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, 128 insertions, 0 deletions

diff --git a/drivers/reset/reset-a10sr.c b/drivers/reset/reset-a10sr.c
new file mode 100644
index 000000000..99b3bc838
--- /dev/null
+++ b/drivers/reset/reset-a10sr.c
@@ -0,0 +1,128 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  Copyright Intel Corporation (C) 2017. All Rights Reserved
+ *
+ * Reset driver for Altera Arria10 MAX5 System Resource Chip
+ *
+ * Adapted from reset-socfpga.c
+ */
+
+#include <linux/err.h>
+#include <linux/mfd/altera-a10sr.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/reset-controller.h>
+
+#include <dt-bindings/reset/altr,rst-mgr-a10sr.h>
+
+struct a10sr_reset {
+	struct reset_controller_dev     rcdev;
+	struct regmap *regmap;
+};
+
+static inline struct a10sr_reset *to_a10sr_rst(struct reset_controller_dev *rc)
+{
+	return container_of(rc, struct a10sr_reset, rcdev);
+}
+
+static inline int a10sr_reset_shift(unsigned long id)
+{
+	switch (id) {
+	case A10SR_RESET_ENET_HPS:
+		return 1;
+	case A10SR_RESET_PCIE:
+	case A10SR_RESET_FILE:
+	case A10SR_RESET_BQSPI:
+	case A10SR_RESET_USB:
+		return id + 11;
+	default:
+		return -EINVAL;
+	}
+}
+
+static int a10sr_reset_update(struct reset_controller_dev *rcdev,
+			      unsigned long id, bool assert)
+{
+	struct a10sr_reset *a10r = to_a10sr_rst(rcdev);
+	int offset = a10sr_reset_shift(id);
+	u8 mask = ALTR_A10SR_REG_BIT_MASK(offset);
+	int index = ALTR_A10SR_HPS_RST_REG + ALTR_A10SR_REG_OFFSET(offset);
+
+	return regmap_update_bits(a10r->regmap, index, mask, assert ? 0 : mask);
+}
+
+static int a10sr_reset_assert(struct reset_controller_dev *rcdev,
+			      unsigned long id)
+{
+	return a10sr_reset_update(rcdev, id, true);
+}
+
+static int a10sr_reset_deassert(struct reset_controller_dev *rcdev,
+				unsigned long id)
+{
+	return a10sr_reset_update(rcdev, id, false);
+}
+
+static int a10sr_reset_status(struct reset_controller_dev *rcdev,
+			      unsigned long id)
+{
+	int ret;
+	struct a10sr_reset *a10r = to_a10sr_rst(rcdev);
+	int offset = a10sr_reset_shift(id);
+	u8 mask = ALTR_A10SR_REG_BIT_MASK(offset);
+	int index = ALTR_A10SR_HPS_RST_REG + ALTR_A10SR_REG_OFFSET(offset);
+	unsigned int value;
+
+	ret = regmap_read(a10r->regmap, index, &value);
+	if (ret < 0)
+		return ret;
+
+	return !!(value & mask);
+}
+
+static const struct reset_control_ops a10sr_reset_ops = {
+	.assert		= a10sr_reset_assert,
+	.deassert	= a10sr_reset_deassert,
+	.status		= a10sr_reset_status,
+};
+
+static int a10sr_reset_probe(struct platform_device *pdev)
+{
+	struct altr_a10sr *a10sr = dev_get_drvdata(pdev->dev.parent);
+	struct a10sr_reset *a10r;
+
+	a10r = devm_kzalloc(&pdev->dev, sizeof(struct a10sr_reset),
+			    GFP_KERNEL);
+	if (!a10r)
+		return -ENOMEM;
+
+	a10r->rcdev.owner = THIS_MODULE;
+	a10r->rcdev.nr_resets = A10SR_RESET_NUM;
+	a10r->rcdev.ops = &a10sr_reset_ops;
+	a10r->rcdev.of_node = pdev->dev.of_node;
+	a10r->regmap = a10sr->regmap;
+
+	platform_set_drvdata(pdev, a10r);
+
+	return devm_reset_controller_register(&pdev->dev, &a10r->rcdev);
+}
+
+static const struct of_device_id a10sr_reset_of_match[] = {
+	{ .compatible = "altr,a10sr-reset" },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, a10sr_reset_of_match);
+
+static struct platform_driver a10sr_reset_driver = {
+	.probe	= a10sr_reset_probe,
+	.driver = {
+		.name		= "altr_a10sr_reset",
+		.of_match_table	= a10sr_reset_of_match,
+	},
+};
+module_platform_driver(a10sr_reset_driver);
+
+MODULE_AUTHOR("Thor Thayer <thor.thayer@linux.intel.com>");
+MODULE_DESCRIPTION("Altera Arria10 System Resource Reset Controller Driver");
+MODULE_LICENSE("GPL v2");