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, 95 insertions, 0 deletions

diff --git a/tools/testing/radix-tree/regression3.c b/tools/testing/radix-tree/regression3.c
new file mode 100644
index 000000000..9f9a3b280
--- /dev/null
+++ b/tools/testing/radix-tree/regression3.c
@@ -0,0 +1,95 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Regression3
+ * Description:
+ * Helper radix_tree_iter_retry resets next_index to the current index.
+ * In following radix_tree_next_slot current chunk size becomes zero.
+ * This isn't checked and it tries to dereference null pointer in slot.
+ *
+ * Helper radix_tree_iter_resume reset slot to NULL and next_index to index + 1,
+ * for tagger iteraction it also must reset cached tags in iterator to abort
+ * next radix_tree_next_slot and go to slow-path into radix_tree_next_chunk.
+ *
+ * Running:
+ * This test should run to completion immediately. The above bug would
+ * cause it to segfault.
+ *
+ * Upstream commit:
+ * Not yet
+ */
+#include <linux/kernel.h>
+#include <linux/gfp.h>
+#include <linux/slab.h>
+#include <linux/radix-tree.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#include "regression.h"
+
+void regression3_test(void)
+{
+	RADIX_TREE(root, GFP_KERNEL);
+	void *ptr0 = (void *)4ul;
+	void *ptr = (void *)8ul;
+	struct radix_tree_iter iter;
+	void **slot;
+	bool first;
+
+	printv(1, "running regression test 3 (should take milliseconds)\n");
+
+	radix_tree_insert(&root, 0, ptr0);
+	radix_tree_tag_set(&root, 0, 0);
+
+	first = true;
+	radix_tree_for_each_tagged(slot, &root, &iter, 0, 0) {
+		printv(2, "tagged %ld %p\n", iter.index, *slot);
+		if (first) {
+			radix_tree_insert(&root, 1, ptr);
+			radix_tree_tag_set(&root, 1, 0);
+			first = false;
+		}
+		if (radix_tree_deref_retry(*slot)) {
+			printv(2, "retry at %ld\n", iter.index);
+			slot = radix_tree_iter_retry(&iter);
+			continue;
+		}
+	}
+	radix_tree_delete(&root, 1);
+
+	first = true;
+	radix_tree_for_each_slot(slot, &root, &iter, 0) {
+		printv(2, "slot %ld %p\n", iter.index, *slot);
+		if (first) {
+			radix_tree_insert(&root, 1, ptr);
+			first = false;
+		}
+		if (radix_tree_deref_retry(*slot)) {
+			printv(2, "retry at %ld\n", iter.index);
+			slot = radix_tree_iter_retry(&iter);
+			continue;
+		}
+	}
+
+	radix_tree_for_each_slot(slot, &root, &iter, 0) {
+		printv(2, "slot %ld %p\n", iter.index, *slot);
+		if (!iter.index) {
+			printv(2, "next at %ld\n", iter.index);
+			slot = radix_tree_iter_resume(slot, &iter);
+		}
+	}
+
+	radix_tree_tag_set(&root, 0, 0);
+	radix_tree_tag_set(&root, 1, 0);
+	radix_tree_for_each_tagged(slot, &root, &iter, 0, 0) {
+		printv(2, "tagged %ld %p\n", iter.index, *slot);
+		if (!iter.index) {
+			printv(2, "next at %ld\n", iter.index);
+			slot = radix_tree_iter_resume(slot, &iter);
+		}
+	}
+
+	radix_tree_delete(&root, 0);
+	radix_tree_delete(&root, 1);
+
+	printv(1, "regression test 3 passed\n");
+}