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, 226 insertions, 0 deletions

diff --git a/drivers/misc/isl29020.c b/drivers/misc/isl29020.c
new file mode 100644
index 000000000..dd966db28
--- /dev/null
+++ b/drivers/misc/isl29020.c
@@ -0,0 +1,226 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * isl29020.c - Intersil  ALS Driver
+ *
+ * Copyright (C) 2008 Intel Corp
+ *
+ *  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ * Data sheet at: http://www.intersil.com/data/fn/fn6505.pdf
+ */
+
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/i2c.h>
+#include <linux/err.h>
+#include <linux/delay.h>
+#include <linux/sysfs.h>
+#include <linux/pm_runtime.h>
+
+static DEFINE_MUTEX(mutex);
+
+static ssize_t als_sensing_range_show(struct device *dev,
+			struct device_attribute *attr,  char *buf)
+{
+	struct i2c_client *client = to_i2c_client(dev);
+	int  val;
+
+	val = i2c_smbus_read_byte_data(client, 0x00);
+
+	if (val < 0)
+		return val;
+	return sprintf(buf, "%d000\n", 1 << (2 * (val & 3)));
+
+}
+
+static ssize_t als_lux_input_data_show(struct device *dev,
+			struct device_attribute *attr, char *buf)
+{
+	struct i2c_client *client = to_i2c_client(dev);
+	int ret_val, val;
+	unsigned long int lux;
+	int temp;
+
+	pm_runtime_get_sync(dev);
+	msleep(100);
+
+	mutex_lock(&mutex);
+	temp = i2c_smbus_read_byte_data(client, 0x02); /* MSB data */
+	if (temp < 0) {
+		pm_runtime_put_sync(dev);
+		mutex_unlock(&mutex);
+		return temp;
+	}
+
+	ret_val = i2c_smbus_read_byte_data(client, 0x01); /* LSB data */
+	mutex_unlock(&mutex);
+
+	if (ret_val < 0) {
+		pm_runtime_put_sync(dev);
+		return ret_val;
+	}
+
+	ret_val |= temp << 8;
+	val = i2c_smbus_read_byte_data(client, 0x00);
+	pm_runtime_put_sync(dev);
+	if (val < 0)
+		return val;
+	lux = ((((1 << (2 * (val & 3))))*1000) * ret_val) / 65536;
+	return sprintf(buf, "%lu\n", lux);
+}
+
+static ssize_t als_sensing_range_store(struct device *dev,
+		struct device_attribute *attr, const  char *buf, size_t count)
+{
+	struct i2c_client *client = to_i2c_client(dev);
+	int ret_val;
+	unsigned long val;
+
+	ret_val = kstrtoul(buf, 10, &val);
+	if (ret_val)
+		return ret_val;
+
+	if (val < 1 || val > 64000)
+		return -EINVAL;
+
+	/* Pick the smallest sensor range that will meet our requirements */
+	if (val <= 1000)
+		val = 1;
+	else if (val <= 4000)
+		val = 2;
+	else if (val <= 16000)
+		val = 3;
+	else
+		val = 4;
+
+	ret_val = i2c_smbus_read_byte_data(client, 0x00);
+	if (ret_val < 0)
+		return ret_val;
+
+	ret_val &= 0xFC; /*reset the bit before setting them */
+	ret_val |= val - 1;
+	ret_val = i2c_smbus_write_byte_data(client, 0x00, ret_val);
+
+	if (ret_val < 0)
+		return ret_val;
+	return count;
+}
+
+static void als_set_power_state(struct i2c_client *client, int enable)
+{
+	int ret_val;
+
+	ret_val = i2c_smbus_read_byte_data(client, 0x00);
+	if (ret_val < 0)
+		return;
+
+	if (enable)
+		ret_val |= 0x80;
+	else
+		ret_val &= 0x7F;
+
+	i2c_smbus_write_byte_data(client, 0x00, ret_val);
+}
+
+static DEVICE_ATTR(lux0_sensor_range, S_IRUGO | S_IWUSR,
+	als_sensing_range_show, als_sensing_range_store);
+static DEVICE_ATTR(lux0_input, S_IRUGO, als_lux_input_data_show, NULL);
+
+static struct attribute *mid_att_als[] = {
+	&dev_attr_lux0_sensor_range.attr,
+	&dev_attr_lux0_input.attr,
+	NULL
+};
+
+static const struct attribute_group m_als_gr = {
+	.name = "isl29020",
+	.attrs = mid_att_als
+};
+
+static int als_set_default_config(struct i2c_client *client)
+{
+	int retval;
+
+	retval = i2c_smbus_write_byte_data(client, 0x00, 0xc0);
+	if (retval < 0) {
+		dev_err(&client->dev, "default write failed.");
+		return retval;
+	}
+	return 0;
+}
+
+static int  isl29020_probe(struct i2c_client *client)
+{
+	int res;
+
+	res = als_set_default_config(client);
+	if (res <  0)
+		return res;
+
+	res = sysfs_create_group(&client->dev.kobj, &m_als_gr);
+	if (res) {
+		dev_err(&client->dev, "isl29020: device create file failed\n");
+		return res;
+	}
+	dev_info(&client->dev, "%s isl29020: ALS chip found\n", client->name);
+	als_set_power_state(client, 0);
+	pm_runtime_enable(&client->dev);
+	return res;
+}
+
+static void isl29020_remove(struct i2c_client *client)
+{
+	pm_runtime_disable(&client->dev);
+	sysfs_remove_group(&client->dev.kobj, &m_als_gr);
+}
+
+static const struct i2c_device_id isl29020_id[] = {
+	{ .name = "isl29020" },
+	{ }
+};
+
+MODULE_DEVICE_TABLE(i2c, isl29020_id);
+
+#ifdef CONFIG_PM
+
+static int isl29020_runtime_suspend(struct device *dev)
+{
+	struct i2c_client *client = to_i2c_client(dev);
+	als_set_power_state(client, 0);
+	return 0;
+}
+
+static int isl29020_runtime_resume(struct device *dev)
+{
+	struct i2c_client *client = to_i2c_client(dev);
+	als_set_power_state(client, 1);
+	return 0;
+}
+
+static const struct dev_pm_ops isl29020_pm_ops = {
+	.runtime_suspend = isl29020_runtime_suspend,
+	.runtime_resume = isl29020_runtime_resume,
+};
+
+#define ISL29020_PM_OPS (&isl29020_pm_ops)
+#else	/* CONFIG_PM */
+#define ISL29020_PM_OPS NULL
+#endif	/* CONFIG_PM */
+
+static struct i2c_driver isl29020_driver = {
+	.driver = {
+		.name = "isl29020",
+		.pm = ISL29020_PM_OPS,
+	},
+	.probe = isl29020_probe,
+	.remove = isl29020_remove,
+	.id_table = isl29020_id,
+};
+
+module_i2c_driver(isl29020_driver);
+
+MODULE_AUTHOR("Kalhan Trisal <kalhan.trisal@intel.com>");
+MODULE_DESCRIPTION("Intersil isl29020 ALS Driver");
+MODULE_LICENSE("GPL v2");