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, 56 insertions, 0 deletions

diff --git a/net/wireless/tests/util.c b/net/wireless/tests/util.c
new file mode 100644
index 000000000..118f7fb42
--- /dev/null
+++ b/net/wireless/tests/util.c
@@ -0,0 +1,56 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * KUnit fixture to have a (configurable) wiphy
+ *
+ * Copyright (C) 2023 Intel Corporation
+ */
+#include <linux/ieee80211.h>
+#include <net/cfg80211.h>
+#include <kunit/test.h>
+#include <kunit/test-bug.h>
+#include "util.h"
+
+int t_wiphy_init(struct kunit_resource *resource, void *ctx)
+{
+	struct kunit *test = kunit_get_current_test();
+	struct cfg80211_ops *ops;
+	struct wiphy *wiphy;
+	struct t_wiphy_priv *priv;
+
+	ops = kzalloc_obj(*ops);
+	KUNIT_ASSERT_NOT_NULL(test, ops);
+
+	wiphy = wiphy_new_nm(ops, sizeof(*priv), "kunit");
+	KUNIT_ASSERT_NOT_NULL(test, wiphy);
+
+	priv = wiphy_priv(wiphy);
+	priv->ctx = ctx;
+	priv->ops = ops;
+
+	/* Initialize channels, feel free to add more here channels/bands */
+	memcpy(priv->channels_2ghz, channels_2ghz, sizeof(channels_2ghz));
+	wiphy->bands[NL80211_BAND_2GHZ] = &priv->band_2ghz;
+	priv->band_2ghz.channels = priv->channels_2ghz;
+	priv->band_2ghz.n_channels = ARRAY_SIZE(channels_2ghz);
+
+	resource->data = wiphy;
+	resource->name = "wiphy";
+
+	return 0;
+}
+
+void t_wiphy_exit(struct kunit_resource *resource)
+{
+	struct t_wiphy_priv *priv;
+	struct cfg80211_ops *ops;
+
+	priv = wiphy_priv(resource->data);
+	ops = priv->ops;
+
+	/* Should we ensure anything about the state here?
+	 * e.g. full destruction or no calls to any ops on destruction?
+	 */
+
+	wiphy_free(resource->data);
+	kfree(ops);
+}