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, 99 insertions, 0 deletions

diff --git a/net/sched/em_cmp.c b/net/sched/em_cmp.c
new file mode 100644
index 000000000..48c1bce74
--- /dev/null
+++ b/net/sched/em_cmp.c
@@ -0,0 +1,99 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * net/sched/em_cmp.c	Simple packet data comparison ematch
+ *
+ * Authors:	Thomas Graf <tgraf@suug.ch>
+ */
+
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/skbuff.h>
+#include <linux/tc_ematch/tc_em_cmp.h>
+#include <linux/unaligned.h>
+#include <net/pkt_cls.h>
+
+static inline int cmp_needs_transformation(struct tcf_em_cmp *cmp)
+{
+	return unlikely(cmp->flags & TCF_EM_CMP_TRANS);
+}
+
+static int em_cmp_match(struct sk_buff *skb, struct tcf_ematch *em,
+			struct tcf_pkt_info *info)
+{
+	struct tcf_em_cmp *cmp = (struct tcf_em_cmp *) em->data;
+	unsigned char *ptr = tcf_get_base_ptr(skb, cmp->layer);
+	u32 val = 0;
+
+	if (!ptr)
+		return 0;
+	ptr += cmp->off;
+	if (!tcf_valid_offset(skb, ptr, cmp->align))
+		return 0;
+
+	switch (cmp->align) {
+	case TCF_EM_ALIGN_U8:
+		val = *ptr;
+		break;
+
+	case TCF_EM_ALIGN_U16:
+		val = get_unaligned_be16(ptr);
+
+		if (cmp_needs_transformation(cmp))
+			val = be16_to_cpu(val);
+		break;
+
+	case TCF_EM_ALIGN_U32:
+		/* Worth checking boundaries? The branching seems
+		 * to get worse. Visit again.
+		 */
+		val = get_unaligned_be32(ptr);
+
+		if (cmp_needs_transformation(cmp))
+			val = be32_to_cpu(val);
+		break;
+
+	default:
+		return 0;
+	}
+
+	if (cmp->mask)
+		val &= cmp->mask;
+
+	switch (cmp->opnd) {
+	case TCF_EM_OPND_EQ:
+		return val == cmp->val;
+	case TCF_EM_OPND_LT:
+		return val < cmp->val;
+	case TCF_EM_OPND_GT:
+		return val > cmp->val;
+	}
+
+	return 0;
+}
+
+static struct tcf_ematch_ops em_cmp_ops = {
+	.kind	  = TCF_EM_CMP,
+	.datalen  = sizeof(struct tcf_em_cmp),
+	.match	  = em_cmp_match,
+	.owner	  = THIS_MODULE,
+	.link	  = LIST_HEAD_INIT(em_cmp_ops.link)
+};
+
+static int __init init_em_cmp(void)
+{
+	return tcf_em_register(&em_cmp_ops);
+}
+
+static void __exit exit_em_cmp(void)
+{
+	tcf_em_unregister(&em_cmp_ops);
+}
+
+MODULE_DESCRIPTION("ematch classifier for basic data types(8/16/32 bit) against skb data");
+MODULE_LICENSE("GPL");
+
+module_init(init_em_cmp);
+module_exit(exit_em_cmp);
+
+MODULE_ALIAS_TCF_EMATCH(TCF_EM_CMP);