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

1 files changed, 87 insertions, 0 deletions

diff --git a/drivers/acpi/tiny-power-button.c b/drivers/acpi/tiny-power-button.c
new file mode 100644
index 000000000..92516ef84
--- /dev/null
+++ b/drivers/acpi/tiny-power-button.c
@@ -0,0 +1,87 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+#include <linux/acpi.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/sched/signal.h>
+#include <acpi/button.h>
+
+MODULE_AUTHOR("Josh Triplett");
+MODULE_DESCRIPTION("ACPI Tiny Power Button Driver");
+MODULE_LICENSE("GPL");
+
+static int power_signal __read_mostly = CONFIG_ACPI_TINY_POWER_BUTTON_SIGNAL;
+module_param(power_signal, int, 0644);
+MODULE_PARM_DESC(power_signal, "Power button sends this signal to init");
+
+static const struct acpi_device_id tiny_power_button_device_ids[] = {
+	{ ACPI_BUTTON_HID_POWER, 0 },
+	{ ACPI_BUTTON_HID_POWERF, 0 },
+	{ "", 0 },
+};
+MODULE_DEVICE_TABLE(acpi, tiny_power_button_device_ids);
+
+static void acpi_tiny_power_button_notify(acpi_handle handle, u32 event, void *data)
+{
+	kill_cad_pid(power_signal, 1);
+}
+
+static void acpi_tiny_power_button_notify_run(void *not_used)
+{
+	acpi_tiny_power_button_notify(NULL, ACPI_FIXED_HARDWARE_EVENT, NULL);
+}
+
+static u32 acpi_tiny_power_button_event(void *not_used)
+{
+	acpi_os_execute(OSL_NOTIFY_HANDLER, acpi_tiny_power_button_notify_run, NULL);
+	return ACPI_INTERRUPT_HANDLED;
+}
+
+static int acpi_tiny_power_button_probe(struct platform_device *pdev)
+{
+	struct acpi_device *device;
+	acpi_status status;
+
+	device = ACPI_COMPANION(&pdev->dev);
+	if (!device)
+		return -ENODEV;
+
+	if (device->device_type == ACPI_BUS_TYPE_POWER_BUTTON) {
+		status = acpi_install_fixed_event_handler(ACPI_EVENT_POWER_BUTTON,
+							  acpi_tiny_power_button_event,
+							  NULL);
+	} else {
+		status = acpi_install_notify_handler(device->handle,
+						     ACPI_DEVICE_NOTIFY,
+						     acpi_tiny_power_button_notify,
+						     NULL);
+	}
+	if (ACPI_FAILURE(status))
+		return -ENODEV;
+
+	return 0;
+}
+
+static void acpi_tiny_power_button_remove(struct platform_device *pdev)
+{
+	struct acpi_device *device = ACPI_COMPANION(&pdev->dev);
+
+	if (device->device_type == ACPI_BUS_TYPE_POWER_BUTTON) {
+		acpi_remove_fixed_event_handler(ACPI_EVENT_POWER_BUTTON,
+						acpi_tiny_power_button_event);
+	} else {
+		acpi_remove_notify_handler(device->handle, ACPI_DEVICE_NOTIFY,
+					   acpi_tiny_power_button_notify);
+	}
+	acpi_os_wait_events_complete();
+}
+
+static struct platform_driver acpi_tiny_power_button_driver = {
+	.probe = acpi_tiny_power_button_probe,
+	.remove = acpi_tiny_power_button_remove,
+	.driver = {
+		.name = "acpi-tiny-power-button",
+		.acpi_match_table = tiny_power_button_device_ids,
+	},
+};
+
+module_platform_driver(acpi_tiny_power_button_driver);