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, 84 insertions, 0 deletions

diff --git a/drivers/md/dm-audit.c b/drivers/md/dm-audit.c
new file mode 100644
index 000000000..2e979eeb1
--- /dev/null
+++ b/drivers/md/dm-audit.c
@@ -0,0 +1,84 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Creating audit records for mapped devices.
+ *
+ * Copyright (C) 2021 Fraunhofer AISEC. All rights reserved.
+ *
+ * Authors: Michael Weiß <michael.weiss@aisec.fraunhofer.de>
+ */
+
+#include <linux/audit.h>
+#include <linux/module.h>
+#include <linux/device-mapper.h>
+#include <linux/bio.h>
+#include <linux/blkdev.h>
+
+#include "dm-audit.h"
+#include "dm-core.h"
+
+static struct audit_buffer *dm_audit_log_start(int audit_type,
+					       const char *dm_msg_prefix,
+					       const char *op)
+{
+	struct audit_buffer *ab;
+
+	if (audit_enabled == AUDIT_OFF)
+		return NULL;
+
+	ab = audit_log_start(audit_context(), GFP_KERNEL, audit_type);
+	if (unlikely(!ab))
+		return NULL;
+
+	audit_log_format(ab, "module=%s op=%s", dm_msg_prefix, op);
+	return ab;
+}
+
+void dm_audit_log_ti(int audit_type, const char *dm_msg_prefix, const char *op,
+		     struct dm_target *ti, int result)
+{
+	struct audit_buffer *ab = NULL;
+	struct mapped_device *md = dm_table_get_md(ti->table);
+	int dev_major = dm_disk(md)->major;
+	int dev_minor = dm_disk(md)->first_minor;
+
+	switch (audit_type) {
+	case AUDIT_DM_CTRL:
+		ab = dm_audit_log_start(audit_type, dm_msg_prefix, op);
+		if (unlikely(!ab))
+			return;
+		audit_log_task_info(ab);
+		audit_log_format(ab, " dev=%d:%d error_msg='%s'", dev_major,
+				 dev_minor, !result ? ti->error : "success");
+		break;
+	case AUDIT_DM_EVENT:
+		ab = dm_audit_log_start(audit_type, dm_msg_prefix, op);
+		if (unlikely(!ab))
+			return;
+		audit_log_format(ab, " dev=%d:%d sector=?", dev_major,
+				 dev_minor);
+		break;
+	default: /* unintended use */
+		return;
+	}
+
+	audit_log_format(ab, " res=%d", result);
+	audit_log_end(ab);
+}
+EXPORT_SYMBOL_GPL(dm_audit_log_ti);
+
+void dm_audit_log_bio(const char *dm_msg_prefix, const char *op,
+		      struct bio *bio, sector_t sector, int result)
+{
+	struct audit_buffer *ab;
+	int dev_major = MAJOR(bio->bi_bdev->bd_dev);
+	int dev_minor = MINOR(bio->bi_bdev->bd_dev);
+
+	ab = dm_audit_log_start(AUDIT_DM_EVENT, dm_msg_prefix, op);
+	if (unlikely(!ab))
+		return;
+
+	audit_log_format(ab, " dev=%d:%d sector=%llu res=%d",
+			 dev_major, dev_minor, sector, result);
+	audit_log_end(ab);
+}
+EXPORT_SYMBOL_GPL(dm_audit_log_bio);