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, 99 insertions, 0 deletions

diff --git a/drivers/reset/reset-gpio.c b/drivers/reset/reset-gpio.c
new file mode 100644
index 000000000..bf95296a5
--- /dev/null
+++ b/drivers/reset/reset-gpio.c
@@ -0,0 +1,99 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <linux/auxiliary_bus.h>
+#include <linux/gpio/consumer.h>
+#include <linux/module.h>
+#include <linux/property.h>
+#include <linux/reset-controller.h>
+
+struct reset_gpio_priv {
+	struct reset_controller_dev rc;
+	struct gpio_desc *reset;
+};
+
+static inline struct reset_gpio_priv
+*rc_to_reset_gpio(struct reset_controller_dev *rc)
+{
+	return container_of(rc, struct reset_gpio_priv, rc);
+}
+
+static int reset_gpio_assert(struct reset_controller_dev *rc, unsigned long id)
+{
+	struct reset_gpio_priv *priv = rc_to_reset_gpio(rc);
+
+	return gpiod_set_value_cansleep(priv->reset, 1);
+}
+
+static int reset_gpio_deassert(struct reset_controller_dev *rc,
+			       unsigned long id)
+{
+	struct reset_gpio_priv *priv = rc_to_reset_gpio(rc);
+
+	return gpiod_set_value_cansleep(priv->reset, 0);
+}
+
+static int reset_gpio_status(struct reset_controller_dev *rc, unsigned long id)
+{
+	struct reset_gpio_priv *priv = rc_to_reset_gpio(rc);
+
+	return gpiod_get_value_cansleep(priv->reset);
+}
+
+static const struct reset_control_ops reset_gpio_ops = {
+	.assert = reset_gpio_assert,
+	.deassert = reset_gpio_deassert,
+	.status = reset_gpio_status,
+};
+
+static int reset_gpio_fwnode_xlate(struct reset_controller_dev *rcdev,
+				   const struct fwnode_reference_args *reset_spec)
+{
+	return reset_spec->args[0];
+}
+
+static int reset_gpio_probe(struct auxiliary_device *adev,
+			    const struct auxiliary_device_id *id)
+{
+	struct device *dev = &adev->dev;
+	struct reset_gpio_priv *priv;
+
+	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	priv->reset = devm_gpiod_get(dev, "reset", GPIOD_OUT_HIGH);
+	if (IS_ERR(priv->reset))
+		return dev_err_probe(dev, PTR_ERR(priv->reset),
+				     "Could not get reset gpios\n");
+
+	priv->rc.ops = &reset_gpio_ops;
+	priv->rc.owner = THIS_MODULE;
+	priv->rc.dev = dev;
+
+	/* Cells to match GPIO specifier, but it's not really used */
+	priv->rc.fwnode_reset_n_cells = 2;
+	priv->rc.fwnode_xlate = reset_gpio_fwnode_xlate;
+	priv->rc.nr_resets = 1;
+
+	return devm_reset_controller_register(dev, &priv->rc);
+}
+
+static const struct auxiliary_device_id reset_gpio_ids[] = {
+	{ .name = "reset.gpio" },
+	{}
+};
+MODULE_DEVICE_TABLE(auxiliary, reset_gpio_ids);
+
+static struct auxiliary_driver reset_gpio_driver = {
+	.probe		= reset_gpio_probe,
+	.id_table	= reset_gpio_ids,
+	.driver	= {
+		.name = "reset-gpio",
+		.suppress_bind_attrs = true,
+	},
+};
+module_auxiliary_driver(reset_gpio_driver);
+
+MODULE_AUTHOR("Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>");
+MODULE_DESCRIPTION("Generic GPIO reset driver");
+MODULE_LICENSE("GPL");