summaryrefslogtreecommitdiff
path: root/rust
diff options
context:
space:
mode:
Diffstat (limited to 'rust')
-rw-r--r--rust/bindings/bindings_helper.h4
-rw-r--r--rust/helpers/clk.c66
-rw-r--r--rust/helpers/cpufreq.c10
-rw-r--r--rust/helpers/cpumask.c25
-rw-r--r--rust/helpers/helpers.c3
-rw-r--r--rust/helpers/mm.c50
-rw-r--r--rust/kernel/clk.rs334
-rw-r--r--rust/kernel/cpu.rs30
-rw-r--r--rust/kernel/cpufreq.rs1321
-rw-r--r--rust/kernel/cpumask.rs330
-rw-r--r--rust/kernel/lib.rs8
-rw-r--r--rust/kernel/miscdevice.rs45
-rw-r--r--rust/kernel/mm.rs344
-rw-r--r--rust/kernel/mm/virt.rs471
-rw-r--r--rust/kernel/opp.rs1146
-rw-r--r--rust/kernel/task.rs247
-rw-r--r--rust/macros/module.rs20
17 files changed, 4328 insertions, 126 deletions
diff --git a/rust/bindings/bindings_helper.h b/rust/bindings/bindings_helper.h
index f600a1f945f4..a5a6fb45d405 100644
--- a/rust/bindings/bindings_helper.h
+++ b/rust/bindings/bindings_helper.h
@@ -16,7 +16,10 @@
#include <linux/blk-mq.h>
#include <linux/blk_types.h>
#include <linux/blkdev.h>
+#include <linux/clk.h>
#include <linux/configfs.h>
+#include <linux/cpu.h>
+#include <linux/cpufreq.h>
#include <linux/cpumask.h>
#include <linux/cred.h>
#include <linux/device/faux.h>
@@ -35,6 +38,7 @@
#include <linux/phy.h>
#include <linux/pid_namespace.h>
#include <linux/platform_device.h>
+#include <linux/pm_opp.h>
#include <linux/poll.h>
#include <linux/property.h>
#include <linux/refcount.h>
diff --git a/rust/helpers/clk.c b/rust/helpers/clk.c
new file mode 100644
index 000000000000..6d04372c9f3b
--- /dev/null
+++ b/rust/helpers/clk.c
@@ -0,0 +1,66 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <linux/clk.h>
+
+/*
+ * The "inline" implementation of below helpers are only available when
+ * CONFIG_HAVE_CLK or CONFIG_HAVE_CLK_PREPARE aren't set.
+ */
+#ifndef CONFIG_HAVE_CLK
+struct clk *rust_helper_clk_get(struct device *dev, const char *id)
+{
+ return clk_get(dev, id);
+}
+
+void rust_helper_clk_put(struct clk *clk)
+{
+ clk_put(clk);
+}
+
+int rust_helper_clk_enable(struct clk *clk)
+{
+ return clk_enable(clk);
+}
+
+void rust_helper_clk_disable(struct clk *clk)
+{
+ clk_disable(clk);
+}
+
+unsigned long rust_helper_clk_get_rate(struct clk *clk)
+{
+ return clk_get_rate(clk);
+}
+
+int rust_helper_clk_set_rate(struct clk *clk, unsigned long rate)
+{
+ return clk_set_rate(clk, rate);
+}
+#endif
+
+#ifndef CONFIG_HAVE_CLK_PREPARE
+int rust_helper_clk_prepare(struct clk *clk)
+{
+ return clk_prepare(clk);
+}
+
+void rust_helper_clk_unprepare(struct clk *clk)
+{
+ clk_unprepare(clk);
+}
+#endif
+
+struct clk *rust_helper_clk_get_optional(struct device *dev, const char *id)
+{
+ return clk_get_optional(dev, id);
+}
+
+int rust_helper_clk_prepare_enable(struct clk *clk)
+{
+ return clk_prepare_enable(clk);
+}
+
+void rust_helper_clk_disable_unprepare(struct clk *clk)
+{
+ clk_disable_unprepare(clk);
+}
diff --git a/rust/helpers/cpufreq.c b/rust/helpers/cpufreq.c
new file mode 100644
index 000000000000..7c1343c4d65e
--- /dev/null
+++ b/rust/helpers/cpufreq.c
@@ -0,0 +1,10 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <linux/cpufreq.h>
+
+#ifdef CONFIG_CPU_FREQ
+void rust_helper_cpufreq_register_em_with_opp(struct cpufreq_policy *policy)
+{
+ cpufreq_register_em_with_opp(policy);
+}
+#endif
diff --git a/rust/helpers/cpumask.c b/rust/helpers/cpumask.c
index 2d380a86c34a..eb10598a0242 100644
--- a/rust/helpers/cpumask.c
+++ b/rust/helpers/cpumask.c
@@ -7,16 +7,41 @@ void rust_helper_cpumask_set_cpu(unsigned int cpu, struct cpumask *dstp)
cpumask_set_cpu(cpu, dstp);
}
+void rust_helper___cpumask_set_cpu(unsigned int cpu, struct cpumask *dstp)
+{
+ __cpumask_set_cpu(cpu, dstp);
+}
+
void rust_helper_cpumask_clear_cpu(int cpu, struct cpumask *dstp)
{
cpumask_clear_cpu(cpu, dstp);
}
+void rust_helper___cpumask_clear_cpu(int cpu, struct cpumask *dstp)
+{
+ __cpumask_clear_cpu(cpu, dstp);
+}
+
+bool rust_helper_cpumask_test_cpu(int cpu, struct cpumask *srcp)
+{
+ return cpumask_test_cpu(cpu, srcp);
+}
+
void rust_helper_cpumask_setall(struct cpumask *dstp)
{
cpumask_setall(dstp);
}
+bool rust_helper_cpumask_empty(struct cpumask *srcp)
+{
+ return cpumask_empty(srcp);
+}
+
+bool rust_helper_cpumask_full(struct cpumask *srcp)
+{
+ return cpumask_full(srcp);
+}
+
unsigned int rust_helper_cpumask_weight(struct cpumask *srcp)
{
return cpumask_weight(srcp);
diff --git a/rust/helpers/helpers.c b/rust/helpers/helpers.c
index c228a5363212..805307018f0e 100644
--- a/rust/helpers/helpers.c
+++ b/rust/helpers/helpers.c
@@ -12,6 +12,8 @@
#include "bug.c"
#include "build_assert.c"
#include "build_bug.c"
+#include "clk.c"
+#include "cpufreq.c"
#include "cpumask.c"
#include "cred.c"
#include "device.c"
@@ -22,6 +24,7 @@
#include "io.c"
#include "jump_label.c"
#include "kunit.c"
+#include "mm.c"
#include "mutex.c"
#include "page.c"
#include "platform.c"
diff --git a/rust/helpers/mm.c b/rust/helpers/mm.c
new file mode 100644
index 000000000000..81b510c96fd2
--- /dev/null
+++ b/rust/helpers/mm.c
@@ -0,0 +1,50 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <linux/mm.h>
+#include <linux/sched/mm.h>
+
+void rust_helper_mmgrab(struct mm_struct *mm)
+{
+ mmgrab(mm);
+}
+
+void rust_helper_mmdrop(struct mm_struct *mm)
+{
+ mmdrop(mm);
+}
+
+void rust_helper_mmget(struct mm_struct *mm)
+{
+ mmget(mm);
+}
+
+bool rust_helper_mmget_not_zero(struct mm_struct *mm)
+{
+ return mmget_not_zero(mm);
+}
+
+void rust_helper_mmap_read_lock(struct mm_struct *mm)
+{
+ mmap_read_lock(mm);
+}
+
+bool rust_helper_mmap_read_trylock(struct mm_struct *mm)
+{
+ return mmap_read_trylock(mm);
+}
+
+void rust_helper_mmap_read_unlock(struct mm_struct *mm)
+{
+ mmap_read_unlock(mm);
+}
+
+struct vm_area_struct *rust_helper_vma_lookup(struct mm_struct *mm,
+ unsigned long addr)
+{
+ return vma_lookup(mm, addr);
+}
+
+void rust_helper_vma_end_read(struct vm_area_struct *vma)
+{
+ vma_end_read(vma);
+}
diff --git a/rust/kernel/clk.rs b/rust/kernel/clk.rs
new file mode 100644
index 000000000000..6041c6d07527
--- /dev/null
+++ b/rust/kernel/clk.rs
@@ -0,0 +1,334 @@
+// SPDX-License-Identifier: GPL-2.0
+
+//! Clock abstractions.
+//!
+//! C header: [`include/linux/clk.h`](srctree/include/linux/clk.h)
+//!
+//! Reference: <https://docs.kernel.org/driver-api/clk.html>
+
+use crate::ffi::c_ulong;
+
+/// The frequency unit.
+///
+/// Represents a frequency in hertz, wrapping a [`c_ulong`] value.
+///
+/// ## Examples
+///
+/// ```
+/// use kernel::clk::Hertz;
+///
+/// let hz = 1_000_000_000;
+/// let rate = Hertz(hz);
+///
+/// assert_eq!(rate.as_hz(), hz);
+/// assert_eq!(rate, Hertz(hz));
+/// assert_eq!(rate, Hertz::from_khz(hz / 1_000));
+/// assert_eq!(rate, Hertz::from_mhz(hz / 1_000_000));
+/// assert_eq!(rate, Hertz::from_ghz(hz / 1_000_000_000));
+/// ```
+#[derive(Copy, Clone, PartialEq, Eq, Debug)]
+pub struct Hertz(pub c_ulong);
+
+impl Hertz {
+ /// Create a new instance from kilohertz (kHz)
+ pub fn from_khz(khz: c_ulong) -> Self {
+ Self(khz * 1_000)
+ }
+
+ /// Create a new instance from megahertz (MHz)
+ pub fn from_mhz(mhz: c_ulong) -> Self {
+ Self(mhz * 1_000_000)
+ }
+
+ /// Create a new instance from gigahertz (GHz)
+ pub fn from_ghz(ghz: c_ulong) -> Self {
+ Self(ghz * 1_000_000_000)
+ }
+
+ /// Get the frequency in hertz
+ pub fn as_hz(&self) -> c_ulong {
+ self.0
+ }
+
+ /// Get the frequency in kilohertz
+ pub fn as_khz(&self) -> c_ulong {
+ self.0 / 1_000
+ }
+
+ /// Get the frequency in megahertz
+ pub fn as_mhz(&self) -> c_ulong {
+ self.0 / 1_000_000
+ }
+
+ /// Get the frequency in gigahertz
+ pub fn as_ghz(&self) -> c_ulong {
+ self.0 / 1_000_000_000
+ }
+}
+
+impl From<Hertz> for c_ulong {
+ fn from(freq: Hertz) -> Self {
+ freq.0
+ }
+}
+
+#[cfg(CONFIG_COMMON_CLK)]
+mod common_clk {
+ use super::Hertz;
+ use crate::{
+ device::Device,
+ error::{from_err_ptr, to_result, Result},
+ prelude::*,
+ };
+
+ use core::{ops::Deref, ptr};
+
+ /// A reference-counted clock.
+ ///
+ /// Rust abstraction for the C [`struct clk`].
+ ///
+ /// # Invariants
+ ///
+ /// A [`Clk`] instance holds either a pointer to a valid [`struct clk`] created by the C
+ /// portion of the kernel or a NULL pointer.
+ ///
+ /// Instances of this type are reference-counted. Calling [`Clk::get`] ensures that the
+ /// allocation remains valid for the lifetime of the [`Clk`].
+ ///
+ /// ## Examples
+ ///
+ /// The following example demonstrates how to obtain and configure a clock for a device.
+ ///
+ /// ```
+ /// use kernel::c_str;
+ /// use kernel::clk::{Clk, Hertz};
+ /// use kernel::device::Device;
+ /// use kernel::error::Result;
+ ///
+ /// fn configure_clk(dev: &Device) -> Result {
+ /// let clk = Clk::get(dev, Some(c_str!("apb_clk")))?;
+ ///
+ /// clk.prepare_enable()?;
+ ///
+ /// let expected_rate = Hertz::from_ghz(1);
+ ///
+ /// if clk.rate() != expected_rate {
+ /// clk.set_rate(expected_rate)?;
+ /// }
+ ///
+ /// clk.disable_unprepare();
+ /// Ok(())
+ /// }
+ /// ```
+ ///
+ /// [`struct clk`]: https://docs.kernel.org/driver-api/clk.html
+ #[repr(transparent)]
+ pub struct Clk(*mut bindings::clk);
+
+ impl Clk {
+ /// Gets [`Clk`] corresponding to a [`Device`] and a connection id.
+ ///
+ /// Equivalent to the kernel's [`clk_get`] API.
+ ///
+ /// [`clk_get`]: https://docs.kernel.org/core-api/kernel-api.html#c.clk_get
+ pub fn get(dev: &Device, name: Option<&CStr>) -> Result<Self> {
+ let con_id = if let Some(name) = name {
+ name.as_ptr()
+ } else {
+ ptr::null()
+ };
+
+ // SAFETY: It is safe to call [`clk_get`] for a valid device pointer.
+ //
+ // INVARIANT: The reference-count is decremented when [`Clk`] goes out of scope.
+ Ok(Self(from_err_ptr(unsafe {
+ bindings::clk_get(dev.as_raw(), con_id)
+ })?))
+ }
+
+ /// Obtain the raw [`struct clk`] pointer.
+ #[inline]
+ pub fn as_raw(&self) -> *mut bindings::clk {
+ self.0
+ }
+
+ /// Enable the clock.
+ ///
+ /// Equivalent to the kernel's [`clk_enable`] API.
+ ///
+ /// [`clk_enable`]: https://docs.kernel.org/core-api/kernel-api.html#c.clk_enable
+ #[inline]
+ pub fn enable(&self) -> Result {
+ // SAFETY: By the type invariants, self.as_raw() is a valid argument for
+ // [`clk_enable`].
+ to_result(unsafe { bindings::clk_enable(self.as_raw()) })
+ }
+
+ /// Disable the clock.
+ ///
+ /// Equivalent to the kernel's [`clk_disable`] API.
+ ///
+ /// [`clk_disable`]: https://docs.kernel.org/core-api/kernel-api.html#c.clk_disable
+ #[inline]
+ pub fn disable(&self) {
+ // SAFETY: By the type invariants, self.as_raw() is a valid argument for
+ // [`clk_disable`].
+ unsafe { bindings::clk_disable(self.as_raw()) };
+ }
+
+ /// Prepare the clock.
+ ///
+ /// Equivalent to the kernel's [`clk_prepare`] API.
+ ///
+ /// [`clk_prepare`]: https://docs.kernel.org/core-api/kernel-api.html#c.clk_prepare
+ #[inline]
+ pub fn prepare(&self) -> Result {
+ // SAFETY: By the type invariants, self.as_raw() is a valid argument for
+ // [`clk_prepare`].
+ to_result(unsafe { bindings::clk_prepare(self.as_raw()) })
+ }
+
+ /// Unprepare the clock.
+ ///
+ /// Equivalent to the kernel's [`clk_unprepare`] API.
+ ///
+ /// [`clk_unprepare`]: https://docs.kernel.org/core-api/kernel-api.html#c.clk_unprepare
+ #[inline]
+ pub fn unprepare(&self) {
+ // SAFETY: By the type invariants, self.as_raw() is a valid argument for
+ // [`clk_unprepare`].
+ unsafe { bindings::clk_unprepare(self.as_raw()) };
+ }
+
+ /// Prepare and enable the clock.
+ ///
+ /// Equivalent to calling [`Clk::prepare`] followed by [`Clk::enable`].
+ #[inline]
+ pub fn prepare_enable(&self) -> Result {
+ // SAFETY: By the type invariants, self.as_raw() is a valid argument for
+ // [`clk_prepare_enable`].
+ to_result(unsafe { bindings::clk_prepare_enable(self.as_raw()) })
+ }
+
+ /// Disable and unprepare the clock.
+ ///
+ /// Equivalent to calling [`Clk::disable`] followed by [`Clk::unprepare`].
+ #[inline]
+ pub fn disable_unprepare(&self) {
+ // SAFETY: By the type invariants, self.as_raw() is a valid argument for
+ // [`clk_disable_unprepare`].
+ unsafe { bindings::clk_disable_unprepare(self.as_raw()) };
+ }
+
+ /// Get clock's rate.
+ ///
+ /// Equivalent to the kernel's [`clk_get_rate`] API.
+ ///
+ /// [`clk_get_rate`]: https://docs.kernel.org/core-api/kernel-api.html#c.clk_get_rate
+ #[inline]
+ pub fn rate(&self) -> Hertz {
+ // SAFETY: By the type invariants, self.as_raw() is a valid argument for
+ // [`clk_get_rate`].
+ Hertz(unsafe { bindings::clk_get_rate(self.as_raw()) })
+ }
+
+ /// Set clock's rate.
+ ///
+ /// Equivalent to the kernel's [`clk_set_rate`] API.
+ ///
+ /// [`clk_set_rate`]: https://docs.kernel.org/core-api/kernel-api.html#c.clk_set_rate
+ #[inline]
+ pub fn set_rate(&self, rate: Hertz) -> Result {
+ // SAFETY: By the type invariants, self.as_raw() is a valid argument for
+ // [`clk_set_rate`].
+ to_result(unsafe { bindings::clk_set_rate(self.as_raw(), rate.as_hz()) })
+ }
+ }
+
+ impl Drop for Clk {
+ fn drop(&mut self) {
+ // SAFETY: By the type invariants, self.as_raw() is a valid argument for [`clk_put`].
+ unsafe { bindings::clk_put(self.as_raw()) };
+ }
+ }
+
+ /// A reference-counted optional clock.
+ ///
+ /// A lightweight wrapper around an optional [`Clk`]. An [`OptionalClk`] represents a [`Clk`]
+ /// that a driver can function without but may improve performance or enable additional
+ /// features when available.
+ ///
+ /// # Invariants
+ ///
+ /// An [`OptionalClk`] instance encapsulates a [`Clk`] with either a valid [`struct clk`] or
+ /// `NULL` pointer.
+ ///
+ /// Instances of this type are reference-counted. Calling [`OptionalClk::get`] ensures that the
+ /// allocation remains valid for the lifetime of the [`OptionalClk`].
+ ///
+ /// ## Examples
+ ///
+ /// The following example demonstrates how to obtain and configure an optional clock for a
+ /// device. The code functions correctly whether or not the clock is available.
+ ///
+ /// ```
+ /// use kernel::c_str;
+ /// use kernel::clk::{OptionalClk, Hertz};
+ /// use kernel::device::Device;
+ /// use kernel::error::Result;
+ ///
+ /// fn configure_clk(dev: &Device) -> Result {
+ /// let clk = OptionalClk::get(dev, Some(c_str!("apb_clk")))?;
+ ///
+ /// clk.prepare_enable()?;
+ ///
+ /// let expected_rate = Hertz::from_ghz(1);
+ ///
+ /// if clk.rate() != expected_rate {
+ /// clk.set_rate(expected_rate)?;
+ /// }
+ ///
+ /// clk.disable_unprepare();
+ /// Ok(())
+ /// }
+ /// ```
+ ///
+ /// [`struct clk`]: https://docs.kernel.org/driver-api/clk.html
+ pub struct OptionalClk(Clk);
+
+ impl OptionalClk {
+ /// Gets [`OptionalClk`] corresponding to a [`Device`] and a connection id.
+ ///
+ /// Equivalent to the kernel's [`clk_get_optional`] API.
+ ///
+ /// [`clk_get_optional`]:
+ /// https://docs.kernel.org/core-api/kernel-api.html#c.clk_get_optional
+ pub fn get(dev: &Device, name: Option<&CStr>) -> Result<Self> {
+ let con_id = if let Some(name) = name {
+ name.as_ptr()
+ } else {
+ ptr::null()
+ };
+
+ // SAFETY: It is safe to call [`clk_get_optional`] for a valid device pointer.
+ //
+ // INVARIANT: The reference-count is decremented when [`OptionalClk`] goes out of
+ // scope.
+ Ok(Self(Clk(from_err_ptr(unsafe {
+ bindings::clk_get_optional(dev.as_raw(), con_id)
+ })?)))
+ }
+ }
+
+ // Make [`OptionalClk`] behave like [`Clk`].
+ impl Deref for OptionalClk {
+ type Target = Clk;
+
+ fn deref(&self) -> &Clk {
+ &self.0
+ }
+ }
+}
+
+#[cfg(CONFIG_COMMON_CLK)]
+pub use common_clk::*;
diff --git a/rust/kernel/cpu.rs b/rust/kernel/cpu.rs
new file mode 100644
index 000000000000..10c5c3b25873
--- /dev/null
+++ b/rust/kernel/cpu.rs
@@ -0,0 +1,30 @@
+// SPDX-License-Identifier: GPL-2.0
+
+//! Generic CPU definitions.
+//!
+//! C header: [`include/linux/cpu.h`](srctree/include/linux/cpu.h)
+
+use crate::{bindings, device::Device, error::Result, prelude::ENODEV};
+
+/// Creates a new instance of CPU's device.
+///
+/// # Safety
+///
+/// Reference counting is not implemented for the CPU device in the C code. When a CPU is
+/// hot-unplugged, the corresponding CPU device is unregistered, but its associated memory
+/// is not freed.
+///
+/// Callers must ensure that the CPU device is not used after it has been unregistered.
+/// This can be achieved, for example, by registering a CPU hotplug notifier and removing
+/// any references to the CPU device within the notifier's callback.
+pub unsafe fn from_cpu(cpu: u32) -> Result<&'static Device> {
+ // SAFETY: It is safe to call `get_cpu_device()` for any CPU.
+ let ptr = unsafe { bindings::get_cpu_device(cpu) };
+ if ptr.is_null() {
+ return Err(ENODEV);
+ }
+
+ // SAFETY: The pointer returned by `get_cpu_device()`, if not `NULL`, is a valid pointer to
+ // a `struct device` and is never freed by the C code.
+ Ok(unsafe { Device::as_ref(ptr) })
+}
diff --git a/rust/kernel/cpufreq.rs b/rust/kernel/cpufreq.rs
new file mode 100644
index 000000000000..09b856bb297b
--- /dev/null
+++ b/rust/kernel/cpufreq.rs
@@ -0,0 +1,1321 @@
+// SPDX-License-Identifier: GPL-2.0
+
+//! CPU frequency scaling.
+//!
+//! This module provides rust abstractions for interacting with the cpufreq subsystem.
+//!
+//! C header: [`include/linux/cpufreq.h`](srctree/include/linux/cpufreq.h)
+//!
+//! Reference: <https://docs.kernel.org/admin-guide/pm/cpufreq.html>
+
+use crate::{
+ clk::Hertz,
+ cpumask,
+ device::{Bound, Device},
+ devres::Devres,
+ error::{code::*, from_err_ptr, from_result, to_result, Result, VTABLE_DEFAULT_ERROR},
+ ffi::{c_char, c_ulong},
+ prelude::*,
+ types::ForeignOwnable,
+ types::Opaque,
+};
+
+#[cfg(CONFIG_COMMON_CLK)]
+use crate::clk::Clk;
+
+use core::{
+ cell::UnsafeCell,
+ marker::PhantomData,
+ mem::MaybeUninit,
+ ops::{Deref, DerefMut},
+ pin::Pin,
+ ptr,
+};
+
+use macros::vtable;
+
+/// Maximum length of CPU frequency driver's name.
+const CPUFREQ_NAME_LEN: usize = bindings::CPUFREQ_NAME_LEN as usize;
+
+/// Default transition latency value in nanoseconds.
+pub const ETERNAL_LATENCY_NS: u32 = bindings::CPUFREQ_ETERNAL as u32;
+
+/// CPU frequency driver flags.
+pub mod flags {
+ /// Driver needs to update internal limits even if frequency remains unchanged.
+ pub const NEED_UPDATE_LIMITS: u16 = 1 << 0;
+
+ /// Platform where constants like `loops_per_jiffy` are unaffected by frequency changes.
+ pub const CONST_LOOPS: u16 = 1 << 1;
+
+ /// Register driver as a thermal cooling device automatically.
+ pub const IS_COOLING_DEV: u16 = 1 << 2;
+
+ /// Supports multiple clock domains with per-policy governors in `cpu/cpuN/cpufreq/`.
+ pub const HAVE_GOVERNOR_PER_POLICY: u16 = 1 << 3;
+
+ /// Allows post-change notifications outside of the `target()` routine.
+ pub const ASYNC_NOTIFICATION: u16 = 1 << 4;
+
+ /// Ensure CPU starts at a valid frequency from the driver's freq-table.
+ pub const NEED_INITIAL_FREQ_CHECK: u16 = 1 << 5;
+
+ /// Disallow governors with `dynamic_switching` capability.
+ pub const NO_AUTO_DYNAMIC_SWITCHING: u16 = 1 << 6;
+}
+
+/// Relations from the C code.
+const CPUFREQ_RELATION_L: u32 = 0;
+const CPUFREQ_RELATION_H: u32 = 1;
+const CPUFREQ_RELATION_C: u32 = 2;
+
+/// Can be used with any of the above values.
+const CPUFREQ_RELATION_E: u32 = 1 << 2;
+
+/// CPU frequency selection relations.
+///
+/// CPU frequency selection relations, each optionally marked as "efficient".
+#[derive(Copy, Clone, Debug, Eq, PartialEq)]
+pub enum Relation {
+ /// Select the lowest frequency at or above target.
+ Low(bool),
+ /// Select the highest frequency below or at target.
+ High(bool),
+ /// Select the closest frequency to the target.
+ Close(bool),
+}
+
+impl Relation {
+ // Construct from a C-compatible `u32` value.
+ fn new(val: u32) -> Result<Self> {
+ let efficient = val & CPUFREQ_RELATION_E != 0;
+
+ Ok(match val & !CPUFREQ_RELATION_E {
+ CPUFREQ_RELATION_L => Self::Low(efficient),
+ CPUFREQ_RELATION_H => Self::High(efficient),
+ CPUFREQ_RELATION_C => Self::Close(efficient),
+ _ => return Err(EINVAL),
+ })
+ }
+}
+
+impl From<Relation> for u32 {
+ // Convert to a C-compatible `u32` value.
+ fn from(rel: Relation) -> Self {
+ let (mut val, efficient) = match rel {
+ Relation::Low(e) => (CPUFREQ_RELATION_L, e),
+ Relation::High(e) => (CPUFREQ_RELATION_H, e),
+ Relation::Close(e) => (CPUFREQ_RELATION_C, e),
+ };
+
+ if efficient {
+ val |= CPUFREQ_RELATION_E;
+ }
+
+ val
+ }
+}
+
+/// Policy data.
+///
+/// Rust abstraction for the C `struct cpufreq_policy_data`.
+///
+/// # Invariants
+///
+/// A [`PolicyData`] instance always corresponds to a valid C `struct cpufreq_policy_data`.
+///
+/// The callers must ensure that the `struct cpufreq_policy_data` is valid for access and remains
+/// valid for the lifetime of the returned reference.
+#[repr(transparent)]
+pub struct PolicyData(Opaque<bindings::cpufreq_policy_data>);
+
+impl PolicyData {
+ /// Creates a mutable reference to an existing `struct cpufreq_policy_data` pointer.
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that `ptr` is valid for writing and remains valid for the lifetime
+ /// of the returned reference.
+ #[inline]
+ pub unsafe fn from_raw_mut<'a>(ptr: *mut bindings::cpufreq_policy_data) -> &'a mut Self {
+ // SAFETY: Guaranteed by the safety requirements of the function.
+ //
+ // INVARIANT: The caller ensures that `ptr` is valid for writing and remains valid for the
+ // lifetime of the returned reference.
+ unsafe { &mut *ptr.cast() }
+ }
+
+ /// Returns a raw pointer to the underlying C `cpufreq_policy_data`.
+ #[inline]
+ pub fn as_raw(&self) -> *mut bindings::cpufreq_policy_data {
+ let this: *const Self = self;
+ this.cast_mut().cast()
+ }
+
+ /// Wrapper for `cpufreq_generic_frequency_table_verify`.
+ #[inline]
+ pub fn generic_verify(&self) -> Result {
+ // SAFETY: By the type invariant, the pointer stored in `self` is valid.
+ to_result(unsafe { bindings::cpufreq_generic_frequency_table_verify(self.as_raw()) })
+ }
+}
+
+/// The frequency table index.
+///
+/// Represents index with a frequency table.
+///
+/// # Invariants
+///
+/// The index must correspond to a valid entry in the [`Table`] it is used for.
+#[derive(Copy, Clone, PartialEq, Eq, Debug)]
+pub struct TableIndex(usize);
+
+impl TableIndex {
+ /// Creates an instance of [`TableIndex`].
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that `index` correspond to a valid entry in the [`Table`] it is used
+ /// for.
+ pub unsafe fn new(index: usize) -> Self {
+ // INVARIANT: The caller ensures that `index` correspond to a valid entry in the [`Table`].
+ Self(index)
+ }
+}
+
+impl From<TableIndex> for usize {
+ #[inline]
+ fn from(index: TableIndex) -> Self {
+ index.0
+ }
+}
+
+/// CPU frequency table.
+///
+/// Rust abstraction for the C `struct cpufreq_frequency_table`.
+///
+/// # Invariants
+///
+/// A [`Table`] instance always corresponds to a valid C `struct cpufreq_frequency_table`.
+///
+/// The callers must ensure that the `struct cpufreq_frequency_table` is valid for access and
+/// remains valid for the lifetime of the returned reference.
+///
+/// ## Examples
+///
+/// The following example demonstrates how to read a frequency value from [`Table`].
+///
+/// ```
+/// use kernel::cpufreq::{Policy, TableIndex};
+///
+/// fn show_freq(policy: &Policy) -> Result {
+/// let table = policy.freq_table()?;
+///
+/// // SAFETY: Index is a valid entry in the table.
+/// let index = unsafe { TableIndex::new(0) };
+///
+/// pr_info!("The frequency at index 0 is: {:?}\n", table.freq(index)?);
+/// pr_info!("The flags at index 0 is: {}\n", table.flags(index));
+/// pr_info!("The data at index 0 is: {}\n", table.data(index));
+/// Ok(())
+/// }
+/// ```
+#[repr(transparent)]
+pub struct Table(Opaque<bindings::cpufreq_frequency_table>);
+
+impl Table {
+ /// Creates a reference to an existing C `struct cpufreq_frequency_table` pointer.
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that `ptr` is valid for reading and remains valid for the lifetime
+ /// of the returned reference.
+ #[inline]
+ pub unsafe fn from_raw<'a>(ptr: *const bindings::cpufreq_frequency_table) -> &'a Self {
+ // SAFETY: Guaranteed by the safety requirements of the function.
+ //
+ // INVARIANT: The caller ensures that `ptr` is valid for reading and remains valid for the
+ // lifetime of the returned reference.
+ unsafe { &*ptr.cast() }
+ }
+
+ /// Returns the raw mutable pointer to the C `struct cpufreq_frequency_table`.
+ #[inline]
+ pub fn as_raw(&self) -> *mut bindings::cpufreq_frequency_table {
+ let this: *const Self = self;
+ this.cast_mut().cast()
+ }
+
+ /// Returns frequency at `index` in the [`Table`].
+ #[inline]
+ pub fn freq(&self, index: TableIndex) -> Result<Hertz> {
+ // SAFETY: By the type invariant, the pointer stored in `self` is valid and `index` is
+ // guaranteed to be valid by its safety requirements.
+ Ok(Hertz::from_khz(unsafe {
+ (*self.as_raw().add(index.into())).frequency.try_into()?
+ }))
+ }
+
+ /// Returns flags at `index` in the [`Table`].
+ #[inline]
+ pub fn flags(&self, index: TableIndex) -> u32 {
+ // SAFETY: By the type invariant, the pointer stored in `self` is valid and `index` is
+ // guaranteed to be valid by its safety requirements.
+ unsafe { (*self.as_raw().add(index.into())).flags }
+ }
+
+ /// Returns data at `index` in the [`Table`].
+ #[inline]
+ pub fn data(&self, index: TableIndex) -> u32 {
+ // SAFETY: By the type invariant, the pointer stored in `self` is valid and `index` is
+ // guaranteed to be valid by its safety requirements.
+ unsafe { (*self.as_raw().add(index.into())).driver_data }
+ }
+}
+
+/// CPU frequency table owned and pinned in memory, created from a [`TableBuilder`].
+pub struct TableBox {
+ entries: Pin<KVec<bindings::cpufreq_frequency_table>>,
+}
+
+impl TableBox {
+ /// Constructs a new [`TableBox`] from a [`KVec`] of entries.
+ ///
+ /// # Errors
+ ///
+ /// Returns `EINVAL` if the entries list is empty.
+ #[inline]
+ fn new(entries: KVec<bindings::cpufreq_frequency_table>) -> Result<Self> {
+ if entries.is_empty() {
+ return Err(EINVAL);
+ }
+
+ Ok(Self {
+ // Pin the entries to memory, since we are passing its pointer to the C code.
+ entries: Pin::new(entries),
+ })
+ }
+
+ /// Returns a raw pointer to the underlying C `cpufreq_frequency_table`.
+ #[inline]
+ fn as_raw(&self) -> *const bindings::cpufreq_frequency_table {
+ // The pointer is valid until the table gets dropped.
+ self.entries.as_ptr()
+ }
+}
+
+impl Deref for TableBox {
+ type Target = Table;
+
+ fn deref(&self) -> &Self::Target {
+ // SAFETY: The caller owns TableBox, it is safe to deref.
+ unsafe { Self::Target::from_raw(self.as_raw()) }
+ }
+}
+
+/// CPU frequency table builder.
+///
+/// This is used by the CPU frequency drivers to build a frequency table dynamically.
+///
+/// ## Examples
+///
+/// The following example demonstrates how to create a CPU frequency table.
+///
+/// ```
+/// use kernel::cpufreq::{TableBuilder, TableIndex};
+/// use kernel::clk::Hertz;
+///
+/// let mut builder = TableBuilder::new();
+///
+/// // Adds few entries to the table.
+/// builder.add(Hertz::from_mhz(700), 0, 1).unwrap();
+/// builder.add(Hertz::from_mhz(800), 2, 3).unwrap();
+/// builder.add(Hertz::from_mhz(900), 4, 5).unwrap();
+/// builder.add(Hertz::from_ghz(1), 6, 7).unwrap();
+///
+/// let table = builder.to_table().unwrap();
+///
+/// // SAFETY: Index values correspond to valid entries in the table.
+/// let (index0, index2) = unsafe { (TableIndex::new(0), TableIndex::new(2)) };
+///
+/// assert_eq!(table.freq(index0), Ok(Hertz::from_mhz(700)));
+/// assert_eq!(table.flags(index0), 0);
+/// assert_eq!(table.data(index0), 1);
+///
+/// assert_eq!(table.freq(index2), Ok(Hertz::from_mhz(900)));
+/// assert_eq!(table.flags(index2), 4);
+/// assert_eq!(table.data(index2), 5);
+/// ```
+#[derive(Default)]
+#[repr(transparent)]
+pub struct TableBuilder {
+ entries: KVec<bindings::cpufreq_frequency_table>,
+}
+
+impl TableBuilder {
+ /// Creates a new instance of [`TableBuilder`].
+ #[inline]
+ pub fn new() -> Self {
+ Self {
+ entries: KVec::new(),
+ }
+ }
+
+ /// Adds a new entry to the table.
+ pub fn add(&mut self, freq: Hertz, flags: u32, driver_data: u32) -> Result {
+ // Adds the new entry at the end of the vector.
+ Ok(self.entries.push(
+ bindings::cpufreq_frequency_table {
+ flags,
+ driver_data,
+ frequency: freq.as_khz() as u32,
+ },
+ GFP_KERNEL,
+ )?)
+ }
+
+ /// Consumes the [`TableBuilder`] and returns [`TableBox`].
+ pub fn to_table(mut self) -> Result<TableBox> {
+ // Add last entry to the table.
+ self.add(Hertz(c_ulong::MAX), 0, 0)?;
+
+ TableBox::new(self.entries)
+ }
+}
+
+/// CPU frequency policy.
+///
+/// Rust abstraction for the C `struct cpufreq_policy`.
+///
+/// # Invariants
+///
+/// A [`Policy`] instance always corresponds to a valid C `struct cpufreq_policy`.
+///
+/// The callers must ensure that the `struct cpufreq_policy` is valid for access and remains valid
+/// for the lifetime of the returned reference.
+///
+/// ## Examples
+///
+/// The following example demonstrates how to create a CPU frequency table.
+///
+/// ```
+/// use kernel::cpufreq::{ETERNAL_LATENCY_NS, Policy};
+///
+/// fn update_policy(policy: &mut Policy) {
+/// policy
+/// .set_dvfs_possible_from_any_cpu(true)
+/// .set_fast_switch_possible(true)
+/// .set_transition_latency_ns(ETERNAL_LATENCY_NS);
+///
+/// pr_info!("The policy details are: {:?}\n", (policy.cpu(), policy.cur()));
+/// }
+/// ```
+#[repr(transparent)]
+pub struct Policy(Opaque<bindings::cpufreq_policy>);
+
+impl Policy {
+ /// Creates a reference to an existing `struct cpufreq_policy` pointer.
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that `ptr` is valid for reading and remains valid for the lifetime
+ /// of the returned reference.
+ #[inline]
+ pub unsafe fn from_raw<'a>(ptr: *const bindings::cpufreq_policy) -> &'a Self {
+ // SAFETY: Guaranteed by the safety requirements of the function.
+ //
+ // INVARIANT: The caller ensures that `ptr` is valid for reading and remains valid for the
+ // lifetime of the returned reference.
+ unsafe { &*ptr.cast() }
+ }
+
+ /// Creates a mutable reference to an existing `struct cpufreq_policy` pointer.
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that `ptr` is valid for writing and remains valid for the lifetime
+ /// of the returned reference.
+ #[inline]
+ pub unsafe fn from_raw_mut<'a>(ptr: *mut bindings::cpufreq_policy) -> &'a mut Self {
+ // SAFETY: Guaranteed by the safety requirements of the function.
+ //
+ // INVARIANT: The caller ensures that `ptr` is valid for writing and remains valid for the
+ // lifetime of the returned reference.
+ unsafe { &mut *ptr.cast() }
+ }
+
+ /// Returns a raw mutable pointer to the C `struct cpufreq_policy`.
+ #[inline]
+ fn as_raw(&self) -> *mut bindings::cpufreq_policy {
+ let this: *const Self = self;
+ this.cast_mut().cast()
+ }
+
+ #[inline]
+ fn as_ref(&self) -> &bindings::cpufreq_policy {
+ // SAFETY: By the type invariant, the pointer stored in `self` is valid.
+ unsafe { &*self.as_raw() }
+ }
+
+ #[inline]
+ fn as_mut_ref(&mut self) -> &mut bindings::cpufreq_policy {
+ // SAFETY: By the type invariant, the pointer stored in `self` is valid.
+ unsafe { &mut *self.as_raw() }
+ }
+
+ /// Returns the primary CPU for the [`Policy`].
+ #[inline]
+ pub fn cpu(&self) -> u32 {
+ self.as_ref().cpu
+ }
+
+ /// Returns the minimum frequency for the [`Policy`].
+ #[inline]
+ pub fn min(&self) -> Hertz {
+ Hertz::from_khz(self.as_ref().min as usize)
+ }
+
+ /// Set the minimum frequency for the [`Policy`].
+ #[inline]
+ pub fn set_min(&mut self, min: Hertz) -> &mut Self {
+ self.as_mut_ref().min = min.as_khz() as u32;
+ self
+ }
+
+ /// Returns the maximum frequency for the [`Policy`].
+ #[inline]
+ pub fn max(&self) -> Hertz {
+ Hertz::from_khz(self.as_ref().max as usize)
+ }
+
+ /// Set the maximum frequency for the [`Policy`].
+ #[inline]
+ pub fn set_max(&mut self, max: Hertz) -> &mut Self {
+ self.as_mut_ref().max = max.as_khz() as u32;
+ self
+ }
+
+ /// Returns the current frequency for the [`Policy`].
+ #[inline]
+ pub fn cur(&self) -> Hertz {
+ Hertz::from_khz(self.as_ref().cur as usize)
+ }
+
+ /// Returns the suspend frequency for the [`Policy`].
+ #[inline]
+ pub fn suspend_freq(&self) -> Hertz {
+ Hertz::from_khz(self.as_ref().suspend_freq as usize)
+ }
+
+ /// Sets the suspend frequency for the [`Policy`].
+ #[inline]
+ pub fn set_suspend_freq(&mut self, freq: Hertz) -> &mut Self {
+ self.as_mut_ref().suspend_freq = freq.as_khz() as u32;
+ self
+ }
+
+ /// Provides a wrapper to the generic suspend routine.
+ #[inline]
+ pub fn generic_suspend(&mut self) -> Result {
+ // SAFETY: By the type invariant, the pointer stored in `self` is valid.
+ to_result(unsafe { bindings::cpufreq_generic_suspend(self.as_mut_ref()) })
+ }
+
+ /// Provides a wrapper to the generic get routine.
+ #[inline]
+ pub fn generic_get(&self) -> Result<u32> {
+ // SAFETY: By the type invariant, the pointer stored in `self` is valid.
+ Ok(unsafe { bindings::cpufreq_generic_get(self.cpu()) })
+ }
+
+ /// Provides a wrapper to the register with energy model using the OPP core.
+ #[cfg(CONFIG_PM_OPP)]
+ #[inline]
+ pub fn register_em_opp(&mut self) {
+ // SAFETY: By the type invariant, the pointer stored in `self` is valid.
+ unsafe { bindings::cpufreq_register_em_with_opp(self.as_mut_ref()) };
+ }
+
+ /// Gets [`cpumask::Cpumask`] for a cpufreq [`Policy`].
+ #[inline]
+ pub fn cpus(&mut self) -> &mut cpumask::Cpumask {
+ // SAFETY: The pointer to `cpus` is valid for writing and remains valid for the lifetime of
+ // the returned reference.
+ unsafe { cpumask::CpumaskVar::as_mut_ref(&mut self.as_mut_ref().cpus) }
+ }
+
+ /// Sets clock for the [`Policy`].
+ ///
+ /// # Safety
+ ///
+ /// The caller must guarantee that the returned [`Clk`] is not dropped while it is getting used
+ /// by the C code.
+ #[cfg(CONFIG_COMMON_CLK)]
+ pub unsafe fn set_clk(&mut self, dev: &Device, name: Option<&CStr>) -> Result<Clk> {
+ let clk = Clk::get(dev, name)?;
+ self.as_mut_ref().clk = clk.as_raw();
+ Ok(clk)
+ }
+
+ /// Allows / disallows frequency switching code to run on any CPU.
+ #[inline]
+ pub fn set_dvfs_possible_from_any_cpu(&mut self, val: bool) -> &mut Self {
+ self.as_mut_ref().dvfs_possible_from_any_cpu = val;
+ self
+ }
+
+ /// Returns if fast switching of frequencies is possible or not.
+ #[inline]
+ pub fn fast_switch_possible(&self) -> bool {
+ self.as_ref().fast_switch_possible
+ }
+
+ /// Enables / disables fast frequency switching.
+ #[inline]
+ pub fn set_fast_switch_possible(&mut self, val: bool) -> &mut Self {
+ self.as_mut_ref().fast_switch_possible = val;
+ self
+ }
+
+ /// Sets transition latency (in nanoseconds) for the [`Policy`].
+ #[inline]
+ pub fn set_transition_latency_ns(&mut self, latency_ns: u32) -> &mut Self {
+ self.as_mut_ref().cpuinfo.transition_latency = latency_ns;
+ self
+ }
+
+ /// Sets cpuinfo `min_freq`.
+ #[inline]
+ pub fn set_cpuinfo_min_freq(&mut self, min_freq: Hertz) -> &mut Self {
+ self.as_mut_ref().cpuinfo.min_freq = min_freq.as_khz() as u32;
+ self
+ }
+
+ /// Sets cpuinfo `max_freq`.
+ #[inline]
+ pub fn set_cpuinfo_max_freq(&mut self, max_freq: Hertz) -> &mut Self {
+ self.as_mut_ref().cpuinfo.max_freq = max_freq.as_khz() as u32;
+ self
+ }
+
+ /// Set `transition_delay_us`, i.e. the minimum time between successive frequency change
+ /// requests.
+ #[inline]
+ pub fn set_transition_delay_us(&mut self, transition_delay_us: u32) -> &mut Self {
+ self.as_mut_ref().transition_delay_us = transition_delay_us;
+ self
+ }
+
+ /// Returns reference to the CPU frequency [`Table`] for the [`Policy`].
+ pub fn freq_table(&self) -> Result<&Table> {
+ if self.as_ref().freq_table.is_null() {
+ return Err(EINVAL);
+ }
+
+ // SAFETY: The `freq_table` is guaranteed to be valid for reading and remains valid for the
+ // lifetime of the returned reference.
+ Ok(unsafe { Table::from_raw(self.as_ref().freq_table) })
+ }
+
+ /// Sets the CPU frequency [`Table`] for the [`Policy`].
+ ///
+ /// # Safety
+ ///
+ /// The caller must guarantee that the [`Table`] is not dropped while it is getting used by the
+ /// C code.
+ #[inline]
+ pub unsafe fn set_freq_table(&mut self, table: &Table) -> &mut Self {
+ self.as_mut_ref().freq_table = table.as_raw();
+ self
+ }
+
+ /// Returns the [`Policy`]'s private data.
+ pub fn data<T: ForeignOwnable>(&mut self) -> Option<<T>::Borrowed<'_>> {
+ if self.as_ref().driver_data.is_null() {
+ None
+ } else {
+ // SAFETY: The data is earlier set from [`set_data`].
+ Some(unsafe { T::borrow(self.as_ref().driver_data) })
+ }
+ }
+
+ /// Sets the private data of the [`Policy`] using a foreign-ownable wrapper.
+ ///
+ /// # Errors
+ ///
+ /// Returns `EBUSY` if private data is already set.
+ fn set_data<T: ForeignOwnable>(&mut self, data: T) -> Result {
+ if self.as_ref().driver_data.is_null() {
+ // Transfer the ownership of the data to the foreign interface.
+ self.as_mut_ref().driver_data = <T as ForeignOwnable>::into_foreign(data) as _;
+ Ok(())
+ } else {
+ Err(EBUSY)
+ }
+ }
+
+ /// Clears and returns ownership of the private data.
+ fn clear_data<T: ForeignOwnable>(&mut self) -> Option<T> {
+ if self.as_ref().driver_data.is_null() {
+ None
+ } else {
+ let data = Some(
+ // SAFETY: The data is earlier set by us from [`set_data`]. It is safe to take
+ // back the ownership of the data from the foreign interface.
+ unsafe { <T as ForeignOwnable>::from_foreign(self.as_ref().driver_data) },
+ );
+ self.as_mut_ref().driver_data = ptr::null_mut();
+ data
+ }
+ }
+}
+
+/// CPU frequency policy created from a CPU number.
+///
+/// This struct represents the CPU frequency policy obtained for a specific CPU, providing safe
+/// access to the underlying `cpufreq_policy` and ensuring proper cleanup when the `PolicyCpu` is
+/// dropped.
+struct PolicyCpu<'a>(&'a mut Policy);
+
+impl<'a> PolicyCpu<'a> {
+ fn from_cpu(cpu: u32) -> Result<Self> {
+ // SAFETY: It is safe to call `cpufreq_cpu_get` for any valid CPU.
+ let ptr = from_err_ptr(unsafe { bindings::cpufreq_cpu_get(cpu) })?;
+
+ Ok(Self(
+ // SAFETY: The `ptr` is guaranteed to be valid and remains valid for the lifetime of
+ // the returned reference.
+ unsafe { Policy::from_raw_mut(ptr) },
+ ))
+ }
+}
+
+impl<'a> Deref for PolicyCpu<'a> {
+ type Target = Policy;
+
+ fn deref(&self) -> &Self::Target {
+ self.0
+ }
+}
+
+impl<'a> DerefMut for PolicyCpu<'a> {
+ fn deref_mut(&mut self) -> &mut Policy {
+ self.0
+ }
+}
+
+impl<'a> Drop for PolicyCpu<'a> {
+ fn drop(&mut self) {
+ // SAFETY: The underlying pointer is guaranteed to be valid for the lifetime of `self`.
+ unsafe { bindings::cpufreq_cpu_put(self.0.as_raw()) };
+ }
+}
+
+/// CPU frequency driver.
+///
+/// Implement this trait to provide a CPU frequency driver and its callbacks.
+///
+/// Reference: <https://docs.kernel.org/cpu-freq/cpu-drivers.html>
+#[vtable]
+pub trait Driver {
+ /// Driver's name.
+ const NAME: &'static CStr;
+
+ /// Driver's flags.
+ const FLAGS: u16;
+
+ /// Boost support.
+ const BOOST_ENABLED: bool;
+
+ /// Policy specific data.
+ ///
+ /// Require that `PData` implements `ForeignOwnable`. We guarantee to never move the underlying
+ /// wrapped data structure.
+ type PData: ForeignOwnable;
+
+ /// Driver's `init` callback.
+ fn init(policy: &mut Policy) -> Result<Self::PData>;
+
+ /// Driver's `exit` callback.
+ fn exit(_policy: &mut Policy, _data: Option<Self::PData>) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `online` callback.
+ fn online(_policy: &mut Policy) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `offline` callback.
+ fn offline(_policy: &mut Policy) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `suspend` callback.
+ fn suspend(_policy: &mut Policy) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `resume` callback.
+ fn resume(_policy: &mut Policy) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `ready` callback.
+ fn ready(_policy: &mut Policy) {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `verify` callback.
+ fn verify(data: &mut PolicyData) -> Result;
+
+ /// Driver's `setpolicy` callback.
+ fn setpolicy(_policy: &mut Policy) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `target` callback.
+ fn target(_policy: &mut Policy, _target_freq: u32, _relation: Relation) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `target_index` callback.
+ fn target_index(_policy: &mut Policy, _index: TableIndex) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `fast_switch` callback.
+ fn fast_switch(_policy: &mut Policy, _target_freq: u32) -> u32 {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `adjust_perf` callback.
+ fn adjust_perf(_policy: &mut Policy, _min_perf: usize, _target_perf: usize, _capacity: usize) {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `get_intermediate` callback.
+ fn get_intermediate(_policy: &mut Policy, _index: TableIndex) -> u32 {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `target_intermediate` callback.
+ fn target_intermediate(_policy: &mut Policy, _index: TableIndex) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `get` callback.
+ fn get(_policy: &mut Policy) -> Result<u32> {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `update_limits` callback.
+ fn update_limits(_policy: &mut Policy) {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `bios_limit` callback.
+ fn bios_limit(_policy: &mut Policy, _limit: &mut u32) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `set_boost` callback.
+ fn set_boost(_policy: &mut Policy, _state: i32) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// Driver's `register_em` callback.
+ fn register_em(_policy: &mut Policy) {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+}
+
+/// CPU frequency driver Registration.
+///
+/// ## Examples
+///
+/// The following example demonstrates how to register a cpufreq driver.
+///
+/// ```
+/// use kernel::{
+/// cpufreq,
+/// c_str,
+/// device::{Core, Device},
+/// macros::vtable,
+/// of, platform,
+/// sync::Arc,
+/// };
+/// struct SampleDevice;
+///
+/// #[derive(Default)]
+/// struct SampleDriver;
+///
+/// #[vtable]
+/// impl cpufreq::Driver for SampleDriver {
+/// const NAME: &'static CStr = c_str!("cpufreq-sample");
+/// const FLAGS: u16 = cpufreq::flags::NEED_INITIAL_FREQ_CHECK | cpufreq::flags::IS_COOLING_DEV;
+/// const BOOST_ENABLED: bool = true;
+///
+/// type PData = Arc<SampleDevice>;
+///
+/// fn init(policy: &mut cpufreq::Policy) -> Result<Self::PData> {
+/// // Initialize here
+/// Ok(Arc::new(SampleDevice, GFP_KERNEL)?)
+/// }
+///
+/// fn exit(_policy: &mut cpufreq::Policy, _data: Option<Self::PData>) -> Result {
+/// Ok(())
+/// }
+///
+/// fn suspend(policy: &mut cpufreq::Policy) -> Result {
+/// policy.generic_suspend()
+/// }
+///
+/// fn verify(data: &mut cpufreq::PolicyData) -> Result {
+/// data.generic_verify()
+/// }
+///
+/// fn target_index(policy: &mut cpufreq::Policy, index: cpufreq::TableIndex) -> Result {
+/// // Update CPU frequency
+/// Ok(())
+/// }
+///
+/// fn get(policy: &mut cpufreq::Policy) -> Result<u32> {
+/// policy.generic_get()
+/// }
+/// }
+///
+/// impl platform::Driver for SampleDriver {
+/// type IdInfo = ();
+/// const OF_ID_TABLE: Option<of::IdTable<Self::IdInfo>> = None;
+///
+/// fn probe(
+/// pdev: &platform::Device<Core>,
+/// _id_info: Option<&Self::IdInfo>,
+/// ) -> Result<Pin<KBox<Self>>> {
+/// cpufreq::Registration::<SampleDriver>::new_foreign_owned(pdev.as_ref())?;
+/// Ok(KBox::new(Self {}, GFP_KERNEL)?.into())
+/// }
+/// }
+/// ```
+#[repr(transparent)]
+pub struct Registration<T: Driver>(KBox<UnsafeCell<bindings::cpufreq_driver>>, PhantomData<T>);
+
+/// SAFETY: `Registration` doesn't offer any methods or access to fields when shared between threads
+/// or CPUs, so it is safe to share it.
+unsafe impl<T: Driver> Sync for Registration<T> {}
+
+#[allow(clippy::non_send_fields_in_send_ty)]
+/// SAFETY: Registration with and unregistration from the cpufreq subsystem can happen from any
+/// thread.
+unsafe impl<T: Driver> Send for Registration<T> {}
+
+impl<T: Driver> Registration<T> {
+ const VTABLE: bindings::cpufreq_driver = bindings::cpufreq_driver {
+ name: Self::copy_name(T::NAME),
+ boost_enabled: T::BOOST_ENABLED,
+ flags: T::FLAGS,
+
+ // Initialize mandatory callbacks.
+ init: Some(Self::init_callback),
+ verify: Some(Self::verify_callback),
+
+ // Initialize optional callbacks based on the traits of `T`.
+ setpolicy: if T::HAS_SETPOLICY {
+ Some(Self::setpolicy_callback)
+ } else {
+ None
+ },
+ target: if T::HAS_TARGET {
+ Some(Self::target_callback)
+ } else {
+ None
+ },
+ target_index: if T::HAS_TARGET_INDEX {
+ Some(Self::target_index_callback)
+ } else {
+ None
+ },
+ fast_switch: if T::HAS_FAST_SWITCH {
+ Some(Self::fast_switch_callback)
+ } else {
+ None
+ },
+ adjust_perf: if T::HAS_ADJUST_PERF {
+ Some(Self::adjust_perf_callback)
+ } else {
+ None
+ },
+ get_intermediate: if T::HAS_GET_INTERMEDIATE {
+ Some(Self::get_intermediate_callback)
+ } else {
+ None
+ },
+ target_intermediate: if T::HAS_TARGET_INTERMEDIATE {
+ Some(Self::target_intermediate_callback)
+ } else {
+ None
+ },
+ get: if T::HAS_GET {
+ Some(Self::get_callback)
+ } else {
+ None
+ },
+ update_limits: if T::HAS_UPDATE_LIMITS {
+ Some(Self::update_limits_callback)
+ } else {
+ None
+ },
+ bios_limit: if T::HAS_BIOS_LIMIT {
+ Some(Self::bios_limit_callback)
+ } else {
+ None
+ },
+ online: if T::HAS_ONLINE {
+ Some(Self::online_callback)
+ } else {
+ None
+ },
+ offline: if T::HAS_OFFLINE {
+ Some(Self::offline_callback)
+ } else {
+ None
+ },
+ exit: if T::HAS_EXIT {
+ Some(Self::exit_callback)
+ } else {
+ None
+ },
+ suspend: if T::HAS_SUSPEND {
+ Some(Self::suspend_callback)
+ } else {
+ None
+ },
+ resume: if T::HAS_RESUME {
+ Some(Self::resume_callback)
+ } else {
+ None
+ },
+ ready: if T::HAS_READY {
+ Some(Self::ready_callback)
+ } else {
+ None
+ },
+ set_boost: if T::HAS_SET_BOOST {
+ Some(Self::set_boost_callback)
+ } else {
+ None
+ },
+ register_em: if T::HAS_REGISTER_EM {
+ Some(Self::register_em_callback)
+ } else {
+ None
+ },
+ // SAFETY: All zeros is a valid value for `bindings::cpufreq_driver`.
+ ..unsafe { MaybeUninit::zeroed().assume_init() }
+ };
+
+ const fn copy_name(name: &'static CStr) -> [c_char; CPUFREQ_NAME_LEN] {
+ let src = name.as_bytes_with_nul();
+ let mut dst = [0; CPUFREQ_NAME_LEN];
+
+ build_assert!(src.len() <= CPUFREQ_NAME_LEN);
+
+ let mut i = 0;
+ while i < src.len() {
+ dst[i] = src[i];
+ i += 1;
+ }
+
+ dst
+ }
+
+ /// Registers a CPU frequency driver with the cpufreq core.
+ pub fn new() -> Result<Self> {
+ // We can't use `&Self::VTABLE` directly because the cpufreq core modifies some fields in
+ // the C `struct cpufreq_driver`, which requires a mutable reference.
+ let mut drv = KBox::new(UnsafeCell::new(Self::VTABLE), GFP_KERNEL)?;
+
+ // SAFETY: `drv` is guaranteed to be valid for the lifetime of `Registration`.
+ to_result(unsafe { bindings::cpufreq_register_driver(drv.get_mut()) })?;
+
+ Ok(Self(drv, PhantomData))
+ }
+
+ /// Same as [`Registration::new`], but does not return a [`Registration`] instance.
+ ///
+ /// Instead the [`Registration`] is owned by [`Devres`] and will be revoked / dropped, once the
+ /// device is detached.
+ pub fn new_foreign_owned(dev: &Device<Bound>) -> Result {
+ Devres::new_foreign_owned(dev, Self::new()?, GFP_KERNEL)
+ }
+}
+
+/// CPU frequency driver callbacks.
+impl<T: Driver> Registration<T> {
+ /// Driver's `init` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn init_callback(ptr: *mut bindings::cpufreq_policy) -> kernel::ffi::c_int {
+ from_result(|| {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+
+ let data = T::init(policy)?;
+ policy.set_data(data)?;
+ Ok(0)
+ })
+ }
+
+ /// Driver's `exit` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn exit_callback(ptr: *mut bindings::cpufreq_policy) {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+
+ let data = policy.clear_data();
+ let _ = T::exit(policy, data);
+ }
+
+ /// Driver's `online` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn online_callback(ptr: *mut bindings::cpufreq_policy) -> kernel::ffi::c_int {
+ from_result(|| {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+ T::online(policy).map(|()| 0)
+ })
+ }
+
+ /// Driver's `offline` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn offline_callback(ptr: *mut bindings::cpufreq_policy) -> kernel::ffi::c_int {
+ from_result(|| {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+ T::offline(policy).map(|()| 0)
+ })
+ }
+
+ /// Driver's `suspend` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn suspend_callback(ptr: *mut bindings::cpufreq_policy) -> kernel::ffi::c_int {
+ from_result(|| {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+ T::suspend(policy).map(|()| 0)
+ })
+ }
+
+ /// Driver's `resume` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn resume_callback(ptr: *mut bindings::cpufreq_policy) -> kernel::ffi::c_int {
+ from_result(|| {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+ T::resume(policy).map(|()| 0)
+ })
+ }
+
+ /// Driver's `ready` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn ready_callback(ptr: *mut bindings::cpufreq_policy) {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+ T::ready(policy);
+ }
+
+ /// Driver's `verify` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn verify_callback(ptr: *mut bindings::cpufreq_policy_data) -> kernel::ffi::c_int {
+ from_result(|| {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let data = unsafe { PolicyData::from_raw_mut(ptr) };
+ T::verify(data).map(|()| 0)
+ })
+ }
+
+ /// Driver's `setpolicy` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn setpolicy_callback(ptr: *mut bindings::cpufreq_policy) -> kernel::ffi::c_int {
+ from_result(|| {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+ T::setpolicy(policy).map(|()| 0)
+ })
+ }
+
+ /// Driver's `target` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn target_callback(
+ ptr: *mut bindings::cpufreq_policy,
+ target_freq: u32,
+ relation: u32,
+ ) -> kernel::ffi::c_int {
+ from_result(|| {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+ T::target(policy, target_freq, Relation::new(relation)?).map(|()| 0)
+ })
+ }
+
+ /// Driver's `target_index` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn target_index_callback(
+ ptr: *mut bindings::cpufreq_policy,
+ index: u32,
+ ) -> kernel::ffi::c_int {
+ from_result(|| {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+
+ // SAFETY: The C code guarantees that `index` corresponds to a valid entry in the
+ // frequency table.
+ let index = unsafe { TableIndex::new(index as usize) };
+
+ T::target_index(policy, index).map(|()| 0)
+ })
+ }
+
+ /// Driver's `fast_switch` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn fast_switch_callback(
+ ptr: *mut bindings::cpufreq_policy,
+ target_freq: u32,
+ ) -> kernel::ffi::c_uint {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+ T::fast_switch(policy, target_freq)
+ }
+
+ /// Driver's `adjust_perf` callback.
+ extern "C" fn adjust_perf_callback(
+ cpu: u32,
+ min_perf: usize,
+ target_perf: usize,
+ capacity: usize,
+ ) {
+ if let Ok(mut policy) = PolicyCpu::from_cpu(cpu) {
+ T::adjust_perf(&mut policy, min_perf, target_perf, capacity);
+ }
+ }
+
+ /// Driver's `get_intermediate` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn get_intermediate_callback(
+ ptr: *mut bindings::cpufreq_policy,
+ index: u32,
+ ) -> kernel::ffi::c_uint {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+
+ // SAFETY: The C code guarantees that `index` corresponds to a valid entry in the
+ // frequency table.
+ let index = unsafe { TableIndex::new(index as usize) };
+
+ T::get_intermediate(policy, index)
+ }
+
+ /// Driver's `target_intermediate` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn target_intermediate_callback(
+ ptr: *mut bindings::cpufreq_policy,
+ index: u32,
+ ) -> kernel::ffi::c_int {
+ from_result(|| {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+
+ // SAFETY: The C code guarantees that `index` corresponds to a valid entry in the
+ // frequency table.
+ let index = unsafe { TableIndex::new(index as usize) };
+
+ T::target_intermediate(policy, index).map(|()| 0)
+ })
+ }
+
+ /// Driver's `get` callback.
+ extern "C" fn get_callback(cpu: u32) -> kernel::ffi::c_uint {
+ PolicyCpu::from_cpu(cpu).map_or(0, |mut policy| T::get(&mut policy).map_or(0, |f| f))
+ }
+
+ /// Driver's `update_limit` callback.
+ extern "C" fn update_limits_callback(ptr: *mut bindings::cpufreq_policy) {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+ T::update_limits(policy);
+ }
+
+ /// Driver's `bios_limit` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn bios_limit_callback(cpu: i32, limit: *mut u32) -> kernel::ffi::c_int {
+ from_result(|| {
+ let mut policy = PolicyCpu::from_cpu(cpu as u32)?;
+
+ // SAFETY: `limit` is guaranteed by the C code to be valid.
+ T::bios_limit(&mut policy, &mut (unsafe { *limit })).map(|()| 0)
+ })
+ }
+
+ /// Driver's `set_boost` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn set_boost_callback(
+ ptr: *mut bindings::cpufreq_policy,
+ state: i32,
+ ) -> kernel::ffi::c_int {
+ from_result(|| {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+ T::set_boost(policy, state).map(|()| 0)
+ })
+ }
+
+ /// Driver's `register_em` callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn register_em_callback(ptr: *mut bindings::cpufreq_policy) {
+ // SAFETY: The `ptr` is guaranteed to be valid by the contract with the C code for the
+ // lifetime of `policy`.
+ let policy = unsafe { Policy::from_raw_mut(ptr) };
+ T::register_em(policy);
+ }
+}
+
+impl<T: Driver> Drop for Registration<T> {
+ /// Unregisters with the cpufreq core.
+ fn drop(&mut self) {
+ // SAFETY: `self.0` is guaranteed to be valid for the lifetime of `Registration`.
+ unsafe { bindings::cpufreq_unregister_driver(self.0.get_mut()) };
+ }
+}
diff --git a/rust/kernel/cpumask.rs b/rust/kernel/cpumask.rs
new file mode 100644
index 000000000000..c90bfac9346a
--- /dev/null
+++ b/rust/kernel/cpumask.rs
@@ -0,0 +1,330 @@
+// SPDX-License-Identifier: GPL-2.0
+
+//! CPU Mask abstractions.
+//!
+//! C header: [`include/linux/cpumask.h`](srctree/include/linux/cpumask.h)
+
+use crate::{
+ alloc::{AllocError, Flags},
+ prelude::*,
+ types::Opaque,
+};
+
+#[cfg(CONFIG_CPUMASK_OFFSTACK)]
+use core::ptr::{self, NonNull};
+
+#[cfg(not(CONFIG_CPUMASK_OFFSTACK))]
+use core::mem::MaybeUninit;
+
+use core::ops::{Deref, DerefMut};
+
+/// A CPU Mask.
+///
+/// Rust abstraction for the C `struct cpumask`.
+///
+/// # Invariants
+///
+/// A [`Cpumask`] instance always corresponds to a valid C `struct cpumask`.
+///
+/// The callers must ensure that the `struct cpumask` is valid for access and
+/// remains valid for the lifetime of the returned reference.
+///
+/// ## Examples
+///
+/// The following example demonstrates how to update a [`Cpumask`].
+///
+/// ```
+/// use kernel::bindings;
+/// use kernel::cpumask::Cpumask;
+///
+/// fn set_clear_cpu(ptr: *mut bindings::cpumask, set_cpu: u32, clear_cpu: i32) {
+/// // SAFETY: The `ptr` is valid for writing and remains valid for the lifetime of the
+/// // returned reference.
+/// let mask = unsafe { Cpumask::as_mut_ref(ptr) };
+///
+/// mask.set(set_cpu);
+/// mask.clear(clear_cpu);
+/// }
+/// ```
+#[repr(transparent)]
+pub struct Cpumask(Opaque<bindings::cpumask>);
+
+impl Cpumask {
+ /// Creates a mutable reference to an existing `struct cpumask` pointer.
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that `ptr` is valid for writing and remains valid for the lifetime
+ /// of the returned reference.
+ pub unsafe fn as_mut_ref<'a>(ptr: *mut bindings::cpumask) -> &'a mut Self {
+ // SAFETY: Guaranteed by the safety requirements of the function.
+ //
+ // INVARIANT: The caller ensures that `ptr` is valid for writing and remains valid for the
+ // lifetime of the returned reference.
+ unsafe { &mut *ptr.cast() }
+ }
+
+ /// Creates a reference to an existing `struct cpumask` pointer.
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that `ptr` is valid for reading and remains valid for the lifetime
+ /// of the returned reference.
+ pub unsafe fn as_ref<'a>(ptr: *const bindings::cpumask) -> &'a Self {
+ // SAFETY: Guaranteed by the safety requirements of the function.
+ //
+ // INVARIANT: The caller ensures that `ptr` is valid for reading and remains valid for the
+ // lifetime of the returned reference.
+ unsafe { &*ptr.cast() }
+ }
+
+ /// Obtain the raw `struct cpumask` pointer.
+ pub fn as_raw(&self) -> *mut bindings::cpumask {
+ let this: *const Self = self;
+ this.cast_mut().cast()
+ }
+
+ /// Set `cpu` in the cpumask.
+ ///
+ /// ATTENTION: Contrary to C, this Rust `set()` method is non-atomic.
+ /// This mismatches kernel naming convention and corresponds to the C
+ /// function `__cpumask_set_cpu()`.
+ #[inline]
+ pub fn set(&mut self, cpu: u32) {
+ // SAFETY: By the type invariant, `self.as_raw` is a valid argument to `__cpumask_set_cpu`.
+ unsafe { bindings::__cpumask_set_cpu(cpu, self.as_raw()) };
+ }
+
+ /// Clear `cpu` in the cpumask.
+ ///
+ /// ATTENTION: Contrary to C, this Rust `clear()` method is non-atomic.
+ /// This mismatches kernel naming convention and corresponds to the C
+ /// function `__cpumask_clear_cpu()`.
+ #[inline]
+ pub fn clear(&mut self, cpu: i32) {
+ // SAFETY: By the type invariant, `self.as_raw` is a valid argument to
+ // `__cpumask_clear_cpu`.
+ unsafe { bindings::__cpumask_clear_cpu(cpu, self.as_raw()) };
+ }
+
+ /// Test `cpu` in the cpumask.
+ ///
+ /// Equivalent to the kernel's `cpumask_test_cpu` API.
+ #[inline]
+ pub fn test(&self, cpu: i32) -> bool {
+ // SAFETY: By the type invariant, `self.as_raw` is a valid argument to `cpumask_test_cpu`.
+ unsafe { bindings::cpumask_test_cpu(cpu, self.as_raw()) }
+ }
+
+ /// Set all CPUs in the cpumask.
+ ///
+ /// Equivalent to the kernel's `cpumask_setall` API.
+ #[inline]
+ pub fn setall(&mut self) {
+ // SAFETY: By the type invariant, `self.as_raw` is a valid argument to `cpumask_setall`.
+ unsafe { bindings::cpumask_setall(self.as_raw()) };
+ }
+
+ /// Checks if cpumask is empty.
+ ///
+ /// Equivalent to the kernel's `cpumask_empty` API.
+ #[inline]
+ pub fn empty(&self) -> bool {
+ // SAFETY: By the type invariant, `self.as_raw` is a valid argument to `cpumask_empty`.
+ unsafe { bindings::cpumask_empty(self.as_raw()) }
+ }
+
+ /// Checks if cpumask is full.
+ ///
+ /// Equivalent to the kernel's `cpumask_full` API.
+ #[inline]
+ pub fn full(&self) -> bool {
+ // SAFETY: By the type invariant, `self.as_raw` is a valid argument to `cpumask_full`.
+ unsafe { bindings::cpumask_full(self.as_raw()) }
+ }
+
+ /// Get weight of the cpumask.
+ ///
+ /// Equivalent to the kernel's `cpumask_weight` API.
+ #[inline]
+ pub fn weight(&self) -> u32 {
+ // SAFETY: By the type invariant, `self.as_raw` is a valid argument to `cpumask_weight`.
+ unsafe { bindings::cpumask_weight(self.as_raw()) }
+ }
+
+ /// Copy cpumask.
+ ///
+ /// Equivalent to the kernel's `cpumask_copy` API.
+ #[inline]
+ pub fn copy(&self, dstp: &mut Self) {
+ // SAFETY: By the type invariant, `Self::as_raw` is a valid argument to `cpumask_copy`.
+ unsafe { bindings::cpumask_copy(dstp.as_raw(), self.as_raw()) };
+ }
+}
+
+/// A CPU Mask pointer.
+///
+/// Rust abstraction for the C `struct cpumask_var_t`.
+///
+/// # Invariants
+///
+/// A [`CpumaskVar`] instance always corresponds to a valid C `struct cpumask_var_t`.
+///
+/// The callers must ensure that the `struct cpumask_var_t` is valid for access and remains valid
+/// for the lifetime of [`CpumaskVar`].
+///
+/// ## Examples
+///
+/// The following example demonstrates how to create and update a [`CpumaskVar`].
+///
+/// ```
+/// use kernel::cpumask::CpumaskVar;
+///
+/// let mut mask = CpumaskVar::new_zero(GFP_KERNEL).unwrap();
+///
+/// assert!(mask.empty());
+/// mask.set(2);
+/// assert!(mask.test(2));
+/// mask.set(3);
+/// assert!(mask.test(3));
+/// assert_eq!(mask.weight(), 2);
+///
+/// let mask2 = CpumaskVar::try_clone(&mask).unwrap();
+/// assert!(mask2.test(2));
+/// assert!(mask2.test(3));
+/// assert_eq!(mask2.weight(), 2);
+/// ```
+pub struct CpumaskVar {
+ #[cfg(CONFIG_CPUMASK_OFFSTACK)]
+ ptr: NonNull<Cpumask>,
+ #[cfg(not(CONFIG_CPUMASK_OFFSTACK))]
+ mask: Cpumask,
+}
+
+impl CpumaskVar {
+ /// Creates a zero-initialized instance of the [`CpumaskVar`].
+ pub fn new_zero(_flags: Flags) -> Result<Self, AllocError> {
+ Ok(Self {
+ #[cfg(CONFIG_CPUMASK_OFFSTACK)]
+ ptr: {
+ let mut ptr: *mut bindings::cpumask = ptr::null_mut();
+
+ // SAFETY: It is safe to call this method as the reference to `ptr` is valid.
+ //
+ // INVARIANT: The associated memory is freed when the `CpumaskVar` goes out of
+ // scope.
+ unsafe { bindings::zalloc_cpumask_var(&mut ptr, _flags.as_raw()) };
+ NonNull::new(ptr.cast()).ok_or(AllocError)?
+ },
+
+ #[cfg(not(CONFIG_CPUMASK_OFFSTACK))]
+ // SAFETY: FFI type is valid to be zero-initialized.
+ //
+ // INVARIANT: The associated memory is freed when the `CpumaskVar` goes out of scope.
+ mask: unsafe { core::mem::zeroed() },
+ })
+ }
+
+ /// Creates an instance of the [`CpumaskVar`].
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that the returned [`CpumaskVar`] is properly initialized before
+ /// getting used.
+ pub unsafe fn new(_flags: Flags) -> Result<Self, AllocError> {
+ Ok(Self {
+ #[cfg(CONFIG_CPUMASK_OFFSTACK)]
+ ptr: {
+ let mut ptr: *mut bindings::cpumask = ptr::null_mut();
+
+ // SAFETY: It is safe to call this method as the reference to `ptr` is valid.
+ //
+ // INVARIANT: The associated memory is freed when the `CpumaskVar` goes out of
+ // scope.
+ unsafe { bindings::alloc_cpumask_var(&mut ptr, _flags.as_raw()) };
+ NonNull::new(ptr.cast()).ok_or(AllocError)?
+ },
+ #[cfg(not(CONFIG_CPUMASK_OFFSTACK))]
+ // SAFETY: Guaranteed by the safety requirements of the function.
+ //
+ // INVARIANT: The associated memory is freed when the `CpumaskVar` goes out of scope.
+ mask: unsafe { MaybeUninit::uninit().assume_init() },
+ })
+ }
+
+ /// Creates a mutable reference to an existing `struct cpumask_var_t` pointer.
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that `ptr` is valid for writing and remains valid for the lifetime
+ /// of the returned reference.
+ pub unsafe fn as_mut_ref<'a>(ptr: *mut bindings::cpumask_var_t) -> &'a mut Self {
+ // SAFETY: Guaranteed by the safety requirements of the function.
+ //
+ // INVARIANT: The caller ensures that `ptr` is valid for writing and remains valid for the
+ // lifetime of the returned reference.
+ unsafe { &mut *ptr.cast() }
+ }
+
+ /// Creates a reference to an existing `struct cpumask_var_t` pointer.
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that `ptr` is valid for reading and remains valid for the lifetime
+ /// of the returned reference.
+ pub unsafe fn as_ref<'a>(ptr: *const bindings::cpumask_var_t) -> &'a Self {
+ // SAFETY: Guaranteed by the safety requirements of the function.
+ //
+ // INVARIANT: The caller ensures that `ptr` is valid for reading and remains valid for the
+ // lifetime of the returned reference.
+ unsafe { &*ptr.cast() }
+ }
+
+ /// Clones cpumask.
+ pub fn try_clone(cpumask: &Cpumask) -> Result<Self> {
+ // SAFETY: The returned cpumask_var is initialized right after this call.
+ let mut cpumask_var = unsafe { Self::new(GFP_KERNEL) }?;
+
+ cpumask.copy(&mut cpumask_var);
+ Ok(cpumask_var)
+ }
+}
+
+// Make [`CpumaskVar`] behave like a pointer to [`Cpumask`].
+impl Deref for CpumaskVar {
+ type Target = Cpumask;
+
+ #[cfg(CONFIG_CPUMASK_OFFSTACK)]
+ fn deref(&self) -> &Self::Target {
+ // SAFETY: The caller owns CpumaskVar, so it is safe to deref the cpumask.
+ unsafe { &*self.ptr.as_ptr() }
+ }
+
+ #[cfg(not(CONFIG_CPUMASK_OFFSTACK))]
+ fn deref(&self) -> &Self::Target {
+ &self.mask
+ }
+}
+
+impl DerefMut for CpumaskVar {
+ #[cfg(CONFIG_CPUMASK_OFFSTACK)]
+ fn deref_mut(&mut self) -> &mut Cpumask {
+ // SAFETY: The caller owns CpumaskVar, so it is safe to deref the cpumask.
+ unsafe { self.ptr.as_mut() }
+ }
+
+ #[cfg(not(CONFIG_CPUMASK_OFFSTACK))]
+ fn deref_mut(&mut self) -> &mut Cpumask {
+ &mut self.mask
+ }
+}
+
+impl Drop for CpumaskVar {
+ fn drop(&mut self) {
+ #[cfg(CONFIG_CPUMASK_OFFSTACK)]
+ // SAFETY: By the type invariant, `self.as_raw` is a valid argument to `free_cpumask_var`.
+ unsafe {
+ bindings::free_cpumask_var(self.as_raw())
+ };
+ }
+}
diff --git a/rust/kernel/lib.rs b/rust/kernel/lib.rs
index 1cf1bd5b73fc..911c72a0fc21 100644
--- a/rust/kernel/lib.rs
+++ b/rust/kernel/lib.rs
@@ -44,8 +44,13 @@ pub mod auxiliary;
pub mod block;
#[doc(hidden)]
pub mod build_assert;
+pub mod clk;
#[cfg(CONFIG_CONFIGFS_FS)]
pub mod configfs;
+pub mod cpu;
+#[cfg(CONFIG_CPU_FREQ)]
+pub mod cpufreq;
+pub mod cpumask;
pub mod cred;
pub mod device;
pub mod device_id;
@@ -67,9 +72,12 @@ pub mod jump_label;
pub mod kunit;
pub mod list;
pub mod miscdevice;
+pub mod mm;
#[cfg(CONFIG_NET)]
pub mod net;
pub mod of;
+#[cfg(CONFIG_PM_OPP)]
+pub mod opp;
pub mod page;
#[cfg(CONFIG_PCI)]
pub mod pci;
diff --git a/rust/kernel/miscdevice.rs b/rust/kernel/miscdevice.rs
index fa9ecc42602a..9d9771247c38 100644
--- a/rust/kernel/miscdevice.rs
+++ b/rust/kernel/miscdevice.rs
@@ -14,6 +14,7 @@ use crate::{
error::{to_result, Error, Result, VTABLE_DEFAULT_ERROR},
ffi::{c_int, c_long, c_uint, c_ulong},
fs::File,
+ mm::virt::VmaNew,
prelude::*,
seq_file::SeqFile,
str::CStr,
@@ -119,6 +120,22 @@ pub trait MiscDevice: Sized {
drop(device);
}
+ /// Handle for mmap.
+ ///
+ /// This function is invoked when a user space process invokes the `mmap` system call on
+ /// `file`. The function is a callback that is part of the VMA initializer. The kernel will do
+ /// initial setup of the VMA before calling this function. The function can then interact with
+ /// the VMA initialization by calling methods of `vma`. If the function does not return an
+ /// error, the kernel will complete initialization of the VMA according to the properties of
+ /// `vma`.
+ fn mmap(
+ _device: <Self::Ptr as ForeignOwnable>::Borrowed<'_>,
+ _file: &File,
+ _vma: &VmaNew,
+ ) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
/// Handler for ioctls.
///
/// The `cmd` argument is usually manipulated using the utilties in [`kernel::ioctl`].
@@ -226,6 +243,33 @@ impl<T: MiscDevice> MiscdeviceVTable<T> {
/// # Safety
///
/// `file` must be a valid file that is associated with a `MiscDeviceRegistration<T>`.
+ /// `vma` must be a vma that is currently being mmap'ed with this file.
+ unsafe extern "C" fn mmap(
+ file: *mut bindings::file,
+ vma: *mut bindings::vm_area_struct,
+ ) -> c_int {
+ // SAFETY: The mmap call of a file can access the private data.
+ let private = unsafe { (*file).private_data };
+ // SAFETY: This is a Rust Miscdevice, so we call `into_foreign` in `open` and
+ // `from_foreign` in `release`, and `fops_mmap` is guaranteed to be called between those
+ // two operations.
+ let device = unsafe { <T::Ptr as ForeignOwnable>::borrow(private) };
+ // SAFETY: The caller provides a vma that is undergoing initial VMA setup.
+ let area = unsafe { VmaNew::from_raw(vma) };
+ // SAFETY:
+ // * The file is valid for the duration of this call.
+ // * There is no active fdget_pos region on the file on this thread.
+ let file = unsafe { File::from_raw_file(file) };
+
+ match T::mmap(device, file, area) {
+ Ok(()) => 0,
+ Err(err) => err.to_errno(),
+ }
+ }
+
+ /// # Safety
+ ///
+ /// `file` must be a valid file that is associated with a `MiscDeviceRegistration<T>`.
unsafe extern "C" fn ioctl(file: *mut bindings::file, cmd: c_uint, arg: c_ulong) -> c_long {
// SAFETY: The ioctl call of a file can access the private data.
let private = unsafe { (*file).private_data };
@@ -291,6 +335,7 @@ impl<T: MiscDevice> MiscdeviceVTable<T> {
const VTABLE: bindings::file_operations = bindings::file_operations {
open: Some(Self::open),
release: Some(Self::release),
+ mmap: if T::HAS_MMAP { Some(Self::mmap) } else { None },
unlocked_ioctl: if T::HAS_IOCTL {
Some(Self::ioctl)
} else {
diff --git a/rust/kernel/mm.rs b/rust/kernel/mm.rs
new file mode 100644
index 000000000000..615907a0f3b4
--- /dev/null
+++ b/rust/kernel/mm.rs
@@ -0,0 +1,344 @@
+// SPDX-License-Identifier: GPL-2.0
+
+// Copyright (C) 2024 Google LLC.
+
+//! Memory management.
+//!
+//! This module deals with managing the address space of userspace processes. Each process has an
+//! instance of [`Mm`], which keeps track of multiple VMAs (virtual memory areas). Each VMA
+//! corresponds to a region of memory that the userspace process can access, and the VMA lets you
+//! control what happens when userspace reads or writes to that region of memory.
+//!
+//! C header: [`include/linux/mm.h`](srctree/include/linux/mm.h)
+#![cfg(CONFIG_MMU)]
+
+use crate::{
+ bindings,
+ types::{ARef, AlwaysRefCounted, NotThreadSafe, Opaque},
+};
+use core::{ops::Deref, ptr::NonNull};
+
+pub mod virt;
+use virt::VmaRef;
+
+/// A wrapper for the kernel's `struct mm_struct`.
+///
+/// This represents the address space of a userspace process, so each process has one `Mm`
+/// instance. It may hold many VMAs internally.
+///
+/// There is a counter called `mm_users` that counts the users of the address space; this includes
+/// the userspace process itself, but can also include kernel threads accessing the address space.
+/// Once `mm_users` reaches zero, this indicates that the address space can be destroyed. To access
+/// the address space, you must prevent `mm_users` from reaching zero while you are accessing it.
+/// The [`MmWithUser`] type represents an address space where this is guaranteed, and you can
+/// create one using [`mmget_not_zero`].
+///
+/// The `ARef<Mm>` smart pointer holds an `mmgrab` refcount. Its destructor may sleep.
+///
+/// # Invariants
+///
+/// Values of this type are always refcounted using `mmgrab`.
+///
+/// [`mmget_not_zero`]: Mm::mmget_not_zero
+#[repr(transparent)]
+pub struct Mm {
+ mm: Opaque<bindings::mm_struct>,
+}
+
+// SAFETY: It is safe to call `mmdrop` on another thread than where `mmgrab` was called.
+unsafe impl Send for Mm {}
+// SAFETY: All methods on `Mm` can be called in parallel from several threads.
+unsafe impl Sync for Mm {}
+
+// SAFETY: By the type invariants, this type is always refcounted.
+unsafe impl AlwaysRefCounted for Mm {
+ #[inline]
+ fn inc_ref(&self) {
+ // SAFETY: The pointer is valid since self is a reference.
+ unsafe { bindings::mmgrab(self.as_raw()) };
+ }
+
+ #[inline]
+ unsafe fn dec_ref(obj: NonNull<Self>) {
+ // SAFETY: The caller is giving up their refcount.
+ unsafe { bindings::mmdrop(obj.cast().as_ptr()) };
+ }
+}
+
+/// A wrapper for the kernel's `struct mm_struct`.
+///
+/// This type is like [`Mm`], but with non-zero `mm_users`. It can only be used when `mm_users` can
+/// be proven to be non-zero at compile-time, usually because the relevant code holds an `mmget`
+/// refcount. It can be used to access the associated address space.
+///
+/// The `ARef<MmWithUser>` smart pointer holds an `mmget` refcount. Its destructor may sleep.
+///
+/// # Invariants
+///
+/// Values of this type are always refcounted using `mmget`. The value of `mm_users` is non-zero.
+#[repr(transparent)]
+pub struct MmWithUser {
+ mm: Mm,
+}
+
+// SAFETY: It is safe to call `mmput` on another thread than where `mmget` was called.
+unsafe impl Send for MmWithUser {}
+// SAFETY: All methods on `MmWithUser` can be called in parallel from several threads.
+unsafe impl Sync for MmWithUser {}
+
+// SAFETY: By the type invariants, this type is always refcounted.
+unsafe impl AlwaysRefCounted for MmWithUser {
+ #[inline]
+ fn inc_ref(&self) {
+ // SAFETY: The pointer is valid since self is a reference.
+ unsafe { bindings::mmget(self.as_raw()) };
+ }
+
+ #[inline]
+ unsafe fn dec_ref(obj: NonNull<Self>) {
+ // SAFETY: The caller is giving up their refcount.
+ unsafe { bindings::mmput(obj.cast().as_ptr()) };
+ }
+}
+
+// Make all `Mm` methods available on `MmWithUser`.
+impl Deref for MmWithUser {
+ type Target = Mm;
+
+ #[inline]
+ fn deref(&self) -> &Mm {
+ &self.mm
+ }
+}
+
+/// A wrapper for the kernel's `struct mm_struct`.
+///
+/// This type is identical to `MmWithUser` except that it uses `mmput_async` when dropping a
+/// refcount. This means that the destructor of `ARef<MmWithUserAsync>` is safe to call in atomic
+/// context.
+///
+/// # Invariants
+///
+/// Values of this type are always refcounted using `mmget`. The value of `mm_users` is non-zero.
+#[repr(transparent)]
+pub struct MmWithUserAsync {
+ mm: MmWithUser,
+}
+
+// SAFETY: It is safe to call `mmput_async` on another thread than where `mmget` was called.
+unsafe impl Send for MmWithUserAsync {}
+// SAFETY: All methods on `MmWithUserAsync` can be called in parallel from several threads.
+unsafe impl Sync for MmWithUserAsync {}
+
+// SAFETY: By the type invariants, this type is always refcounted.
+unsafe impl AlwaysRefCounted for MmWithUserAsync {
+ #[inline]
+ fn inc_ref(&self) {
+ // SAFETY: The pointer is valid since self is a reference.
+ unsafe { bindings::mmget(self.as_raw()) };
+ }
+
+ #[inline]
+ unsafe fn dec_ref(obj: NonNull<Self>) {
+ // SAFETY: The caller is giving up their refcount.
+ unsafe { bindings::mmput_async(obj.cast().as_ptr()) };
+ }
+}
+
+// Make all `MmWithUser` methods available on `MmWithUserAsync`.
+impl Deref for MmWithUserAsync {
+ type Target = MmWithUser;
+
+ #[inline]
+ fn deref(&self) -> &MmWithUser {
+ &self.mm
+ }
+}
+
+// These methods are safe to call even if `mm_users` is zero.
+impl Mm {
+ /// Returns a raw pointer to the inner `mm_struct`.
+ #[inline]
+ pub fn as_raw(&self) -> *mut bindings::mm_struct {
+ self.mm.get()
+ }
+
+ /// Obtain a reference from a raw pointer.
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that `ptr` points at an `mm_struct`, and that it is not deallocated
+ /// during the lifetime 'a.
+ #[inline]
+ pub unsafe fn from_raw<'a>(ptr: *const bindings::mm_struct) -> &'a Mm {
+ // SAFETY: Caller promises that the pointer is valid for 'a. Layouts are compatible due to
+ // repr(transparent).
+ unsafe { &*ptr.cast() }
+ }
+
+ /// Calls `mmget_not_zero` and returns a handle if it succeeds.
+ #[inline]
+ pub fn mmget_not_zero(&self) -> Option<ARef<MmWithUser>> {
+ // SAFETY: The pointer is valid since self is a reference.
+ let success = unsafe { bindings::mmget_not_zero(self.as_raw()) };
+
+ if success {
+ // SAFETY: We just created an `mmget` refcount.
+ Some(unsafe { ARef::from_raw(NonNull::new_unchecked(self.as_raw().cast())) })
+ } else {
+ None
+ }
+ }
+}
+
+// These methods require `mm_users` to be non-zero.
+impl MmWithUser {
+ /// Obtain a reference from a raw pointer.
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that `ptr` points at an `mm_struct`, and that `mm_users` remains
+ /// non-zero for the duration of the lifetime 'a.
+ #[inline]
+ pub unsafe fn from_raw<'a>(ptr: *const bindings::mm_struct) -> &'a MmWithUser {
+ // SAFETY: Caller promises that the pointer is valid for 'a. The layout is compatible due
+ // to repr(transparent).
+ unsafe { &*ptr.cast() }
+ }
+
+ /// Use `mmput_async` when dropping this refcount.
+ #[inline]
+ pub fn into_mmput_async(me: ARef<MmWithUser>) -> ARef<MmWithUserAsync> {
+ // SAFETY: The layouts and invariants are compatible.
+ unsafe { ARef::from_raw(ARef::into_raw(me).cast()) }
+ }
+
+ /// Attempt to access a vma using the vma read lock.
+ ///
+ /// This is an optimistic trylock operation, so it may fail if there is contention. In that
+ /// case, you should fall back to taking the mmap read lock.
+ ///
+ /// When per-vma locks are disabled, this always returns `None`.
+ #[inline]
+ pub fn lock_vma_under_rcu(&self, vma_addr: usize) -> Option<VmaReadGuard<'_>> {
+ #[cfg(CONFIG_PER_VMA_LOCK)]
+ {
+ // SAFETY: Calling `bindings::lock_vma_under_rcu` is always okay given an mm where
+ // `mm_users` is non-zero.
+ let vma = unsafe { bindings::lock_vma_under_rcu(self.as_raw(), vma_addr) };
+ if !vma.is_null() {
+ return Some(VmaReadGuard {
+ // SAFETY: If `lock_vma_under_rcu` returns a non-null ptr, then it points at a
+ // valid vma. The vma is stable for as long as the vma read lock is held.
+ vma: unsafe { VmaRef::from_raw(vma) },
+ _nts: NotThreadSafe,
+ });
+ }
+ }
+
+ // Silence warnings about unused variables.
+ #[cfg(not(CONFIG_PER_VMA_LOCK))]
+ let _ = vma_addr;
+
+ None
+ }
+
+ /// Lock the mmap read lock.
+ #[inline]
+ pub fn mmap_read_lock(&self) -> MmapReadGuard<'_> {
+ // SAFETY: The pointer is valid since self is a reference.
+ unsafe { bindings::mmap_read_lock(self.as_raw()) };
+
+ // INVARIANT: We just acquired the read lock.
+ MmapReadGuard {
+ mm: self,
+ _nts: NotThreadSafe,
+ }
+ }
+
+ /// Try to lock the mmap read lock.
+ #[inline]
+ pub fn mmap_read_trylock(&self) -> Option<MmapReadGuard<'_>> {
+ // SAFETY: The pointer is valid since self is a reference.
+ let success = unsafe { bindings::mmap_read_trylock(self.as_raw()) };
+
+ if success {
+ // INVARIANT: We just acquired the read lock.
+ Some(MmapReadGuard {
+ mm: self,
+ _nts: NotThreadSafe,
+ })
+ } else {
+ None
+ }
+ }
+}
+
+/// A guard for the mmap read lock.
+///
+/// # Invariants
+///
+/// This `MmapReadGuard` guard owns the mmap read lock.
+pub struct MmapReadGuard<'a> {
+ mm: &'a MmWithUser,
+ // `mmap_read_lock` and `mmap_read_unlock` must be called on the same thread
+ _nts: NotThreadSafe,
+}
+
+impl<'a> MmapReadGuard<'a> {
+ /// Look up a vma at the given address.
+ #[inline]
+ pub fn vma_lookup(&self, vma_addr: usize) -> Option<&virt::VmaRef> {
+ // SAFETY: By the type invariants we hold the mmap read guard, so we can safely call this
+ // method. Any value is okay for `vma_addr`.
+ let vma = unsafe { bindings::vma_lookup(self.mm.as_raw(), vma_addr) };
+
+ if vma.is_null() {
+ None
+ } else {
+ // SAFETY: We just checked that a vma was found, so the pointer references a valid vma.
+ //
+ // Furthermore, the returned vma is still under the protection of the read lock guard
+ // and can be used while the mmap read lock is still held. That the vma is not used
+ // after the MmapReadGuard gets dropped is enforced by the borrow-checker.
+ unsafe { Some(virt::VmaRef::from_raw(vma)) }
+ }
+ }
+}
+
+impl Drop for MmapReadGuard<'_> {
+ #[inline]
+ fn drop(&mut self) {
+ // SAFETY: We hold the read lock by the type invariants.
+ unsafe { bindings::mmap_read_unlock(self.mm.as_raw()) };
+ }
+}
+
+/// A guard for the vma read lock.
+///
+/// # Invariants
+///
+/// This `VmaReadGuard` guard owns the vma read lock.
+pub struct VmaReadGuard<'a> {
+ vma: &'a VmaRef,
+ // `vma_end_read` must be called on the same thread as where the lock was taken
+ _nts: NotThreadSafe,
+}
+
+// Make all `VmaRef` methods available on `VmaReadGuard`.
+impl Deref for VmaReadGuard<'_> {
+ type Target = VmaRef;
+
+ #[inline]
+ fn deref(&self) -> &VmaRef {
+ self.vma
+ }
+}
+
+impl Drop for VmaReadGuard<'_> {
+ #[inline]
+ fn drop(&mut self) {
+ // SAFETY: We hold the read lock by the type invariants.
+ unsafe { bindings::vma_end_read(self.vma.as_ptr()) };
+ }
+}
diff --git a/rust/kernel/mm/virt.rs b/rust/kernel/mm/virt.rs
new file mode 100644
index 000000000000..31803674aecc
--- /dev/null
+++ b/rust/kernel/mm/virt.rs
@@ -0,0 +1,471 @@
+// SPDX-License-Identifier: GPL-2.0
+
+// Copyright (C) 2024 Google LLC.
+
+//! Virtual memory.
+//!
+//! This module deals with managing a single VMA in the address space of a userspace process. Each
+//! VMA corresponds to a region of memory that the userspace process can access, and the VMA lets
+//! you control what happens when userspace reads or writes to that region of memory.
+//!
+//! The module has several different Rust types that all correspond to the C type called
+//! `vm_area_struct`. The different structs represent what kind of access you have to the VMA, e.g.
+//! [`VmaRef`] is used when you hold the mmap or vma read lock. Using the appropriate struct
+//! ensures that you can't, for example, accidentally call a function that requires holding the
+//! write lock when you only hold the read lock.
+
+use crate::{
+ bindings,
+ error::{code::EINVAL, to_result, Result},
+ mm::MmWithUser,
+ page::Page,
+ types::Opaque,
+};
+
+use core::ops::Deref;
+
+/// A wrapper for the kernel's `struct vm_area_struct` with read access.
+///
+/// It represents an area of virtual memory.
+///
+/// # Invariants
+///
+/// The caller must hold the mmap read lock or the vma read lock.
+#[repr(transparent)]
+pub struct VmaRef {
+ vma: Opaque<bindings::vm_area_struct>,
+}
+
+// Methods you can call when holding the mmap or vma read lock (or stronger). They must be usable
+// no matter what the vma flags are.
+impl VmaRef {
+ /// Access a virtual memory area given a raw pointer.
+ ///
+ /// # Safety
+ ///
+ /// Callers must ensure that `vma` is valid for the duration of 'a, and that the mmap or vma
+ /// read lock (or stronger) is held for at least the duration of 'a.
+ #[inline]
+ pub unsafe fn from_raw<'a>(vma: *const bindings::vm_area_struct) -> &'a Self {
+ // SAFETY: The caller ensures that the invariants are satisfied for the duration of 'a.
+ unsafe { &*vma.cast() }
+ }
+
+ /// Returns a raw pointer to this area.
+ #[inline]
+ pub fn as_ptr(&self) -> *mut bindings::vm_area_struct {
+ self.vma.get()
+ }
+
+ /// Access the underlying `mm_struct`.
+ #[inline]
+ pub fn mm(&self) -> &MmWithUser {
+ // SAFETY: By the type invariants, this `vm_area_struct` is valid and we hold the mmap/vma
+ // read lock or stronger. This implies that the underlying mm has a non-zero value of
+ // `mm_users`.
+ unsafe { MmWithUser::from_raw((*self.as_ptr()).vm_mm) }
+ }
+
+ /// Returns the flags associated with the virtual memory area.
+ ///
+ /// The possible flags are a combination of the constants in [`flags`].
+ #[inline]
+ pub fn flags(&self) -> vm_flags_t {
+ // SAFETY: By the type invariants, the caller holds at least the mmap read lock, so this
+ // access is not a data race.
+ unsafe { (*self.as_ptr()).__bindgen_anon_2.vm_flags }
+ }
+
+ /// Returns the (inclusive) start address of the virtual memory area.
+ #[inline]
+ pub fn start(&self) -> usize {
+ // SAFETY: By the type invariants, the caller holds at least the mmap read lock, so this
+ // access is not a data race.
+ unsafe { (*self.as_ptr()).__bindgen_anon_1.__bindgen_anon_1.vm_start }
+ }
+
+ /// Returns the (exclusive) end address of the virtual memory area.
+ #[inline]
+ pub fn end(&self) -> usize {
+ // SAFETY: By the type invariants, the caller holds at least the mmap read lock, so this
+ // access is not a data race.
+ unsafe { (*self.as_ptr()).__bindgen_anon_1.__bindgen_anon_1.vm_end }
+ }
+
+ /// Zap pages in the given page range.
+ ///
+ /// This clears page table mappings for the range at the leaf level, leaving all other page
+ /// tables intact, and freeing any memory referenced by the VMA in this range. That is,
+ /// anonymous memory is completely freed, file-backed memory has its reference count on page
+ /// cache folio's dropped, any dirty data will still be written back to disk as usual.
+ ///
+ /// It may seem odd that we clear at the leaf level, this is however a product of the page
+ /// table structure used to map physical memory into a virtual address space - each virtual
+ /// address actually consists of a bitmap of array indices into page tables, which form a
+ /// hierarchical page table level structure.
+ ///
+ /// As a result, each page table level maps a multiple of page table levels below, and thus
+ /// span ever increasing ranges of pages. At the leaf or PTE level, we map the actual physical
+ /// memory.
+ ///
+ /// It is here where a zap operates, as it the only place we can be certain of clearing without
+ /// impacting any other virtual mappings. It is an implementation detail as to whether the
+ /// kernel goes further in freeing unused page tables, but for the purposes of this operation
+ /// we must only assume that the leaf level is cleared.
+ #[inline]
+ pub fn zap_page_range_single(&self, address: usize, size: usize) {
+ let (end, did_overflow) = address.overflowing_add(size);
+ if did_overflow || address < self.start() || self.end() < end {
+ // TODO: call WARN_ONCE once Rust version of it is added
+ return;
+ }
+
+ // SAFETY: By the type invariants, the caller has read access to this VMA, which is
+ // sufficient for this method call. This method has no requirements on the vma flags. The
+ // address range is checked to be within the vma.
+ unsafe {
+ bindings::zap_page_range_single(self.as_ptr(), address, size, core::ptr::null_mut())
+ };
+ }
+
+ /// If the [`VM_MIXEDMAP`] flag is set, returns a [`VmaMixedMap`] to this VMA, otherwise
+ /// returns `None`.
+ ///
+ /// This can be used to access methods that require [`VM_MIXEDMAP`] to be set.
+ ///
+ /// [`VM_MIXEDMAP`]: flags::MIXEDMAP
+ #[inline]
+ pub fn as_mixedmap_vma(&self) -> Option<&VmaMixedMap> {
+ if self.flags() & flags::MIXEDMAP != 0 {
+ // SAFETY: We just checked that `VM_MIXEDMAP` is set. All other requirements are
+ // satisfied by the type invariants of `VmaRef`.
+ Some(unsafe { VmaMixedMap::from_raw(self.as_ptr()) })
+ } else {
+ None
+ }
+ }
+}
+
+/// A wrapper for the kernel's `struct vm_area_struct` with read access and [`VM_MIXEDMAP`] set.
+///
+/// It represents an area of virtual memory.
+///
+/// This struct is identical to [`VmaRef`] except that it must only be used when the
+/// [`VM_MIXEDMAP`] flag is set on the vma.
+///
+/// # Invariants
+///
+/// The caller must hold the mmap read lock or the vma read lock. The `VM_MIXEDMAP` flag must be
+/// set.
+///
+/// [`VM_MIXEDMAP`]: flags::MIXEDMAP
+#[repr(transparent)]
+pub struct VmaMixedMap {
+ vma: VmaRef,
+}
+
+// Make all `VmaRef` methods available on `VmaMixedMap`.
+impl Deref for VmaMixedMap {
+ type Target = VmaRef;
+
+ #[inline]
+ fn deref(&self) -> &VmaRef {
+ &self.vma
+ }
+}
+
+impl VmaMixedMap {
+ /// Access a virtual memory area given a raw pointer.
+ ///
+ /// # Safety
+ ///
+ /// Callers must ensure that `vma` is valid for the duration of 'a, and that the mmap read lock
+ /// (or stronger) is held for at least the duration of 'a. The `VM_MIXEDMAP` flag must be set.
+ #[inline]
+ pub unsafe fn from_raw<'a>(vma: *const bindings::vm_area_struct) -> &'a Self {
+ // SAFETY: The caller ensures that the invariants are satisfied for the duration of 'a.
+ unsafe { &*vma.cast() }
+ }
+
+ /// Maps a single page at the given address within the virtual memory area.
+ ///
+ /// This operation does not take ownership of the page.
+ #[inline]
+ pub fn vm_insert_page(&self, address: usize, page: &Page) -> Result {
+ // SAFETY: By the type invariant of `Self` caller has read access and has verified that
+ // `VM_MIXEDMAP` is set. By invariant on `Page` the page has order 0.
+ to_result(unsafe { bindings::vm_insert_page(self.as_ptr(), address, page.as_ptr()) })
+ }
+}
+
+/// A configuration object for setting up a VMA in an `f_ops->mmap()` hook.
+///
+/// The `f_ops->mmap()` hook is called when a new VMA is being created, and the hook is able to
+/// configure the VMA in various ways to fit the driver that owns it. Using `VmaNew` indicates that
+/// you are allowed to perform operations on the VMA that can only be performed before the VMA is
+/// fully initialized.
+///
+/// # Invariants
+///
+/// For the duration of 'a, the referenced vma must be undergoing initialization in an
+/// `f_ops->mmap()` hook.
+pub struct VmaNew {
+ vma: VmaRef,
+}
+
+// Make all `VmaRef` methods available on `VmaNew`.
+impl Deref for VmaNew {
+ type Target = VmaRef;
+
+ #[inline]
+ fn deref(&self) -> &VmaRef {
+ &self.vma
+ }
+}
+
+impl VmaNew {
+ /// Access a virtual memory area given a raw pointer.
+ ///
+ /// # Safety
+ ///
+ /// Callers must ensure that `vma` is undergoing initial vma setup for the duration of 'a.
+ #[inline]
+ pub unsafe fn from_raw<'a>(vma: *mut bindings::vm_area_struct) -> &'a Self {
+ // SAFETY: The caller ensures that the invariants are satisfied for the duration of 'a.
+ unsafe { &*vma.cast() }
+ }
+
+ /// Internal method for updating the vma flags.
+ ///
+ /// # Safety
+ ///
+ /// This must not be used to set the flags to an invalid value.
+ #[inline]
+ unsafe fn update_flags(&self, set: vm_flags_t, unset: vm_flags_t) {
+ let mut flags = self.flags();
+ flags |= set;
+ flags &= !unset;
+
+ // SAFETY: This is not a data race: the vma is undergoing initial setup, so it's not yet
+ // shared. Additionally, `VmaNew` is `!Sync`, so it cannot be used to write in parallel.
+ // The caller promises that this does not set the flags to an invalid value.
+ unsafe { (*self.as_ptr()).__bindgen_anon_2.__vm_flags = flags };
+ }
+
+ /// Set the `VM_MIXEDMAP` flag on this vma.
+ ///
+ /// This enables the vma to contain both `struct page` and pure PFN pages. Returns a reference
+ /// that can be used to call `vm_insert_page` on the vma.
+ #[inline]
+ pub fn set_mixedmap(&self) -> &VmaMixedMap {
+ // SAFETY: We don't yet provide a way to set VM_PFNMAP, so this cannot put the flags in an
+ // invalid state.
+ unsafe { self.update_flags(flags::MIXEDMAP, 0) };
+
+ // SAFETY: We just set `VM_MIXEDMAP` on the vma.
+ unsafe { VmaMixedMap::from_raw(self.vma.as_ptr()) }
+ }
+
+ /// Set the `VM_IO` flag on this vma.
+ ///
+ /// This is used for memory mapped IO and similar. The flag tells other parts of the kernel to
+ /// avoid looking at the pages. For memory mapped IO this is useful as accesses to the pages
+ /// could have side effects.
+ #[inline]
+ pub fn set_io(&self) {
+ // SAFETY: Setting the VM_IO flag is always okay.
+ unsafe { self.update_flags(flags::IO, 0) };
+ }
+
+ /// Set the `VM_DONTEXPAND` flag on this vma.
+ ///
+ /// This prevents the vma from being expanded with `mremap()`.
+ #[inline]
+ pub fn set_dontexpand(&self) {
+ // SAFETY: Setting the VM_DONTEXPAND flag is always okay.
+ unsafe { self.update_flags(flags::DONTEXPAND, 0) };
+ }
+
+ /// Set the `VM_DONTCOPY` flag on this vma.
+ ///
+ /// This prevents the vma from being copied on fork. This option is only permanent if `VM_IO`
+ /// is set.
+ #[inline]
+ pub fn set_dontcopy(&self) {
+ // SAFETY: Setting the VM_DONTCOPY flag is always okay.
+ unsafe { self.update_flags(flags::DONTCOPY, 0) };
+ }
+
+ /// Set the `VM_DONTDUMP` flag on this vma.
+ ///
+ /// This prevents the vma from being included in core dumps. This option is only permanent if
+ /// `VM_IO` is set.
+ #[inline]
+ pub fn set_dontdump(&self) {
+ // SAFETY: Setting the VM_DONTDUMP flag is always okay.
+ unsafe { self.update_flags(flags::DONTDUMP, 0) };
+ }
+
+ /// Returns whether `VM_READ` is set.
+ ///
+ /// This flag indicates whether userspace is mapping this vma as readable.
+ #[inline]
+ pub fn readable(&self) -> bool {
+ (self.flags() & flags::READ) != 0
+ }
+
+ /// Try to clear the `VM_MAYREAD` flag, failing if `VM_READ` is set.
+ ///
+ /// This flag indicates whether userspace is allowed to make this vma readable with
+ /// `mprotect()`.
+ ///
+ /// Note that this operation is irreversible. Once `VM_MAYREAD` has been cleared, it can never
+ /// be set again.
+ #[inline]
+ pub fn try_clear_mayread(&self) -> Result {
+ if self.readable() {
+ return Err(EINVAL);
+ }
+ // SAFETY: Clearing `VM_MAYREAD` is okay when `VM_READ` is not set.
+ unsafe { self.update_flags(0, flags::MAYREAD) };
+ Ok(())
+ }
+
+ /// Returns whether `VM_WRITE` is set.
+ ///
+ /// This flag indicates whether userspace is mapping this vma as writable.
+ #[inline]
+ pub fn writable(&self) -> bool {
+ (self.flags() & flags::WRITE) != 0
+ }
+
+ /// Try to clear the `VM_MAYWRITE` flag, failing if `VM_WRITE` is set.
+ ///
+ /// This flag indicates whether userspace is allowed to make this vma writable with
+ /// `mprotect()`.
+ ///
+ /// Note that this operation is irreversible. Once `VM_MAYWRITE` has been cleared, it can never
+ /// be set again.
+ #[inline]
+ pub fn try_clear_maywrite(&self) -> Result {
+ if self.writable() {
+ return Err(EINVAL);
+ }
+ // SAFETY: Clearing `VM_MAYWRITE` is okay when `VM_WRITE` is not set.
+ unsafe { self.update_flags(0, flags::MAYWRITE) };
+ Ok(())
+ }
+
+ /// Returns whether `VM_EXEC` is set.
+ ///
+ /// This flag indicates whether userspace is mapping this vma as executable.
+ #[inline]
+ pub fn executable(&self) -> bool {
+ (self.flags() & flags::EXEC) != 0
+ }
+
+ /// Try to clear the `VM_MAYEXEC` flag, failing if `VM_EXEC` is set.
+ ///
+ /// This flag indicates whether userspace is allowed to make this vma executable with
+ /// `mprotect()`.
+ ///
+ /// Note that this operation is irreversible. Once `VM_MAYEXEC` has been cleared, it can never
+ /// be set again.
+ #[inline]
+ pub fn try_clear_mayexec(&self) -> Result {
+ if self.executable() {
+ return Err(EINVAL);
+ }
+ // SAFETY: Clearing `VM_MAYEXEC` is okay when `VM_EXEC` is not set.
+ unsafe { self.update_flags(0, flags::MAYEXEC) };
+ Ok(())
+ }
+}
+
+/// The integer type used for vma flags.
+#[doc(inline)]
+pub use bindings::vm_flags_t;
+
+/// All possible flags for [`VmaRef`].
+pub mod flags {
+ use super::vm_flags_t;
+ use crate::bindings;
+
+ /// No flags are set.
+ pub const NONE: vm_flags_t = bindings::VM_NONE as _;
+
+ /// Mapping allows reads.
+ pub const READ: vm_flags_t = bindings::VM_READ as _;
+
+ /// Mapping allows writes.
+ pub const WRITE: vm_flags_t = bindings::VM_WRITE as _;
+
+ /// Mapping allows execution.
+ pub const EXEC: vm_flags_t = bindings::VM_EXEC as _;
+
+ /// Mapping is shared.
+ pub const SHARED: vm_flags_t = bindings::VM_SHARED as _;
+
+ /// Mapping may be updated to allow reads.
+ pub const MAYREAD: vm_flags_t = bindings::VM_MAYREAD as _;
+
+ /// Mapping may be updated to allow writes.
+ pub const MAYWRITE: vm_flags_t = bindings::VM_MAYWRITE as _;
+
+ /// Mapping may be updated to allow execution.
+ pub const MAYEXEC: vm_flags_t = bindings::VM_MAYEXEC as _;
+
+ /// Mapping may be updated to be shared.
+ pub const MAYSHARE: vm_flags_t = bindings::VM_MAYSHARE as _;
+
+ /// Page-ranges managed without `struct page`, just pure PFN.
+ pub const PFNMAP: vm_flags_t = bindings::VM_PFNMAP as _;
+
+ /// Memory mapped I/O or similar.
+ pub const IO: vm_flags_t = bindings::VM_IO as _;
+
+ /// Do not copy this vma on fork.
+ pub const DONTCOPY: vm_flags_t = bindings::VM_DONTCOPY as _;
+
+ /// Cannot expand with mremap().
+ pub const DONTEXPAND: vm_flags_t = bindings::VM_DONTEXPAND as _;
+
+ /// Lock the pages covered when they are faulted in.
+ pub const LOCKONFAULT: vm_flags_t = bindings::VM_LOCKONFAULT as _;
+
+ /// Is a VM accounted object.
+ pub const ACCOUNT: vm_flags_t = bindings::VM_ACCOUNT as _;
+
+ /// Should the VM suppress accounting.
+ pub const NORESERVE: vm_flags_t = bindings::VM_NORESERVE as _;
+
+ /// Huge TLB Page VM.
+ pub const HUGETLB: vm_flags_t = bindings::VM_HUGETLB as _;
+
+ /// Synchronous page faults. (DAX-specific)
+ pub const SYNC: vm_flags_t = bindings::VM_SYNC as _;
+
+ /// Architecture-specific flag.
+ pub const ARCH_1: vm_flags_t = bindings::VM_ARCH_1 as _;
+
+ /// Wipe VMA contents in child on fork.
+ pub const WIPEONFORK: vm_flags_t = bindings::VM_WIPEONFORK as _;
+
+ /// Do not include in the core dump.
+ pub const DONTDUMP: vm_flags_t = bindings::VM_DONTDUMP as _;
+
+ /// Not soft dirty clean area.
+ pub const SOFTDIRTY: vm_flags_t = bindings::VM_SOFTDIRTY as _;
+
+ /// Can contain `struct page` and pure PFN pages.
+ pub const MIXEDMAP: vm_flags_t = bindings::VM_MIXEDMAP as _;
+
+ /// MADV_HUGEPAGE marked this vma.
+ pub const HUGEPAGE: vm_flags_t = bindings::VM_HUGEPAGE as _;
+
+ /// MADV_NOHUGEPAGE marked this vma.
+ pub const NOHUGEPAGE: vm_flags_t = bindings::VM_NOHUGEPAGE as _;
+
+ /// KSM may merge identical pages.
+ pub const MERGEABLE: vm_flags_t = bindings::VM_MERGEABLE as _;
+}
diff --git a/rust/kernel/opp.rs b/rust/kernel/opp.rs
new file mode 100644
index 000000000000..a566fc3e7dcb
--- /dev/null
+++ b/rust/kernel/opp.rs
@@ -0,0 +1,1146 @@
+// SPDX-License-Identifier: GPL-2.0
+
+//! Operating performance points.
+//!
+//! This module provides rust abstractions for interacting with the OPP subsystem.
+//!
+//! C header: [`include/linux/pm_opp.h`](srctree/include/linux/pm_opp.h)
+//!
+//! Reference: <https://docs.kernel.org/power/opp.html>
+
+use crate::{
+ clk::Hertz,
+ cpumask::{Cpumask, CpumaskVar},
+ device::Device,
+ error::{code::*, from_err_ptr, from_result, to_result, Error, Result, VTABLE_DEFAULT_ERROR},
+ ffi::c_ulong,
+ prelude::*,
+ str::CString,
+ types::{ARef, AlwaysRefCounted, Opaque},
+};
+
+#[cfg(CONFIG_CPU_FREQ)]
+/// Frequency table implementation.
+mod freq {
+ use super::*;
+ use crate::cpufreq;
+ use core::ops::Deref;
+
+ /// OPP frequency table.
+ ///
+ /// A [`cpufreq::Table`] created from [`Table`].
+ pub struct FreqTable {
+ dev: ARef<Device>,
+ ptr: *mut bindings::cpufreq_frequency_table,
+ }
+
+ impl FreqTable {
+ /// Creates a new instance of [`FreqTable`] from [`Table`].
+ pub(crate) fn new(table: &Table) -> Result<Self> {
+ let mut ptr: *mut bindings::cpufreq_frequency_table = ptr::null_mut();
+
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ to_result(unsafe {
+ bindings::dev_pm_opp_init_cpufreq_table(table.dev.as_raw(), &mut ptr)
+ })?;
+
+ Ok(Self {
+ dev: table.dev.clone(),
+ ptr,
+ })
+ }
+
+ /// Returns a reference to the underlying [`cpufreq::Table`].
+ #[inline]
+ fn table(&self) -> &cpufreq::Table {
+ // SAFETY: The `ptr` is guaranteed by the C code to be valid.
+ unsafe { cpufreq::Table::from_raw(self.ptr) }
+ }
+ }
+
+ impl Deref for FreqTable {
+ type Target = cpufreq::Table;
+
+ #[inline]
+ fn deref(&self) -> &Self::Target {
+ self.table()
+ }
+ }
+
+ impl Drop for FreqTable {
+ fn drop(&mut self) {
+ // SAFETY: The pointer was created via `dev_pm_opp_init_cpufreq_table`, and is only
+ // freed here.
+ unsafe {
+ bindings::dev_pm_opp_free_cpufreq_table(self.dev.as_raw(), &mut self.as_raw())
+ };
+ }
+ }
+}
+
+#[cfg(CONFIG_CPU_FREQ)]
+pub use freq::FreqTable;
+
+use core::{marker::PhantomData, ptr};
+
+use macros::vtable;
+
+/// Creates a null-terminated slice of pointers to [`Cstring`]s.
+fn to_c_str_array(names: &[CString]) -> Result<KVec<*const u8>> {
+ // Allocated a null-terminated vector of pointers.
+ let mut list = KVec::with_capacity(names.len() + 1, GFP_KERNEL)?;
+
+ for name in names.iter() {
+ list.push(name.as_ptr() as _, GFP_KERNEL)?;
+ }
+
+ list.push(ptr::null(), GFP_KERNEL)?;
+ Ok(list)
+}
+
+/// The voltage unit.
+///
+/// Represents voltage in microvolts, wrapping a [`c_ulong`] value.
+///
+/// ## Examples
+///
+/// ```
+/// use kernel::opp::MicroVolt;
+///
+/// let raw = 90500;
+/// let volt = MicroVolt(raw);
+///
+/// assert_eq!(usize::from(volt), raw);
+/// assert_eq!(volt, MicroVolt(raw));
+/// ```
+#[derive(Copy, Clone, PartialEq, Eq, Debug)]
+pub struct MicroVolt(pub c_ulong);
+
+impl From<MicroVolt> for c_ulong {
+ #[inline]
+ fn from(volt: MicroVolt) -> Self {
+ volt.0
+ }
+}
+
+/// The power unit.
+///
+/// Represents power in microwatts, wrapping a [`c_ulong`] value.
+///
+/// ## Examples
+///
+/// ```
+/// use kernel::opp::MicroWatt;
+///
+/// let raw = 1000000;
+/// let power = MicroWatt(raw);
+///
+/// assert_eq!(usize::from(power), raw);
+/// assert_eq!(power, MicroWatt(raw));
+/// ```
+#[derive(Copy, Clone, PartialEq, Eq, Debug)]
+pub struct MicroWatt(pub c_ulong);
+
+impl From<MicroWatt> for c_ulong {
+ #[inline]
+ fn from(power: MicroWatt) -> Self {
+ power.0
+ }
+}
+
+/// Handle for a dynamically created [`OPP`].
+///
+/// The associated [`OPP`] is automatically removed when the [`Token`] is dropped.
+///
+/// ## Examples
+///
+/// The following example demonstrates how to create an [`OPP`] dynamically.
+///
+/// ```
+/// use kernel::clk::Hertz;
+/// use kernel::device::Device;
+/// use kernel::error::Result;
+/// use kernel::opp::{Data, MicroVolt, Token};
+/// use kernel::types::ARef;
+///
+/// fn create_opp(dev: &ARef<Device>, freq: Hertz, volt: MicroVolt, level: u32) -> Result<Token> {
+/// let data = Data::new(freq, volt, level, false);
+///
+/// // OPP is removed once token goes out of scope.
+/// data.add_opp(dev)
+/// }
+/// ```
+pub struct Token {
+ dev: ARef<Device>,
+ freq: Hertz,
+}
+
+impl Token {
+ /// Dynamically adds an [`OPP`] and returns a [`Token`] that removes it on drop.
+ fn new(dev: &ARef<Device>, mut data: Data) -> Result<Self> {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ to_result(unsafe { bindings::dev_pm_opp_add_dynamic(dev.as_raw(), &mut data.0) })?;
+ Ok(Self {
+ dev: dev.clone(),
+ freq: data.freq(),
+ })
+ }
+}
+
+impl Drop for Token {
+ fn drop(&mut self) {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ unsafe { bindings::dev_pm_opp_remove(self.dev.as_raw(), self.freq.into()) };
+ }
+}
+
+/// OPP data.
+///
+/// Rust abstraction for the C `struct dev_pm_opp_data`, used to define operating performance
+/// points (OPPs) dynamically.
+///
+/// ## Examples
+///
+/// The following example demonstrates how to create an [`OPP`] with [`Data`].
+///
+/// ```
+/// use kernel::clk::Hertz;
+/// use kernel::device::Device;
+/// use kernel::error::Result;
+/// use kernel::opp::{Data, MicroVolt, Token};
+/// use kernel::types::ARef;
+///
+/// fn create_opp(dev: &ARef<Device>, freq: Hertz, volt: MicroVolt, level: u32) -> Result<Token> {
+/// let data = Data::new(freq, volt, level, false);
+///
+/// // OPP is removed once token goes out of scope.
+/// data.add_opp(dev)
+/// }
+/// ```
+#[repr(transparent)]
+pub struct Data(bindings::dev_pm_opp_data);
+
+impl Data {
+ /// Creates a new instance of [`Data`].
+ ///
+ /// This can be used to define a dynamic OPP to be added to a device.
+ pub fn new(freq: Hertz, volt: MicroVolt, level: u32, turbo: bool) -> Self {
+ Self(bindings::dev_pm_opp_data {
+ turbo,
+ freq: freq.into(),
+ u_volt: volt.into(),
+ level,
+ })
+ }
+
+ /// Adds an [`OPP`] dynamically.
+ ///
+ /// Returns a [`Token`] that ensures the OPP is automatically removed
+ /// when it goes out of scope.
+ #[inline]
+ pub fn add_opp(self, dev: &ARef<Device>) -> Result<Token> {
+ Token::new(dev, self)
+ }
+
+ /// Returns the frequency associated with this OPP data.
+ #[inline]
+ fn freq(&self) -> Hertz {
+ Hertz(self.0.freq)
+ }
+}
+
+/// [`OPP`] search options.
+///
+/// ## Examples
+///
+/// Defines how to search for an [`OPP`] in a [`Table`] relative to a frequency.
+///
+/// ```
+/// use kernel::clk::Hertz;
+/// use kernel::error::Result;
+/// use kernel::opp::{OPP, SearchType, Table};
+/// use kernel::types::ARef;
+///
+/// fn find_opp(table: &Table, freq: Hertz) -> Result<ARef<OPP>> {
+/// let opp = table.opp_from_freq(freq, Some(true), None, SearchType::Exact)?;
+///
+/// pr_info!("OPP frequency is: {:?}\n", opp.freq(None));
+/// pr_info!("OPP voltage is: {:?}\n", opp.voltage());
+/// pr_info!("OPP level is: {}\n", opp.level());
+/// pr_info!("OPP power is: {:?}\n", opp.power());
+///
+/// Ok(opp)
+/// }
+/// ```
+#[derive(Copy, Clone, Debug, Eq, PartialEq)]
+pub enum SearchType {
+ /// Match the exact frequency.
+ Exact,
+ /// Find the highest frequency less than or equal to the given value.
+ Floor,
+ /// Find the lowest frequency greater than or equal to the given value.
+ Ceil,
+}
+
+/// OPP configuration callbacks.
+///
+/// Implement this trait to customize OPP clock and regulator setup for your device.
+#[vtable]
+pub trait ConfigOps {
+ /// This is typically used to scale clocks when transitioning between OPPs.
+ #[inline]
+ fn config_clks(_dev: &Device, _table: &Table, _opp: &OPP, _scaling_down: bool) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+
+ /// This provides access to the old and new OPPs, allowing for safe regulator adjustments.
+ #[inline]
+ fn config_regulators(
+ _dev: &Device,
+ _opp_old: &OPP,
+ _opp_new: &OPP,
+ _data: *mut *mut bindings::regulator,
+ _count: u32,
+ ) -> Result {
+ build_error!(VTABLE_DEFAULT_ERROR)
+ }
+}
+
+/// OPP configuration token.
+///
+/// Returned by the OPP core when configuration is applied to a [`Device`]. The associated
+/// configuration is automatically cleared when the token is dropped.
+pub struct ConfigToken(i32);
+
+impl Drop for ConfigToken {
+ fn drop(&mut self) {
+ // SAFETY: This is the same token value returned by the C code via `dev_pm_opp_set_config`.
+ unsafe { bindings::dev_pm_opp_clear_config(self.0) };
+ }
+}
+
+/// OPP configurations.
+///
+/// Rust abstraction for the C `struct dev_pm_opp_config`.
+///
+/// ## Examples
+///
+/// The following example demonstrates how to set OPP property-name configuration for a [`Device`].
+///
+/// ```
+/// use kernel::device::Device;
+/// use kernel::error::Result;
+/// use kernel::opp::{Config, ConfigOps, ConfigToken};
+/// use kernel::str::CString;
+/// use kernel::types::ARef;
+/// use kernel::macros::vtable;
+///
+/// #[derive(Default)]
+/// struct Driver;
+///
+/// #[vtable]
+/// impl ConfigOps for Driver {}
+///
+/// fn configure(dev: &ARef<Device>) -> Result<ConfigToken> {
+/// let name = CString::try_from_fmt(fmt!("{}", "slow"))?;
+///
+/// // The OPP configuration is cleared once the [`ConfigToken`] goes out of scope.
+/// Config::<Driver>::new()
+/// .set_prop_name(name)?
+/// .set(dev)
+/// }
+/// ```
+#[derive(Default)]
+pub struct Config<T: ConfigOps>
+where
+ T: Default,
+{
+ clk_names: Option<KVec<CString>>,
+ prop_name: Option<CString>,
+ regulator_names: Option<KVec<CString>>,
+ supported_hw: Option<KVec<u32>>,
+
+ // Tuple containing (required device, index)
+ required_dev: Option<(ARef<Device>, u32)>,
+ _data: PhantomData<T>,
+}
+
+impl<T: ConfigOps + Default> Config<T> {
+ /// Creates a new instance of [`Config`].
+ #[inline]
+ pub fn new() -> Self {
+ Self::default()
+ }
+
+ /// Initializes clock names.
+ pub fn set_clk_names(mut self, names: KVec<CString>) -> Result<Self> {
+ if self.clk_names.is_some() {
+ return Err(EBUSY);
+ }
+
+ if names.is_empty() {
+ return Err(EINVAL);
+ }
+
+ self.clk_names = Some(names);
+ Ok(self)
+ }
+
+ /// Initializes property name.
+ pub fn set_prop_name(mut self, name: CString) -> Result<Self> {
+ if self.prop_name.is_some() {
+ return Err(EBUSY);
+ }
+
+ self.prop_name = Some(name);
+ Ok(self)
+ }
+
+ /// Initializes regulator names.
+ pub fn set_regulator_names(mut self, names: KVec<CString>) -> Result<Self> {
+ if self.regulator_names.is_some() {
+ return Err(EBUSY);
+ }
+
+ if names.is_empty() {
+ return Err(EINVAL);
+ }
+
+ self.regulator_names = Some(names);
+
+ Ok(self)
+ }
+
+ /// Initializes required devices.
+ pub fn set_required_dev(mut self, dev: ARef<Device>, index: u32) -> Result<Self> {
+ if self.required_dev.is_some() {
+ return Err(EBUSY);
+ }
+
+ self.required_dev = Some((dev, index));
+ Ok(self)
+ }
+
+ /// Initializes supported hardware.
+ pub fn set_supported_hw(mut self, hw: KVec<u32>) -> Result<Self> {
+ if self.supported_hw.is_some() {
+ return Err(EBUSY);
+ }
+
+ if hw.is_empty() {
+ return Err(EINVAL);
+ }
+
+ self.supported_hw = Some(hw);
+ Ok(self)
+ }
+
+ /// Sets the configuration with the OPP core.
+ ///
+ /// The returned [`ConfigToken`] will remove the configuration when dropped.
+ pub fn set(self, dev: &Device) -> Result<ConfigToken> {
+ let (_clk_list, clk_names) = match &self.clk_names {
+ Some(x) => {
+ let list = to_c_str_array(x)?;
+ let ptr = list.as_ptr();
+ (Some(list), ptr)
+ }
+ None => (None, ptr::null()),
+ };
+
+ let (_regulator_list, regulator_names) = match &self.regulator_names {
+ Some(x) => {
+ let list = to_c_str_array(x)?;
+ let ptr = list.as_ptr();
+ (Some(list), ptr)
+ }
+ None => (None, ptr::null()),
+ };
+
+ let prop_name = self
+ .prop_name
+ .as_ref()
+ .map_or(ptr::null(), |p| p.as_char_ptr());
+
+ let (supported_hw, supported_hw_count) = self
+ .supported_hw
+ .as_ref()
+ .map_or((ptr::null(), 0), |hw| (hw.as_ptr(), hw.len() as u32));
+
+ let (required_dev, required_dev_index) = self
+ .required_dev
+ .as_ref()
+ .map_or((ptr::null_mut(), 0), |(dev, idx)| (dev.as_raw(), *idx));
+
+ let mut config = bindings::dev_pm_opp_config {
+ clk_names,
+ config_clks: if T::HAS_CONFIG_CLKS {
+ Some(Self::config_clks)
+ } else {
+ None
+ },
+ prop_name,
+ regulator_names,
+ config_regulators: if T::HAS_CONFIG_REGULATORS {
+ Some(Self::config_regulators)
+ } else {
+ None
+ },
+ supported_hw,
+ supported_hw_count,
+
+ required_dev,
+ required_dev_index,
+ };
+
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements. The OPP core guarantees not to access fields of [`Config`] after this call
+ // and so we don't need to save a copy of them for future use.
+ let ret = unsafe { bindings::dev_pm_opp_set_config(dev.as_raw(), &mut config) };
+ if ret < 0 {
+ Err(Error::from_errno(ret))
+ } else {
+ Ok(ConfigToken(ret))
+ }
+ }
+
+ /// Config's clk callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn config_clks(
+ dev: *mut bindings::device,
+ opp_table: *mut bindings::opp_table,
+ opp: *mut bindings::dev_pm_opp,
+ _data: *mut kernel::ffi::c_void,
+ scaling_down: bool,
+ ) -> kernel::ffi::c_int {
+ from_result(|| {
+ // SAFETY: 'dev' is guaranteed by the C code to be valid.
+ let dev = unsafe { Device::get_device(dev) };
+ T::config_clks(
+ &dev,
+ // SAFETY: 'opp_table' is guaranteed by the C code to be valid.
+ &unsafe { Table::from_raw_table(opp_table, &dev) },
+ // SAFETY: 'opp' is guaranteed by the C code to be valid.
+ unsafe { OPP::from_raw_opp(opp)? },
+ scaling_down,
+ )
+ .map(|()| 0)
+ })
+ }
+
+ /// Config's regulator callback.
+ ///
+ /// SAFETY: Called from C. Inputs must be valid pointers.
+ extern "C" fn config_regulators(
+ dev: *mut bindings::device,
+ old_opp: *mut bindings::dev_pm_opp,
+ new_opp: *mut bindings::dev_pm_opp,
+ regulators: *mut *mut bindings::regulator,
+ count: kernel::ffi::c_uint,
+ ) -> kernel::ffi::c_int {
+ from_result(|| {
+ // SAFETY: 'dev' is guaranteed by the C code to be valid.
+ let dev = unsafe { Device::get_device(dev) };
+ T::config_regulators(
+ &dev,
+ // SAFETY: 'old_opp' is guaranteed by the C code to be valid.
+ unsafe { OPP::from_raw_opp(old_opp)? },
+ // SAFETY: 'new_opp' is guaranteed by the C code to be valid.
+ unsafe { OPP::from_raw_opp(new_opp)? },
+ regulators,
+ count,
+ )
+ .map(|()| 0)
+ })
+ }
+}
+
+/// A reference-counted OPP table.
+///
+/// Rust abstraction for the C `struct opp_table`.
+///
+/// # Invariants
+///
+/// The pointer stored in `Self` is non-null and valid for the lifetime of the [`Table`].
+///
+/// Instances of this type are reference-counted.
+///
+/// ## Examples
+///
+/// The following example demonstrates how to get OPP [`Table`] for a [`Cpumask`] and set its
+/// frequency.
+///
+/// ```
+/// # #![cfg(CONFIG_OF)]
+/// use kernel::clk::Hertz;
+/// use kernel::cpumask::Cpumask;
+/// use kernel::device::Device;
+/// use kernel::error::Result;
+/// use kernel::opp::Table;
+/// use kernel::types::ARef;
+///
+/// fn get_table(dev: &ARef<Device>, mask: &mut Cpumask, freq: Hertz) -> Result<Table> {
+/// let mut opp_table = Table::from_of_cpumask(dev, mask)?;
+///
+/// if opp_table.opp_count()? == 0 {
+/// return Err(EINVAL);
+/// }
+///
+/// pr_info!("Max transition latency is: {} ns\n", opp_table.max_transition_latency_ns());
+/// pr_info!("Suspend frequency is: {:?}\n", opp_table.suspend_freq());
+///
+/// opp_table.set_rate(freq)?;
+/// Ok(opp_table)
+/// }
+/// ```
+pub struct Table {
+ ptr: *mut bindings::opp_table,
+ dev: ARef<Device>,
+ #[allow(dead_code)]
+ em: bool,
+ #[allow(dead_code)]
+ of: bool,
+ cpus: Option<CpumaskVar>,
+}
+
+/// SAFETY: It is okay to send ownership of [`Table`] across thread boundaries.
+unsafe impl Send for Table {}
+
+/// SAFETY: It is okay to access [`Table`] through shared references from other threads because
+/// we're either accessing properties that don't change or that are properly synchronised by C code.
+unsafe impl Sync for Table {}
+
+impl Table {
+ /// Creates a new reference-counted [`Table`] from a raw pointer.
+ ///
+ /// # Safety
+ ///
+ /// Callers must ensure that `ptr` is valid and non-null.
+ unsafe fn from_raw_table(ptr: *mut bindings::opp_table, dev: &ARef<Device>) -> Self {
+ // SAFETY: By the safety requirements, ptr is valid and its refcount will be incremented.
+ //
+ // INVARIANT: The reference-count is decremented when [`Table`] goes out of scope.
+ unsafe { bindings::dev_pm_opp_get_opp_table_ref(ptr) };
+
+ Self {
+ ptr,
+ dev: dev.clone(),
+ em: false,
+ of: false,
+ cpus: None,
+ }
+ }
+
+ /// Creates a new reference-counted [`Table`] instance for a [`Device`].
+ pub fn from_dev(dev: &Device) -> Result<Self> {
+ // SAFETY: The requirements are satisfied by the existence of the [`Device`] and its safety
+ // requirements.
+ //
+ // INVARIANT: The reference-count is incremented by the C code and is decremented when
+ // [`Table`] goes out of scope.
+ let ptr = from_err_ptr(unsafe { bindings::dev_pm_opp_get_opp_table(dev.as_raw()) })?;
+
+ Ok(Self {
+ ptr,
+ dev: dev.into(),
+ em: false,
+ of: false,
+ cpus: None,
+ })
+ }
+
+ /// Creates a new reference-counted [`Table`] instance for a [`Device`] based on device tree
+ /// entries.
+ #[cfg(CONFIG_OF)]
+ pub fn from_of(dev: &ARef<Device>, index: i32) -> Result<Self> {
+ // SAFETY: The requirements are satisfied by the existence of the [`Device`] and its safety
+ // requirements.
+ //
+ // INVARIANT: The reference-count is incremented by the C code and is decremented when
+ // [`Table`] goes out of scope.
+ to_result(unsafe { bindings::dev_pm_opp_of_add_table_indexed(dev.as_raw(), index) })?;
+
+ // Get the newly created [`Table`].
+ let mut table = Self::from_dev(dev)?;
+ table.of = true;
+
+ Ok(table)
+ }
+
+ /// Remove device tree based [`Table`].
+ #[cfg(CONFIG_OF)]
+ #[inline]
+ fn remove_of(&self) {
+ // SAFETY: The requirements are satisfied by the existence of the [`Device`] and its safety
+ // requirements. We took the reference from [`from_of`] earlier, it is safe to drop the
+ // same now.
+ unsafe { bindings::dev_pm_opp_of_remove_table(self.dev.as_raw()) };
+ }
+
+ /// Creates a new reference-counted [`Table`] instance for a [`Cpumask`] based on device tree
+ /// entries.
+ #[cfg(CONFIG_OF)]
+ pub fn from_of_cpumask(dev: &Device, cpumask: &mut Cpumask) -> Result<Self> {
+ // SAFETY: The cpumask is valid and the returned pointer will be owned by the [`Table`]
+ // instance.
+ //
+ // INVARIANT: The reference-count is incremented by the C code and is decremented when
+ // [`Table`] goes out of scope.
+ to_result(unsafe { bindings::dev_pm_opp_of_cpumask_add_table(cpumask.as_raw()) })?;
+
+ // Fetch the newly created table.
+ let mut table = Self::from_dev(dev)?;
+ table.cpus = Some(CpumaskVar::try_clone(cpumask)?);
+
+ Ok(table)
+ }
+
+ /// Remove device tree based [`Table`] for a [`Cpumask`].
+ #[cfg(CONFIG_OF)]
+ #[inline]
+ fn remove_of_cpumask(&self, cpumask: &Cpumask) {
+ // SAFETY: The cpumask is valid and we took the reference from [`from_of_cpumask`] earlier,
+ // it is safe to drop the same now.
+ unsafe { bindings::dev_pm_opp_of_cpumask_remove_table(cpumask.as_raw()) };
+ }
+
+ /// Returns the number of [`OPP`]s in the [`Table`].
+ pub fn opp_count(&self) -> Result<u32> {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ let ret = unsafe { bindings::dev_pm_opp_get_opp_count(self.dev.as_raw()) };
+ if ret < 0 {
+ Err(Error::from_errno(ret))
+ } else {
+ Ok(ret as u32)
+ }
+ }
+
+ /// Returns max clock latency (in nanoseconds) of the [`OPP`]s in the [`Table`].
+ #[inline]
+ pub fn max_clock_latency_ns(&self) -> usize {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ unsafe { bindings::dev_pm_opp_get_max_clock_latency(self.dev.as_raw()) }
+ }
+
+ /// Returns max volt latency (in nanoseconds) of the [`OPP`]s in the [`Table`].
+ #[inline]
+ pub fn max_volt_latency_ns(&self) -> usize {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ unsafe { bindings::dev_pm_opp_get_max_volt_latency(self.dev.as_raw()) }
+ }
+
+ /// Returns max transition latency (in nanoseconds) of the [`OPP`]s in the [`Table`].
+ #[inline]
+ pub fn max_transition_latency_ns(&self) -> usize {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ unsafe { bindings::dev_pm_opp_get_max_transition_latency(self.dev.as_raw()) }
+ }
+
+ /// Returns the suspend [`OPP`]'s frequency.
+ #[inline]
+ pub fn suspend_freq(&self) -> Hertz {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ Hertz(unsafe { bindings::dev_pm_opp_get_suspend_opp_freq(self.dev.as_raw()) })
+ }
+
+ /// Synchronizes regulators used by the [`Table`].
+ #[inline]
+ pub fn sync_regulators(&self) -> Result {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ to_result(unsafe { bindings::dev_pm_opp_sync_regulators(self.dev.as_raw()) })
+ }
+
+ /// Gets sharing CPUs.
+ #[inline]
+ pub fn sharing_cpus(dev: &Device, cpumask: &mut Cpumask) -> Result {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ to_result(unsafe { bindings::dev_pm_opp_get_sharing_cpus(dev.as_raw(), cpumask.as_raw()) })
+ }
+
+ /// Sets sharing CPUs.
+ pub fn set_sharing_cpus(&mut self, cpumask: &mut Cpumask) -> Result {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ to_result(unsafe {
+ bindings::dev_pm_opp_set_sharing_cpus(self.dev.as_raw(), cpumask.as_raw())
+ })?;
+
+ if let Some(mask) = self.cpus.as_mut() {
+ // Update the cpumask as this will be used while removing the table.
+ cpumask.copy(mask);
+ }
+
+ Ok(())
+ }
+
+ /// Gets sharing CPUs from device tree.
+ #[cfg(CONFIG_OF)]
+ #[inline]
+ pub fn of_sharing_cpus(dev: &Device, cpumask: &mut Cpumask) -> Result {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ to_result(unsafe {
+ bindings::dev_pm_opp_of_get_sharing_cpus(dev.as_raw(), cpumask.as_raw())
+ })
+ }
+
+ /// Updates the voltage value for an [`OPP`].
+ #[inline]
+ pub fn adjust_voltage(
+ &self,
+ freq: Hertz,
+ volt: MicroVolt,
+ volt_min: MicroVolt,
+ volt_max: MicroVolt,
+ ) -> Result {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ to_result(unsafe {
+ bindings::dev_pm_opp_adjust_voltage(
+ self.dev.as_raw(),
+ freq.into(),
+ volt.into(),
+ volt_min.into(),
+ volt_max.into(),
+ )
+ })
+ }
+
+ /// Creates [`FreqTable`] from [`Table`].
+ #[cfg(CONFIG_CPU_FREQ)]
+ #[inline]
+ pub fn cpufreq_table(&mut self) -> Result<FreqTable> {
+ FreqTable::new(self)
+ }
+
+ /// Configures device with [`OPP`] matching the frequency value.
+ #[inline]
+ pub fn set_rate(&self, freq: Hertz) -> Result {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ to_result(unsafe { bindings::dev_pm_opp_set_rate(self.dev.as_raw(), freq.into()) })
+ }
+
+ /// Configures device with [`OPP`].
+ #[inline]
+ pub fn set_opp(&self, opp: &OPP) -> Result {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ to_result(unsafe { bindings::dev_pm_opp_set_opp(self.dev.as_raw(), opp.as_raw()) })
+ }
+
+ /// Finds [`OPP`] based on frequency.
+ pub fn opp_from_freq(
+ &self,
+ freq: Hertz,
+ available: Option<bool>,
+ index: Option<u32>,
+ stype: SearchType,
+ ) -> Result<ARef<OPP>> {
+ let raw_dev = self.dev.as_raw();
+ let index = index.unwrap_or(0);
+ let mut rate = freq.into();
+
+ let ptr = from_err_ptr(match stype {
+ SearchType::Exact => {
+ if let Some(available) = available {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and
+ // its safety requirements. The returned pointer will be owned by the new
+ // [`OPP`] instance.
+ unsafe {
+ bindings::dev_pm_opp_find_freq_exact_indexed(
+ raw_dev, rate, index, available,
+ )
+ }
+ } else {
+ return Err(EINVAL);
+ }
+ }
+
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements. The returned pointer will be owned by the new [`OPP`] instance.
+ SearchType::Ceil => unsafe {
+ bindings::dev_pm_opp_find_freq_ceil_indexed(raw_dev, &mut rate, index)
+ },
+
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements. The returned pointer will be owned by the new [`OPP`] instance.
+ SearchType::Floor => unsafe {
+ bindings::dev_pm_opp_find_freq_floor_indexed(raw_dev, &mut rate, index)
+ },
+ })?;
+
+ // SAFETY: The `ptr` is guaranteed by the C code to be valid.
+ unsafe { OPP::from_raw_opp_owned(ptr) }
+ }
+
+ /// Finds [`OPP`] based on level.
+ pub fn opp_from_level(&self, mut level: u32, stype: SearchType) -> Result<ARef<OPP>> {
+ let raw_dev = self.dev.as_raw();
+
+ let ptr = from_err_ptr(match stype {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements. The returned pointer will be owned by the new [`OPP`] instance.
+ SearchType::Exact => unsafe { bindings::dev_pm_opp_find_level_exact(raw_dev, level) },
+
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements. The returned pointer will be owned by the new [`OPP`] instance.
+ SearchType::Ceil => unsafe {
+ bindings::dev_pm_opp_find_level_ceil(raw_dev, &mut level)
+ },
+
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements. The returned pointer will be owned by the new [`OPP`] instance.
+ SearchType::Floor => unsafe {
+ bindings::dev_pm_opp_find_level_floor(raw_dev, &mut level)
+ },
+ })?;
+
+ // SAFETY: The `ptr` is guaranteed by the C code to be valid.
+ unsafe { OPP::from_raw_opp_owned(ptr) }
+ }
+
+ /// Finds [`OPP`] based on bandwidth.
+ pub fn opp_from_bw(&self, mut bw: u32, index: i32, stype: SearchType) -> Result<ARef<OPP>> {
+ let raw_dev = self.dev.as_raw();
+
+ let ptr = from_err_ptr(match stype {
+ // The OPP core doesn't support this yet.
+ SearchType::Exact => return Err(EINVAL),
+
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements. The returned pointer will be owned by the new [`OPP`] instance.
+ SearchType::Ceil => unsafe {
+ bindings::dev_pm_opp_find_bw_ceil(raw_dev, &mut bw, index)
+ },
+
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements. The returned pointer will be owned by the new [`OPP`] instance.
+ SearchType::Floor => unsafe {
+ bindings::dev_pm_opp_find_bw_floor(raw_dev, &mut bw, index)
+ },
+ })?;
+
+ // SAFETY: The `ptr` is guaranteed by the C code to be valid.
+ unsafe { OPP::from_raw_opp_owned(ptr) }
+ }
+
+ /// Enables the [`OPP`].
+ #[inline]
+ pub fn enable_opp(&self, freq: Hertz) -> Result {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ to_result(unsafe { bindings::dev_pm_opp_enable(self.dev.as_raw(), freq.into()) })
+ }
+
+ /// Disables the [`OPP`].
+ #[inline]
+ pub fn disable_opp(&self, freq: Hertz) -> Result {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ to_result(unsafe { bindings::dev_pm_opp_disable(self.dev.as_raw(), freq.into()) })
+ }
+
+ /// Registers with the Energy model.
+ #[cfg(CONFIG_OF)]
+ pub fn of_register_em(&mut self, cpumask: &mut Cpumask) -> Result {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements.
+ to_result(unsafe {
+ bindings::dev_pm_opp_of_register_em(self.dev.as_raw(), cpumask.as_raw())
+ })?;
+
+ self.em = true;
+ Ok(())
+ }
+
+ /// Unregisters with the Energy model.
+ #[cfg(all(CONFIG_OF, CONFIG_ENERGY_MODEL))]
+ #[inline]
+ fn of_unregister_em(&self) {
+ // SAFETY: The requirements are satisfied by the existence of [`Device`] and its safety
+ // requirements. We registered with the EM framework earlier, it is safe to unregister now.
+ unsafe { bindings::em_dev_unregister_perf_domain(self.dev.as_raw()) };
+ }
+}
+
+impl Drop for Table {
+ fn drop(&mut self) {
+ // SAFETY: By the type invariants, we know that `self` owns a reference, so it is safe
+ // to relinquish it now.
+ unsafe { bindings::dev_pm_opp_put_opp_table(self.ptr) };
+
+ #[cfg(CONFIG_OF)]
+ {
+ #[cfg(CONFIG_ENERGY_MODEL)]
+ if self.em {
+ self.of_unregister_em();
+ }
+
+ if self.of {
+ self.remove_of();
+ } else if let Some(cpumask) = self.cpus.take() {
+ self.remove_of_cpumask(&cpumask);
+ }
+ }
+ }
+}
+
+/// A reference-counted Operating performance point (OPP).
+///
+/// Rust abstraction for the C `struct dev_pm_opp`.
+///
+/// # Invariants
+///
+/// The pointer stored in `Self` is non-null and valid for the lifetime of the [`OPP`].
+///
+/// Instances of this type are reference-counted. The reference count is incremented by the
+/// `dev_pm_opp_get` function and decremented by `dev_pm_opp_put`. The Rust type `ARef<OPP>`
+/// represents a pointer that owns a reference count on the [`OPP`].
+///
+/// A reference to the [`OPP`], &[`OPP`], isn't refcounted by the Rust code.
+///
+/// ## Examples
+///
+/// The following example demonstrates how to get [`OPP`] corresponding to a frequency value and
+/// configure the device with it.
+///
+/// ```
+/// use kernel::clk::Hertz;
+/// use kernel::error::Result;
+/// use kernel::opp::{SearchType, Table};
+///
+/// fn configure_opp(table: &Table, freq: Hertz) -> Result {
+/// let opp = table.opp_from_freq(freq, Some(true), None, SearchType::Exact)?;
+///
+/// if opp.freq(None) != freq {
+/// return Err(EINVAL);
+/// }
+///
+/// table.set_opp(&opp)
+/// }
+/// ```
+#[repr(transparent)]
+pub struct OPP(Opaque<bindings::dev_pm_opp>);
+
+/// SAFETY: It is okay to send the ownership of [`OPP`] across thread boundaries.
+unsafe impl Send for OPP {}
+
+/// SAFETY: It is okay to access [`OPP`] through shared references from other threads because we're
+/// either accessing properties that don't change or that are properly synchronised by C code.
+unsafe impl Sync for OPP {}
+
+/// SAFETY: The type invariants guarantee that [`OPP`] is always refcounted.
+unsafe impl AlwaysRefCounted for OPP {
+ fn inc_ref(&self) {
+ // SAFETY: The existence of a shared reference means that the refcount is nonzero.
+ unsafe { bindings::dev_pm_opp_get(self.0.get()) };
+ }
+
+ unsafe fn dec_ref(obj: ptr::NonNull<Self>) {
+ // SAFETY: The safety requirements guarantee that the refcount is nonzero.
+ unsafe { bindings::dev_pm_opp_put(obj.cast().as_ptr()) }
+ }
+}
+
+impl OPP {
+ /// Creates an owned reference to a [`OPP`] from a valid pointer.
+ ///
+ /// The refcount is incremented by the C code and will be decremented by `dec_ref` when the
+ /// [`ARef`] object is dropped.
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that `ptr` is valid and the refcount of the [`OPP`] is incremented.
+ /// The caller must also ensure that it doesn't explicitly drop the refcount of the [`OPP`], as
+ /// the returned [`ARef`] object takes over the refcount increment on the underlying object and
+ /// the same will be dropped along with it.
+ pub unsafe fn from_raw_opp_owned(ptr: *mut bindings::dev_pm_opp) -> Result<ARef<Self>> {
+ let ptr = ptr::NonNull::new(ptr).ok_or(ENODEV)?;
+
+ // SAFETY: The safety requirements guarantee the validity of the pointer.
+ //
+ // INVARIANT: The reference-count is decremented when [`OPP`] goes out of scope.
+ Ok(unsafe { ARef::from_raw(ptr.cast()) })
+ }
+
+ /// Creates a reference to a [`OPP`] from a valid pointer.
+ ///
+ /// The refcount is not updated by the Rust API unless the returned reference is converted to
+ /// an [`ARef`] object.
+ ///
+ /// # Safety
+ ///
+ /// The caller must ensure that `ptr` is valid and remains valid for the duration of `'a`.
+ #[inline]
+ pub unsafe fn from_raw_opp<'a>(ptr: *mut bindings::dev_pm_opp) -> Result<&'a Self> {
+ // SAFETY: The caller guarantees that the pointer is not dangling and stays valid for the
+ // duration of 'a. The cast is okay because [`OPP`] is `repr(transparent)`.
+ Ok(unsafe { &*ptr.cast() })
+ }
+
+ #[inline]
+ fn as_raw(&self) -> *mut bindings::dev_pm_opp {
+ self.0.get()
+ }
+
+ /// Returns the frequency of an [`OPP`].
+ pub fn freq(&self, index: Option<u32>) -> Hertz {
+ let index = index.unwrap_or(0);
+
+ // SAFETY: By the type invariants, we know that `self` owns a reference, so it is safe to
+ // use it.
+ Hertz(unsafe { bindings::dev_pm_opp_get_freq_indexed(self.as_raw(), index) })
+ }
+
+ /// Returns the voltage of an [`OPP`].
+ #[inline]
+ pub fn voltage(&self) -> MicroVolt {
+ // SAFETY: By the type invariants, we know that `self` owns a reference, so it is safe to
+ // use it.
+ MicroVolt(unsafe { bindings::dev_pm_opp_get_voltage(self.as_raw()) })
+ }
+
+ /// Returns the level of an [`OPP`].
+ #[inline]
+ pub fn level(&self) -> u32 {
+ // SAFETY: By the type invariants, we know that `self` owns a reference, so it is safe to
+ // use it.
+ unsafe { bindings::dev_pm_opp_get_level(self.as_raw()) }
+ }
+
+ /// Returns the power of an [`OPP`].
+ #[inline]
+ pub fn power(&self) -> MicroWatt {
+ // SAFETY: By the type invariants, we know that `self` owns a reference, so it is safe to
+ // use it.
+ MicroWatt(unsafe { bindings::dev_pm_opp_get_power(self.as_raw()) })
+ }
+
+ /// Returns the required pstate of an [`OPP`].
+ #[inline]
+ pub fn required_pstate(&self, index: u32) -> u32 {
+ // SAFETY: By the type invariants, we know that `self` owns a reference, so it is safe to
+ // use it.
+ unsafe { bindings::dev_pm_opp_get_required_pstate(self.as_raw(), index) }
+ }
+
+ /// Returns true if the [`OPP`] is turbo.
+ #[inline]
+ pub fn is_turbo(&self) -> bool {
+ // SAFETY: By the type invariants, we know that `self` owns a reference, so it is safe to
+ // use it.
+ unsafe { bindings::dev_pm_opp_is_turbo(self.as_raw()) }
+ }
+}
diff --git a/rust/kernel/task.rs b/rust/kernel/task.rs
index 9e6f6854948d..927413d85484 100644
--- a/rust/kernel/task.rs
+++ b/rust/kernel/task.rs
@@ -7,6 +7,7 @@
use crate::{
bindings,
ffi::{c_int, c_long, c_uint},
+ mm::MmWithUser,
pid_namespace::PidNamespace,
types::{ARef, NotThreadSafe, Opaque},
};
@@ -33,22 +34,20 @@ pub const TASK_NORMAL: c_uint = bindings::TASK_NORMAL as c_uint;
#[macro_export]
macro_rules! current {
() => {
- // SAFETY: Deref + addr-of below create a temporary `TaskRef` that cannot outlive the
- // caller.
+ // SAFETY: This expression creates a temporary value that is dropped at the end of the
+ // caller's scope. The following mechanisms ensure that the resulting `&CurrentTask` cannot
+ // leave current task context:
+ //
+ // * To return to userspace, the caller must leave the current scope.
+ // * Operations such as `begin_new_exec()` are necessarily unsafe and the caller of
+ // `begin_new_exec()` is responsible for safety.
+ // * Rust abstractions for things such as a `kthread_use_mm()` scope must require the
+ // closure to be `Send`, so the `NotThreadSafe` field of `CurrentTask` ensures that the
+ // `&CurrentTask` cannot cross the scope in either direction.
unsafe { &*$crate::task::Task::current() }
};
}
-/// Returns the currently running task's pid namespace.
-#[macro_export]
-macro_rules! current_pid_ns {
- () => {
- // SAFETY: Deref + addr-of below create a temporary `PidNamespaceRef` that cannot outlive
- // the caller.
- unsafe { &*$crate::task::Task::current_pid_ns() }
- };
-}
-
/// Wraps the kernel's `struct task_struct`.
///
/// # Invariants
@@ -87,7 +86,7 @@ macro_rules! current_pid_ns {
/// impl State {
/// fn new() -> Self {
/// Self {
-/// creator: current!().into(),
+/// creator: ARef::from(&**current!()),
/// index: 0,
/// }
/// }
@@ -107,6 +106,44 @@ unsafe impl Send for Task {}
// synchronised by C code (e.g., `signal_pending`).
unsafe impl Sync for Task {}
+/// Represents the [`Task`] in the `current` global.
+///
+/// This type exists to provide more efficient operations that are only valid on the current task.
+/// For example, to retrieve the pid-namespace of a task, you must use rcu protection unless it is
+/// the current task.
+///
+/// # Invariants
+///
+/// Each value of this type must only be accessed from the task context it was created within.
+///
+/// Of course, every thread is in a different task context, but for the purposes of this invariant,
+/// these operations also permanently leave the task context:
+///
+/// * Returning to userspace from system call context.
+/// * Calling `release_task()`.
+/// * Calling `begin_new_exec()` in a binary format loader.
+///
+/// Other operations temporarily create a new sub-context:
+///
+/// * Calling `kthread_use_mm()` creates a new context, and `kthread_unuse_mm()` returns to the
+/// old context.
+///
+/// This means that a `CurrentTask` obtained before a `kthread_use_mm()` call may be used again
+/// once `kthread_unuse_mm()` is called, but it must not be used between these two calls.
+/// Conversely, a `CurrentTask` obtained between a `kthread_use_mm()`/`kthread_unuse_mm()` pair
+/// must not be used after `kthread_unuse_mm()`.
+#[repr(transparent)]
+pub struct CurrentTask(Task, NotThreadSafe);
+
+// Make all `Task` methods available on `CurrentTask`.
+impl Deref for CurrentTask {
+ type Target = Task;
+ #[inline]
+ fn deref(&self) -> &Task {
+ &self.0
+ }
+}
+
/// The type of process identifiers (PIDs).
pub type Pid = bindings::pid_t;
@@ -133,119 +170,29 @@ impl Task {
///
/// # Safety
///
- /// Callers must ensure that the returned object doesn't outlive the current task/thread.
- pub unsafe fn current() -> impl Deref<Target = Task> {
- struct TaskRef<'a> {
- task: &'a Task,
- _not_send: NotThreadSafe,
+ /// Callers must ensure that the returned object is only used to access a [`CurrentTask`]
+ /// within the task context that was active when this function was called. For more details,
+ /// see the invariants section for [`CurrentTask`].
+ pub unsafe fn current() -> impl Deref<Target = CurrentTask> {
+ struct TaskRef {
+ task: *const CurrentTask,
}
- impl Deref for TaskRef<'_> {
- type Target = Task;
+ impl Deref for TaskRef {
+ type Target = CurrentTask;
fn deref(&self) -> &Self::Target {
- self.task
+ // SAFETY: The returned reference borrows from this `TaskRef`, so it cannot outlive
+ // the `TaskRef`, which the caller of `Task::current()` has promised will not
+ // outlive the task/thread for which `self.task` is the `current` pointer. Thus, it
+ // is okay to return a `CurrentTask` reference here.
+ unsafe { &*self.task }
}
}
- let current = Task::current_raw();
TaskRef {
- // SAFETY: If the current thread is still running, the current task is valid. Given
- // that `TaskRef` is not `Send`, we know it cannot be transferred to another thread
- // (where it could potentially outlive the caller).
- task: unsafe { &*current.cast() },
- _not_send: NotThreadSafe,
- }
- }
-
- /// Returns a PidNamespace reference for the currently executing task's/thread's pid namespace.
- ///
- /// This function can be used to create an unbounded lifetime by e.g., storing the returned
- /// PidNamespace in a global variable which would be a bug. So the recommended way to get the
- /// current task's/thread's pid namespace is to use the [`current_pid_ns`] macro because it is
- /// safe.
- ///
- /// # Safety
- ///
- /// Callers must ensure that the returned object doesn't outlive the current task/thread.
- pub unsafe fn current_pid_ns() -> impl Deref<Target = PidNamespace> {
- struct PidNamespaceRef<'a> {
- task: &'a PidNamespace,
- _not_send: NotThreadSafe,
- }
-
- impl Deref for PidNamespaceRef<'_> {
- type Target = PidNamespace;
-
- fn deref(&self) -> &Self::Target {
- self.task
- }
- }
-
- // The lifetime of `PidNamespace` is bound to `Task` and `struct pid`.
- //
- // The `PidNamespace` of a `Task` doesn't ever change once the `Task` is alive. A
- // `unshare(CLONE_NEWPID)` or `setns(fd_pidns/pidfd, CLONE_NEWPID)` will not have an effect
- // on the calling `Task`'s pid namespace. It will only effect the pid namespace of children
- // created by the calling `Task`. This invariant guarantees that after having acquired a
- // reference to a `Task`'s pid namespace it will remain unchanged.
- //
- // When a task has exited and been reaped `release_task()` will be called. This will set
- // the `PidNamespace` of the task to `NULL`. So retrieving the `PidNamespace` of a task
- // that is dead will return `NULL`. Note, that neither holding the RCU lock nor holding a
- // referencing count to
- // the `Task` will prevent `release_task()` being called.
- //
- // In order to retrieve the `PidNamespace` of a `Task` the `task_active_pid_ns()` function
- // can be used. There are two cases to consider:
- //
- // (1) retrieving the `PidNamespace` of the `current` task
- // (2) retrieving the `PidNamespace` of a non-`current` task
- //
- // From system call context retrieving the `PidNamespace` for case (1) is always safe and
- // requires neither RCU locking nor a reference count to be held. Retrieving the
- // `PidNamespace` after `release_task()` for current will return `NULL` but no codepath
- // like that is exposed to Rust.
- //
- // Retrieving the `PidNamespace` from system call context for (2) requires RCU protection.
- // Accessing `PidNamespace` outside of RCU protection requires a reference count that
- // must've been acquired while holding the RCU lock. Note that accessing a non-`current`
- // task means `NULL` can be returned as the non-`current` task could have already passed
- // through `release_task()`.
- //
- // To retrieve (1) the `current_pid_ns!()` macro should be used which ensure that the
- // returned `PidNamespace` cannot outlive the calling scope. The associated
- // `current_pid_ns()` function should not be called directly as it could be abused to
- // created an unbounded lifetime for `PidNamespace`. The `current_pid_ns!()` macro allows
- // Rust to handle the common case of accessing `current`'s `PidNamespace` without RCU
- // protection and without having to acquire a reference count.
- //
- // For (2) the `task_get_pid_ns()` method must be used. This will always acquire a
- // reference on `PidNamespace` and will return an `Option` to force the caller to
- // explicitly handle the case where `PidNamespace` is `None`, something that tends to be
- // forgotten when doing the equivalent operation in `C`. Missing RCU primitives make it
- // difficult to perform operations that are otherwise safe without holding a reference
- // count as long as RCU protection is guaranteed. But it is not important currently. But we
- // do want it in the future.
- //
- // Note for (2) the required RCU protection around calling `task_active_pid_ns()`
- // synchronizes against putting the last reference of the associated `struct pid` of
- // `task->thread_pid`. The `struct pid` stored in that field is used to retrieve the
- // `PidNamespace` of the caller. When `release_task()` is called `task->thread_pid` will be
- // `NULL`ed and `put_pid()` on said `struct pid` will be delayed in `free_pid()` via
- // `call_rcu()` allowing everyone with an RCU protected access to the `struct pid` acquired
- // from `task->thread_pid` to finish.
- //
- // SAFETY: The current task's pid namespace is valid as long as the current task is running.
- let pidns = unsafe { bindings::task_active_pid_ns(Task::current_raw()) };
- PidNamespaceRef {
- // SAFETY: If the current thread is still running, the current task and its associated
- // pid namespace are valid. `PidNamespaceRef` is not `Send`, so we know it cannot be
- // transferred to another thread (where it could potentially outlive the current
- // `Task`). The caller needs to ensure that the PidNamespaceRef doesn't outlive the
- // current task/thread.
- task: unsafe { PidNamespace::from_ptr(pidns) },
- _not_send: NotThreadSafe,
+ // CAST: The layout of `struct task_struct` and `CurrentTask` is identical.
+ task: Task::current_raw().cast(),
}
}
@@ -328,6 +275,70 @@ impl Task {
}
}
+impl CurrentTask {
+ /// Access the address space of the current task.
+ ///
+ /// This function does not touch the refcount of the mm.
+ #[inline]
+ pub fn mm(&self) -> Option<&MmWithUser> {
+ // SAFETY: The `mm` field of `current` is not modified from other threads, so reading it is
+ // not a data race.
+ let mm = unsafe { (*self.as_ptr()).mm };
+
+ if mm.is_null() {
+ return None;
+ }
+
+ // SAFETY: If `current->mm` is non-null, then it references a valid mm with a non-zero
+ // value of `mm_users`. Furthermore, the returned `&MmWithUser` borrows from this
+ // `CurrentTask`, so it cannot escape the scope in which the current pointer was obtained.
+ //
+ // This is safe even if `kthread_use_mm()`/`kthread_unuse_mm()` are used. There are two
+ // relevant cases:
+ // * If the `&CurrentTask` was created before `kthread_use_mm()`, then it cannot be
+ // accessed during the `kthread_use_mm()`/`kthread_unuse_mm()` scope due to the
+ // `NotThreadSafe` field of `CurrentTask`.
+ // * If the `&CurrentTask` was created within a `kthread_use_mm()`/`kthread_unuse_mm()`
+ // scope, then the `&CurrentTask` cannot escape that scope, so the returned `&MmWithUser`
+ // also cannot escape that scope.
+ // In either case, it's not possible to read `current->mm` and keep using it after the
+ // scope is ended with `kthread_unuse_mm()`.
+ Some(unsafe { MmWithUser::from_raw(mm) })
+ }
+
+ /// Access the pid namespace of the current task.
+ ///
+ /// This function does not touch the refcount of the namespace or use RCU protection.
+ ///
+ /// To access the pid namespace of another task, see [`Task::get_pid_ns`].
+ #[doc(alias = "task_active_pid_ns")]
+ #[inline]
+ pub fn active_pid_ns(&self) -> Option<&PidNamespace> {
+ // SAFETY: It is safe to call `task_active_pid_ns` without RCU protection when calling it
+ // on the current task.
+ let active_ns = unsafe { bindings::task_active_pid_ns(self.as_ptr()) };
+
+ if active_ns.is_null() {
+ return None;
+ }
+
+ // The lifetime of `PidNamespace` is bound to `Task` and `struct pid`.
+ //
+ // The `PidNamespace` of a `Task` doesn't ever change once the `Task` is alive.
+ //
+ // From system call context retrieving the `PidNamespace` for the current task is always
+ // safe and requires neither RCU locking nor a reference count to be held. Retrieving the
+ // `PidNamespace` after `release_task()` for current will return `NULL` but no codepath
+ // like that is exposed to Rust.
+ //
+ // SAFETY: If `current`'s pid ns is non-null, then it references a valid pid ns.
+ // Furthermore, the returned `&PidNamespace` borrows from this `CurrentTask`, so it cannot
+ // escape the scope in which the current pointer was obtained, e.g. it cannot live past a
+ // `release_task()` call.
+ Some(unsafe { PidNamespace::from_ptr(active_ns) })
+ }
+}
+
// SAFETY: The type invariants guarantee that `Task` is always refcounted.
unsafe impl crate::types::AlwaysRefCounted for Task {
fn inc_ref(&self) {
diff --git a/rust/macros/module.rs b/rust/macros/module.rs
index 2f66107847f7..c4afdd69e490 100644
--- a/rust/macros/module.rs
+++ b/rust/macros/module.rs
@@ -176,7 +176,9 @@ pub(crate) fn module(ts: TokenStream) -> TokenStream {
let info = ModuleInfo::parse(&mut it);
- let mut modinfo = ModInfoBuilder::new(info.name.as_ref());
+ // Rust does not allow hyphens in identifiers, use underscore instead.
+ let ident = info.name.replace('-', "_");
+ let mut modinfo = ModInfoBuilder::new(ident.as_ref());
if let Some(author) = info.author {
modinfo.emit("author", &author);
}
@@ -301,14 +303,15 @@ pub(crate) fn module(ts: TokenStream) -> TokenStream {
#[doc(hidden)]
#[link_section = \"{initcall_section}\"]
#[used]
- pub static __{name}_initcall: extern \"C\" fn() -> kernel::ffi::c_int = __{name}_init;
+ pub static __{ident}_initcall: extern \"C\" fn() ->
+ kernel::ffi::c_int = __{ident}_init;
#[cfg(not(MODULE))]
#[cfg(CONFIG_HAVE_ARCH_PREL32_RELOCATIONS)]
core::arch::global_asm!(
r#\".section \"{initcall_section}\", \"a\"
- __{name}_initcall:
- .long __{name}_init - .
+ __{ident}_initcall:
+ .long __{ident}_init - .
.previous
\"#
);
@@ -316,7 +319,7 @@ pub(crate) fn module(ts: TokenStream) -> TokenStream {
#[cfg(not(MODULE))]
#[doc(hidden)]
#[no_mangle]
- pub extern \"C\" fn __{name}_init() -> kernel::ffi::c_int {{
+ pub extern \"C\" fn __{ident}_init() -> kernel::ffi::c_int {{
// SAFETY: This function is inaccessible to the outside due to the double
// module wrapping it. It is called exactly once by the C side via its
// placement above in the initcall section.
@@ -326,13 +329,13 @@ pub(crate) fn module(ts: TokenStream) -> TokenStream {
#[cfg(not(MODULE))]
#[doc(hidden)]
#[no_mangle]
- pub extern \"C\" fn __{name}_exit() {{
+ pub extern \"C\" fn __{ident}_exit() {{
// SAFETY:
// - This function is inaccessible to the outside due to the double
// module wrapping it. It is called exactly once by the C side via its
// unique name,
- // - furthermore it is only called after `__{name}_init` has returned `0`
- // (which delegates to `__init`).
+ // - furthermore it is only called after `__{ident}_init` has
+ // returned `0` (which delegates to `__init`).
unsafe {{ __exit() }}
}}
@@ -372,6 +375,7 @@ pub(crate) fn module(ts: TokenStream) -> TokenStream {
",
type_ = info.type_,
name = info.name,
+ ident = ident,
modinfo = modinfo.buffer,
initcall_section = ".initcall6.init"
)