sched/cache: Keep nr_pref_llc_running in the runnable domain, to fix LLC mis-scheduling buggrafted

alb_break_llc() decides whether to break LLC preference during active
load balance. It does so by testing that every runnable fair task on the
source rq prefers its LLC:

	env->src_rq->nr_pref_llc_running == env->src_rq->cfs.h_nr_runnable

But the two counters cover different sets. nr_pref_llc_running is updated
in account_llc_enqueue()/account_llc_dequeue(), next to cfs_rq->nr_queued,
so it follows queued tasks. h_nr_runnable is updated in set_delayed()/
clear_delayed() and drops delay-dequeued tasks.

So under DELAY_DEQUEUE, a preferring task that goes to sleep stays counted
in nr_pref_llc_running while h_nr_runnable falls. The equality then breaks,
alb_break_llc() returns false, and active balance is free to pull a task
off its preferred LLC. Active balance only moves runnable tasks, and this
is the only LLC check it consults: once the stopper runs, LBF_ACTIVE_LB
skips the per-task test in can_migrate_task(). The runnable set is the one
we want.

Fix it on the counter side. A task should be counted in
nr_pref_llc_running exactly while it is both queued on its preferred LLC
(pref_llc_queued) and runnable (!sched_delayed). Define that membership
once in task_pref_llc_runnable(), and adjust the counter only through
pref_llc_running_inc()/pref_llc_running_dec() from the four sites that
change either input: account_llc_enqueue(), account_llc_dequeue(),
set_delayed() and clear_delayed(). Gating every update on the same
predicate keeps the delay, wake and dequeue paths from double-counting
or underflowing; see the comments at those sites for the ordering.

nr_llc_running and sd->llc_counts are not touched and stay on queued
semantics.

Fixes: 714059f79ff0 ("sched/cache: Handle moving single tasks to/from their preferred LLC")
Closes: https://lore.kernel.org/lkml/20260827135000.735138-1-zhanxusheng@xiaomi.com/
Reported-by: Zhan Xusheng <zhanxusheng@xiaomi.com>
Suggested-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Kayra Cizmeci <kayracizmeci@gmail.com>
Cc: <stable@kernel.org> # v7.2.x
Link: https://patch.msgid.link/06af61afedac32e6477f57feb4d658f6c411c3af.1790035273.git.tim.c.chen@linux.intel.com

4 files changed, 377 insertions, 0 deletions

diff --git a/drivers/block/rnull/Kconfig b/drivers/block/rnull/Kconfig
new file mode 100644
index 000000000..7bc5b376c
--- /dev/null
+++ b/drivers/block/rnull/Kconfig
@@ -0,0 +1,13 @@
+# SPDX-License-Identifier: GPL-2.0
+#
+# Rust null block device driver configuration
+
+config BLK_DEV_RUST_NULL
+	tristate "Rust null block driver (Experimental)"
+	depends on RUST && CONFIGFS_FS
+	help
+	  This is the Rust implementation of the null block driver. Like
+	  the C version, the driver allows the user to create virutal block
+	  devices that can be configured via various configuration options.
+
+	  If unsure, say N.
diff --git a/drivers/block/rnull/Makefile b/drivers/block/rnull/Makefile
new file mode 100644
index 000000000..11cfa5e61
--- /dev/null
+++ b/drivers/block/rnull/Makefile
@@ -0,0 +1,3 @@
+
+obj-$(CONFIG_BLK_DEV_RUST_NULL) += rnull_mod.o
+rnull_mod-y := rnull.o
diff --git a/drivers/block/rnull/configfs.rs b/drivers/block/rnull/configfs.rs
new file mode 100644
index 000000000..32c10c3f4
--- /dev/null
+++ b/drivers/block/rnull/configfs.rs
@@ -0,0 +1,261 @@
+// SPDX-License-Identifier: GPL-2.0
+
+use super::NullBlkDevice;
+use kernel::{
+    block::mq::gen_disk::{GenDisk, GenDiskBuilder},
+    configfs::{self, AttributeOperations},
+    configfs_attrs,
+    fmt::{self, Write as _},
+    new_mutex,
+    page::PAGE_SIZE,
+    prelude::*,
+    str::{kstrtobool_bytes, CString},
+    sync::Mutex,
+};
+
+pub(crate) fn subsystem() -> impl PinInit<kernel::configfs::Subsystem<Config>, Error> {
+    let item_type = configfs_attrs! {
+        container: configfs::Subsystem<Config>,
+        data: Config,
+        child: DeviceConfig,
+        attributes: [
+            features: 0,
+        ],
+    };
+
+    kernel::configfs::Subsystem::new(c"rnull", item_type, try_pin_init!(Config {}))
+}
+
+#[pin_data]
+pub(crate) struct Config {}
+
+#[vtable]
+impl AttributeOperations<0> for Config {
+    type Data = Config;
+
+    fn show(_this: &Config, page: &mut [u8; PAGE_SIZE]) -> Result<usize> {
+        let mut writer = kernel::str::Formatter::new(page);
+        writer.write_str("blocksize,size,rotational,irqmode\n")?;
+        Ok(writer.bytes_written())
+    }
+}
+
+#[vtable]
+impl configfs::GroupOperations for Config {
+    type Child = DeviceConfig;
+
+    fn make_group(
+        &self,
+        name: &CStr,
+    ) -> Result<impl PinInit<configfs::Group<DeviceConfig>, Error>> {
+        let item_type = configfs_attrs! {
+            container: configfs::Group<DeviceConfig>,
+            data: DeviceConfig,
+            attributes: [
+                // Named for compatibility with C null_blk
+                power: 0,
+                blocksize: 1,
+                rotational: 2,
+                size: 3,
+                irqmode: 4,
+            ],
+        };
+
+        Ok(configfs::Group::new(
+            name.try_into()?,
+            item_type,
+            // TODO: cannot coerce new_mutex!() to impl PinInit<_, Error>, so put mutex inside
+            try_pin_init!( DeviceConfig {
+                data <- new_mutex!(DeviceConfigInner {
+                    powered: false,
+                    block_size: 4096,
+                    rotational: false,
+                    disk: None,
+                    capacity_mib: 4096,
+                    irq_mode: IRQMode::None,
+                    name: name.try_into()?,
+                }),
+            }),
+        ))
+    }
+}
+
+#[derive(Debug, Clone, Copy)]
+pub(crate) enum IRQMode {
+    None,
+    Soft,
+}
+
+impl TryFrom<u8> for IRQMode {
+    type Error = kernel::error::Error;
+
+    fn try_from(value: u8) -> Result<Self> {
+        match value {
+            0 => Ok(Self::None),
+            1 => Ok(Self::Soft),
+            _ => Err(EINVAL),
+        }
+    }
+}
+
+impl fmt::Display for IRQMode {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        match self {
+            Self::None => f.write_str("0")?,
+            Self::Soft => f.write_str("1")?,
+        }
+        Ok(())
+    }
+}
+
+#[pin_data]
+pub(crate) struct DeviceConfig {
+    #[pin]
+    data: Mutex<DeviceConfigInner>,
+}
+
+#[pin_data]
+struct DeviceConfigInner {
+    powered: bool,
+    name: CString,
+    block_size: u32,
+    rotational: bool,
+    capacity_mib: u64,
+    irq_mode: IRQMode,
+    disk: Option<GenDisk<NullBlkDevice>>,
+}
+
+#[vtable]
+impl configfs::AttributeOperations<0> for DeviceConfig {
+    type Data = DeviceConfig;
+
+    fn show(this: &DeviceConfig, page: &mut [u8; PAGE_SIZE]) -> Result<usize> {
+        let mut writer = kernel::str::Formatter::new(page);
+
+        if this.data.lock().powered {
+            writer.write_str("1\n")?;
+        } else {
+            writer.write_str("0\n")?;
+        }
+
+        Ok(writer.bytes_written())
+    }
+
+    fn store(this: &DeviceConfig, page: &[u8]) -> Result {
+        let power_op = kstrtobool_bytes(page)?;
+        let mut guard = this.data.lock();
+
+        if !guard.powered && power_op {
+            guard.disk = Some(NullBlkDevice::new(
+                &guard.name,
+                guard.block_size,
+                guard.rotational,
+                guard.capacity_mib,
+                guard.irq_mode,
+            )?);
+            guard.powered = true;
+        } else if guard.powered && !power_op {
+            drop(guard.disk.take());
+            guard.powered = false;
+        }
+
+        Ok(())
+    }
+}
+
+#[vtable]
+impl configfs::AttributeOperations<1> for DeviceConfig {
+    type Data = DeviceConfig;
+
+    fn show(this: &DeviceConfig, page: &mut [u8; PAGE_SIZE]) -> Result<usize> {
+        let mut writer = kernel::str::Formatter::new(page);
+        writer.write_fmt(fmt!("{}\n", this.data.lock().block_size))?;
+        Ok(writer.bytes_written())
+    }
+
+    fn store(this: &DeviceConfig, page: &[u8]) -> Result {
+        if this.data.lock().powered {
+            return Err(EBUSY);
+        }
+
+        let text = core::str::from_utf8(page)?.trim();
+        let value = text.parse::<u32>().map_err(|_| EINVAL)?;
+
+        GenDiskBuilder::validate_block_size(value)?;
+        this.data.lock().block_size = value;
+        Ok(())
+    }
+}
+
+#[vtable]
+impl configfs::AttributeOperations<2> for DeviceConfig {
+    type Data = DeviceConfig;
+
+    fn show(this: &DeviceConfig, page: &mut [u8; PAGE_SIZE]) -> Result<usize> {
+        let mut writer = kernel::str::Formatter::new(page);
+
+        if this.data.lock().rotational {
+            writer.write_str("1\n")?;
+        } else {
+            writer.write_str("0\n")?;
+        }
+
+        Ok(writer.bytes_written())
+    }
+
+    fn store(this: &DeviceConfig, page: &[u8]) -> Result {
+        if this.data.lock().powered {
+            return Err(EBUSY);
+        }
+
+        this.data.lock().rotational = kstrtobool_bytes(page)?;
+
+        Ok(())
+    }
+}
+
+#[vtable]
+impl configfs::AttributeOperations<3> for DeviceConfig {
+    type Data = DeviceConfig;
+
+    fn show(this: &DeviceConfig, page: &mut [u8; PAGE_SIZE]) -> Result<usize> {
+        let mut writer = kernel::str::Formatter::new(page);
+        writer.write_fmt(fmt!("{}\n", this.data.lock().capacity_mib))?;
+        Ok(writer.bytes_written())
+    }
+
+    fn store(this: &DeviceConfig, page: &[u8]) -> Result {
+        if this.data.lock().powered {
+            return Err(EBUSY);
+        }
+
+        let text = core::str::from_utf8(page)?.trim();
+        let value = text.parse::<u64>().map_err(|_| EINVAL)?;
+
+        this.data.lock().capacity_mib = value;
+        Ok(())
+    }
+}
+
+#[vtable]
+impl configfs::AttributeOperations<4> for DeviceConfig {
+    type Data = DeviceConfig;
+
+    fn show(this: &DeviceConfig, page: &mut [u8; PAGE_SIZE]) -> Result<usize> {
+        let mut writer = kernel::str::Formatter::new(page);
+        writer.write_fmt(fmt!("{}\n", this.data.lock().irq_mode))?;
+        Ok(writer.bytes_written())
+    }
+
+    fn store(this: &DeviceConfig, page: &[u8]) -> Result {
+        if this.data.lock().powered {
+            return Err(EBUSY);
+        }
+
+        let text = core::str::from_utf8(page)?.trim();
+        let value = text.parse::<u8>().map_err(|_| EINVAL)?;
+
+        this.data.lock().irq_mode = IRQMode::try_from(value)?;
+        Ok(())
+    }
+}
diff --git a/drivers/block/rnull/rnull.rs b/drivers/block/rnull/rnull.rs
new file mode 100644
index 000000000..0ca8715fe
--- /dev/null
+++ b/drivers/block/rnull/rnull.rs
@@ -0,0 +1,100 @@
+// SPDX-License-Identifier: GPL-2.0
+
+//! This is a Rust implementation of the C null block driver.
+
+mod configfs;
+
+use configfs::IRQMode;
+use kernel::{
+    block::{
+        self,
+        mq::{
+            self,
+            gen_disk::{self, GenDisk},
+            Operations, TagSet,
+        },
+    },
+    prelude::*,
+    sync::{aref::ARef, Arc},
+};
+
+module! {
+    type: NullBlkModule,
+    name: "rnull_mod",
+    authors: ["Andreas Hindborg"],
+    description: "Rust implementation of the C null block driver",
+    license: "GPL v2",
+}
+
+#[pin_data]
+struct NullBlkModule {
+    #[pin]
+    configfs_subsystem: kernel::configfs::Subsystem<configfs::Config>,
+}
+
+impl kernel::InPlaceModule for NullBlkModule {
+    fn init(_module: &'static ThisModule) -> impl PinInit<Self, Error> {
+        pr_info!("Rust null_blk loaded\n");
+
+        try_pin_init!(Self {
+            configfs_subsystem <- configfs::subsystem(),
+        })
+    }
+}
+
+struct NullBlkDevice;
+
+impl NullBlkDevice {
+    fn new(
+        name: &CStr,
+        block_size: u32,
+        rotational: bool,
+        capacity_mib: u64,
+        irq_mode: IRQMode,
+    ) -> Result<GenDisk<Self>> {
+        let tagset = Arc::pin_init(TagSet::new(1, 256, 1), GFP_KERNEL)?;
+
+        let queue_data = Box::new(QueueData { irq_mode }, GFP_KERNEL)?;
+
+        gen_disk::GenDiskBuilder::new()
+            .capacity_sectors(capacity_mib << (20 - block::SECTOR_SHIFT))
+            .logical_block_size(block_size)?
+            .physical_block_size(block_size)?
+            .rotational(rotational)
+            .build(fmt!("{}", name.to_str()?), tagset, queue_data)
+    }
+}
+
+struct QueueData {
+    irq_mode: IRQMode,
+}
+
+#[vtable]
+impl Operations for NullBlkDevice {
+    type QueueData = KBox<QueueData>;
+
+    #[inline(always)]
+    fn queue_rq(queue_data: &QueueData, rq: ARef<mq::Request<Self>>, _is_last: bool) -> Result {
+        match queue_data.irq_mode {
+            IRQMode::None => mq::Request::end_ok(rq)
+                .map_err(|_e| kernel::error::code::EIO)
+                // We take no refcounts on the request, so we expect to be able to
+                // end the request. The request reference must be unique at this
+                // point, and so `end_ok` cannot fail.
+                .expect("Fatal error - expected to be able to end request"),
+            IRQMode::Soft => mq::Request::complete(rq),
+        }
+        Ok(())
+    }
+
+    fn commit_rqs(_queue_data: &QueueData) {}
+
+    fn complete(rq: ARef<mq::Request<Self>>) {
+        mq::Request::end_ok(rq)
+            .map_err(|_e| kernel::error::code::EIO)
+            // We take no refcounts on the request, so we expect to be able to
+            // end the request. The request reference must be unique at this
+            // point, and so `end_ok` cannot fail.
+            .expect("Fatal error - expected to be able to end request");
+    }
+}