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, 356 insertions, 0 deletions

diff --git a/include/trace/events/compaction.h b/include/trace/events/compaction.h
new file mode 100644
index 000000000..d05759d18
--- /dev/null
+++ b/include/trace/events/compaction.h
@@ -0,0 +1,356 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM compaction
+
+#if !defined(_TRACE_COMPACTION_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_COMPACTION_H
+
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/tracepoint.h>
+#include <trace/events/mmflags.h>
+
+
+DECLARE_EVENT_CLASS(mm_compaction_isolate_template,
+
+	TP_PROTO(
+		unsigned long start_pfn,
+		unsigned long end_pfn,
+		unsigned long nr_scanned,
+		unsigned long nr_taken),
+
+	TP_ARGS(start_pfn, end_pfn, nr_scanned, nr_taken),
+
+	TP_STRUCT__entry(
+		__field(unsigned long, start_pfn)
+		__field(unsigned long, end_pfn)
+		__field(unsigned long, nr_scanned)
+		__field(unsigned long, nr_taken)
+	),
+
+	TP_fast_assign(
+		__entry->start_pfn = start_pfn;
+		__entry->end_pfn = end_pfn;
+		__entry->nr_scanned = nr_scanned;
+		__entry->nr_taken = nr_taken;
+	),
+
+	TP_printk("range=(0x%lx ~ 0x%lx) nr_scanned=%lu nr_taken=%lu",
+		__entry->start_pfn,
+		__entry->end_pfn,
+		__entry->nr_scanned,
+		__entry->nr_taken)
+);
+
+DEFINE_EVENT(mm_compaction_isolate_template, mm_compaction_isolate_migratepages,
+
+	TP_PROTO(
+		unsigned long start_pfn,
+		unsigned long end_pfn,
+		unsigned long nr_scanned,
+		unsigned long nr_taken),
+
+	TP_ARGS(start_pfn, end_pfn, nr_scanned, nr_taken)
+);
+
+DEFINE_EVENT(mm_compaction_isolate_template, mm_compaction_isolate_freepages,
+
+	TP_PROTO(
+		unsigned long start_pfn,
+		unsigned long end_pfn,
+		unsigned long nr_scanned,
+		unsigned long nr_taken),
+
+	TP_ARGS(start_pfn, end_pfn, nr_scanned, nr_taken)
+);
+
+DEFINE_EVENT(mm_compaction_isolate_template, mm_compaction_fast_isolate_freepages,
+
+	TP_PROTO(
+		unsigned long start_pfn,
+		unsigned long end_pfn,
+		unsigned long nr_scanned,
+		unsigned long nr_taken),
+
+	TP_ARGS(start_pfn, end_pfn, nr_scanned, nr_taken)
+);
+
+#ifdef CONFIG_COMPACTION
+TRACE_EVENT(mm_compaction_migratepages,
+
+	TP_PROTO(unsigned int nr_migratepages,
+		unsigned int nr_succeeded),
+
+	TP_ARGS(nr_migratepages, nr_succeeded),
+
+	TP_STRUCT__entry(
+		__field(unsigned long, nr_migrated)
+		__field(unsigned long, nr_failed)
+	),
+
+	TP_fast_assign(
+		__entry->nr_migrated = nr_succeeded;
+		__entry->nr_failed = nr_migratepages - nr_succeeded;
+	),
+
+	TP_printk("nr_migrated=%lu nr_failed=%lu",
+		__entry->nr_migrated,
+		__entry->nr_failed)
+);
+
+TRACE_EVENT(mm_compaction_begin,
+	TP_PROTO(struct compact_control *cc, unsigned long zone_start,
+		unsigned long zone_end, bool sync),
+
+	TP_ARGS(cc, zone_start, zone_end, sync),
+
+	TP_STRUCT__entry(
+		__field(unsigned long, zone_start)
+		__field(unsigned long, migrate_pfn)
+		__field(unsigned long, free_pfn)
+		__field(unsigned long, zone_end)
+		__field(bool, sync)
+	),
+
+	TP_fast_assign(
+		__entry->zone_start = zone_start;
+		__entry->migrate_pfn = cc->migrate_pfn;
+		__entry->free_pfn = cc->free_pfn;
+		__entry->zone_end = zone_end;
+		__entry->sync = sync;
+	),
+
+	TP_printk("zone_start=0x%lx migrate_pfn=0x%lx free_pfn=0x%lx zone_end=0x%lx, mode=%s",
+		__entry->zone_start,
+		__entry->migrate_pfn,
+		__entry->free_pfn,
+		__entry->zone_end,
+		__entry->sync ? "sync" : "async")
+);
+
+TRACE_EVENT(mm_compaction_end,
+	TP_PROTO(struct compact_control *cc, unsigned long zone_start,
+		unsigned long zone_end, bool sync,
+		int status),
+
+	TP_ARGS(cc, zone_start, zone_end, sync, status),
+
+	TP_STRUCT__entry(
+		__field(unsigned long, zone_start)
+		__field(unsigned long, migrate_pfn)
+		__field(unsigned long, free_pfn)
+		__field(unsigned long, zone_end)
+		__field(bool, sync)
+		__field(int, status)
+	),
+
+	TP_fast_assign(
+		__entry->zone_start = zone_start;
+		__entry->migrate_pfn = cc->migrate_pfn;
+		__entry->free_pfn = cc->free_pfn;
+		__entry->zone_end = zone_end;
+		__entry->sync = sync;
+		__entry->status = status;
+	),
+
+	TP_printk("zone_start=0x%lx migrate_pfn=0x%lx free_pfn=0x%lx zone_end=0x%lx, mode=%s status=%s",
+		__entry->zone_start,
+		__entry->migrate_pfn,
+		__entry->free_pfn,
+		__entry->zone_end,
+		__entry->sync ? "sync" : "async",
+		__print_symbolic(__entry->status, COMPACTION_STATUS))
+);
+
+TRACE_EVENT(mm_compaction_try_to_compact_pages,
+
+	TP_PROTO(
+		int order,
+		gfp_t gfp_mask,
+		int prio),
+
+	TP_ARGS(order, gfp_mask, prio),
+
+	TP_STRUCT__entry(
+		__field(int, order)
+		__field(unsigned long, gfp_mask)
+		__field(int, prio)
+	),
+
+	TP_fast_assign(
+		__entry->order = order;
+		__entry->gfp_mask = (__force unsigned long)gfp_mask;
+		__entry->prio = prio;
+	),
+
+	TP_printk("order=%d gfp_mask=%s priority=%d",
+		__entry->order,
+		show_gfp_flags(__entry->gfp_mask),
+		__entry->prio)
+);
+
+DECLARE_EVENT_CLASS(mm_compaction_suitable_template,
+
+	TP_PROTO(struct zone *zone,
+		int order,
+		int ret),
+
+	TP_ARGS(zone, order, ret),
+
+	TP_STRUCT__entry(
+		__field(int, nid)
+		__field(enum zone_type, idx)
+		__field(int, order)
+		__field(int, ret)
+	),
+
+	TP_fast_assign(
+		__entry->nid = zone_to_nid(zone);
+		__entry->idx = zone_idx(zone);
+		__entry->order = order;
+		__entry->ret = ret;
+	),
+
+	TP_printk("node=%d zone=%-8s order=%d ret=%s",
+		__entry->nid,
+		__print_symbolic(__entry->idx, ZONE_TYPE),
+		__entry->order,
+		__print_symbolic(__entry->ret, COMPACTION_STATUS))
+);
+
+DEFINE_EVENT(mm_compaction_suitable_template, mm_compaction_finished,
+
+	TP_PROTO(struct zone *zone,
+		int order,
+		int ret),
+
+	TP_ARGS(zone, order, ret)
+);
+
+DEFINE_EVENT(mm_compaction_suitable_template, mm_compaction_suitable,
+
+	TP_PROTO(struct zone *zone,
+		int order,
+		int ret),
+
+	TP_ARGS(zone, order, ret)
+);
+
+DECLARE_EVENT_CLASS(mm_compaction_defer_template,
+
+	TP_PROTO(struct zone *zone, int order),
+
+	TP_ARGS(zone, order),
+
+	TP_STRUCT__entry(
+		__field(int, nid)
+		__field(enum zone_type, idx)
+		__field(int, order)
+		__field(unsigned int, considered)
+		__field(unsigned int, defer_shift)
+		__field(int, order_failed)
+	),
+
+	TP_fast_assign(
+		__entry->nid = zone_to_nid(zone);
+		__entry->idx = zone_idx(zone);
+		__entry->order = order;
+		__entry->considered = zone->compact_considered;
+		__entry->defer_shift = zone->compact_defer_shift;
+		__entry->order_failed = zone->compact_order_failed;
+	),
+
+	TP_printk("node=%d zone=%-8s order=%d order_failed=%d consider=%u limit=%lu",
+		__entry->nid,
+		__print_symbolic(__entry->idx, ZONE_TYPE),
+		__entry->order,
+		__entry->order_failed,
+		__entry->considered,
+		1UL << __entry->defer_shift)
+);
+
+DEFINE_EVENT(mm_compaction_defer_template, mm_compaction_deferred,
+
+	TP_PROTO(struct zone *zone, int order),
+
+	TP_ARGS(zone, order)
+);
+
+DEFINE_EVENT(mm_compaction_defer_template, mm_compaction_defer_compaction,
+
+	TP_PROTO(struct zone *zone, int order),
+
+	TP_ARGS(zone, order)
+);
+
+DEFINE_EVENT(mm_compaction_defer_template, mm_compaction_defer_reset,
+
+	TP_PROTO(struct zone *zone, int order),
+
+	TP_ARGS(zone, order)
+);
+
+TRACE_EVENT(mm_compaction_kcompactd_sleep,
+
+	TP_PROTO(int nid),
+
+	TP_ARGS(nid),
+
+	TP_STRUCT__entry(
+		__field(int, nid)
+	),
+
+	TP_fast_assign(
+		__entry->nid = nid;
+	),
+
+	TP_printk("nid=%d", __entry->nid)
+);
+
+DECLARE_EVENT_CLASS(kcompactd_wake_template,
+
+	TP_PROTO(int nid, int order, enum zone_type highest_zoneidx),
+
+	TP_ARGS(nid, order, highest_zoneidx),
+
+	TP_STRUCT__entry(
+		__field(int, nid)
+		__field(int, order)
+		__field(enum zone_type, highest_zoneidx)
+	),
+
+	TP_fast_assign(
+		__entry->nid = nid;
+		__entry->order = order;
+		__entry->highest_zoneidx = highest_zoneidx;
+	),
+
+	/*
+	 * classzone_idx is previous name of the highest_zoneidx.
+	 * Reason not to change it is the ABI requirement of the tracepoint.
+	 */
+	TP_printk("nid=%d order=%d classzone_idx=%-8s",
+		__entry->nid,
+		__entry->order,
+		__print_symbolic(__entry->highest_zoneidx, ZONE_TYPE))
+);
+
+DEFINE_EVENT(kcompactd_wake_template, mm_compaction_wakeup_kcompactd,
+
+	TP_PROTO(int nid, int order, enum zone_type highest_zoneidx),
+
+	TP_ARGS(nid, order, highest_zoneidx)
+);
+
+DEFINE_EVENT(kcompactd_wake_template, mm_compaction_kcompactd_wake,
+
+	TP_PROTO(int nid, int order, enum zone_type highest_zoneidx),
+
+	TP_ARGS(nid, order, highest_zoneidx)
+);
+#endif
+
+#endif /* _TRACE_COMPACTION_H */
+
+/* This part must be outside protection */
+#include <trace/define_trace.h>