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

2 files changed, 111 insertions, 0 deletions

diff --git a/Documentation/devicetree/bindings/csky/cpus.txt b/Documentation/devicetree/bindings/csky/cpus.txt
new file mode 100644
index 000000000..ae79412f2
--- /dev/null
+++ b/Documentation/devicetree/bindings/csky/cpus.txt
@@ -0,0 +1,73 @@
+==================
+C-SKY CPU Bindings
+==================
+
+The device tree allows to describe the layout of CPUs in a system through
+the "cpus" node, which in turn contains a number of subnodes (ie "cpu")
+defining properties for every cpu.
+
+Only SMP system need to care about the cpus node and single processor
+needn't define cpus node at all.
+
+=====================================
+cpus and cpu node bindings definition
+=====================================
+
+- cpus node
+
+	Description: Container of cpu nodes
+
+	The node name must be "cpus".
+
+	A cpus node must define the following properties:
+
+	- #address-cells
+		Usage: required
+		Value type: <u32>
+		Definition: must be set to 1
+	- #size-cells
+		Usage: required
+		Value type: <u32>
+		Definition: must be set to 0
+
+- cpu node
+
+	Description: Describes one of SMP cores
+
+	PROPERTIES
+
+	- device_type
+		Usage: required
+		Value type: <string>
+		Definition: must be "cpu"
+	- reg
+		Usage: required
+		Value type: <u32>
+		Definition: CPU index
+	- compatible:
+		Usage: required
+		Value type: <string>
+		Definition: must contain "csky", eg:
+			"csky,610"
+			"csky,807"
+			"csky,810"
+			"csky,860"
+
+Example:
+--------
+
+	cpus {
+		#address-cells = <1>;
+		#size-cells = <0>;
+		cpu@0 {
+			device_type = "cpu";
+			reg = <0>;
+			status = "ok";
+		};
+
+		cpu@1 {
+			device_type = "cpu";
+			reg = <1>;
+			status = "ok";
+		};
+	};
diff --git a/Documentation/devicetree/bindings/csky/pmu.txt b/Documentation/devicetree/bindings/csky/pmu.txt
new file mode 100644
index 000000000..728d05ca6
--- /dev/null
+++ b/Documentation/devicetree/bindings/csky/pmu.txt
@@ -0,0 +1,38 @@
+===============================
+C-SKY Performance Monitor Units
+===============================
+
+C-SKY Performance Monitor is designed for ck807/ck810/ck860 SMP soc and
+it could count cpu's events for helping analysis performance issues.
+
+============================
+PMU node bindings definition
+============================
+
+	Description: Describes PMU
+
+	PROPERTIES
+
+	- compatible
+		Usage: required
+		Value type: <string>
+		Definition: must be "csky,csky-pmu"
+	- interrupts
+		Usage: required
+		Value type: <u32 IRQ_TYPE_XXX>
+		Definition: must be pmu irq num defined by soc
+	- count-width
+		Usage: optional
+		Value type: <u32>
+		Definition: the width of pmu counter
+
+Examples:
+---------
+#include <dt-bindings/interrupt-controller/irq.h>
+
+	pmu: performace-monitor {
+		compatible = "csky,csky-pmu";
+		interrupts = <23 IRQ_TYPE_EDGE_RISING>;
+		interrupt-parent = <&intc>;
+		count-width = <48>;
+        };