"
Thomas Gleixner [Fri, 4 Apr 2025 15:10:52 +0000 (17:10 +0200)]
Revert "timekeeping: Fix possible inconsistencies in _COARSE clockids"

This reverts commit 757b000f7b936edf79311ab0971fe465bbda75ea.

Miroslav reported that the changes for handling the inconsistencies in the
coarse time getters result in a regression on the adjtimex() side.

There are two issues:

  1) The forwarding of the base time moves the update out of the original
     period and establishes a new one.

  2) The clearing of the accumulated NTP error is changing the behaviour as
     well.

Userspace expects that multiplier/frequency updates are in effect, when the
syscall returns, so delaying the update to the next tick is not solving the
problem either.

Revert the change, so that the established expectations of user space
implementations (ntpd, chronyd) are restored. The re-introduced
inconsistency of the coarse time getters will be addressed in a subsequent
fix.

Fixes: 757b000f7b93 ("timekeeping: Fix possible inconsistencies in _COARSE clockids")
Reported-by: Miroslav Lichvar <mlichvar@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/Z-qsg6iDGlcIJulJ@localhost
2 months agoMerge tag 'riscv-for-linus-6.15-mw1' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Fri, 4 Apr 2025 16:49:17 +0000 (09:49 -0700)]
Merge tag 'riscv-for-linus-6.15-mw1' of git://git./linux/kernel/git/riscv/linux

Pull RISC-V updates from Palmer Dabbelt:

 - The sub-architecture selection Kconfig system has been cleaned up,
   the documentation has been improved, and various detections have been
   fixed

 - The vector-related extensions dependencies are now validated when
   parsing from device tree and in the DT bindings

 - Misaligned access probing can be overridden via a kernel command-line
   parameter, along with various fixes to misalign access handling

 - Support for relocatable !MMU kernels builds

 - Support for hpge pfnmaps, which should improve TLB utilization

 - Support for runtime constants, which improves the d_hash()
   performance

 - Support for bfloat16, Zicbom, Zaamo, Zalrsc, Zicntr, Zihpm

 - Various fixes, including:
      - We were missing a secondary mmu notifier call when flushing the
        tlb which is required for IOMMU
      - Fix ftrace panics by saving the registers as expected by ftrace
      - Fix a couple of stimecmp usage related to cpu hotplug
      - purgatory_start is now aligned as per the STVEC requirements
      - A fix for hugetlb when calculating the size of non-present PTEs

* tag 'riscv-for-linus-6.15-mw1' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux: (65 commits)
  riscv: Add norvc after .option arch in runtime const
  riscv: Make sure toolchain supports zba before using zba instructions
  riscv/purgatory: 4B align purgatory_start
  riscv/kexec_file: Handle R_RISCV_64 in purgatory relocator
  selftests: riscv: fix v_exec_initval_nolibc.c
  riscv: Fix hugetlb retrieval of number of ptes in case of !present pte
  riscv: print hartid on bringup
  riscv: Add norvc after .option arch in runtime const
  riscv: Remove CONFIG_PAGE_OFFSET
  riscv: Support CONFIG_RELOCATABLE on riscv32
  asm-generic: Always define Elf_Rel and Elf_Rela
  riscv: Support CONFIG_RELOCATABLE on NOMMU
  riscv: Allow NOMMU kernels to access all of RAM
  riscv: Remove duplicate CONFIG_PAGE_OFFSET definition
  RISC-V: errata: Use medany for relocatable builds
  dt-bindings: riscv: document vector crypto requirements
  dt-bindings: riscv: add vector sub-extension dependencies
  dt-bindings: riscv: d requires f
  RISC-V: add f & d extension validation checks
  RISC-V: add vector crypto extension validation checks
  ...

2 months agoMerge tag 'net-6.15-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Linus Torvalds [Fri, 4 Apr 2025 16:15:35 +0000 (09:15 -0700)]
Merge tag 'net-6.15-rc1' of git://git./linux/kernel/git/netdev/net

Pull networking fixes from Jakub Kicinski:
 "Including fixes from netfilter.

  Current release - regressions:

   - four fixes for the netdev per-instance locking

  Current release - new code bugs:

   - consolidate more code between existing Rx zero-copy and uring so
     that the latter doesn't miss / have to duplicate the safety checks

  Previous releases - regressions:

   - ipv6: fix omitted Netlink attributes when using SKIP_STATS

  Previous releases - always broken:

   - net: fix geneve_opt length integer overflow

   - udp: fix multiple wrap arounds of sk->sk_rmem_alloc when it
     approaches INT_MAX

   - dsa: mvpp2: add a lock to avoid corruption of the shared TCAM

   - dsa: airoha: fix issues with traffic QoS configuration / offload,
     and flow table offload

  Misc:

   - touch up the Netlink YAML specs of old families to make them usable
     for user space C codegen"

* tag 'net-6.15-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (56 commits)
  selftests: net: amt: indicate progress in the stress test
  netlink: specs: rt_route: pull the ifa- prefix out of the names
  netlink: specs: rt_addr: pull the ifa- prefix out of the names
  netlink: specs: rt_addr: fix get multi command name
  netlink: specs: rt_addr: fix the spec format / schema failures
  net: avoid false positive warnings in __net_mp_close_rxq()
  net: move mp dev config validation to __net_mp_open_rxq()
  net: ibmveth: make veth_pool_store stop hanging
  arcnet: Add NULL check in com20020pci_probe()
  ipv6: Do not consider link down nexthops in path selection
  ipv6: Start path selection from the first nexthop
  usbnet:fix NPE during rx_complete
  net: octeontx2: Handle XDP_ABORTED and XDP invalid as XDP_DROP
  net: fix geneve_opt length integer overflow
  io_uring/zcrx: fix selftests w/ updated netdev Python helpers
  selftests: net: use netdevsim in netns test
  docs: net: document netdev notifier expectations
  net: dummy: request ops lock
  netdevsim: add dummy device notifiers
  net: rename rtnl_net_debug to lock_debug
  ...

2 months agoMerge tag 'spi-fix-v6.15-merge-window' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Fri, 4 Apr 2025 16:09:34 +0000 (09:09 -0700)]
Merge tag 'spi-fix-v6.15-merge-window' of git://git./linux/kernel/git/broonie/spi

Pull spi fixes from Mark Brown:
 "A small collection of fixes that came in during the merge window,
  everything is driver specific with nothing standing out particularly"

* tag 'spi-fix-v6.15-merge-window' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:
  spi: bcm2835: Restore native CS probing when pinctrl-bcm2835 is absent
  spi: bcm2835: Do not call gpiod_put() on invalid descriptor
  spi: cadence-qspi: revert "Improve spi memory performance"
  spi: cadence: Fix out-of-bounds array access in cdns_mrvl_xspi_setup_clock()
  spi: fsl-qspi: use devm function instead of driver remove
  spi: SPI_QPIC_SNAND should be tristate and depend on MTD
  spi-rockchip: Fix register out of bounds access

2 months agoMerge tag 'soc-drivers-6.15-2' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc
Linus Torvalds [Fri, 4 Apr 2025 16:06:32 +0000 (09:06 -0700)]
Merge tag 'soc-drivers-6.15-2' of git://git./linux/kernel/git/soc/soc

Pull more SoC driver updates from Arnd Bergmann:
 "This is the promised follow-up to the soc drivers branch, adding minor
  updates to omap and freescale drivers.

  Most notably, Ioana Ciornei takes over maintenance of the DPAA bus
  driver used in some NXP (originally Freescale) chips"

* tag 'soc-drivers-6.15-2' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc:
  bus: fsl-mc: Remove deadcode
  MAINTAINERS: add the linuppc-dev list to the fsl-mc bus entry
  MAINTAINERS: fix nonexistent dtbinding file name
  MAINTAINERS: add myself as maintainer for the fsl-mc bus
  irqdomain: soc: Switch to irq_find_mapping()
  Input: tsc2007 - accept standard properties

2 months agoMerge tag 'platform-drivers-x86-v6.15-2' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Fri, 4 Apr 2025 16:00:49 +0000 (09:00 -0700)]
Merge tag 'platform-drivers-x86-v6.15-2' of git://git./linux/kernel/git/pdx86/platform-drivers-x86

Pull x86 platform driver fixes from Ilpo Järvinen:

 - thinkpad_acpi:
     - Fix NULL pointer dereferences while probing
     - Disable ACPI fan access for T495* and E560

 - ISST: Correct command storage data length

* tag 'platform-drivers-x86-v6.15-2' of git://git.kernel.org/pub/scm/linux/kernel/git/pdx86/platform-drivers-x86:
  MAINTAINERS: consistently use my dedicated email address
  platform/x86: ISST: Correct command storage data length
  platform/x86: thinkpad_acpi: disable ACPI fan access for T495* and E560
  platform/x86: thinkpad_acpi: Fix NULL pointer dereferences while probing

2 months agogenirq/migration: Use irqd_get_parent_data() in irq_force_complete_move()
Thomas Gleixner [Fri, 4 Apr 2025 14:51:19 +0000 (16:51 +0200)]
genirq/migration: Use irqd_get_parent_data() in irq_force_complete_move()

Frank reported, that the common irq_force_complete_move() breaks the out of
tree build of ia64. The reason is that ia64 uses the migration code, but
does not have hierarchical interrupt domains enabled.

This went unnoticed in mainline as both x86 and RISC-V have hierarchical
domains enabled. Not that it matters for mainline, but it's still
inconsistent.

Use irqd_get_parent_data() instead of accessing the parent_data field
directly. The helper returns NULL when hierarchical domains are disabled
otherwise it accesses the parent_data field of the domain.

No functional change.

Fixes: 751dc837dabd ("genirq: Introduce common irq_force_complete_move() implementation")
Reported-by: Frank Scheiner <frank.scheiner@web.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Frank Scheiner <frank.scheiner@web.de>
Link: https://lore.kernel.org/all/87h634ugig.ffs@tglx
2 months agoselftests: net: amt: indicate progress in the stress test
Jakub Kicinski [Thu, 3 Apr 2025 14:56:36 +0000 (07:56 -0700)]
selftests: net: amt: indicate progress in the stress test

Our CI expects output from the test at least once every 10 minutes.
The AMT test when running on debug kernel is just on the edge
of that time for the stress test. Improve the output:
 - print the name of the test first, before starting it,
 - output a dot every 10% of the way.

Output after:

  TEST: amt discovery                                                 [ OK ]
  TEST: IPv4 amt multicast forwarding                                 [ OK ]
  TEST: IPv6 amt multicast forwarding                                 [ OK ]
  TEST: IPv4 amt traffic forwarding torture               ..........  [ OK ]
  TEST: IPv6 amt traffic forwarding torture               ..........  [ OK ]

Reviewed-by: Taehee Yoo <ap420073@gmail.com>
Link: https://patch.msgid.link/20250403145636.2891166-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 months agoirqdomain: Stop using 'host' for domain
Jiri Slaby (SUSE) [Wed, 19 Mar 2025 09:28:57 +0000 (10:28 +0100)]
irqdomain: Stop using 'host' for domain

It is confusing to see 'host' and 'domain' to be used as 'domain'. Given
this header is all about domains, switch the remaining 'host' uses to
'domain'.

Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/20250319092951.37667-5-jirislaby@kernel.org
2 months agoirqdomain: Rename irq_get_default_host() to irq_get_default_domain()
Jiri Slaby (SUSE) [Wed, 19 Mar 2025 09:28:56 +0000 (10:28 +0100)]
irqdomain: Rename irq_get_default_host() to irq_get_default_domain()

Naming interrupt domains host is confusing at best and the irqdomain code
uses both domain and host inconsistently.

Therefore rename irq_get_default_host() to irq_get_default_domain().

Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/20250319092951.37667-4-jirislaby@kernel.org
2 months agoirqdomain: Rename irq_set_default_host() to irq_set_default_domain()
Jiri Slaby (SUSE) [Wed, 19 Mar 2025 09:28:55 +0000 (10:28 +0100)]
irqdomain: Rename irq_set_default_host() to irq_set_default_domain()

Naming interrupt domains host is confusing at best and the irqdomain code
uses both domain and host inconsistently.

Therefore rename irq_set_default_host() to irq_set_default_domain().

Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/20250319092951.37667-3-jirislaby@kernel.org
2 months agoMerge branch 'netlink-specs-rt_addr-fix-problems-revealed-by-c-codegen'
Jakub Kicinski [Fri, 4 Apr 2025 14:36:11 +0000 (07:36 -0700)]
Merge branch 'netlink-specs-rt_addr-fix-problems-revealed-by-c-codegen'

Jakub Kicinski says:

====================
netlink: specs: rt_addr: fix problems revealed by C codegen

I put together basic YNL C support for classic netlink. This revealed
a few problems in the rt_addr spec.

v1: https://lore.kernel.org/20250401012939.2116915-1-kuba@kernel.org
====================

Link: https://patch.msgid.link/20250403013706.2828322-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 months agonetlink: specs: rt_route: pull the ifa- prefix out of the names
Jakub Kicinski [Thu, 3 Apr 2025 01:37:06 +0000 (18:37 -0700)]
netlink: specs: rt_route: pull the ifa- prefix out of the names

YAML specs don't normally include the C prefix name in the name
of the YAML attr. Remove the ifa- prefix from all attributes
in route-attrs and metrics and specify name-prefix instead.

This is a bit risky, hopefully there aren't many users out there.

Fixes: 023289b4f582 ("doc/netlink: Add spec for rt route messages")
Reviewed-by: Donald Hunter <donald.hunter@gmail.com>
Link: https://patch.msgid.link/20250403013706.2828322-5-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 months agonetlink: specs: rt_addr: pull the ifa- prefix out of the names
Jakub Kicinski [Thu, 3 Apr 2025 01:37:05 +0000 (18:37 -0700)]
netlink: specs: rt_addr: pull the ifa- prefix out of the names

YAML specs don't normally include the C prefix name in the name
of the YAML attr. Remove the ifa- prefix from all attributes
in addr-attrs and specify name-prefix instead.

This is a bit risky, hopefully there aren't many users out there.

Fixes: dfb0f7d9d979 ("doc/netlink: Add spec for rt addr messages")
Reviewed-by: Donald Hunter <donald.hunter@gmail.com>
Link: https://patch.msgid.link/20250403013706.2828322-4-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 months agonetlink: specs: rt_addr: fix get multi command name
Jakub Kicinski [Thu, 3 Apr 2025 01:37:04 +0000 (18:37 -0700)]
netlink: specs: rt_addr: fix get multi command name

Command names should match C defines, codegens may depend on it.

Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Fixes: 4f280376e531 ("selftests/net: Add selftest for IPv4 RTM_GETMULTICAST support")
Reviewed-by: Donald Hunter <donald.hunter@gmail.com>
Link: https://patch.msgid.link/20250403013706.2828322-3-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 months agonetlink: specs: rt_addr: fix the spec format / schema failures
Jakub Kicinski [Thu, 3 Apr 2025 01:37:03 +0000 (18:37 -0700)]
netlink: specs: rt_addr: fix the spec format / schema failures

The spec is mis-formatted, schema validation says:

  Failed validating 'type' in schema['properties']['operations']['properties']['list']['items']['properties']['dump']['properties']['request']['properties']['value']:
    {'minimum': 0, 'type': 'integer'}

  On instance['operations']['list'][3]['dump']['request']['value']:
    '58 - ifa-family'

The ifa-family clearly wants to be part of an attribute list.

Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Donald Hunter <donald.hunter@gmail.com>
Reviewed-by: Yuyang Huang <yuyanghuang@google.com>
Fixes: 4f280376e531 ("selftests/net: Add selftest for IPv4 RTM_GETMULTICAST support")
Link: https://patch.msgid.link/20250403013706.2828322-2-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 months agoMerge branch 'net-make-memory-provider-install-close-paths-more-common'
Jakub Kicinski [Fri, 4 Apr 2025 14:35:42 +0000 (07:35 -0700)]
Merge branch 'net-make-memory-provider-install-close-paths-more-common'

Jakub Kicinski says:

====================
net: make memory provider install / close paths more common

We seem to be fixing bugs in config path for devmem which also exist
in the io_uring ZC path. Let's try to make the two paths more common,
otherwise this is bound to keep happening.

Found by code inspection and compile tested only.

v1: https://lore.kernel.org/20250331194201.2026422-1-kuba@kernel.org
====================

Link: https://patch.msgid.link/20250403013405.2827250-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 months agonet: avoid false positive warnings in __net_mp_close_rxq()
Jakub Kicinski [Thu, 3 Apr 2025 01:34:05 +0000 (18:34 -0700)]
net: avoid false positive warnings in __net_mp_close_rxq()

Commit under Fixes solved the problem of spurious warnings when we
uninstall an MP from a device while its down. The __net_mp_close_rxq()
which is used by io_uring was not fixed. Move the fix over and reuse
__net_mp_close_rxq() in the devmem path.

Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Fixes: a70f891e0fa0 ("net: devmem: do not WARN conditionally after netdev_rx_queue_restart()")
Reviewed-by: Mina Almasry <almasrymina@google.com>
Link: https://patch.msgid.link/20250403013405.2827250-3-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 months agonet: move mp dev config validation to __net_mp_open_rxq()
Jakub Kicinski [Thu, 3 Apr 2025 01:34:04 +0000 (18:34 -0700)]
net: move mp dev config validation to __net_mp_open_rxq()

devmem code performs a number of safety checks to avoid having
to reimplement all of them in the drivers. Move those to
__net_mp_open_rxq() and reuse that function for binding to make
sure that io_uring ZC also benefits from them.

While at it rename the queue ID variable to rxq_idx in
__net_mp_open_rxq(), we touch most of the relevant lines.

The XArray insertion is reordered after the netdev_rx_queue_restart()
call, otherwise we'd need to duplicate the queue index check
or risk inserting an invalid pointer. The XArray allocation
failures should be extremely rare.

Reviewed-by: Mina Almasry <almasrymina@google.com>
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Fixes: 6e18ed929d3b ("net: add helpers for setting a memory provider on an rx queue")
Link: https://patch.msgid.link/20250403013405.2827250-2-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 months agonet: ibmveth: make veth_pool_store stop hanging
Dave Marquardt [Wed, 2 Apr 2025 15:44:03 +0000 (10:44 -0500)]
net: ibmveth: make veth_pool_store stop hanging

v2:
- Created a single error handling unlock and exit in veth_pool_store
- Greatly expanded commit message with previous explanatory-only text

Summary: Use rtnl_mutex to synchronize veth_pool_store with itself,
ibmveth_close and ibmveth_open, preventing multiple calls in a row to
napi_disable.

Background: Two (or more) threads could call veth_pool_store through
writing to /sys/devices/vio/30000002/pool*/*. You can do this easily
with a little shell script. This causes a hang.

I configured LOCKDEP, compiled ibmveth.c with DEBUG, and built a new
kernel. I ran this test again and saw:

    Setting pool0/active to 0
    Setting pool1/active to 1
    [   73.911067][ T4365] ibmveth 30000002 eth0: close starting
    Setting pool1/active to 1
    Setting pool1/active to 0
    [   73.911367][ T4366] ibmveth 30000002 eth0: close starting
    [   73.916056][ T4365] ibmveth 30000002 eth0: close complete
    [   73.916064][ T4365] ibmveth 30000002 eth0: open starting
    [  110.808564][  T712] systemd-journald[712]: Sent WATCHDOG=1 notification.
    [  230.808495][  T712] systemd-journald[712]: Sent WATCHDOG=1 notification.
    [  243.683786][  T123] INFO: task stress.sh:4365 blocked for more than 122 seconds.
    [  243.683827][  T123]       Not tainted 6.14.0-01103-g2df0c02dab82-dirty #8
    [  243.683833][  T123] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
    [  243.683838][  T123] task:stress.sh       state:D stack:28096 pid:4365  tgid:4365  ppid:4364   task_flags:0x400040 flags:0x00042000
    [  243.683852][  T123] Call Trace:
    [  243.683857][  T123] [c00000000c38f690] [0000000000000001] 0x1 (unreliable)
    [  243.683868][  T123] [c00000000c38f840] [c00000000001f908] __switch_to+0x318/0x4e0
    [  243.683878][  T123] [c00000000c38f8a0] [c000000001549a70] __schedule+0x500/0x12a0
    [  243.683888][  T123] [c00000000c38f9a0] [c00000000154a878] schedule+0x68/0x210
    [  243.683896][  T123] [c00000000c38f9d0] [c00000000154ac80] schedule_preempt_disabled+0x30/0x50
    [  243.683904][  T123] [c00000000c38fa00] [c00000000154dbb0] __mutex_lock+0x730/0x10f0
    [  243.683913][  T123] [c00000000c38fb10] [c000000001154d40] napi_enable+0x30/0x60
    [  243.683921][  T123] [c00000000c38fb40] [c000000000f4ae94] ibmveth_open+0x68/0x5dc
    [  243.683928][  T123] [c00000000c38fbe0] [c000000000f4aa20] veth_pool_store+0x220/0x270
    [  243.683936][  T123] [c00000000c38fc70] [c000000000826278] sysfs_kf_write+0x68/0xb0
    [  243.683944][  T123] [c00000000c38fcb0] [c0000000008240b8] kernfs_fop_write_iter+0x198/0x2d0
    [  243.683951][  T123] [c00000000c38fd00] [c00000000071b9ac] vfs_write+0x34c/0x650
    [  243.683958][  T123] [c00000000c38fdc0] [c00000000071bea8] ksys_write+0x88/0x150
    [  243.683966][  T123] [c00000000c38fe10] [c0000000000317f4] system_call_exception+0x124/0x340
    [  243.683973][  T123] [c00000000c38fe50] [c00000000000d05c] system_call_vectored_common+0x15c/0x2ec
    ...
    [  243.684087][  T123] Showing all locks held in the system:
    [  243.684095][  T123] 1 lock held by khungtaskd/123:
    [  243.684099][  T123]  #0: c00000000278e370 (rcu_read_lock){....}-{1:2}, at: debug_show_all_locks+0x50/0x248
    [  243.684114][  T123] 4 locks held by stress.sh/4365:
    [  243.684119][  T123]  #0: c00000003a4cd3f8 (sb_writers#3){.+.+}-{0:0}, at: ksys_write+0x88/0x150
    [  243.684132][  T123]  #1: c000000041aea888 (&of->mutex#2){+.+.}-{3:3}, at: kernfs_fop_write_iter+0x154/0x2d0
    [  243.684143][  T123]  #2: c0000000366fb9a8 (kn->active#64){.+.+}-{0:0}, at: kernfs_fop_write_iter+0x160/0x2d0
    [  243.684155][  T123]  #3: c000000035ff4cb8 (&dev->lock){+.+.}-{3:3}, at: napi_enable+0x30/0x60
    [  243.684166][  T123] 5 locks held by stress.sh/4366:
    [  243.684170][  T123]  #0: c00000003a4cd3f8 (sb_writers#3){.+.+}-{0:0}, at: ksys_write+0x88/0x150
    [  243.684183][  T123]  #1: c00000000aee2288 (&of->mutex#2){+.+.}-{3:3}, at: kernfs_fop_write_iter+0x154/0x2d0
    [  243.684194][  T123]  #2: c0000000366f4ba8 (kn->active#64){.+.+}-{0:0}, at: kernfs_fop_write_iter+0x160/0x2d0
    [  243.684205][  T123]  #3: c000000035ff4cb8 (&dev->lock){+.+.}-{3:3}, at: napi_disable+0x30/0x60
    [  243.684216][  T123]  #4: c0000003ff9bbf18 (&rq->__lock){-.-.}-{2:2}, at: __schedule+0x138/0x12a0

From the ibmveth debug, two threads are calling veth_pool_store, which
calls ibmveth_close and ibmveth_open. Here's the sequence:

  T4365             T4366
  ----------------- ----------------- ---------
  veth_pool_store   veth_pool_store
                    ibmveth_close
  ibmveth_close
  napi_disable
                    napi_disable
  ibmveth_open
  napi_enable                         <- HANG

ibmveth_close calls napi_disable at the top and ibmveth_open calls
napi_enable at the top.

https://docs.kernel.org/networking/napi.html]] says

  The control APIs are not idempotent. Control API calls are safe
  against concurrent use of datapath APIs but an incorrect sequence of
  control API calls may result in crashes, deadlocks, or race
  conditions. For example, calling napi_disable() multiple times in a
  row will deadlock.

In the normal open and close paths, rtnl_mutex is acquired to prevent
other callers. This is missing from veth_pool_store. Use rtnl_mutex in
veth_pool_store fixes these hangs.

Signed-off-by: Dave Marquardt <davemarq@linux.ibm.com>
Fixes: 860f242eb534 ("[PATCH] ibmveth change buffer pools dynamically")
Reviewed-by: Nick Child <nnac123@linux.ibm.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20250402154403.386744-1-davemarq@linux.ibm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 months agoarcnet: Add NULL check in com20020pci_probe()
Henry Martin [Wed, 2 Apr 2025 13:50:36 +0000 (21:50 +0800)]
arcnet: Add NULL check in com20020pci_probe()

devm_kasprintf() returns NULL when memory allocation fails. Currently,
com20020pci_probe() does not check for this case, which results in a
NULL pointer dereference.

Add NULL check after devm_kasprintf() to prevent this issue and ensure
no resources are left allocated.

Fixes: 6b17a597fc2f ("arcnet: restoring support for multiple Sohard Arcnet cards")
Signed-off-by: Henry Martin <bsdhenrymartin@gmail.com>
Link: https://patch.msgid.link/20250402135036.44697-1-bsdhenrymartin@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 months agoMerge branch 'ipv6-multipath-routing-fixes'
Jakub Kicinski [Fri, 4 Apr 2025 14:30:09 +0000 (07:30 -0700)]
Merge branch 'ipv6-multipath-routing-fixes'

Ido Schimmel says:

====================
ipv6: Multipath routing fixes

This patchset contains two fixes for IPv6 multipath routing. See the
commit messages for more details.
====================

Link: https://patch.msgid.link/20250402114224.293392-1-idosch@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>