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, 124 insertions, 0 deletions

diff --git a/include/drm/spsc_queue.h b/include/drm/spsc_queue.h
new file mode 100644
index 000000000..ee9df8cc6
--- /dev/null
+++ b/include/drm/spsc_queue.h
@@ -0,0 +1,124 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef DRM_SCHEDULER_SPSC_QUEUE_H_
+#define DRM_SCHEDULER_SPSC_QUEUE_H_
+
+#include <linux/atomic.h>
+#include <linux/preempt.h>
+
+/** SPSC lockless queue */
+
+struct spsc_node {
+
+	/* Stores spsc_node* */
+	struct spsc_node *next;
+};
+
+struct spsc_queue {
+
+	 struct spsc_node *head;
+
+	/* atomic pointer to struct spsc_node* */
+	atomic_long_t tail;
+
+	atomic_t job_count;
+};
+
+static inline void spsc_queue_init(struct spsc_queue *queue)
+{
+	queue->head = NULL;
+	atomic_long_set(&queue->tail, (long)&queue->head);
+	atomic_set(&queue->job_count, 0);
+}
+
+static inline struct spsc_node *spsc_queue_peek(struct spsc_queue *queue)
+{
+	return queue->head;
+}
+
+static inline int spsc_queue_count(struct spsc_queue *queue)
+{
+	return atomic_read(&queue->job_count);
+}
+
+static inline bool spsc_queue_push(struct spsc_queue *queue, struct spsc_node *node)
+{
+	struct spsc_node **tail;
+
+	node->next = NULL;
+
+	preempt_disable();
+
+	atomic_inc(&queue->job_count);
+	smp_mb__after_atomic();
+
+	tail = (struct spsc_node **)atomic_long_xchg(&queue->tail, (long)&node->next);
+	WRITE_ONCE(*tail, node);
+
+	/*
+	 * In case of first element verify new node will be visible to the consumer
+	 * thread when we ping the kernel thread that there is new work to do.
+	 */
+	smp_wmb();
+
+	preempt_enable();
+
+	return tail == &queue->head;
+}
+
+
+static inline struct spsc_node *spsc_queue_pop(struct spsc_queue *queue)
+{
+	struct spsc_node *next, *node;
+
+	/* Verify reading from memory and not the cache */
+	smp_rmb();
+
+	node = READ_ONCE(queue->head);
+
+	if (!node)
+		return NULL;
+
+	next = READ_ONCE(node->next);
+	WRITE_ONCE(queue->head, next);
+
+	if (unlikely(!next)) {
+		/* slowpath for the last element in the queue */
+
+		if (atomic_long_cmpxchg(&queue->tail,
+				(long)&node->next, (long) &queue->head) != (long)&node->next) {
+			/* Updating tail failed wait for new next to appear */
+			do {
+				smp_rmb();
+			} while (unlikely(!(queue->head = READ_ONCE(node->next))));
+		}
+	}
+
+	atomic_dec(&queue->job_count);
+	return node;
+}
+
+
+
+#endif /* DRM_SCHEDULER_SPSC_QUEUE_H_ */