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, 66 insertions, 0 deletions

diff --git a/include/trace/events/siox.h b/include/trace/events/siox.h
new file mode 100644
index 000000000..68a43fc2c
--- /dev/null
+++ b/include/trace/events/siox.h
@@ -0,0 +1,66 @@
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM siox
+
+#if !defined(_TRACE_SIOX_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_SIOX_H
+
+#include <linux/tracepoint.h>
+
+TRACE_EVENT(siox_set_data,
+	    TP_PROTO(const struct siox_master *smaster,
+		     const struct siox_device *sdevice,
+		     unsigned int devno, size_t bufoffset),
+	    TP_ARGS(smaster, sdevice, devno, bufoffset),
+	    TP_STRUCT__entry(
+			     __field(int, busno)
+			     __field(unsigned int, devno)
+			     __field(size_t, inbytes)
+			     __dynamic_array(u8, buf, sdevice->inbytes)
+			    ),
+	    TP_fast_assign(
+			   __entry->busno = smaster->busno;
+			   __entry->devno = devno;
+			   __entry->inbytes = sdevice->inbytes;
+			   memcpy(__get_dynamic_array(buf),
+				  smaster->buf + bufoffset, sdevice->inbytes);
+			  ),
+	    TP_printk("siox-%d-%u [%*phD]",
+		      __entry->busno,
+		      __entry->devno,
+		      (int)__entry->inbytes, __get_dynamic_array(buf)
+		     )
+);
+
+TRACE_EVENT(siox_get_data,
+	    TP_PROTO(const struct siox_master *smaster,
+		     const struct siox_device *sdevice,
+		     unsigned int devno, u8 status_clean,
+		     size_t bufoffset),
+	    TP_ARGS(smaster, sdevice, devno, status_clean, bufoffset),
+	    TP_STRUCT__entry(
+			     __field(int, busno)
+			     __field(unsigned int, devno)
+			     __field(u8, status_clean)
+			     __field(size_t, outbytes)
+			     __dynamic_array(u8, buf, sdevice->outbytes)
+			    ),
+	    TP_fast_assign(
+			   __entry->busno = smaster->busno;
+			   __entry->devno = devno;
+			   __entry->status_clean = status_clean;
+			   __entry->outbytes = sdevice->outbytes;
+			   memcpy(__get_dynamic_array(buf),
+				  smaster->buf + bufoffset, sdevice->outbytes);
+			  ),
+	    TP_printk("siox-%d-%u (%02hhx) [%*phD]",
+		      __entry->busno,
+		      __entry->devno,
+		      __entry->status_clean,
+		      (int)__entry->outbytes, __get_dynamic_array(buf)
+		     )
+);
+
+#endif /* if !defined(_TRACE_SIOX_H) || defined(TRACE_HEADER_MULTI_READ) */
+
+/* This part must be outside protection */
+#include <trace/define_trace.h>