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, 73 insertions, 0 deletions

diff --git a/include/net/dst_ops.h b/include/net/dst_ops.h
new file mode 100644
index 000000000..3a9001a04
--- /dev/null
+++ b/include/net/dst_ops.h
@@ -0,0 +1,73 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _NET_DST_OPS_H
+#define _NET_DST_OPS_H
+#include <linux/types.h>
+#include <linux/percpu_counter.h>
+#include <linux/cache.h>
+
+struct dst_entry;
+struct kmem_cachep;
+struct net_device;
+struct sk_buff;
+struct sock;
+struct net;
+
+struct dst_ops {
+	unsigned short		family;
+	unsigned int		gc_thresh;
+
+	void			(*gc)(struct dst_ops *ops);
+	struct dst_entry *	(*check)(struct dst_entry *, __u32 cookie);
+	unsigned int		(*default_advmss)(const struct dst_entry *);
+	unsigned int		(*mtu)(const struct dst_entry *);
+	u32 *			(*cow_metrics)(struct dst_entry *, unsigned long);
+	void			(*destroy)(struct dst_entry *);
+	void			(*ifdown)(struct dst_entry *,
+					  struct net_device *dev);
+	void			(*negative_advice)(struct sock *sk, struct dst_entry *);
+	void			(*link_failure)(struct sk_buff *);
+	void			(*update_pmtu)(struct dst_entry *dst, struct sock *sk,
+					       struct sk_buff *skb, u32 mtu,
+					       bool confirm_neigh);
+	void			(*redirect)(struct dst_entry *dst, struct sock *sk,
+					    struct sk_buff *skb);
+	int			(*local_out)(struct net *net, struct sock *sk, struct sk_buff *skb);
+	struct neighbour *	(*neigh_lookup)(const struct dst_entry *dst,
+						struct sk_buff *skb,
+						const void *daddr);
+	void			(*confirm_neigh)(const struct dst_entry *dst,
+						 const void *daddr);
+
+	struct kmem_cache	*kmem_cachep;
+
+	struct percpu_counter	pcpuc_entries ____cacheline_aligned_in_smp;
+};
+
+static inline int dst_entries_get_fast(struct dst_ops *dst)
+{
+	return percpu_counter_read_positive(&dst->pcpuc_entries);
+}
+
+static inline int dst_entries_get_slow(struct dst_ops *dst)
+{
+	return percpu_counter_sum_positive(&dst->pcpuc_entries);
+}
+
+#define DST_PERCPU_COUNTER_BATCH 32
+static inline void dst_entries_add(struct dst_ops *dst, int val)
+{
+	percpu_counter_add_batch(&dst->pcpuc_entries, val,
+				 DST_PERCPU_COUNTER_BATCH);
+}
+
+static inline int dst_entries_init(struct dst_ops *dst)
+{
+	return percpu_counter_init(&dst->pcpuc_entries, 0, GFP_KERNEL);
+}
+
+static inline void dst_entries_destroy(struct dst_ops *dst)
+{
+	percpu_counter_destroy(&dst->pcpuc_entries);
+}
+
+#endif