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, 49 insertions, 0 deletions

diff --git a/include/uapi/asm-generic/msgbuf.h b/include/uapi/asm-generic/msgbuf.h
new file mode 100644
index 000000000..6504d7b74
--- /dev/null
+++ b/include/uapi/asm-generic/msgbuf.h
@@ -0,0 +1,49 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+#ifndef __ASM_GENERIC_MSGBUF_H
+#define __ASM_GENERIC_MSGBUF_H
+
+#include <asm/bitsperlong.h>
+#include <asm/ipcbuf.h>
+
+/*
+ * generic msqid64_ds structure.
+ *
+ * Note extra padding because this structure is passed back and forth
+ * between kernel and user space.
+ *
+ * msqid64_ds was originally meant to be architecture specific, but
+ * everyone just ended up making identical copies without specific
+ * optimizations, so we may just as well all use the same one.
+ *
+ * 64 bit architectures use a 64-bit long time field here, while
+ * 32 bit architectures have a pair of unsigned long values.
+ * On big-endian systems, the lower half is in the wrong place.
+ *
+ * Pad space is left for:
+ * - 2 miscellaneous 32-bit values
+ */
+
+struct msqid64_ds {
+	struct ipc64_perm msg_perm;
+#if __BITS_PER_LONG == 64
+	long		 msg_stime;	/* last msgsnd time */
+	long		 msg_rtime;	/* last msgrcv time */
+	long		 msg_ctime;	/* last change time */
+#else
+	unsigned long	msg_stime;	/* last msgsnd time */
+	unsigned long	msg_stime_high;
+	unsigned long	msg_rtime;	/* last msgrcv time */
+	unsigned long	msg_rtime_high;
+	unsigned long	msg_ctime;	/* last change time */
+	unsigned long	msg_ctime_high;
+#endif
+	unsigned long	msg_cbytes;	/* current number of bytes on queue */
+	unsigned long	msg_qnum;	/* number of messages in queue */
+	unsigned long	 msg_qbytes;	/* max number of bytes on queue */
+	__kernel_pid_t msg_lspid;	/* pid of last msgsnd */
+	__kernel_pid_t msg_lrpid;	/* last receive pid */
+	unsigned long	 __unused4;
+	unsigned long	 __unused5;
+};
+
+#endif /* __ASM_GENERIC_MSGBUF_H */