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, 57 insertions, 0 deletions

diff --git a/tools/testing/vma/main.c b/tools/testing/vma/main.c
new file mode 100644
index 000000000..18338f5d2
--- /dev/null
+++ b/tools/testing/vma/main.c
@@ -0,0 +1,57 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#include "shared.h"
+/*
+ * Directly import the VMA implementation here. Our vma_internal.h wrapper
+ * provides userland-equivalent functionality for everything vma.c uses.
+ */
+#include "../../../mm/vma_init.c"
+#include "../../../mm/vma_exec.c"
+#include "../../../mm/vma.c"
+
+/* Tests are included directly so they can test static functions in mm/vma.c. */
+#include "tests/merge.c"
+#include "tests/mmap.c"
+#include "tests/vma.c"
+
+int sysctl_max_map_count __read_mostly = DEFAULT_MAX_MAP_COUNT;
+
+/* Helper functions which utilise static kernel functions. */
+
+struct vm_area_struct *merge_existing(struct vma_merge_struct *vmg)
+{
+	struct vm_area_struct *vma;
+
+	vma = vma_merge_existing_range(vmg);
+	if (vma)
+		vma_assert_attached(vma);
+	return vma;
+}
+
+int attach_vma(struct mm_struct *mm, struct vm_area_struct *vma)
+{
+	int res;
+
+	res = vma_link(mm, vma);
+	if (!res)
+		vma_assert_attached(vma);
+	return res;
+}
+
+/* Main test running which invokes tests/ *.c runners. */
+int main(void)
+{
+	int num_tests = 0, num_fail = 0;
+
+	maple_tree_init();
+	vma_state_init();
+
+	run_merge_tests(&num_tests, &num_fail);
+	run_mmap_tests(&num_tests, &num_fail);
+	run_vma_tests(&num_tests, &num_fail);
+
+	printf("%d tests run, %d passed, %d failed.\n",
+	       num_tests, num_tests - num_fail, num_fail);
+
+	return num_fail == 0 ? EXIT_SUCCESS : EXIT_FAILURE;
+}