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, 114 insertions, 0 deletions

diff --git a/include/net/scm.h b/include/net/scm.h
new file mode 100644
index 000000000..86ae6bc10
--- /dev/null
+++ b/include/net/scm.h
@@ -0,0 +1,114 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __LINUX_NET_SCM_H
+#define __LINUX_NET_SCM_H
+
+#include <linux/limits.h>
+#include <linux/net.h>
+#include <linux/cred.h>
+#include <linux/file.h>
+#include <linux/security.h>
+#include <linux/pid.h>
+#include <linux/nsproxy.h>
+#include <linux/sched/signal.h>
+#include <net/compat.h>
+
+/* Well, we should have at least one descriptor open
+ * to accept passed FDs 8)
+ */
+#define SCM_MAX_FD	253
+
+struct scm_creds {
+	u32	pid;
+	kuid_t	uid;
+	kgid_t	gid;
+};
+
+#ifdef CONFIG_UNIX
+struct unix_edge;
+#endif
+
+struct scm_fp_list {
+	short			count;
+	short			count_unix;
+	short			max;
+#ifdef CONFIG_UNIX
+	bool			inflight;
+	bool			dead;
+	struct list_head	vertices;
+	struct unix_edge	*edges;
+#endif
+	struct user_struct	*user;
+	struct file		*fp[SCM_MAX_FD];
+};
+
+struct scm_cookie {
+	struct pid		*pid;		/* Skb credentials */
+	struct scm_fp_list	*fp;		/* Passed files		*/
+	struct scm_creds	creds;		/* Skb credentials	*/
+#ifdef CONFIG_SECURITY_NETWORK
+	u32			secid;		/* Passed security ID 	*/
+#endif
+};
+
+void scm_detach_fds(struct msghdr *msg, struct scm_cookie *scm, bool notrunc);
+void scm_detach_fds_compat(struct msghdr *msg, struct scm_cookie *scm, bool notrunc);
+int __scm_send(struct socket *sock, struct msghdr *msg, struct scm_cookie *scm);
+void __scm_destroy(struct scm_cookie *scm);
+struct scm_fp_list *scm_fp_dup(struct scm_fp_list *fpl);
+
+#ifdef CONFIG_SECURITY_NETWORK
+static __inline__ void unix_get_peersec_dgram(struct socket *sock, struct scm_cookie *scm)
+{
+	security_socket_getpeersec_dgram(sock, NULL, &scm->secid);
+}
+#else
+static __inline__ void unix_get_peersec_dgram(struct socket *sock, struct scm_cookie *scm)
+{ }
+#endif /* CONFIG_SECURITY_NETWORK */
+
+static __inline__ void scm_set_cred(struct scm_cookie *scm,
+				    struct pid *pid, kuid_t uid, kgid_t gid)
+{
+	scm->pid = get_pid(pid);
+	scm->creds.pid = pid_vnr(pid);
+	scm->creds.uid = uid;
+	scm->creds.gid = gid;
+}
+
+static __inline__ void scm_destroy_cred(struct scm_cookie *scm)
+{
+	put_pid(scm->pid);
+	scm->pid = NULL;
+}
+
+static __inline__ void scm_destroy(struct scm_cookie *scm)
+{
+	scm_destroy_cred(scm);
+	if (scm->fp)
+		__scm_destroy(scm);
+}
+
+static __inline__ int scm_send(struct socket *sock, struct msghdr *msg,
+			       struct scm_cookie *scm, bool forcecreds)
+{
+	memset(scm, 0, sizeof(*scm));
+	scm->creds.uid = INVALID_UID;
+	scm->creds.gid = INVALID_GID;
+	if (forcecreds)
+		scm_set_cred(scm, task_tgid(current), current_uid(), current_gid());
+	unix_get_peersec_dgram(sock, scm);
+	if (msg->msg_controllen <= 0)
+		return 0;
+	return __scm_send(sock, msg, scm);
+}
+
+void scm_recv(struct socket *sock, struct msghdr *msg,
+	      struct scm_cookie *scm, int flags);
+void scm_recv_unix(struct socket *sock, struct msghdr *msg,
+		   struct scm_cookie *scm, int flags);
+
+int scm_recv_one_fd(struct file *f, int __user *ufd, unsigned int flags,
+		    bool notrunc);
+
+#endif /* __LINUX_NET_SCM_H */
+