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, 134 insertions, 0 deletions

diff --git a/include/trace/events/module.h b/include/trace/events/module.h
new file mode 100644
index 000000000..e5a006be9
--- /dev/null
+++ b/include/trace/events/module.h
@@ -0,0 +1,134 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Because linux/module.h has tracepoints in the header, and ftrace.h
+ * used to include this file, define_trace.h includes linux/module.h
+ * But we do not want the module.h to override the TRACE_SYSTEM macro
+ * variable that define_trace.h is processing, so we only set it
+ * when module events are being processed, which would happen when
+ * CREATE_TRACE_POINTS is defined.
+ */
+#ifdef CREATE_TRACE_POINTS
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM module
+#endif
+
+#if !defined(_TRACE_MODULE_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_MODULE_H
+
+#include <linux/tracepoint.h>
+
+#ifdef CONFIG_MODULES
+
+struct module;
+
+#define show_module_flags(flags) __print_flags(flags, "",	\
+	{ (1UL << TAINT_PROPRIETARY_MODULE),	"P" },		\
+	{ (1UL << TAINT_OOT_MODULE),		"O" },		\
+	{ (1UL << TAINT_FORCED_MODULE),		"F" },		\
+	{ (1UL << TAINT_CRAP),			"C" },		\
+	{ (1UL << TAINT_UNSIGNED_MODULE),	"E" })
+
+TRACE_EVENT(module_load,
+
+	TP_PROTO(struct module *mod),
+
+	TP_ARGS(mod),
+
+	TP_STRUCT__entry(
+		__field(	unsigned int,	taints		)
+		__string(	name,		mod->name	)
+	),
+
+	TP_fast_assign(
+		__entry->taints = mod->taints;
+		__assign_str(name);
+	),
+
+	TP_printk("%s %s", __get_str(name), show_module_flags(__entry->taints))
+);
+
+TRACE_EVENT(module_free,
+
+	TP_PROTO(struct module *mod),
+
+	TP_ARGS(mod),
+
+	TP_STRUCT__entry(
+		__string(	name,		mod->name	)
+	),
+
+	TP_fast_assign(
+		__assign_str(name);
+	),
+
+	TP_printk("%s", __get_str(name))
+);
+
+#ifdef CONFIG_MODULE_UNLOAD
+/* trace_module_get/put are only used if CONFIG_MODULE_UNLOAD is defined */
+
+DECLARE_EVENT_CLASS(module_refcnt,
+
+	TP_PROTO(struct module *mod, unsigned long ip),
+
+	TP_ARGS(mod, ip),
+
+	TP_STRUCT__entry(
+		__field(	unsigned long,	ip		)
+		__field(	int,		refcnt		)
+		__string(	name,		mod->name	)
+	),
+
+	TP_fast_assign(
+		__entry->ip	= ip;
+		__entry->refcnt	= atomic_read(&mod->refcnt);
+		__assign_str(name);
+	),
+
+	TP_printk("%s call_site=%ps refcnt=%d",
+		  __get_str(name), (void *)__entry->ip, __entry->refcnt)
+);
+
+DEFINE_EVENT(module_refcnt, module_get,
+
+	TP_PROTO(struct module *mod, unsigned long ip),
+
+	TP_ARGS(mod, ip)
+);
+
+DEFINE_EVENT(module_refcnt, module_put,
+
+	TP_PROTO(struct module *mod, unsigned long ip),
+
+	TP_ARGS(mod, ip)
+);
+#endif /* CONFIG_MODULE_UNLOAD */
+
+TRACE_EVENT(module_request,
+
+	TP_PROTO(char *name, bool wait, unsigned long ip),
+
+	TP_ARGS(name, wait, ip),
+
+	TP_STRUCT__entry(
+		__field(	unsigned long,	ip		)
+		__field(	bool,		wait		)
+		__string(	name,		name		)
+	),
+
+	TP_fast_assign(
+		__entry->ip	= ip;
+		__entry->wait	= wait;
+		__assign_str(name);
+	),
+
+	TP_printk("%s wait=%d call_site=%ps",
+		  __get_str(name), (int)__entry->wait, (void *)__entry->ip)
+);
+
+#endif /* CONFIG_MODULES */
+
+#endif /* _TRACE_MODULE_H */
+
+/* This part must be outside protection */
+#include <trace/define_trace.h>