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, 119 insertions, 0 deletions

diff --git a/include/net/protocol.h b/include/net/protocol.h
new file mode 100644
index 000000000..b2499f88f
--- /dev/null
+++ b/include/net/protocol.h
@@ -0,0 +1,119 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * INET		An implementation of the TCP/IP protocol suite for the LINUX
+ *		operating system.  INET is implemented using the  BSD Socket
+ *		interface as the means of communication with the user level.
+ *
+ *		Definitions for the protocol dispatcher.
+ *
+ * Version:	@(#)protocol.h	1.0.2	05/07/93
+ *
+ * Author:	Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
+ *
+ *	Changes:
+ *		Alan Cox	:	Added a name field and a frag handler
+ *					field for later.
+ *		Alan Cox	:	Cleaned up, and sorted types.
+ *		Pedro Roque	:	inet6 protocols
+ */
+ 
+#ifndef _PROTOCOL_H
+#define _PROTOCOL_H
+
+#include <linux/in6.h>
+#include <linux/skbuff.h>
+#if IS_ENABLED(CONFIG_IPV6)
+#include <linux/ipv6.h>
+#endif
+#include <linux/netdevice.h>
+
+/* This is one larger than the largest protocol value that can be
+ * found in an ipv4 or ipv6 header.  Since in both cases the protocol
+ * value is presented in a __u8, this is defined to be 256.
+ */
+#define MAX_INET_PROTOS		256
+
+/* This is used to register protocols. */
+struct net_protocol {
+	int			(*handler)(struct sk_buff *skb);
+
+	/* This returns an error if we weren't able to handle the error. */
+	int			(*err_handler)(struct sk_buff *skb, u32 info);
+
+	unsigned int		no_policy:1,
+				/* does the protocol do more stringent
+				 * icmp tag validation than simple
+				 * socket lookup?
+				 */
+				icmp_strict_tag_validation:1;
+	u32			secret;
+};
+
+#if IS_ENABLED(CONFIG_IPV6)
+struct inet6_protocol {
+	int	(*handler)(struct sk_buff *skb);
+
+	/* This returns an error if we weren't able to handle the error. */
+	int	(*err_handler)(struct sk_buff *skb,
+			       struct inet6_skb_parm *opt,
+			       u8 type, u8 code, int offset,
+			       __be32 info);
+
+	unsigned int	flags;	/* INET6_PROTO_xxx */
+	u32		secret;
+};
+
+#define INET6_PROTO_NOPOLICY	0x1
+#define INET6_PROTO_FINAL	0x2
+#endif
+
+struct net_offload {
+	struct offload_callbacks callbacks;
+	unsigned int		 flags;	/* Flags used by IPv6 for now */
+	u32			 secret;
+};
+/* This should be set for any extension header which is compatible with GSO. */
+#define INET6_PROTO_GSO_EXTHDR	0x1
+
+/* This is used to register socket interfaces for IP protocols.  */
+struct inet_protosw {
+	struct list_head list;
+
+        /* These two fields form the lookup key.  */
+	unsigned short	 type;	   /* This is the 2nd argument to socket(2). */
+	unsigned short	 protocol; /* This is the L4 protocol number.  */
+
+	struct proto	 *prot;
+	const struct proto_ops *ops;
+  
+	unsigned char	 flags;      /* See INET_PROTOSW_* below.  */
+};
+#define INET_PROTOSW_REUSE 0x01	     /* Are ports automatically reusable? */
+#define INET_PROTOSW_PERMANENT 0x02  /* Permanent protocols are unremovable. */
+#define INET_PROTOSW_ICSK      0x04  /* Is this an inet_connection_sock? */
+
+extern struct net_protocol __rcu *inet_protos[MAX_INET_PROTOS];
+extern const struct net_offload __rcu *inet_offloads[MAX_INET_PROTOS];
+extern const struct net_offload __rcu *inet6_offloads[MAX_INET_PROTOS];
+
+#if IS_ENABLED(CONFIG_IPV6)
+extern struct inet6_protocol __rcu *inet6_protos[MAX_INET_PROTOS];
+#endif
+
+int inet_add_protocol(const struct net_protocol *prot, unsigned char num);
+int inet_del_protocol(const struct net_protocol *prot, unsigned char num);
+int inet_add_offload(const struct net_offload *prot, unsigned char num);
+int inet_del_offload(const struct net_offload *prot, unsigned char num);
+void inet_register_protosw(struct inet_protosw *p);
+void inet_unregister_protosw(struct inet_protosw *p);
+
+#if IS_ENABLED(CONFIG_IPV6)
+int inet6_add_protocol(const struct inet6_protocol *prot, unsigned char num);
+int inet6_del_protocol(const struct inet6_protocol *prot, unsigned char num);
+int inet6_register_protosw(struct inet_protosw *p);
+void inet6_unregister_protosw(struct inet_protosw *p);
+#endif
+int inet6_add_offload(const struct net_offload *prot, unsigned char num);
+int inet6_del_offload(const struct net_offload *prot, unsigned char num);
+
+#endif	/* _PROTOCOL_H */