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, 85 insertions, 0 deletions

diff --git a/drivers/clk/sprd/div.h b/drivers/clk/sprd/div.h
new file mode 100644
index 000000000..a0bcad804
--- /dev/null
+++ b/drivers/clk/sprd/div.h
@@ -0,0 +1,85 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+//
+// Spreadtrum divider clock driver
+//
+// Copyright (C) 2017 Spreadtrum, Inc.
+// Author: Chunyan Zhang <chunyan.zhang@spreadtrum.com>
+
+#ifndef _SPRD_DIV_H_
+#define _SPRD_DIV_H_
+
+#include "common.h"
+
+/**
+ * struct sprd_div_internal - Internal divider description
+ * @shift: Bit offset of the divider in its register
+ * @width: Width of the divider field in its register
+ *
+ * That structure represents a single divider, and is meant to be
+ * embedded in other structures representing the various clock
+ * classes.
+ */
+struct sprd_div_internal {
+	s32	offset;
+	u8	shift;
+	u8	width;
+};
+
+#define _SPRD_DIV_CLK(_offset, _shift, _width)	\
+	{				\
+		.offset = _offset,	\
+		.shift	= _shift,	\
+		.width	= _width,	\
+	}
+
+struct sprd_div {
+	struct sprd_div_internal	div;
+	struct sprd_clk_common	common;
+};
+
+#define SPRD_DIV_CLK_HW_INIT_FN(_struct, _name, _parent, _reg, _offset,	\
+				_shift, _width, _flags, _fn)		\
+	struct sprd_div _struct = {					\
+		.div	= _SPRD_DIV_CLK(_offset, _shift, _width),	\
+		.common	= {						\
+			.regmap		= NULL,				\
+			.reg		= _reg,				\
+			.hw.init	= _fn(_name, _parent,		\
+					      &sprd_div_ops, _flags),	\
+		}							\
+	}
+
+#define SPRD_DIV_CLK(_struct, _name, _parent, _reg,			\
+		     _shift, _width, _flags)				\
+	SPRD_DIV_CLK_HW_INIT_FN(_struct, _name, _parent, _reg, 0x0,	\
+				_shift, _width, _flags, CLK_HW_INIT)
+
+#define SPRD_DIV_CLK_FW_NAME(_struct, _name, _parent, _reg,		\
+			_shift, _width, _flags)				\
+	SPRD_DIV_CLK_HW_INIT_FN(_struct, _name, _parent, _reg, 0x0,	\
+				_shift, _width, _flags, CLK_HW_INIT_FW_NAME)
+
+#define SPRD_DIV_CLK_HW(_struct, _name, _parent, _reg,			\
+			_shift, _width, _flags)				\
+	SPRD_DIV_CLK_HW_INIT_FN(_struct, _name, _parent, _reg, 0x0,	\
+				_shift, _width, _flags, CLK_HW_INIT_HW)
+
+static inline struct sprd_div *hw_to_sprd_div(const struct clk_hw *hw)
+{
+	struct sprd_clk_common *common = hw_to_sprd_clk_common(hw);
+
+	return container_of(common, struct sprd_div, common);
+}
+
+unsigned long sprd_div_helper_recalc_rate(struct sprd_clk_common *common,
+					  const struct sprd_div_internal *div,
+					  unsigned long parent_rate);
+
+int sprd_div_helper_set_rate(const struct sprd_clk_common *common,
+			     const struct sprd_div_internal *div,
+			     unsigned long rate,
+			     unsigned long parent_rate);
+
+extern const struct clk_ops sprd_div_ops;
+
+#endif /* _SPRD_DIV_H_ */