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, 329 insertions, 0 deletions

diff --git a/net/ipv6/xfrm6_protocol.c b/net/ipv6/xfrm6_protocol.c
new file mode 100644
index 000000000..9b586fcec
--- /dev/null
+++ b/net/ipv6/xfrm6_protocol.c
@@ -0,0 +1,329 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* xfrm6_protocol.c - Generic xfrm protocol multiplexer for ipv6.
+ *
+ * Copyright (C) 2013 secunet Security Networks AG
+ *
+ * Author:
+ * Steffen Klassert <steffen.klassert@secunet.com>
+ *
+ * Based on:
+ * net/ipv4/xfrm4_protocol.c
+ */
+
+#include <linux/init.h>
+#include <linux/mutex.h>
+#include <linux/skbuff.h>
+#include <linux/icmpv6.h>
+#include <net/ip6_route.h>
+#include <net/ipv6.h>
+#include <net/protocol.h>
+#include <net/xfrm.h>
+
+static struct xfrm6_protocol __rcu *esp6_handlers __read_mostly;
+static struct xfrm6_protocol __rcu *ah6_handlers __read_mostly;
+static struct xfrm6_protocol __rcu *ipcomp6_handlers __read_mostly;
+static DEFINE_MUTEX(xfrm6_protocol_mutex);
+
+static inline struct xfrm6_protocol __rcu **proto_handlers(u8 protocol)
+{
+	switch (protocol) {
+	case IPPROTO_ESP:
+		return &esp6_handlers;
+	case IPPROTO_AH:
+		return &ah6_handlers;
+	case IPPROTO_COMP:
+		return &ipcomp6_handlers;
+	}
+
+	return NULL;
+}
+
+#define for_each_protocol_rcu(head, handler)		\
+	for (handler = rcu_dereference(head);		\
+	     handler != NULL;				\
+	     handler = rcu_dereference(handler->next))	\
+
+static int xfrm6_rcv_cb(struct sk_buff *skb, u8 protocol, int err)
+{
+	int ret;
+	struct xfrm6_protocol *handler;
+	struct xfrm6_protocol __rcu **head = proto_handlers(protocol);
+
+	if (!head)
+		return 0;
+
+	for_each_protocol_rcu(*proto_handlers(protocol), handler)
+		if ((ret = handler->cb_handler(skb, err)) <= 0)
+			return ret;
+
+	return 0;
+}
+
+int xfrm6_rcv_encap(struct sk_buff *skb, int nexthdr, __be32 spi,
+		    int encap_type)
+{
+	int ret;
+	struct xfrm6_protocol *handler;
+	struct xfrm6_protocol __rcu **head = proto_handlers(nexthdr);
+
+	XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6 = NULL;
+	XFRM_SPI_SKB_CB(skb)->family = AF_INET6;
+	XFRM_SPI_SKB_CB(skb)->daddroff = offsetof(struct ipv6hdr, daddr);
+
+	if (!head)
+		goto out;
+
+	if (!skb_dst(skb)) {
+		const struct ipv6hdr *ip6h = ipv6_hdr(skb);
+		int flags = RT6_LOOKUP_F_HAS_SADDR;
+		struct dst_entry *dst;
+		struct flowi6 fl6 = {
+			.flowi6_iif   = skb->dev->ifindex,
+			.daddr        = ip6h->daddr,
+			.saddr        = ip6h->saddr,
+			.flowlabel    = ip6_flowinfo(ip6h),
+			.flowi6_mark  = skb->mark,
+			.flowi6_proto = ip6h->nexthdr,
+		};
+
+		dst = ip6_route_input_lookup(dev_net(skb->dev), skb->dev, &fl6,
+					     skb, flags);
+		if (dst->error) {
+			dst_release(dst);
+			goto drop;
+		}
+		skb_dst_set(skb, dst);
+	}
+
+	for_each_protocol_rcu(*head, handler)
+		if ((ret = handler->input_handler(skb, nexthdr, spi, encap_type)) != -EINVAL)
+			return ret;
+
+out:
+	icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0);
+
+drop:
+	kfree_skb(skb);
+	return 0;
+}
+EXPORT_SYMBOL(xfrm6_rcv_encap);
+
+static int xfrm6_esp_rcv(struct sk_buff *skb)
+{
+	int ret;
+	struct xfrm6_protocol *handler;
+
+	XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6 = NULL;
+
+	for_each_protocol_rcu(esp6_handlers, handler)
+		if ((ret = handler->handler(skb)) != -EINVAL)
+			return ret;
+
+	icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0);
+
+	kfree_skb(skb);
+	return 0;
+}
+
+static int xfrm6_esp_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
+			  u8 type, u8 code, int offset, __be32 info)
+{
+	struct xfrm6_protocol *handler;
+
+	for_each_protocol_rcu(esp6_handlers, handler)
+		if (!handler->err_handler(skb, opt, type, code, offset, info))
+			return 0;
+
+	return -ENOENT;
+}
+
+static int xfrm6_ah_rcv(struct sk_buff *skb)
+{
+	int ret;
+	struct xfrm6_protocol *handler;
+
+	XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6 = NULL;
+
+	for_each_protocol_rcu(ah6_handlers, handler)
+		if ((ret = handler->handler(skb)) != -EINVAL)
+			return ret;
+
+	icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0);
+
+	kfree_skb(skb);
+	return 0;
+}
+
+static int xfrm6_ah_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
+			 u8 type, u8 code, int offset, __be32 info)
+{
+	struct xfrm6_protocol *handler;
+
+	for_each_protocol_rcu(ah6_handlers, handler)
+		if (!handler->err_handler(skb, opt, type, code, offset, info))
+			return 0;
+
+	return -ENOENT;
+}
+
+static int xfrm6_ipcomp_rcv(struct sk_buff *skb)
+{
+	int ret;
+	struct xfrm6_protocol *handler;
+
+	XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6 = NULL;
+
+	for_each_protocol_rcu(ipcomp6_handlers, handler)
+		if ((ret = handler->handler(skb)) != -EINVAL)
+			return ret;
+
+	icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0);
+
+	kfree_skb(skb);
+	return 0;
+}
+
+static int xfrm6_ipcomp_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
+			     u8 type, u8 code, int offset, __be32 info)
+{
+	struct xfrm6_protocol *handler;
+
+	for_each_protocol_rcu(ipcomp6_handlers, handler)
+		if (!handler->err_handler(skb, opt, type, code, offset, info))
+			return 0;
+
+	return -ENOENT;
+}
+
+static const struct inet6_protocol esp6_protocol = {
+	.handler	=	xfrm6_esp_rcv,
+	.err_handler	=	xfrm6_esp_err,
+	.flags		=	INET6_PROTO_NOPOLICY,
+};
+
+static const struct inet6_protocol ah6_protocol = {
+	.handler	=	xfrm6_ah_rcv,
+	.err_handler	=	xfrm6_ah_err,
+	.flags		=	INET6_PROTO_NOPOLICY,
+};
+
+static const struct inet6_protocol ipcomp6_protocol = {
+	.handler	=	xfrm6_ipcomp_rcv,
+	.err_handler	=	xfrm6_ipcomp_err,
+	.flags		=	INET6_PROTO_NOPOLICY,
+};
+
+static const struct xfrm_input_afinfo xfrm6_input_afinfo = {
+	.family		=	AF_INET6,
+	.callback	=	xfrm6_rcv_cb,
+};
+
+static inline const struct inet6_protocol *netproto(unsigned char protocol)
+{
+	switch (protocol) {
+	case IPPROTO_ESP:
+		return &esp6_protocol;
+	case IPPROTO_AH:
+		return &ah6_protocol;
+	case IPPROTO_COMP:
+		return &ipcomp6_protocol;
+	}
+
+	return NULL;
+}
+
+int xfrm6_protocol_register(struct xfrm6_protocol *handler,
+			    unsigned char protocol)
+{
+	struct xfrm6_protocol __rcu **pprev;
+	struct xfrm6_protocol *t;
+	bool add_netproto = false;
+	int ret = -EEXIST;
+	int priority = handler->priority;
+
+	if (!proto_handlers(protocol) || !netproto(protocol))
+		return -EINVAL;
+
+	mutex_lock(&xfrm6_protocol_mutex);
+
+	if (!rcu_dereference_protected(*proto_handlers(protocol),
+				       lockdep_is_held(&xfrm6_protocol_mutex)))
+		add_netproto = true;
+
+	for (pprev = proto_handlers(protocol);
+	     (t = rcu_dereference_protected(*pprev,
+			lockdep_is_held(&xfrm6_protocol_mutex))) != NULL;
+	     pprev = &t->next) {
+		if (t->priority < priority)
+			break;
+		if (t->priority == priority)
+			goto err;
+	}
+
+	handler->next = *pprev;
+	rcu_assign_pointer(*pprev, handler);
+
+	ret = 0;
+
+err:
+	mutex_unlock(&xfrm6_protocol_mutex);
+
+	if (add_netproto) {
+		if (inet6_add_protocol(netproto(protocol), protocol)) {
+			pr_err("%s: can't add protocol\n", __func__);
+			ret = -EAGAIN;
+		}
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(xfrm6_protocol_register);
+
+int xfrm6_protocol_deregister(struct xfrm6_protocol *handler,
+			      unsigned char protocol)
+{
+	struct xfrm6_protocol __rcu **pprev;
+	struct xfrm6_protocol *t;
+	int ret = -ENOENT;
+
+	if (!proto_handlers(protocol) || !netproto(protocol))
+		return -EINVAL;
+
+	mutex_lock(&xfrm6_protocol_mutex);
+
+	for (pprev = proto_handlers(protocol);
+	     (t = rcu_dereference_protected(*pprev,
+			lockdep_is_held(&xfrm6_protocol_mutex))) != NULL;
+	     pprev = &t->next) {
+		if (t == handler) {
+			*pprev = handler->next;
+			ret = 0;
+			break;
+		}
+	}
+
+	if (!rcu_dereference_protected(*proto_handlers(protocol),
+				       lockdep_is_held(&xfrm6_protocol_mutex))) {
+		if (inet6_del_protocol(netproto(protocol), protocol) < 0) {
+			pr_err("%s: can't remove protocol\n", __func__);
+			ret = -EAGAIN;
+		}
+	}
+
+	mutex_unlock(&xfrm6_protocol_mutex);
+
+	synchronize_net();
+
+	return ret;
+}
+EXPORT_SYMBOL(xfrm6_protocol_deregister);
+
+int __init xfrm6_protocol_init(void)
+{
+	return xfrm_input_register_afinfo(&xfrm6_input_afinfo);
+}
+
+void xfrm6_protocol_fini(void)
+{
+	xfrm_input_unregister_afinfo(&xfrm6_input_afinfo);
+}