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, 82 insertions, 0 deletions

diff --git a/tools/perf/tests/mem2node.c b/tools/perf/tests/mem2node.c
new file mode 100644
index 000000000..7ce1ad7b6
--- /dev/null
+++ b/tools/perf/tests/mem2node.c
@@ -0,0 +1,82 @@
+// SPDX-License-Identifier: GPL-2.0
+#include <linux/compiler.h>
+#include <linux/bitmap.h>
+#include <linux/kernel.h>
+#include <linux/zalloc.h>
+#include <perf/cpumap.h>
+#include <internal/cpumap.h>
+#include "debug.h"
+#include "env.h"
+#include "mem2node.h"
+#include "tests.h"
+
+static struct node {
+	int		 node;
+	const char 	*map;
+} test_nodes[] = {
+	{ .node = 0, .map = "0"     },
+	{ .node = 1, .map = "1-2"   },
+	{ .node = 3, .map = "5-7,9" },
+};
+
+#define T TEST_ASSERT_VAL
+
+static unsigned long *get_bitmap(const char *str, int nbits)
+{
+	struct perf_cpu_map *map = perf_cpu_map__new(str);
+	unsigned long *bm = NULL;
+
+	bm = bitmap_zalloc(nbits);
+
+	if (map && bm) {
+		struct perf_cpu cpu;
+		unsigned int i;
+
+		perf_cpu_map__for_each_cpu(cpu, i, map)
+			__set_bit(cpu.cpu, bm);
+	}
+
+	if (map)
+		perf_cpu_map__put(map);
+	else
+		free(bm);
+
+	return bm && map ? bm : NULL;
+}
+
+static int test__mem2node(struct test_suite *t __maybe_unused, int subtest __maybe_unused)
+{
+	struct mem2node map;
+	struct memory_node nodes[3];
+	struct perf_env env = {
+		.memory_nodes    = (struct memory_node *) &nodes[0],
+		.nr_memory_nodes = ARRAY_SIZE(nodes),
+		.memory_bsize    = 0x100,
+	};
+	unsigned int i;
+
+	for (i = 0; i < ARRAY_SIZE(nodes); i++) {
+		nodes[i].node = test_nodes[i].node;
+		nodes[i].size = 10;
+
+		T("failed: alloc bitmap",
+		  (nodes[i].set = get_bitmap(test_nodes[i].map, 10)));
+	}
+
+	T("failed: mem2node__init", !mem2node__init(&map, &env));
+	T("failed: mem2node__node",  0 == mem2node__node(&map,   0x50));
+	T("failed: mem2node__node",  1 == mem2node__node(&map,  0x100));
+	T("failed: mem2node__node",  1 == mem2node__node(&map,  0x250));
+	T("failed: mem2node__node",  3 == mem2node__node(&map,  0x500));
+	T("failed: mem2node__node",  3 == mem2node__node(&map,  0x650));
+	T("failed: mem2node__node", -1 == mem2node__node(&map,  0x450));
+	T("failed: mem2node__node", -1 == mem2node__node(&map, 0x1050));
+
+	for (i = 0; i < ARRAY_SIZE(nodes); i++)
+		zfree(&nodes[i].set);
+
+	mem2node__exit(&map);
+	return 0;
+}
+
+DEFINE_SUITE("mem2node", mem2node);