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, 75 insertions, 0 deletions

diff --git a/net/atm/raw.c b/net/atm/raw.c
new file mode 100644
index 000000000..87d136c75
--- /dev/null
+++ b/net/atm/raw.c
@@ -0,0 +1,75 @@
+// SPDX-License-Identifier: GPL-2.0
+/* net/atm/raw.c - Raw AAL0 and AAL5 transports */
+
+/* Written 1995-2000 by Werner Almesberger, EPFL LRC/ICA */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ":%s: " fmt, __func__
+
+#include <linux/module.h>
+#include <linux/atmdev.h>
+#include <linux/capability.h>
+#include <linux/kernel.h>
+#include <linux/skbuff.h>
+#include <linux/mm.h>
+#include <linux/slab.h>
+
+#include "common.h"
+#include "protocols.h"
+
+/*
+ * SKB == NULL indicates that the link is being closed
+ */
+
+static void atm_push_raw(struct atm_vcc *vcc, struct sk_buff *skb)
+{
+	if (skb) {
+		struct sock *sk = sk_atm(vcc);
+
+		skb_queue_tail(&sk->sk_receive_queue, skb);
+		sk->sk_data_ready(sk);
+	}
+}
+
+static void atm_pop_raw(struct atm_vcc *vcc, struct sk_buff *skb)
+{
+	struct sock *sk = sk_atm(vcc);
+
+	pr_debug("(%d) %d -= %d\n",
+		 vcc->vci, sk_wmem_alloc_get(sk), ATM_SKB(skb)->acct_truesize);
+	atm_return_tx(vcc, skb);
+	dev_kfree_skb_any(skb);
+	sk->sk_write_space(sk);
+}
+
+static int atm_send_aal0(struct atm_vcc *vcc, struct sk_buff *skb)
+{
+	/*
+	 * Note that if vpi/vci are _ANY or _UNSPEC the below will
+	 * still work
+	 */
+	if (!capable(CAP_NET_ADMIN) &&
+	    (((u32 *)skb->data)[0] & (ATM_HDR_VPI_MASK | ATM_HDR_VCI_MASK)) !=
+	    ((vcc->vpi << ATM_HDR_VPI_SHIFT) |
+	     (vcc->vci << ATM_HDR_VCI_SHIFT))) {
+		kfree_skb(skb);
+		return -EADDRNOTAVAIL;
+	}
+	return vcc->dev->ops->send(vcc, skb);
+}
+
+int atm_init_aal0(struct atm_vcc *vcc)
+{
+	vcc->push = atm_push_raw;
+	vcc->pop = atm_pop_raw;
+	vcc->send = atm_send_aal0;
+	return 0;
+}
+
+int atm_init_aal5(struct atm_vcc *vcc)
+{
+	vcc->push = atm_push_raw;
+	vcc->pop = atm_pop_raw;
+	vcc->send = vcc->dev->ops->send;
+	return 0;
+}
+EXPORT_SYMBOL(atm_init_aal5);