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, 90 insertions, 0 deletions

diff --git a/include/net/pfcp.h b/include/net/pfcp.h
new file mode 100644
index 000000000..639553797
--- /dev/null
+++ b/include/net/pfcp.h
@@ -0,0 +1,90 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _PFCP_H_
+#define _PFCP_H_
+
+#include <uapi/linux/if_ether.h>
+#include <net/dst_metadata.h>
+#include <linux/netdevice.h>
+#include <uapi/linux/ipv6.h>
+#include <net/udp_tunnel.h>
+#include <uapi/linux/udp.h>
+#include <uapi/linux/ip.h>
+#include <linux/string.h>
+#include <linux/types.h>
+#include <linux/bits.h>
+
+#define PFCP_PORT 8805
+
+/* PFCP protocol header */
+struct pfcphdr {
+	u8	flags;
+	u8	message_type;
+	__be16	message_length;
+};
+
+/* PFCP header flags */
+#define PFCP_SEID_FLAG		BIT(0)
+#define PFCP_MP_FLAG		BIT(1)
+
+#define PFCP_VERSION_MASK	GENMASK(4, 0)
+
+#define PFCP_HLEN (sizeof(struct udphdr) + sizeof(struct pfcphdr))
+
+/* PFCP node related messages */
+struct pfcphdr_node {
+	u8	seq_number[3];
+	u8	reserved;
+};
+
+/* PFCP session related messages */
+struct pfcphdr_session {
+	__be64	seid;
+	u8	seq_number[3];
+#ifdef __LITTLE_ENDIAN_BITFIELD
+	u8	message_priority:4,
+		reserved:4;
+#elif defined(__BIG_ENDIAN_BITFIELD)
+	u8	reserved:4,
+		message_priority:4;
+#else
+#error "Please fix <asm/byteorder>"
+#endif
+};
+
+struct pfcp_metadata {
+	u8 type;
+	__be64 seid;
+} __packed;
+
+enum {
+	PFCP_TYPE_NODE		= 0,
+	PFCP_TYPE_SESSION	= 1,
+};
+
+#define PFCP_HEADROOM (sizeof(struct iphdr) + sizeof(struct udphdr) + \
+		       sizeof(struct pfcphdr) + sizeof(struct ethhdr))
+#define PFCP6_HEADROOM (sizeof(struct ipv6hdr) + sizeof(struct udphdr) + \
+			sizeof(struct pfcphdr) + sizeof(struct ethhdr))
+
+static inline struct pfcphdr *pfcp_hdr(struct sk_buff *skb)
+{
+	return (struct pfcphdr *)(udp_hdr(skb) + 1);
+}
+
+static inline struct pfcphdr_node *pfcp_hdr_node(struct sk_buff *skb)
+{
+	return (struct pfcphdr_node *)(pfcp_hdr(skb) + 1);
+}
+
+static inline struct pfcphdr_session *pfcp_hdr_session(struct sk_buff *skb)
+{
+	return (struct pfcphdr_session *)(pfcp_hdr(skb) + 1);
+}
+
+static inline bool netif_is_pfcp(const struct net_device *dev)
+{
+	return dev->rtnl_link_ops &&
+	       !strcmp(dev->rtnl_link_ops->kind, "pfcp");
+}
+
+#endif