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

2 files changed, 103 insertions, 0 deletions

diff --git a/drivers/net/wireguard/generated/netlink.c b/drivers/net/wireguard/generated/netlink.c
new file mode 100644
index 000000000..3ef8c2990
--- /dev/null
+++ b/drivers/net/wireguard/generated/netlink.c
@@ -0,0 +1,73 @@
+// SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause)
+/* Do not edit directly, auto-generated from: */
+/*	Documentation/netlink/specs/wireguard.yaml */
+/* YNL-GEN kernel source */
+/* YNL-ARG --function-prefix wg */
+/* To regenerate run: tools/net/ynl/ynl-regen.sh */
+
+#include <net/netlink.h>
+#include <net/genetlink.h>
+
+#include "netlink.h"
+
+#include <uapi/linux/wireguard.h>
+#include <linux/time_types.h>
+
+/* Common nested types */
+const struct nla_policy wireguard_wgallowedip_nl_policy[WGALLOWEDIP_A_FLAGS + 1] = {
+	[WGALLOWEDIP_A_FAMILY] = { .type = NLA_U16, },
+	[WGALLOWEDIP_A_IPADDR] = NLA_POLICY_MIN_LEN(4),
+	[WGALLOWEDIP_A_CIDR_MASK] = { .type = NLA_U8, },
+	[WGALLOWEDIP_A_FLAGS] = NLA_POLICY_MASK(NLA_U32, 0x1),
+};
+
+const struct nla_policy wireguard_wgpeer_nl_policy[WGPEER_A_PROTOCOL_VERSION + 1] = {
+	[WGPEER_A_PUBLIC_KEY] = NLA_POLICY_EXACT_LEN(WG_KEY_LEN),
+	[WGPEER_A_PRESHARED_KEY] = NLA_POLICY_EXACT_LEN(WG_KEY_LEN),
+	[WGPEER_A_FLAGS] = NLA_POLICY_MASK(NLA_U32, 0x7),
+	[WGPEER_A_ENDPOINT] = NLA_POLICY_MIN_LEN(16),
+	[WGPEER_A_PERSISTENT_KEEPALIVE_INTERVAL] = { .type = NLA_U16, },
+	[WGPEER_A_LAST_HANDSHAKE_TIME] = NLA_POLICY_EXACT_LEN(16),
+	[WGPEER_A_RX_BYTES] = { .type = NLA_U64, },
+	[WGPEER_A_TX_BYTES] = { .type = NLA_U64, },
+	[WGPEER_A_ALLOWEDIPS] = NLA_POLICY_NESTED_ARRAY(wireguard_wgallowedip_nl_policy),
+	[WGPEER_A_PROTOCOL_VERSION] = { .type = NLA_U32, },
+};
+
+/* WG_CMD_GET_DEVICE - dump */
+static const struct nla_policy wireguard_get_device_nl_policy[WGDEVICE_A_IFNAME + 1] = {
+	[WGDEVICE_A_IFINDEX] = { .type = NLA_U32, },
+	[WGDEVICE_A_IFNAME] = { .type = NLA_NUL_STRING, .len = 15, },
+};
+
+/* WG_CMD_SET_DEVICE - do */
+static const struct nla_policy wireguard_set_device_nl_policy[WGDEVICE_A_PEERS + 1] = {
+	[WGDEVICE_A_IFINDEX] = { .type = NLA_U32, },
+	[WGDEVICE_A_IFNAME] = { .type = NLA_NUL_STRING, .len = 15, },
+	[WGDEVICE_A_PRIVATE_KEY] = NLA_POLICY_EXACT_LEN(WG_KEY_LEN),
+	[WGDEVICE_A_PUBLIC_KEY] = NLA_POLICY_EXACT_LEN(WG_KEY_LEN),
+	[WGDEVICE_A_FLAGS] = NLA_POLICY_MASK(NLA_U32, 0x1),
+	[WGDEVICE_A_LISTEN_PORT] = { .type = NLA_U16, },
+	[WGDEVICE_A_FWMARK] = { .type = NLA_U32, },
+	[WGDEVICE_A_PEERS] = NLA_POLICY_NESTED_ARRAY(wireguard_wgpeer_nl_policy),
+};
+
+/* Ops table for wireguard */
+const struct genl_split_ops wireguard_nl_ops[2] = {
+	{
+		.cmd		= WG_CMD_GET_DEVICE,
+		.start		= wg_get_device_start,
+		.dumpit		= wg_get_device_dumpit,
+		.done		= wg_get_device_done,
+		.policy		= wireguard_get_device_nl_policy,
+		.maxattr	= WGDEVICE_A_IFNAME,
+		.flags		= GENL_UNS_ADMIN_PERM | GENL_CMD_CAP_DUMP,
+	},
+	{
+		.cmd		= WG_CMD_SET_DEVICE,
+		.doit		= wg_set_device_doit,
+		.policy		= wireguard_set_device_nl_policy,
+		.maxattr	= WGDEVICE_A_PEERS,
+		.flags		= GENL_UNS_ADMIN_PERM | GENL_CMD_CAP_DO,
+	},
+};
diff --git a/drivers/net/wireguard/generated/netlink.h b/drivers/net/wireguard/generated/netlink.h
new file mode 100644
index 000000000..5dc977ee9
--- /dev/null
+++ b/drivers/net/wireguard/generated/netlink.h
@@ -0,0 +1,30 @@
+/* SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) */
+/* Do not edit directly, auto-generated from: */
+/*	Documentation/netlink/specs/wireguard.yaml */
+/* YNL-GEN kernel header */
+/* YNL-ARG --function-prefix wg */
+/* To regenerate run: tools/net/ynl/ynl-regen.sh */
+
+#ifndef _LINUX_WIREGUARD_GEN_H
+#define _LINUX_WIREGUARD_GEN_H
+
+#include <net/netlink.h>
+#include <net/genetlink.h>
+
+#include <uapi/linux/wireguard.h>
+#include <linux/time_types.h>
+
+/* Common nested types */
+extern const struct nla_policy wireguard_wgallowedip_nl_policy[WGALLOWEDIP_A_FLAGS + 1];
+extern const struct nla_policy wireguard_wgpeer_nl_policy[WGPEER_A_PROTOCOL_VERSION + 1];
+
+/* Ops table for wireguard */
+extern const struct genl_split_ops wireguard_nl_ops[2];
+
+int wg_get_device_start(struct netlink_callback *cb);
+int wg_get_device_done(struct netlink_callback *cb);
+
+int wg_get_device_dumpit(struct sk_buff *skb, struct netlink_callback *cb);
+int wg_set_device_doit(struct sk_buff *skb, struct genl_info *info);
+
+#endif /* _LINUX_WIREGUARD_GEN_H */