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, 127 insertions, 0 deletions

diff --git a/net/wireless/of.c b/net/wireless/of.c
new file mode 100644
index 000000000..99acbea3b
--- /dev/null
+++ b/net/wireless/of.c
@@ -0,0 +1,127 @@
+// SPDX-License-Identifier: ISC
+/*
+ * Copyright (C) 2017 Rafał Miłecki <rafal@milecki.pl>
+ */
+
+#include <linux/of.h>
+#include <net/cfg80211.h>
+#include "core.h"
+
+static bool wiphy_freq_limits_valid_chan(struct wiphy *wiphy,
+					 struct ieee80211_freq_range *freq_limits,
+					 unsigned int n_freq_limits,
+					 struct ieee80211_channel *chan)
+{
+	u32 bw = MHZ_TO_KHZ(20);
+	int i;
+
+	for (i = 0; i < n_freq_limits; i++) {
+		struct ieee80211_freq_range *limit = &freq_limits[i];
+
+		if (cfg80211_does_bw_fit_range(limit,
+					       MHZ_TO_KHZ(chan->center_freq),
+					       bw))
+			return true;
+	}
+
+	return false;
+}
+
+static void wiphy_freq_limits_apply(struct wiphy *wiphy,
+				    struct ieee80211_freq_range *freq_limits,
+				    unsigned int n_freq_limits)
+{
+	enum nl80211_band band;
+	int i;
+
+	if (WARN_ON(!n_freq_limits))
+		return;
+
+	for (band = 0; band < NUM_NL80211_BANDS; band++) {
+		struct ieee80211_supported_band *sband = wiphy->bands[band];
+
+		if (!sband)
+			continue;
+
+		for (i = 0; i < sband->n_channels; i++) {
+			struct ieee80211_channel *chan = &sband->channels[i];
+
+			if (chan->flags & IEEE80211_CHAN_DISABLED)
+				continue;
+
+			if (!wiphy_freq_limits_valid_chan(wiphy, freq_limits,
+							  n_freq_limits,
+							  chan)) {
+				pr_debug("Disabling freq %d MHz as it's out of OF limits\n",
+					 chan->center_freq);
+				chan->flags |= IEEE80211_CHAN_DISABLED;
+			}
+		}
+	}
+}
+
+void wiphy_read_of_freq_limits(struct wiphy *wiphy)
+{
+	struct device *dev = wiphy_dev(wiphy);
+	struct device_node *np;
+	struct property *prop;
+	struct ieee80211_freq_range *freq_limits;
+	unsigned int n_freq_limits;
+	const __be32 *p;
+	int len, i;
+	int err = 0;
+
+	if (!dev)
+		return;
+	np = dev_of_node(dev);
+	if (!np)
+		return;
+
+	prop = of_find_property(np, "ieee80211-freq-limit", &len);
+	if (!prop)
+		return;
+
+	if (!len || len % sizeof(u32) || len / sizeof(u32) % 2) {
+		dev_err(dev, "ieee80211-freq-limit wrong format");
+		return;
+	}
+	n_freq_limits = len / sizeof(u32) / 2;
+
+	freq_limits = kzalloc_objs(*freq_limits, n_freq_limits);
+	if (!freq_limits) {
+		err = -ENOMEM;
+		goto out_kfree;
+	}
+
+	p = NULL;
+	for (i = 0; i < n_freq_limits; i++) {
+		struct ieee80211_freq_range *limit = &freq_limits[i];
+
+		p = of_prop_next_u32(prop, p, &limit->start_freq_khz);
+		if (!p) {
+			err = -EINVAL;
+			goto out_kfree;
+		}
+
+		p = of_prop_next_u32(prop, p, &limit->end_freq_khz);
+		if (!p) {
+			err = -EINVAL;
+			goto out_kfree;
+		}
+
+		if (!limit->start_freq_khz ||
+		    !limit->end_freq_khz ||
+		    limit->start_freq_khz >= limit->end_freq_khz) {
+			err = -EINVAL;
+			goto out_kfree;
+		}
+	}
+
+	wiphy_freq_limits_apply(wiphy, freq_limits, n_freq_limits);
+
+out_kfree:
+	kfree(freq_limits);
+	if (err)
+		dev_err(dev, "Failed to get limits: %d\n", err);
+}
+EXPORT_SYMBOL(wiphy_read_of_freq_limits);